Methods for identifying features associated with clinical response and uses thereof

Abstract

The present disclosure relates to methods for identifying features associated with the clinical response of a subject, e.g., a patient, after treatment with a therapeutic cell composition associated with cell therapy, e.g., attributes of the subject, the therapeutic cell composition, and the input composition used to generate the therapeutic cell composition. Cells of the therapeutic cell composition express a recombinant receptor, e.g., a chimeric receptor, e.g., a chimeric antigen receptor (CAR) or other transgenic receptor, e.g., a T cell receptor (TCR). The methods result in the identification of features associated with the clinical response. In some embodiments, the methods can be used to determine (e.g., predict) a subject's response to treatment with a therapeutic cell composition. TIFF2023528215000025.tif99150

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Application No. 63 / 024,494, filed May 13, 2020, entitled "METHODS OF IDENTIFYING FEATURES ASSOCIATED WITH CLINICAL RESPONSE AND USES THEREOF," and U.S. Provisional Application No. 63 / 037,592, filed June 10, 2020, entitled "METHODS OF IDENTIFYING FEATURES ASSOCIATED WITH CLINICAL RESPONSE AND USES THEREOF," the contents of which are incorporated by reference in their entireties.

[0002] INCORPORATION-BY-REFERENCE TO SEQUENCE LISTING This application is filed with an electronic Sequence Listing. The Sequence Listing is provided as a file entitled 735042023740SeqList.TXT, created on May 8, 2021, and is 51,373 bytes in size. The information in the electronic format of the Sequence Listing is incorporated by reference in its entirety.

[0003] Field The present disclosure relates to methods for identifying features associated with the clinical response of a subject, e.g., a patient, after treatment with a therapeutic cell composition associated with cell therapy, e.g., attributes of the subject, the therapeutic cell composition, and the input composition used to generate the therapeutic cell composition. Cells of the therapeutic cell composition express a recombinant receptor, e.g., a chimeric receptor, e.g., a chimeric antigen receptor (CAR) or other transgenic receptor, e.g., a T cell receptor (TCR). The methods result in the identification of features associated with the clinical response. In some embodiments, the methods can be used to determine (e.g., predict) a subject's response to treatment with a therapeutic cell composition. [Background technology]

[0004] background A variety of immunotherapy and / or cell therapy methods are available for treating disease and pathology.For example, adoptive cell therapy (including the administration of cells that express chimeric receptors, such as chimeric antigen receptors (CAR) and / or other recombinant antigen receptors, that are specific to the disease or disorder of interest, and other adoptive immune cell therapy and adoptive T cell therapy) can be beneficial for the treatment of cancer or other diseases or disorders.There is a need for an improved method for determining whether treatment will bring about beneficial clinical response.The method that addresses this need is provided herein. Summary of the Invention

[0005] overview 1. A method for identifying features associated with a clinical response, the method comprising: (a) (i) subject features determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, the therapeutic cell composition being for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from a sample derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a CAR; and (iii) therapeutic cell composition features determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions generated from one of the plurality of input compositions and expressing the CAR. and (b) receiving features including the therapeutic cell composition features, wherein the therapeutic composition is administered to one of the plurality of subjects; (b) preprocessing the features to identify informative features, wherein the informative features comprise a subset of features including one or more of the subject features, one or more of the input composition features, and one or more of the therapeutic cell composition features; (c) obtaining a clinical response from each of the plurality of subjects after treatment with one of the plurality of therapeutic compositions; (d) applying the informative features and obtained clinical responses from the plurality of subjects as inputs to train a random forest model using supervised learning; and (e) identifying informative features associated with the clinical response from the trained random forest model.

[0006] 1. A method for identifying features associated with a clinical response, the method comprising: (a) (i) subject features determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, the therapeutic cell composition being for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from a sample derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) therapeutic cell composition features determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions generated from one of the plurality of input compositions and expressing the CAR. and (b) receiving features including the therapeutic cell composition features, wherein the therapeutic composition is administered to one of the plurality of subjects; (b) preprocessing the features to identify informative features, wherein the informative features comprise a subset of features including one or more of the subject features, one or more of the input composition features, and one or more of the therapeutic cell composition features; (c) obtaining a clinical response from each of the plurality of subjects over time after treatment with one of the plurality of therapeutic compositions; (d) applying the informative features and clinical responses from the plurality of subjects as inputs to train a random survival forest model using supervised learning; and (e) identifying informative features associated with the clinical response from the trained random survival forest model.

[0007] In some embodiments of any of the methods provided herein, identifying informative features associated with clinical response includes determining an importance measure for each of the informative features. In some embodiments, the importance measure includes a permutation importance measure, an average minimum depth, and / or a random forest, e.g., a trained random forest model, with the root node split by the informative features. In some embodiments, the importance measure includes a permutation importance measure, an average minimum depth, and / or a random survival forest, e.g., a trained random survival forest model, with the root node split by the informative features. In some embodiments, the importance measure is a permutation importance measure. In some embodiments, the importance measure is an average minimum depth. In some embodiments, the importance measure is a random forest, e.g., a trained random forest model, with the root node split by the informative features. In some embodiments, the importance measure is a random survival forest, e.g., a trained random survival forest model, with the root node split by the informative features. In some embodiments, the informative features associated with the clinical response are the first 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 informative features identified by ranking, e.g., ranking the importance scale values ​​of each of the informative features, wherein the importance scale is the same for each informative feature. In some embodiments, the informative features associated with the clinical response are the first 5 informative features identified by ranking the importance scale values ​​of each of the informative features, wherein the importance scale is the same for each informative feature. In some embodiments, the informative features associated with the clinical response are the first informative feature identified by ranking the importance scale values ​​of each of the informative features, wherein the importance scale is the same for each informative feature.

[0008] 1. A method of determining, e.g., predicting, a clinical response, comprising: (a) (i) a subject feature determined from a subject before the subject is treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, the therapeutic cell composition being for treating the disease or condition; (ii) an input composition feature determined from an input composition comprising T cells selected from a sample from the subject, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition determined from the therapeutic cell composition. Provided herein is a method comprising: (a) receiving features comprising the therapeutic cell composition features, wherein the therapeutic cell composition is produced from the input composition and expresses the CAR, and the therapeutic composition is administered to the subject; and (b) applying the features as inputs to a random forest model trained to determine, e.g., predict, the subject's clinical response to treatment with the therapeutic cell composition based on informative features identified by preprocessing prior to treating the subject with the therapeutic cell composition, wherein the features applied as inputs are the same informative features used to train the random forest model.

[0009] A method of determining, e.g., predicting, a clinical response, comprising: (a) (i) a subject feature determined from a subject before the subject is treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, the therapeutic cell composition being for treating the disease or condition; (ii) an input composition feature determined from an input composition comprising T cells selected from a sample from the subject, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from the therapeutic cell composition. and (b) receiving features comprising the therapeutic cell composition features, whereby the therapeutic cell composition is produced from the input composition and expresses the CAR, and the therapeutic composition is administered to the subject; and (b) applying the features as inputs to a random survival forest model that has been trained to determine, e.g., predict, the subject's clinical response to treatment with the therapeutic cell composition based on the informative features identified by preprocessing prior to treating the subject with the therapeutic cell composition, wherein the features applied as inputs are the same informative features used to train the random survival forest model.

[0010] A method of treating a subject, comprising: (a) selecting T cells from a sample from the subject to generate an input composition comprising the T cells; (b) measuring (i) a subject feature determined from the subject before the subject is treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, the therapeutic cell composition being for treating the disease or condition; (ii) an input composition feature determined from an input composition comprising T cells selected from a sample from the subject, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from the therapeutic cell composition, the therapeutic cell composition being generated from the input composition and expressing the CAR, the therapeutic cell composition being administered to the subject; and (c) measuring the feature, including the therapeutic cell composition feature. applying the amount of the therapeutic cell composition as input to a random forest model that has been trained to determine, e.g., predict, the subject's clinical response to treatment with the therapeutic cell composition based on the informative features identified by preprocessing prior to treating the subject with the therapeutic cell composition, wherein the features applied as input are the same informative features used to train the random forest model; and administering a treatment to the subject, wherein: (1) a predetermined treatment regimen comprising the therapeutic cell composition is administered when the subject is determined, e.g., predicted, to have a clinical response selected from the group consisting of complete response (CR), partial response (PR), durable response of greater than 3 months, progression-free survival (PFS) of greater than 3 months, overall response rate (ORR), objective response (OR), a desired pharmacokinetic response that is or is greater than the target pharmacokinetic response, and a no-toxicity response or a mild toxic response (optionally, the toxicity is CRS of grade 2 or less or neurotoxicity of grade 2 or less);Alternatively, (2) if the subject is determined, e.g., predicted, to have a clinical response selected from the group consisting of a toxic response (optionally, the toxic response is severe cytokine release syndrome or severe neurotoxicity), a pharmacokinetic response that is poor compared to the target pharmacokinetic response, progressive disease (PD), a durable response of less than 3 months, and a PFS of less than 3 months, administering to the subject a therapeutic regimen comprising the therapeutic cell composition that is altered compared to the predetermined therapeutic regimen comprising the therapeutic cell composition, the method provided herein comprises the step of:

[0011] A method of treating a subject, comprising: (a) selecting T cells from a sample from the subject to generate an input composition comprising the T cells; (b) measuring (i) subject features determined from the subject before the subject is treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, the therapeutic cell composition being for treating the disease or condition; (ii) input composition features determined from an input composition comprising T cells selected from a sample from the subject, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) therapeutic cell composition features determined from the therapeutic cell composition, the therapeutic cell composition being generated from the input composition and expressing the CAR, the therapeutic cell composition being administered to the subject; and (c) measuring the features, including the therapeutic cell composition features, determined prior to treatment of the subject with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR), the therapeutic cell composition being used to generate the therapeutic cell composition; applying as inputs to a random survival forest model that has been trained to determine, e.g., predict, a clinical response in the subject to be treated with the therapeutic cell composition based on informative features identified by the method prior to treating the subject with the therapeutic cell composition, wherein the features applied as inputs are the same informative features used to train the random survival forest model; and administering a treatment to the subject, wherein: (1) a predetermined treatment regimen comprising the therapeutic cell composition is administered when the subject is determined, e.g., predicted, to have a clinical response selected from the group consisting of complete response (CR), partial response (PR), durable response of greater than 3 months, progression-free survival (PFS) of greater than 3 months, overall response rate (ORR), objective response (OR), a desired pharmacokinetic response that is or is greater than the target pharmacokinetic response, and a no-toxicity response or a mild toxic response (optionally, the toxicity is CRS of grade 2 or less or neurotoxicity of grade 2 or less);Alternatively, (2) when the subject is determined, e.g., predicted, to have a clinical response selected from the group consisting of a toxic response (optionally, the toxic response is severe cytokine release syndrome or severe neurotoxicity), a pharmacokinetic response that is poor compared to the target pharmacokinetic response, progressive disease (PD), a durable response of less than 3 months, and a PFS of less than 3 months, administering to the subject a therapeutic regimen comprising the therapeutic cell composition that is altered compared to the predetermined therapeutic regimen comprising the therapeutic cell composition is provided herein;

[0012] In some of either embodiment, the method further comprises generating a therapeutic cell composition.

[0013] A method of treating a subject, comprising: (a) selecting T cells from a sample derived from the subject to generate an input composition comprising the T cells; (b) generating a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition (i) is for treating the disease or condition, (ii) is generated from the input composition, and (iii) is administered to the subject; (c) measuring features including (i) subject features determined from the subject prior to treatment with the therapeutic cell composition; (ii) input composition features determined from the input composition; and (iii) therapeutic cell composition features determined from the therapeutic cell composition; and (c) assessing a clinical response in the subject treated with the therapeutic cell composition based on the features identified by preprocessing. (d) applying as inputs to a random forest model trained to determine whether the features are the same informative features as those used to train the random forest model prior to treating the subject with the therapeutic cell composition, wherein the features applied as inputs are the same informative features as those used to train the random forest model; and (d) administering a treatment to the subject, wherein (1) if the subject is determined to have a clinical response selected from the group consisting of complete response (CR), partial response (PR), durable response of greater than 3 months, progression-free survival (PFS) of greater than 3 months, objective response (OR), a desired pharmacokinetic response that is or is greater than the target pharmacokinetic response, and a no-toxicity response or a mild toxic response (optionally, a mild toxic response is cytokine release syndrome (CRS) of grade 2 or less or neurotoxicity of grade 2 or less), a predetermined treatment regimen comprising the therapeutic cell composition is administered;Alternatively, (2) if the subject is determined to have a clinical response selected from the group consisting of a toxic response (optionally, the toxic response is severe cytokine release syndrome (CRS) or severe neurotoxicity), a poor pharmacokinetic response compared to the target pharmacokinetic response, progressive disease (PD), a durable response of less than 3 months, and a PFS of less than 3 months, administering to the subject a therapeutic regimen comprising the therapeutic cell composition that is altered compared to the predetermined therapeutic regimen comprising the therapeutic cell composition, is provided herein in some embodiments.

[0014] A method of treating a subject, comprising: (a) selecting T cells from a sample derived from the subject to generate an input composition comprising the T cells; (b) generating a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition (i) is for treating the disease or condition, (ii) is generated from the input composition, and (iii) is administered to the subject; (c) measuring features including (i) subject features determined from the subject prior to treatment with the therapeutic cell composition; (ii) input composition features determined from the input composition; and (iii) therapeutic cell composition features determined from the therapeutic cell composition; and (c) assessing a clinical response in a subject treated with the therapeutic cell composition based on the features identified by preprocessing and the therapeutic cell composition. (d) applying as inputs to a random survival forest model trained to determine whether the subject has a clinical response selected from the group consisting of a complete response (CR), a partial response (PR), a durable response of more than 3 months, a progression-free survival (PFS) of more than 3 months, an objective response (OR), a desired pharmacokinetic response that is or is greater than the target pharmacokinetic response, and a no-toxicity response or a mild toxic response (optionally, a mild toxic response is cytokine release syndrome (CRS) of Grade 2 or less or neurotoxicity of Grade 2 or less), administering a predetermined treatment regimen comprising the therapeutic cell composition;Alternatively, (2) if the subject is determined to have a clinical response selected from the group consisting of a toxic response (optionally, the toxic response is severe cytokine release syndrome (CRS) or severe neurotoxicity), a poor pharmacokinetic response compared to the target pharmacokinetic response, progressive disease (PD), a durable response of less than 3 months, and a PFS of less than 3 months, administering to the subject a therapeutic regimen comprising the therapeutic cell composition that is altered compared to the predetermined therapeutic regimen comprising the therapeutic cell composition, is provided herein in some embodiments.

[0015] In some embodiments, the random forest model is trained to determine whether the subject will have a complete response (CR).In some embodiments, (1) if the subject is determined to have a complete response (CR), the subject is administered the predetermined treatment regimen; or (2) if the subject is determined to have progressive disease (PD), the subject is administered the modified treatment regimen.

[0016] In some embodiments, the random forest model is trained to determine whether the subject will have a partial response (PR).In some embodiments, (1) if the subject is determined to have a partial response (PR), the subject is administered the predetermined treatment regimen; or (2) if the subject is determined to have progressive disease (PD), the subject is administered the modified treatment regimen.

[0017] In some embodiments, Random Forest model is trained to determine whether the subject will have a durable response of more than 3 months.In some embodiments, Random Forest survival model is trained to determine whether the subject will have a durable response of more than 3 months.In some embodiments, (1) if the subject is determined to have a durable response of more than 3 months, the subject is administered the predetermined treatment regimen; or (2) if the subject is determined to have a durable response of less than 3 months, the subject is administered the modified treatment regimen.

[0018] In some embodiments, random forest model is trained to determine whether the subject will have progression-free survival (PFS) of more than 3 months.In some embodiments, random survival forest model is trained to determine whether the subject will have progression-free survival (PFS) of more than 3 months.In some embodiments, (1) if the subject is determined to have progression-free survival (PFS) of more than 3 months, the subject is administered the predetermined treatment regimen; or (2) if the subject is determined to have progression-free survival (PFS) of less than 3 months, the subject is administered the modified treatment regimen.

[0019] In some embodiments, the random forest model is trained to determine whether the subject will have an objective response (OR). In some embodiments, (1) if the subject is determined to have an objective response (OR), the subject is administered the predetermined treatment regimen, or (2) if the subject is determined to have progressive disease (PD), the subject is administered the modified treatment regimen.

[0020] In some embodiments, a random forest model is trained to determine the pharmacokinetic response of a subject.In some embodiments, (1) if the subject is determined to have a desired pharmacokinetic response that is equal to or greater than the target pharmacokinetic response, the subject is administered the predetermined treatment regimen; or (2) if the subject is determined to have a lower pharmacokinetic response than the target pharmacokinetic response, the subject is administered the modified treatment regimen.

[0021] In some embodiments, random forest model is trained to determine whether the subject will have toxic response.In some embodiments, (1) if the subject is determined to have no toxic response or mild toxic response, the subject is administered the predetermined treatment regimen, or (2) if the subject is determined to have toxic response, the subject is administered the modified treatment regimen.In some embodiments, the toxic response is severe CRS.In some embodiments, the toxic response is severe neurotoxicity.

[0022] In some embodiments, the random forest model is trained using supervised training, wherein the supervised training includes: (a) (i) subject features determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from a sample from each of the plurality of subjects, wherein the T cells are used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) therapeutic cell features determined from each of a plurality of therapeutic cell compositions. (b) receiving features including therapeutic cell composition features, each of the plurality of therapeutic cell compositions produced from one of the plurality of input compositions and expressing the CAR, and the therapeutic compositions administered to one of the plurality of subjects; (b) preprocessing the features to identify informative features, the informative features comprising a subset of features including one or more subject features, one or more input composition features, and one or more therapeutic cell composition features; (c) obtaining a clinical response from each of the plurality of subjects after treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features from the plurality of subjects and the obtained clinical response as input to train a random forest model.

[0023] In some embodiments, the Random Survival Forest model is trained using supervised training, wherein the supervised training comprises: (a) (i) subject features determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from a sample from each of the plurality of subjects, wherein the T cells are used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) therapeutic cell composition features determined from each of a plurality of therapeutic cell compositions. The method includes (a) receiving features including therapeutic cell composition features, each of the plurality of therapeutic cell compositions produced from one of the plurality of input compositions and expressing the CAR, and the therapeutic composition being administered to one of the plurality of subjects; (b) preprocessing the features to identify informative features, the informative features comprising a subset of features including one or more subject features, one or more input composition features, and one or more therapeutic cell composition features; (c) obtaining clinical responses from each of the plurality of subjects over time after treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features and clinical responses from the plurality of subjects as inputs to train a random survival forest model using supervised learning.

[0024] (a) (i) a subject feature determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, the therapeutic cell composition being for treating the disease or condition; (ii) an input composition feature determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from a sample derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions being treated with a therapeutic cell composition selected from the plurality of input compositions. Provided herein are methods for developing a random forest model, the methods comprising: (a) receiving features comprising a therapeutic cell composition produced from one of the plurality of therapeutic compositions and expressing the CAR, wherein the therapeutic composition is administered to one of the plurality of subjects; (b) preprocessing the features to identify informative features, wherein the informative features comprise a subset of features comprising one or more subject features, one or more input composition features, and one or more therapeutic cell composition features; (c) obtaining a clinical response from each of the plurality of subjects after treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features from the plurality of subjects and the obtained clinical response as input to train a random forest model.

[0025] (a) (i) subject features determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, the therapeutic cell composition being for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from a sample derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) therapeutic cell composition features determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions being selected from one of the plurality of input compositions. Provided herein is a method for developing a random survival forest model, comprising: (a) receiving features comprising the therapeutic cell composition features, wherein the therapeutic cell composition has been generated and expresses the CAR, and the therapeutic composition is administered to one of the plurality of subjects; (b) preprocessing the features to identify informative features, wherein the informative features comprise a subset of features comprising one or more subject features, one or more input composition features, and one or more therapeutic cell composition features; (c) obtaining clinical responses from each of the plurality of subjects over time after treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features and clinical responses from the plurality of subjects as inputs to train a random survival forest model using supervised learning.

[0026] In some embodiments, each of the plurality of subjects is administered one of the plurality of therapeutic cell compositions, and the one therapeutic cell composition administered to a subject is a therapeutic cell composition made from an input composition of a sample derived from the subject.

[0027] In some embodiments, the preprocessing to identify informative features includes: a) removing target features, input composition features, and therapeutic cell composition features with greater than 50%, greater than about 50%, or 50% missing data; b) removing target features, input composition features, and therapeutic cell composition features with zero variance, or greater than 95%, greater than about 95%, or 95% of data values ​​equal to a single value, and / or fewer than 0.1n unique values, where n = the number of samples; c) imputing missing data for target features, input composition features, and therapeutic cell composition features using multivariate imputation with chained equations. d) identifying a covariate cluster comprising a set of subject features, input composition features, and therapeutic cell composition features, and combinations thereof, having a correlation coefficient greater than, about, or equal to 0.5, and iteratively selecting subject features, input composition features, and therapeutic cell composition features from the covariate cluster, wherein the selected subject features, input composition features, and therapeutic cell composition features have the smallest average absolute correlation with all remaining subject features, input composition features, and therapeutic cell composition features. In some embodiments, the preprocessing to identify informative features includes or is the step of removing subject features, input composition features, and therapeutic cell composition features having greater than, about, or 50% data deficiency. In some embodiments, preprocessing to identify informative features includes or is removing target features, input composition features, and therapeutic cellular composition features that have zero variance or greater than, about, or 95% of data values ​​equal to a single value, and fewer than 0.1n unique values, where n = the number of samples. In some embodiments, preprocessing to identify informative features includes or is imputing missing data for target features, input composition features, and therapeutic cellular composition features by multivariate imputation with chained equations.In some embodiments, the preprocessing to identify informative features includes or is a step of identifying a covariate cluster comprising a set of subject features, input composition features, and therapeutic cell composition features and combinations thereof that have a correlation coefficient greater than, about, or equal to 0.5, and iteratively selecting subject features, input composition features, and therapeutic cell composition features from the covariate cluster, wherein the selected subject features, input composition features, and therapeutic cell composition features have the smallest average absolute correlation with all remaining subject features, input composition features, and therapeutic cell composition features.

[0028] In some embodiments, the random forest model is evaluated using cross-validation. In some embodiments, the random survival forest model is evaluated using cross-validation. In some embodiments, the cross-validation is 10-fold cross-validation or at least 10-fold cross-validation. In some embodiments, the cross-validation is nested cross-validation.

[0029] In some embodiments, the plurality of subjects is 500, 400, 300, 200, 150, 100, 50, 25, 15, or 10 subjects, or about 500, 400, 300, 200, 150, 100, 50, 25, 15, or 10 subjects, or any number between any of the foregoing. In some embodiments, the plurality of subjects is 10, about 10, or more than 10 subjects and less than 250 subjects. In some embodiments, the plurality of subjects is 20, about 20, or more than 20 subjects and less than 200 subjects. In some embodiments, the plurality of subjects is 20, about 20, or more than 20 subjects and less than 150 subjects. In some embodiments, the plurality of subjects is 20, about 20, or more than 20 subjects and less than 150 subjects. In some embodiments, the plurality of subjects is 20, about 20, or more than 20 subjects and less than 100 subjects. In some embodiments, the plurality of subjects is participating in a clinical trial.

[0030] In some embodiments, the subject features include one or more of subject attributes and clinical attributes. In some embodiments, the subject attributes include one or more of age, weight, height, ethnicity, race, sex, and body mass index. In some embodiments, the clinical attributes include one or more of biomarkers, disease diagnosis, disease burden, disease duration, disease grade, and treatment history. In some embodiments, the input composition features include a cell phenotype. In some embodiments, the therapeutic cell composition features include one or more of a cell phenotype, recombinant receptor-dependent activity, and dose. In some embodiments, the clinical response includes one or more of a complete response (CR), a partial response (PR), a durable response, progression-free survival (PFS), overall response rate (ORR), objective response (OR), a pharmacokinetic response that is or is greater than the target pharmacokinetic response, a no toxic response or a mild toxic response, a toxic response, a pharmacokinetic response that is lower than the target response, or one or more of no CR, PR, durable response, ORR, OR, or PFS.

[0031] In some embodiments, the clinical response is or comprises a complete response (CR), a partial response (PR), a durable response, progression-free survival (PFS), an objective response (OR), a pharmacokinetic response that is or is greater than the target pharmacokinetic response, no toxic response or a mild toxic response, a toxic response, a pharmacokinetic response that is lower than the target response, or no CR, PR, durable response, or objective response (OR).

[0032] In some embodiments, the clinical response is a complete response (CR). In some embodiments, the clinical response is no complete response (CR). In some embodiments, the clinical response is a partial response (PR). In some embodiments, the clinical response is no partial response (PR). In some embodiments, the clinical response is an objective response (OR). In some embodiments, the clinical response is no objective response (OR). In some embodiments, the clinical response is a toxic response. In some embodiments, the clinical response is no toxic response. In some embodiments, the toxic response is a mild toxic response. In some embodiments, the toxic response is a severe toxic response. In some embodiments, the toxic response is severe CRS. In some embodiments, the toxic response is severe neurotoxicity. In some embodiments, the clinical response is a durable response. In some embodiments, the clinical response is no durable response. In some embodiments, the clinical response is a duration of response (DOR). In some embodiments, the clinical response is a duration of response (DOR) of at least 3 months or at least about 3 months. In some embodiments, the clinical response is progression-free survival (PFS). In some embodiments, the clinical response is a progression-free survival (PFS) of at least 3 months or at least about 3 months. In some embodiments, the clinical response is a target pharmacokinetic response or a pharmacokinetic response greater than the target pharmacokinetic response. In some embodiments, the pharmacokinetic response is a measure of the expansion of CAR T cells of a therapeutic cell composition after treatment of a subject with the therapeutic cell composition. In some embodiments, the pharmacokinetic response is a measure of the maximum CAR T cell concentration in a subject after treatment of the subject with a therapeutic cell composition. In some embodiments, the pharmacokinetic response is a measure of the time point at which the CAR T cell concentration is maximum in the subject after treatment of the subject with a therapeutic cell composition. In some embodiments, the pharmacokinetic response is a measure of the exposure of the subject to CAR T cells of the therapeutic cell composition after treatment of the subject with the therapeutic cell composition.

[0033] In some embodiments, the sample comprises a whole blood sample, a buffy coat sample, a peripheral blood mononuclear cell (PBMC) sample, an unfractionated T cell sample, a lymphocyte sample, a leukocyte sample, an apheresis product, or a leukocyte apheresis product. In some embodiments, the sample is an apheresis product or a leukocyte apheresis product. In some embodiments, the apheresis product or a leukocyte apheresis product has been previously cryopreserved. In some embodiments, the T cells comprise primary cells obtained from a subject. In some embodiments, the T cells comprise CD3+, CD4+, and / or CD8+.

[0034] In some embodiments, the input composition comprises CD4+, CD8+, or CD4+ and CD8+ T cells, the therapeutic cell composition comprises CD4+, CD8+, or CD4+ and CD8+ T cells expressing a recombinant receptor and is produced from the input composition, the input composition features comprise input composition features derived from the CD4+, CD8+, or CD4+ and CD8+ T cell composition of the input composition, and the therapeutic cell composition features comprise therapeutic cell composition features derived from the CD4+, CD8+, or CD4+ and CD8+ T cells of the therapeutic composition.

[0035] In some embodiments, the input composition comprises separate compositions of CD4+ and CD8+ T cells, the therapeutic cell composition comprises separate compositions of CD4+ and CD8+ T cells expressing a recombinant receptor, and is generated from the respective CD4+ or CD8+ T cell compositions of the input composition, the input composition features comprise input composition features derived from the CD4+ and CD8+ T cell compositions of the input composition, and the therapeutic cell composition features comprise therapeutic cell composition features derived from each of the CD4+ and CD8+ T cells of the separate compositions of the therapeutic composition.

[0036] In some embodiments, the input composition comprises separate compositions of CD4+ and CD8+ T cells, and the therapeutic cell composition comprises a mixed composition of CD4+ and CD8+ T cells expressing a recombinant receptor and is generated from the separate compositions of CD4+ and CD8+ T cells of the input composition, and the input composition features comprise input composition features derived from the separate compositions of CD4+ and CD8+ T cells of the input composition, and the therapeutic cell composition features comprise therapeutic cell composition features derived from the mixed composition of CD4+ and CD8+ cells of the therapeutic composition.

[0037] In some embodiments, the recombinant receptor is a chimeric antigen receptor (CAR). In some embodiments, the predetermined treatment regimen comprises: a) 25×10 6 CD8+CAR+ T cells and 25 × 10 6 a) comprises or is a single treatment comprising administering 50 x 10 CD4+ CAR+ T cells separately to a subject; 6 CD8+CAR+ T cells and 50 × 10 6 or c) comprises or is a single treatment comprising administering 75 x 10 CD4+ CAR+ T cells separately to a subject; or 6 CD8+CAR+ T cells and 75 × 10 6 In some embodiments, the method further comprises or is a single treatment comprising administering 25×10 CD4+ CAR+ T cells to a subject separately. In some embodiments, the method further comprises modifying the predetermined therapeutic regimen, e.g., the modified therapeutic regimen comprises administering 25×10 CD4+ CAR+ T cells to a subject. 6 CD8+CAR+ T cells and 25 × 10 6 or if the single treatment comprises administering 50 x 10 CD4+ CAR+ T cells separately to a subject. 6 CD8+CAR+ T cells and 50 × 10 6 a single treatment comprising administering 50×10 CD4+ CAR+ T cells separately to a subject; 6 CD8+CAR+ T cells and 50 × 10 6or if the single treatment comprises administering 75 x 10 CD4+ CAR+ T cells separately to a subject 6 CD8+CAR+ T cells and 75 × 10 6 or is a single treatment comprising administering 25 x 10 CD4+ CAR+ T cells separately to a subject; or 6 CD8+CAR+ T cells and 25 × 10 6 or if the single treatment comprises administering 75 x 10 CD4+ CAR+ T cells separately to a subject 6 CD8+CAR+ T cells and 75 × 10 6 In some embodiments, the method comprises or is a single treatment comprising administering 50×10 CD4+ CAR+ T cells to a subject separately. In some embodiments, the method comprises modifying the predetermined therapeutic regimen, e.g., the modified therapeutic regimen comprises administering 50×10 CD4+ CAR+ T cells to a subject. 6 CD8+CAR+ T cells and 50 × 10 6 or if the single treatment comprises administering 25 x 10 CD4+ CAR+ T cells separately to a subject 6 CD8+CAR+ T cells and 25 × 10 6 a single treatment comprising administering 75 x 10 CD4+ CAR+ T cells separately to a subject; 6 CD8+CAR+ T cells and 75 × 10 6 or if the single treatment comprises administering 50 x 10 CD4+ CAR+ T cells separately to a subject. 6 CD8+CAR+ T cells and 50 × 10 6 or is a single treatment comprising administering 75 x 10 CD4+ CAR+ T cells separately to a subject; or 6 CD8+CAR+ T cells and 75 × 10 6 or if the single treatment comprises administering 25 x 10 CD4+ CAR+ T cells separately to a subject 6CD8+CAR+ T cells and 25 × 10 6 In some embodiments, the therapeutic cell composition comprises or is a single treatment comprising administering two or more CD4+ CAR+ T cells separately to a subject. In some embodiments, the therapeutic cell composition comprises modifying a predetermined therapeutic regimen, e.g., the modified therapeutic regimen comprises administering the therapeutic cell composition in combination with a second therapeutic agent. [The present invention 1001] 1. A method for determining a clinical response, comprising: (a) Below: (i) a subject characteristic determined from a subject prior to treatment of the subject with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from an input composition, the input composition comprising T cells selected from a sample derived from the subject, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from the therapeutic cell composition, wherein the therapeutic cell composition is produced from the input composition and expresses the CAR, and the therapeutic composition is administered to the subject; receiving a feature including: (b) applying the features as inputs to a random forest model trained to determine the subject's clinical response to treatment with the therapeutic cell composition based on the informative features identified by preprocessing prior to treating the subject with the therapeutic cell composition, wherein the features applied as inputs are the same informative features used to train the random forest model. The method comprising: [The present invention 1002] 1. A method for determining a clinical response, comprising: (a) Below: (i) a subject characteristic determined from a subject prior to treatment of the subject with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from an input composition, the input composition comprising T cells selected from a sample derived from the subject, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from the therapeutic cell composition, wherein the therapeutic cell composition is produced from the input composition and expresses the CAR, and the therapeutic composition is administered to the subject; receiving a feature including: (b) applying the features as inputs to a random survival forest model trained to determine the subject's clinical response to treatment with the therapeutic cell composition based on the informative features identified by preprocessing prior to treating the subject with the therapeutic cell composition, wherein the features applied as inputs are the same informative features used to train the random survival forest model. The method comprising: [The present invention 1003] The method of invention 1001 or invention 1002, wherein the clinical response is or includes a complete response (CR), a partial response (PR), a durable response, progression-free survival (PFS), an objective response (OR), a pharmacokinetic response that is the target pharmacokinetic response or is greater than the target pharmacokinetic response, no toxic response or a mild toxic response, a toxic response, a pharmacokinetic response that is lower than the target response, or no CR, PR, durable response or objective response (OR). [The present invention 1004] 1001. The method of claim 1001, wherein the clinical response is a complete response (CR) or no complete response (CR). [The present invention 1005] 1001. The method of claim 1001, wherein the clinical response is partial response (PR) or no partial response (PR). [The present invention 1006] The method of claim 1001, wherein the clinical response is an objective response (OR) or no objective response (OR). [The present invention 1007] 1001. The method of claim 1001, wherein the clinical response is a toxic response or no toxic response. [The present invention 1008] The method of claim 1003 or claim 1007, wherein the toxic response is severe cytokine release syndrome (CRS) or severe neurotoxicity. [The present invention 1009] The method of invention 1001 or invention 1002, wherein the clinical response is a durable response or no durable response. [The present invention 1010] The method of claim 1001 or claim 1002, wherein the clinical response is duration of response (DOR). [The present invention 1011] The method of claim 1001 or claim 1002, wherein the clinical response is a duration of response (DOR) of at least 3 months or at least about 3 months. [The present invention 1012] The method of claim 1001 or 1002, wherein the clinical response is progression-free survival (PFS). [The present invention 1013] The method of invention 1001 or invention 1002, wherein the clinical response is a progression free survival (PFS) of at least 3 months or at least about 3 months. [The present invention 1014] 1001. The method of claim 1001, wherein the clinical response is a target pharmacokinetic response or a pharmacokinetic response greater than the target pharmacokinetic response. [The present invention 1015] The pharmacokinetic response is (i) expansion of CAR T cells of a therapeutic cell composition following treatment of a subject with the therapeutic cell composition; (ii) the maximum CAR T cell concentration in the subject following treatment of the subject with the therapeutic cell composition; (iii) the time point at which CAR T cell concentration is maximal in the subject following treatment of the subject with a therapeutic cell composition; or (iv) exposing the subject to CAR T cells of the therapeutic cell composition after treatment of the subject with the therapeutic cell composition. The method of the present invention 1003 or 1014, wherein the measurement value is [The present invention 1016] (a) selecting T cells from a sample derived from a subject to generate an input composition comprising T cells; (b) Below: (i) a subject characteristic determined from a subject prior to treatment of the subject with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from the input composition, the input composition comprising T cells selected from the sample from the subject, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from the therapeutic cell composition, wherein the therapeutic cell composition is produced from the input composition and expresses the CAR, and the therapeutic composition is administered to the subject; a step of measuring a feature quantity including the (c) applying the features as inputs to a random forest model trained to determine the subject's clinical response to treatment with the therapeutic cell composition based on the informative features identified by preprocessing prior to treating the subject with the therapeutic cell composition, wherein the features applied as inputs are the same informative features used to train the random forest model; and (d) administering a treatment to the subject, (1) if the subject is determined to have a clinical response selected from the group consisting of a complete response (CR), a partial response (PR), a durable response of greater than 3 months, a progression-free survival (PFS) of greater than 3 months, an objective response (OR), a desired pharmacokinetic response that is or is greater than the target pharmacokinetic response, and a no-toxicity response or a mild toxic response (optionally, a mild toxic response is cytokine release syndrome (CRS) of Grade 2 or less or neurotoxicity of Grade 2 or less), a predetermined therapeutic regimen comprising the therapeutic cell composition is administered; or (2) if the subject is determined to have a clinical response selected from the group consisting of a toxic response (optionally, the toxic response is severe cytokine release syndrome (CRS) or severe neurotoxicity), a poor pharmacokinetic response compared to the target pharmacokinetic response, progressive disease (PD), a durable response of less than 3 months, and a PFS of less than 3 months, administering to the subject a therapeutic regimen comprising the therapeutic cell composition that is altered compared to the predetermined therapeutic regimen comprising the therapeutic cell composition; The process A method of treating a subject, comprising: [The present invention 1017] (a) selecting T cells from a sample derived from a subject to generate an input composition comprising T cells; (b) Below: (i) a subject characteristic determined from a subject prior to treatment of the subject with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from the input composition, the input composition comprising T cells selected from the sample from the subject, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from the therapeutic cell composition, wherein the therapeutic cell composition is produced from the input composition and expresses the CAR, and the therapeutic composition is administered to the subject; a step of measuring a feature quantity including the (c) applying the features as inputs to a random survival forest model trained to determine a clinical response in the subject to be treated with the therapeutic cell composition based on the informative features identified by preprocessing prior to treating the subject with the therapeutic cell composition, wherein the features applied as inputs are the same informative features used to train the random survival forest model; and (d) administering a treatment to the subject, (1) if the subject is determined to have a clinical response selected from the group consisting of a complete response (CR), a partial response (PR), a durable response of greater than 3 months, a progression-free survival (PFS) of greater than 3 months, an objective response (OR), a desired pharmacokinetic response that is or is greater than the target pharmacokinetic response, and a no-toxicity response or a mild toxic response (optionally, a mild toxic response is cytokine release syndrome (CRS) of Grade 2 or less or neurotoxicity of Grade 2 or less), a predetermined therapeutic regimen comprising the therapeutic cell composition is administered; or (2) if the subject is determined to have a clinical response selected from the group consisting of a toxic response (optionally, the toxic response is severe cytokine release syndrome (CRS) or severe neurotoxicity), a poor pharmacokinetic response compared to the target pharmacokinetic response, progressive disease (PD), a durable response of less than 3 months, and a PFS of less than 3 months, administering to the subject a therapeutic regimen comprising the therapeutic cell composition that is altered compared to the predetermined therapeutic regimen comprising the therapeutic cell composition; The process A method of treating a subject, comprising: [The present invention 1018] A random forest model is trained to determine whether a subject will have a complete response (CR), where: (1) if the subject is determined to have a complete response (CR), the subject is administered the predetermined treatment regimen; or (2) if the subject is determined to have progressive disease (PD), the subject is administered the modified treatment regimen; The method of the present invention 1016. [The present invention 1019] A random forest model is trained to determine whether a subject will have a partial response (PR), where: (1) if the subject is determined to have a partial response (PR), the subject is administered the predetermined treatment regimen; or (2) if the subject is determined to have progressive disease (PD), the subject is administered the modified treatment regimen; The method of the present invention 1016. [The present invention 1020] A random forest model is trained to determine whether a subject will have a durable response of greater than 3 months, wherein: (1) if the subject is determined to have a durable response of greater than 3 months, the subject is administered the predetermined treatment regimen; or (2) If the subject is determined to have a durable response of less than 3 months, the subject is administered the modified treatment regimen. The method of the present invention 1016. [The present invention 1021] A random survival forest model is trained to determine whether a subject will have a durable response for more than 3 months, wherein: (1) if the subject is determined to have a durable response of greater than 3 months, the subject is administered the predetermined treatment regimen; or (2) If the subject is determined to have a durable response of less than 3 months, the subject is administered the modified treatment regimen. The method of the present invention 1017. [The present invention 1022] A random forest model was trained to determine whether a subject will have a progression-free survival (PFS) of greater than 3 months, wherein: (1) if the subject is determined to have a progression-free survival (PFS) of greater than 3 months, the subject is administered the predetermined treatment regimen; or (2) If the subject is determined to have a progression-free survival (PFS) of less than 3 months, the subject is administered the modified treatment regimen. The method of the present invention 1016. [The present invention 1023] A random survival forest model is trained to determine whether a subject will have a progression-free survival (PFS) of greater than 3 months, wherein: (1) if the subject is determined to have a progression-free survival (PFS) of greater than 3 months, the subject is administered the predetermined treatment regimen; or (2) If the subject is determined to have a progression-free survival (PFS) of less than 3 months, the subject is administered the modified treatment regimen. The method of the present invention 1017. [The present invention 1024] A random forest model is trained to determine whether a subject will have an objective response (OR), where: (1) if the subject is determined to have an objective response (OR), the subject is administered the predetermined treatment regimen; or (2) if the subject is determined to have progressive disease (PD), the subject is administered the modified treatment regimen; The method of the present invention 1016. [The present invention 1025] A random forest model was trained to determine the pharmacokinetic response of a subject, where: (1) if the subject is determined to have a desired pharmacokinetic response that is at or greater than the target pharmacokinetic response, the subject is administered the predetermined treatment regimen; or (2) If the subject is determined to have a lower pharmacokinetic response compared to the target pharmacokinetic response, the subject is administered the modified treatment regimen. The method of the present invention 1016. [The present invention 1026] The pharmacokinetic response is (i) expansion of CAR T cells of a therapeutic cell composition following treatment of a subject with the therapeutic cell composition; (ii) the maximum CAR T cell concentration in the subject following treatment of the subject with the therapeutic cell composition; (iii) the time point at which CAR T cell concentration is maximal in the subject following treatment of the subject with a therapeutic cell composition; or (iv) exposing the subject to CAR T cells of the therapeutic cell composition after treatment of the subject with the therapeutic cell composition. The method of the present invention 1025 is a measurement of [The present invention 1027] A random forest model is trained to determine whether a subject will have a toxic response, wherein: (1) if the subject is determined to have a non-toxic or mildly toxic response, the subject is administered the predetermined therapeutic regimen; or (2) if the subject is determined to have a toxic response, the subject is administered the modified treatment regimen; The method of the present invention 1016. [The present invention 1028] The method of claim 1027, wherein the toxic response is severe CRS or severe neurotoxicity. [The present invention 1029] The method of any one of claims 1016 to 1028, further comprising the step of producing a therapeutic cell composition. [The present invention 1030] The random forest model is trained using supervised training, the supervised training comprising: (a) Below: (i) a subject feature determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from samples derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions being produced from one of the plurality of input compositions and expressing the CAR, and the therapeutic composition being administered to one of the plurality of subjects; receiving a feature including: (b) preprocessing the features to identify informative features, the informative features comprising a subset of features including one or more target features, one or more input composition features, and one or more therapeutic cell composition features; (c) obtaining a clinical response from each of the plurality of subjects following treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features from the plurality of subjects and the obtained clinical responses as inputs to train a random forest model. Any of the methods of the present invention 1001, 1003 to 1016, 1018 to 1020, 1022 and 1024 to 1029, including the method of [The present invention 1031] A random survival forest model is trained using supervised training, the supervised training comprising: (a) Below: (i) a subject feature determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from samples derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions being produced from one of the plurality of input compositions and expressing the CAR, and the therapeutic composition being administered to one of the plurality of subjects; receiving a feature including: (b) preprocessing the features to identify informative features, the informative features comprising a subset of features including one or more target features, one or more input composition features, and one or more therapeutic cell composition features; (c) obtaining a clinical response over time from each of the plurality of subjects after treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features and clinical responses from the plurality of subjects as inputs to train a random survival forest model using supervised learning. Any of the methods of the present invention 1002, 1003, 1008 to 1013, 1015, 1017, 1021, 1023 and 1029, including the method of [The present invention 1032] (a) Below: (i) a subject feature determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from samples derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions being produced from one of the plurality of input compositions and expressing the CAR, and the therapeutic composition being administered to one of the plurality of subjects; receiving a feature including: (b) preprocessing the features to identify informative features, the informative features comprising a subset of features including one or more target features, one or more input composition features, and one or more therapeutic cell composition features; (c) obtaining a clinical response from each of the plurality of subjects following treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features from the plurality of subjects and the obtained clinical responses as inputs to train a random forest model. How to develop a random forest model, including: [The present invention 1033] (a) Below: (i) a subject feature determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from samples derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions being produced from one of the plurality of input compositions and expressing the CAR, and the therapeutic composition being administered to one of the plurality of subjects; receiving a feature including: (b) preprocessing the features to identify informative features, the informative features comprising a subset of features including one or more target features, one or more input composition features, and one or more therapeutic cell composition features; (c) obtaining a clinical response over time from each of the plurality of subjects after treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features and clinical responses from the plurality of subjects as inputs to train a random survival forest model using supervised learning. A method for developing a random survival forest model, including: [The present invention 1034] 1. A method for identifying features associated with a clinical response, comprising: (a) Below: (i) a subject feature determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from samples derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the CAR; and (iii) a therapeutic cell composition feature determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions being produced from one of the plurality of input compositions and expressing the CAR, and the therapeutic composition being administered to one of the plurality of subjects; receiving a feature including: (b) pre-processing the features to identify informative features, the informative features comprising a subset of features including one or more of the target features, one or more of the input composition features, and one or more of the therapeutic cell composition features; (c) obtaining a clinical response from each of said plurality of subjects following treatment with one of said plurality of therapeutic compositions; (d) applying the informative features and obtained clinical responses from the plurality of subjects as inputs to train a random forest model using supervised learning; and (e) identifying informative features associated with the clinical response from the trained random forest model. The method comprising: [This invention 1035] 1. A method for identifying features associated with a clinical response, comprising: (a) Below: (i) a subject feature determined from each of a plurality of subjects before the subjects are treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with a disease or condition, wherein the therapeutic cell composition is for treating the disease or condition; (ii) input composition features determined from each of a plurality of input compositions, each of the plurality of input compositions comprising T cells selected from samples derived from each of the plurality of subjects, the T cells being used to generate the therapeutic cell composition comprising T cells comprising the chimeric antigen receptor (CAR); and (iii) a therapeutic cell composition feature determined from each of a plurality of therapeutic cell compositions, each of the plurality of therapeutic cell compositions being produced from one of the plurality of input compositions and expressing the CAR, and the therapeutic composition being administered to one of the plurality of subjects; receiving a feature including: (b) pre-processing the features to identify informative features, the informative features comprising a subset of features including one or more of the target features, one or more of the input composition features, and one or more of the therapeutic cell composition features; (c) obtaining a clinical response over time from each of said plurality of subjects after treatment with one of said plurality of therapeutic compositions; (d) applying the informative features and clinical responses from the plurality of subjects as inputs to train a random survival forest model using supervised learning; and (e) identifying informative features associated with the clinical response from the trained random survival forest model. The method comprising: [The present invention 1036] Any of the methods of inventions 1032 to 1035, wherein the clinical response is or comprises a complete response (CR), a partial response (PR), a durable response, progression-free survival (PFS), an objective response (OR), a pharmacokinetic response that is or is greater than the target pharmacokinetic response, a no toxic response or a mild toxic response, a toxic response, a pharmacokinetic response that is lower than the target response, or no CR, PR, durable response, or objective response (OR). [This invention 1037] The method of any of claims 1034 to 1036, wherein identifying informative features associated with the clinical response comprises determining an importance measure for each of said informative features. [The present invention 1038] The method of the present invention 1037, wherein the importance measure includes a permutation importance measure, an average minimum depth and / or a total number of trees in a trained random forest model, and wherein the root node is split by informative features. [This invention 1039] The method of the present invention 1037, wherein the importance measure includes a permutation importance measure, an average minimum depth and / or a total number of trees in a trained random survival forest model, and wherein the root node is split by informative features. [The present invention 1040] Any of the methods of inventions 1037 to 1039, wherein the informative features associated with the clinical response are the first 10, 9, 8, 7, 6, 5, 4, 3, 2 or 1 informative features identified by ranking the values ​​of the importance scale for each of the informative features, and the importance scale is the same for each informative feature. [The present invention 1041] Any of the methods of the present inventions 1030 to 1040, wherein one of the plurality of therapeutic cell compositions is administered to each of the plurality of subjects, and the one therapeutic cell composition administered to the subject is a therapeutic cell composition prepared from an input composition of a sample derived from the subject. [The present invention 1042] The preprocessing for identifying informative features is a) removing target features, input composition features, and therapeutic cell composition features that have greater than 50%, greater than about 50%, or 50% data loss; b) removing target features, input composition features, and therapeutic cell composition features that have (i) zero variance, (ii) more than, about, or 95% of data values ​​equal to a single value, and / or (iii) fewer than 0.1n unique values, where n=the number of samples; c) imputing missing data for the target features, input composition features, and therapeutic cell composition features using multivariate imputation with chained equations; and d) identifying a covariate cluster including a set of target features, input composition features, therapeutic cell composition features, and combinations thereof having correlation coefficients with absolute values ​​greater than, about, or equal to 0.5, and iteratively selecting target features, input composition features, and therapeutic cell composition features from the covariate cluster, wherein the selected target features, input composition features, and therapeutic cell composition features have the smallest average absolute correlations with all remaining target features, input composition features, and therapeutic cell composition features. Any of the methods of the present invention 1030 to 1041, including one or more of the following: [This invention 1043] Any of the methods of the present inventions 1030 to 1042, wherein the preprocessing for identifying informative features includes or is a step of deleting target features, input composition features, and therapeutic cell composition features having data loss of more than 50%, approximately 50%, or 50%. [This invention 1044] Any of the methods of inventions 1030 to 1043, wherein the preprocessing for identifying informative features includes or is a step of deleting target features, input composition features, and therapeutic cell composition features having data loss of more than 60%, about 60%, or 60%. [This invention 1045] Any of the methods of inventions 1030 to 1044, wherein the preprocessing to identify informative features includes or is a step of removing target features, input composition features, and therapeutic cell composition features that have (i) zero variance, or (ii) more than 95%, about 95%, or 95% of data values ​​equal to a single value, and fewer than 0.1n unique values, where n = the number of samples. [The present invention 1046] Any of the methods of the present inventions 1030 to 1045, wherein the preprocessing for identifying informative features includes or is a step of complementing missing data of target features, input composition features, and therapeutic cell composition features by multivariate imputation using chain equations. [This invention 1047] Any of the methods of inventions 1030 to 1046, wherein the preprocessing for identifying informative features includes or is a step of identifying a covariate cluster including a set of target features, input composition features, therapeutic cell composition features, and combinations thereof, having correlation coefficients with absolute values ​​greater than 0.5, approximately 0.5, or equal to 0.5, and iteratively selecting target features, input composition features, and therapeutic cell composition features from the covariate cluster, wherein the selected target features, input composition features, and therapeutic cell composition features have the smallest average absolute correlation with all remaining target features, input composition features, and therapeutic cell composition features. [This invention 1048] 8. Any of the methods of claims 1030 to 1047, wherein said plurality of subjects is, or is about, 500, 400, 300, 200, 150, 100, 50, 25, 15, or 10 subjects, or any number between any of the foregoing values. [This invention 1049] 1049. The method of any of claims 1030 to 1048, wherein said plurality of subjects is 10, about 10, or more than 10 subjects and less than 250 subjects. [The present invention 1050] 1049. The method of any of claims 1030 to 1049, wherein said plurality of subjects is 20, about 20, or more than 20 and less than 200 subjects. [This invention 1051] 1050. The method of any of claims 1030 to 1050, wherein said plurality of subjects is 20, about 20, or more than 20 and less than 150 subjects. [This invention 1052] 1052. The method of any of claims 1030 to 1051, wherein said plurality of subjects is 20, about 20, or more than 20 and less than 100 subjects. [This invention 1053] The method of any of claims 1001 to 1052, wherein the object features include one or more of object attributes and clinical attributes. [This invention 1054] The method of claim 1053, wherein the subject attributes include one or more of age, weight, height, ethnicity, race, sex, and body mass index. [This invention 1055] The method of claim 1053 or claim 1054, wherein the clinical attributes comprise one or more of biomarkers, disease diagnosis, disease burden, disease duration, disease grade and treatment history. [The present invention 1056] The target features were: medication group, bridging chemotherapy, bridging chemotherapy and radiotherapy, bridging chemotherapy systemic treatment, cell origin, relapse or refractory after chemotherapy, type of diagnosis, disease cohort, disease burden, relapsed or refractory disease, disease origin, gender, route of administration of therapeutic cell composition, fold change in LDH, height, lesion count, oxygen saturation, body temperature (°C), maximum diameter of tumor before treatment with therapeutic cell composition, fold change in SPD, SPD value before lymphocyte-depleting chemotherapy, BMI, weight, sex, ethnicity, race, age, IPI score, ECOG score, disease severity. Stage, disease burden based on LDH before lymphocyte-depleting chemotherapy, disease burden based on SPD before lymphocyte-depleting chemotherapy, subject having active CNS disease at the time of treatment, disease burden based on extranodal disease classification, number of extranodal sites, disease burden based on bulky disease classification, medical history, number of prior lines of therapy, number of prior lines of systemic therapy, prior allogeneic hematopoietic stem cell transplant (allo-HSCT), prior autologous hematopoietic stem cell transplant (auto-HSCT), chemotherapy-refractory or chemotherapy-sensitive disease type, bridging anticancer therapy for disease control, leukocyte apheresis Number of days from date of leukapheresis to first infusion, number of months from diagnosis to treatment with therapeutic cell composition, baseline C-reactive protein (CRP), lymphocyte count (10^9 / L) before leukapheresis, gene double expressor, gene double hit, gene triple hit, gene double hit or triple hit, gene double hit or triple hit or double expressor, albumin level, alkaline phosphatase level, basophil count, absolute basophil count, direct bilirubin, total bilirubin, blood urea nitrogen level, calcium level, carbon dioxide level, chloride level, creatinine level, eosinophil count, absolute eosinophil count, glucose level, hematocrit level, hemoglobin level, LDH level, lesion count, lymphocyte count, absolute lymphocyte count, magnesium level, absolute monocyte count, monocyte count, absolute neutrophil count, neutrophil count, phosphate level, platelet count, potassium level, total protein, red blood cell count, aspartate aminotransferase level, alanine aminotransferase level, sodium level, two-way product sum, triglycerides, tumor maximum diameter, tumor perpendicular diameter, uric acid level,and one or more of the following: a white blood cell count; [This invention 1057] The method of any one of claims 1001 to 1056, wherein the input composition feature comprises a cell phenotype. [This invention 1058] Input composition features are CAS3- / CCR7- / CD27- / CD4+, CAS3- / CCR7- / CD27+ / CD4+, CAS3- / CCR7+ / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CA S3- / CCR7+ / CD27+ / CD4+, CAS3- / CD27+ / CD4+, CAS3- / CD28- / CD27- / CD4+, CAS3- / CD28- / CD27+ / CD4+, CAS3- / CD28+ / CD4+ , CAS3- / CD28+ / CD27- / CD4+, CAS3- / CD28+ / CD27+ / CD4+, CAS3- / CCR7- / CD45RA- / CD4+, CAS3- / CCR7- / CD4+, CD45RA+ / CD4 +, CAS3- / CCR7+ / CD45RA- / CD4+, CAS3- / CCR7+ / CD45RA+ / CD4+, CAS+ / CD4+, CAS+ / CD3+ / CD4+, CD4+ clonality, CAS3- / CCR7- / CD27- / CD8+, CAS3- / CCR7- / CD27+ / CD8+, CAS3- / CCR7+ / CD8+, CAS3- / CCR7+ / CD27- / CD8+, CAS3- / CCR7+ / CD27+ / CD8+, CAS 3- / CD27+ / CD8+, CAS3- / CD28- / CD27- / CD8+, CAS3- / CD28- / CD27+ / CD8+, CAS3- / CD28+ / CD8+, CAS3- / CD28+ / CD27- / CD8+, The method of any of claims 1001 to 1057, comprising one or more of the following clonalities: CAS3- / CD28+ / CD27+ / CD8+, CAS3- / CCR7- / CD85RA- / CD8+, CAS3- / CCR7- / CD8+, CD85RA+ / CD8+, CAS3- / CCR7+ / CD85RA- / CD8+, CAS3- / CCR7+ / CD85RA+ / CD8+, CAS+ / CD8+, CAS+ / CD3+ / CD8+ and CD8+. [This invention 1059] The method of any of claims 1001 to 1058, wherein the therapeutic cell composition characteristics include one or more of cell phenotype, recombinant receptor-dependent activity, and dosage. [The present invention 1060] Therapeutic cell composition features are CAS3- / CCR7- / CD27- / CD8+, CAS3- / CCR7- / CD27+ / CD8+, CAS3- / CCR7+ / CD8+, CAS3- / CCR7+ / CD27- / CD8+, CAS3- / CCR7+ / CD27+ / CD8+, CAS 3- / CD27+ / CD8+, CAS3- / CD28+ / CD8+, CAS3- / CD28+ / CD27- / CD8+, CAS3- / CD28+ / CD27+ / CD8+, CAS3- / CCR7- / CD45RA- / CD8+, CAS3- / CCR7- / CD45RA+ / CD 8+, CAS3- / CCR7+ / CD45RA- / CD8+, CAS3- / CCR7+ / CD45RA+ / CD8+, CAS+ / CD3+ / CAR+ / CD8+, CD3+ / CAR+ / CD8+, CD3+ / CD8+, CAR+ / CD8+, CD8+ cell clonality, EGFRt+ / CD8+, cytokine- / CD8+, IFNG+ / CD8+, IFNg+ / IL2 / CD8+, IFNg+ / IL17+ / TNFa+ / CD8+, IFNg+ / IL2+ / IL17+ / TNFa+ / CD8+, IFNg+ / IL2+ / TNFa+ / CD8+, CAR + / IFNg+ / CD8+, IFNg+ / TNFa+ / CD8+, CAR+ / IL2+ / CD8+, IL2+ / TNFa+ / CD8+, CD8+ cytolysis, CAR+ / TNFa+ / CD8+, viable cell concentration of CD8+ cells, vector copy number in CD8+ cells, EGFRt+ vector copy number in CD8+, CD8+ viability, GMCSF+ / CD8+, IFNG+ / CD8+, IL10+ / CD8+, IL13+ / CD8+, IL2+ / CD8+, IL4+ / CD8+, IL5+ / CD8+, IL6+ / CD8+, MIP1A+ / CD8+, MI P1B+ / CD8+, sCD137+ / CD8+, TNFa+ / CD8+, CD8+ cell dose, CD8+ cell dose level, percent viable administered CD8+ cells, total non-viable administered CD8+ cells, total viable administered CD8+ cells, total dose of CD8+ cells, CAS3- / CCR7- / CD27- / CD4+, CAS3- / CCR7- / CD27+ / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CAS3- / CCR7+ / CD27+ / CD4+, CAS3- / CD27+ / CD4+,CAS3- / CD28+ / CD4+, CAS3- / CD28+ / CD27- / CD4+, CAS3- / CD28+ / CD27+ / CD4+, CAS3- / CCR7- / CD45RA- / CD4+, CAS3- / C CR7- / CD45RA+ / CD4+, CAS3- / CCR7+ / CD45RA- / CD4+, CAS3- / CCR7+ / CD45RA+ / CD4+, CAS+ / CD3+ / CAR+ / CD4+, CD3+ / CAR + / CD4+, CD3+ / CD4+, CAR+ / CD4+, CD4+ cell clonality, EGFRt+ / CD4+, cytokine- / CD4+, IFNG+ / CD4+, IFNg+ / IL2 / CD4+, IFNg+ / IL17+ / TNFa+ / CD4+, IFNg+ / IL2+ / IL17+ / TNFa+ / CD4+, IFNg+ / IL2+ / TNFa+ / CD4+, CAR+ / IFNg+ / CD4+, IFNg+ / TNFa+ / CD4+, CAR+ / IL2+ / CD4+, IL2+ / TNFa+ / CD4+, cytolysis by CD4+, CAR+ / TNFa+ / CD4+, viable cell concentration in CD4+ cells, vector copy number in CD4+ cells, EGFRt+ vector copy number in CD4+, viability of CD4+, GMCSF+ / CD4+, IFNG+ / CD4+, IL10+ / CD4+, IL13+ / CD4+, IL2+ / CD4+, IL4+ / CD4+, IL Any of the methods of claims 1001-1059, comprising one or more of: 5+ / CD4+, IL6+ / CD4+, MIP1A+ / CD4+, MIP1B+ / CD4+, sCD137+ / CD4+, TNFa+ / CD4+, CD4+ cell dose, CD4+ cell dose level, percent viable administered CD4+ cells, total non-viable administered CD4+ cells, total viable administered CD4+ cells, and total dose of CD4+ cells. [The present invention 1061] 1060. The method of any of claims 1001 to 1060, wherein the sample comprises a whole blood sample, a buffy coat sample, a peripheral blood mononuclear cell (PBMC) sample, an unfractionated T cell sample, a lymphocyte sample, a leukocyte sample, an apheresis product, or a leukocyte apheresis product. [The present invention 1062] The method of any one of claims 1001 to 1061, wherein the sample is an apheresis product or a leukapheresis product. [The present invention 1063] The method of claim 1062, wherein the apheresis product or leukapheresis product has been previously cryopreserved. [The present invention 1064] The method of any of claims 1001 to 1063, wherein the T cells comprise primary cells obtained from the subject. [This invention 1065] The method of any of claims 1001 to 1064, wherein the T cells comprise CD3+, CD4+ and / or CD8+ T cells. [The present invention 1066] Any of the methods of inventions 1001 to 1065, wherein the input composition comprises CD4+, CD8+, or CD4+ and CD8+ T cells, the therapeutic cell composition comprises CD4+, CD8+, or CD4+ and CD8+ T cells expressing the CAR, and is produced from the input composition, the input composition feature comprises an input composition feature derived from the CD4+, CD8+, or CD4+ and CD8+ T cell composition of the input composition, and the therapeutic cell composition feature comprises a therapeutic cell composition feature derived from the CD4+, CD8+, or CD4+ and CD8+ T cells of the therapeutic composition. [This invention 1067] Any of the methods of inventions 1001 to 1065, wherein the input composition comprises separate compositions of CD4+ and CD8+ T cells, the therapeutic cell composition comprises separate compositions of CD4+ and CD8+ T cells expressing the CAR, and is prepared from each of the CD4+ or CD8+ T cell compositions of the input composition, the input composition feature comprises an input composition feature derived from the CD4+ and CD8+ T cell compositions of the input composition, and the therapeutic cell composition feature comprises a therapeutic cell composition feature derived from each of the CD4+ and CD8+ T cells of the separate compositions of the therapeutic composition. [The present invention 1068] Any of the methods of inventions 1001 to 1065, wherein the input composition comprises separate compositions of CD4+ and CD8+ T cells, the therapeutic cell composition comprises a mixed composition of CD4+ and CD8+ T cells expressing the CAR, and is produced from the separate compositions of CD4+ and CD8+ T cells of the input composition, the input composition feature comprises an input composition feature derived from the separate compositions of CD4+ and CD8+ T cells of the input composition, and the therapeutic cell composition feature comprises a therapeutic cell composition feature derived from the mixed composition of CD4+ and CD8+ cells of the therapeutic composition. [The present invention 1069] The prescribed treatment regimen a)25×10 6 CD8+CAR+ T cells and 25 × 10 6 comprises or is a single treatment comprising administering CD4+ CAR+ T cells separately to a subject; b)50×10 6 CD8+CAR+ T cells and 50 × 10 6 comprises or is a single treatment comprising administering CD4+ CAR+ T cells separately to a subject; or c)75×10 6 CD8+CAR+ T cells and 75 × 10 6 or is a single treatment comprising administering to a subject separate CD4+ CAR+ T cells; Any of the methods of the present invention 1016 to 1031 and 1041 to 1068. [The present invention 1070] The altered treatment regimen comprises: a) the predetermined treatment regimen is 25 x 10 6 CD8+CAR+ T cells and 25 × 10 6 or if the single treatment comprises administering 50 x 10 CD4+ CAR+ T cells separately to a subject. 6 CD8+CAR+ T cells and 50 × 10 6 comprises or is a single treatment comprising administering CD4+ CAR+ T cells separately to a subject; b) said predetermined treatment regimen is 50 x 10 6 CD8+CAR+ T cells and 50 × 10 6 or if the single treatment comprises administering 75 x 10 CD4+ CAR+ T cells separately to a subject 6 CD8+CAR+ T cells and 75 × 10 6 comprises or is a single treatment comprising administering CD4+ CAR+ T cells separately to a subject; or c) said predetermined treatment regimen is 25 x 10 6 CD8+CAR+ T cells and 25 × 10 6 or if the single treatment comprises administering 75 x 10 CD4+ CAR+ T cells separately to a subject 6 CD8+CAR+ T cells and 75 × 10 6 or is a single treatment comprising administering to a subject separate CD4+ CAR+ T cells; Any of the methods of the present invention 1016 to 1031 and 1041 to 1068. [This invention 1071] The altered treatment regimen comprises: a) the predetermined treatment regimen is 50 x 10 6 CD8+CAR+ T cells and 50 × 10 6 or if the single treatment comprises administering 25 x 10 CD4+ CAR+ T cells separately to a subject 6 CD8+CAR+ T cells and 25 × 10 6 comprises or is a single treatment comprising administering CD4+ CAR+ T cells separately to a subject; b) said prescribed treatment regimen is 75 x 10 6 CD8+CAR+ T cells and 75 × 10 6 or if the single treatment comprises administering 50 x 10 CD4+ CAR+ T cells separately to a subject. 6 CD8+CAR+ T cells and 50 × 10 6 comprises or is a single treatment comprising administering CD4+ CAR+ T cells separately to a subject; or c) said prescribed treatment regimen is 75 x 10 6 CD8+CAR+ T cells and 75 × 10 6 or if the single treatment comprises administering 25 x 10 CD4+ CAR+ T cells separately to a subject 6 CD8+CAR+ T cells and 25 × 10 6 or is a single treatment comprising administering to a subject separate CD4+ CAR+ T cells; Any of the methods of the present invention 1016 to 1031 and 1041 to 1068. [This invention 1072] The method of any of claims 1016 to 1031 and 1041 to 1068, wherein said altered treatment regimen comprises administering a therapeutic cell composition in combination with a second therapeutic agent. [Brief explanation of the drawings]

[0038] [Figure 1A] 1A and 1B show exemplary decision trees included in a random forest model. [Figure 1B] 1A and 1B show exemplary decision trees included in a random forest model. [Figure 2A] Figure 2A shows exemplary significant feature clusters identified using Random Forest and Random Survival Forest as being important for correlation with log AUC (AUC, area under the concentration-time curve over 28 days post-infusion) following treatment with a therapeutic cell composition. Arrows indicate the direction of correlation. [Figure 2B] FIG. 2B shows the correlation between log10AUC and patient age. [Figure 2C] Figure 2C shows the cumulative local effect of patient age on log10AUC independent of all other features. [Figure 2D] FIG. 2D shows the correlation between log10 AUC and the total number of prior treatments the patient had received. [Figure 2E] Figure 2E shows the cumulative local effect of the number of prior treatments a patient has received on the log10 AUC independent of all other features. [Figure 2F] FIG. 2F shows the correlation between log10 AUC and CD8+ T cell effector cytokine secretion of therapeutic cell compositions. [Figure 2G] Figure 2G shows the cumulative local effect of therapeutic cell compositions on CD8+ T cell effector cytokine secretion on log10 AUC independent of all other features. [Figure 3A] Figure 3A shows exemplary significant feature clusters identified using Random Forest and Random Survival Forest as being important for correlation with progression-free survival (PFS) after treatment with the therapeutic cell composition. Arrows indicate the direction of correlation. [Figure 3B] Figure 3B shows the cumulative local effect of antigen-specific cytokine production by CD4+ T cells in the therapeutic cell composition on PFS independent of all other characteristics. [Figure 3C] Figure 3C shows the cumulative local effect of pre-treatment (pre-LDC) lactate dehydrogenase (LDH) levels with lymphocyte-depleting chemotherapy on PFS independent of all other features. [Figure 4A] Figure 4A shows exemplary significant feature clusters identified using Random Forest and Random Survival Forest as being important in correlation with complete response (CR) after treatment with the therapeutic cell composition. Arrows indicate the direction of correlation. [Figure 4B] FIG. 4B shows the cumulative local effect of therapeutic cell compositions on CD4+ T cell antigen-specific cytokine production on CR independent of all other characteristics. [Figure 4C] FIG. 4C shows the cumulative local effect of therapeutic cell compositions on CD8+ T cell antigen-specific cytokine production on CR independent of all other characteristics. [Figure 4D]Figure 4D shows the cumulative local effect of tumor burden, as measured by the two-way product sum (SPD) before treatment with lymphodepleting chemotherapy (pre-LDC), on CR independent of all other features. [Figure 4E] Figure 4E shows the cumulative local effect of tumor burden, as measured by pre-treatment (pre-LDC) LDH with lymphocyte-depleting chemotherapy, on CR independent of all other features. [Figure 5A] Figure 5A shows exemplary significant feature clusters identified using Random Forest and Random Survival Forest as being important for correlation with neurological events (NE) following treatment with the therapeutic cell composition. Arrows indicate the direction of correlation. [Figure 5B] Figure 5B shows the identified significant feature clusters that correlate with cytokine release syndrome (CRS) after treatment with the therapeutic cell composition, with arrows indicating the direction of correlation. [Figure 5C] Figure 5C shows the cumulative local effect of tumor burden as measured by LDH before treatment with lymphocyte-depleting chemotherapy (pre-LDC) on neurological events (NE), independent of all other features. [Figure 5D] Figure 5D shows the cumulative local effect of tumor burden, as measured by pre-treatment (pre-LDC) LDH with lymphocyte-depleting chemotherapy, on CRS independent of all other features. [Figure 5E] Figure 5E shows the cumulative local effect of bridging therapy on CRS independent of all other features. DETAILED DESCRIPTION OF THE INVENTION

[0039] Detailed Description Provided herein are methods for identifying features associated with clinical response in a subject following treatment with a cell therapy, e.g., an engineered T cell therapy (e.g., a therapeutic cell composition), for the treatment of diseases and conditions, e.g., various cancers. In some embodiments, the methods determine, e.g., predict, a subject's clinical response to treatment with a cell therapy (e.g., a therapeutic cell composition) before the subject is treated. Aspects of the methods provided herein, e.g., aspects thereof, relate to determining effective dosing and administration of a cell therapy, e.g., a therapeutic cell composition.

[0040] Numerous subject (patient) attributes, attributes of the starting materials (e.g., input composition characteristics) used to generate drug formulations, and attributes of drug formulations (e.g., therapeutic cell compositions) have demonstrated nominally significant univariate relationships with clinical endpoints, e.g., response, in cell therapy clinical trials. However, clinical response to cell therapy can depend on many factors, including, but not limited to, subject characteristics, therapeutic cell composition characteristics, and input composition characteristics from which the generated therapeutic cell composition is derived. Quantifying the multifactorial contributions of subject characteristics, starting material (e.g., input composition) characteristics, and drug formulation (e.g., therapeutic cell composition) characteristics to efficacy, safety, and pharmacokinetic (PK) response is a challenge in the field of cell therapy.

[0041] The methods provided herein address the challenge of evaluating multivariate features using supervised machine learning. For example, the machine learning models provided herein can evaluate how multiple diverse features may contribute to (e.g., determine or predict) clinical response. The model can be queried or interrogated to identify features, e.g., feature groups, that correlate with clinical response. In some cases, the methods herein can also rank the importance of each feature in a feature set in determining clinical response. In some cases, this information is useful for optimizing clinical experience in heterogeneous patient populations. In some cases, this information is useful for optimizing the production and manufacturing of drug formulations.

[0042] The methods provided herein include machine learning models trained to determine (e.g., predict) a subject's clinical response to a cellular therapy, e.g., a therapeutic composition, e.g., complete response (CR), partial response (PR), durable response (e.g., durability of response, DOR), toxicity response, and / or pharmacokinetic response based on features, e.g., subject attributes (e.g., subject features), attributes of the therapeutic cell composition (e.g., therapeutic cell composition features), and attributes of input compositions used to generate the therapeutic cell composition (e.g., input composition features). In some embodiments, subject features include subject attributes, e.g., age and weight, and clinical attributes, e.g., expression of biomarkers and combinations of biomarkers, disease burden (e.g., measures of tumor burden), treatment history, and combinations thereof. In some embodiments, therapeutic cell composition features include, but are not limited to, cell phenotype, e.g., cell health (e.g., number of viable cells, number of dead cells), presence and / or expression of surface markers, absence or lack of expression of surface markers, presence and / or expression of cytokines, absence or lack of expression of cytokines, recombinant receptor expression (e.g., CAR+), recombinant receptor-dependent activity (e.g., cytolytic activity, cytokine production), and combinations thereof. In some embodiments, input composition features include, but are not limited to, cell phenotype, e.g., cell health (e.g., concentration of viable cells, number of dead cells), presence and / or expression of surface markers, absence or lack of expression of surface markers, and combinations thereof.

[0043] In some aspects, the methods provided herein include a machine learning model trained to determine (e.g., predict) a subject's clinical response to a cellular therapy, e.g., a therapeutic composition, based on attributes of the subject (e.g., subject features), attributes of the therapeutic cellular composition (e.g., therapeutic cellular composition features), and attributes of an input composition (e.g., input composition features) used to generate the therapeutic cellular composition. In some aspects, the therapeutic cellular composition is generated using the input composition as a starting material. In some aspects, training a machine learning model using subject features, therapeutic cellular composition features, and input composition features provides certain advantages over training using only a subset of such feature sets. Such advantages include a more accurate prediction of a subject's clinical response or a more complete identification of features associated with an informative clinical response. In some aspects, the provided methods are based on the recognition that, even when an input composition is used as a starting material to generate a therapeutic cell composition, features of the pre-manufactured input composition may contain information related to the clinical response to the therapeutic cell composition that is not contained in or accounted for by features of the manufactured therapeutic cell composition. Thus, in some aspects, including input composition features in training a model can improve model performance or identification of informative features compared to that obtained when training using target features and / or therapeutic cell composition features alone.

[0044] In some aspects, the machine learning model that is expected to be used by the methods provided herein is a transparent machine learning model.The use of a transparent machine learning model is particularly advantageous because it allows the feature associated with the clinical response of the subject to be identified.In some embodiments, the feature identified as associated with clinical response can be evaluated in the subject before treating the subject with cell therapy drug to determine, for example, predict whether the subject will have a desirable or favorable clinical response to treatment.

[0045] In some cases, a model, such as a traditional "black box" model, may be considered transparent if it can be queried or interrogated in a manner that allows for understanding of how the model reached a particular decision. In some embodiments, understanding how the model reached a particular decision is or includes identifying a feature or features, e.g., one or more variables, that contributed to the decision. For example, in light of the methods provided herein, in some embodiments, a model is considered transparent if it can be interrogated or queried to identify features associated with a clinical response (e.g., measure feature importance for a clinical response). In some embodiments, the contribution of each feature to reaching a particular decision is quantified. In some cases, feature identification and / or quantification is determined by, for example, systematically manipulating the model or under controlled, known conditions and evaluating the model's accuracy (e.g., predictive accuracy) and its change upon manipulation.

[0046] In some embodiments, the machine learning model is a random forest model. In some embodiments, the machine learning model is a random survival forest model. As described above, an advantage of using random forests and random survival forests is their transparency. For example, random forest models and random survival forest models can be interrogated to identify features used to predict a subject's clinical response to a cell therapy, e.g., a therapeutic cell composition. In some embodiments, the features used to determine, e.g., predict, a clinical response are those considered to be associated with a clinical response. In some embodiments, identifying features used to determine, e.g., predict, a subject's clinical response includes assessing feature importance, e.g., as described herein (see Sections IB1a and IB2a).

[0047] In some embodiments, the random forest model provided herein is interrogated to identify features associated with clinical response. In some embodiments, the random forest model provided herein is used to determine (e.g., classify or predict) which clinical response a subject who has not yet been treated with a cell therapy drug (e.g., a therapeutic cell composition) will have. In some embodiments, by determining, e.g., predicting, before treatment, which clinical response a subject will have, the subject may be treated according to a predetermined therapeutic regimen, or may be treated according to a therapeutic regimen that is different (e.g., modified) from the predetermined therapeutic regimen. In some embodiments, modifying the predetermined therapeutic regimen in light of the determined, e.g., predicted, clinical response may result in an improved or favorable clinical response, or may increase the probability or likelihood that the subject will have an improved or favorable clinical response.

[0048] The random survival forest model can handle right-censored survival data and avoid restrictive assumptions, such as proportional hazards or parametric assumptions. In some embodiments, the random survival forest model can handle nonlinear effects and interactions between multiple variables. Such characteristics are advantageous for building risk prediction models, such as risk prediction models for clinical response. In some embodiments, the random survival forest models provided herein are interrogated to identify features associated with clinical response. For example, features associated with the probability of having a clinical response within a given amount of time can be identified. In some embodiments, the random survival forest models provided herein are used to determine the probability of a subject having a clinical response after treatment with a cell therapy, e.g., a therapeutic cell composition, before the subject is treated. In some embodiments, the random survival forest models provided herein are used to determine a subject's clinical response function and cumulative hazard function. In some embodiments, the random survival forest model can estimate the risk of a subject having a clinical response. In some embodiments, the random survival forest model can estimate the risk of a subject not having a clinical response. In some embodiments, determining (e.g., estimating or predicting) whether a subject will or will not have a clinical response after treatment before treatment can result in the subject being treated according to a predetermined therapeutic regimen, or according to a therapeutic regimen that is different (e.g., modified) from the predetermined therapeutic regimen. In some embodiments, modifying the predetermined therapeutic regimen can result in an improved or favorable clinical response, or can increase the probability or likelihood that the subject will have an improved or favorable clinical response.

[0049] The machine learning models provided herein, such as random forests and random survival forests, are trained to predict clinical responses based on various features associated with the subject (e.g., patient) being treated, the therapeutic cell composition administered to the subject, and the input composition (e.g., starting material derived from the subject) for producing the therapeutic cell composition. In some embodiments, the machine learning models provided herein are trained using features associated with the subject being treated (e.g., subject features, pre-treatment), the therapeutic cell composition administered to the subject (e.g., therapeutic cell composition features), and the input composition (e.g., starting material derived from the subject) for producing the therapeutic cell composition (e.g., input composition features).

[0050] In some embodiments, the training is supervised learning. When a model is trained using supervised learning, the clinical response of a subject treated with a therapeutic cell composition (the features thereof, for example, subject features, therapeutic cell composition features, and input composition features, have been obtained) is determined, obtained, or otherwise received. As described above, in some embodiments, the clinical response includes efficacy outcomes such as overall response; complete response (CR); partial response (PR); durable response (e.g., durability of response, DOR), such as a response that is durable for at least 3 months, 6 months, or longer; safety outcomes such as toxicity, for example, the occurrence of neurotoxicity or CRS; and pharmacokinetic responses, such as the maximum serum concentration of cells (C max ) and exposure (e.g., area under the curve (AUC)).

[0051] In addition to the clinical response after treatment with the therapeutic cell composition, a model can be trained using labeled data by measuring, obtaining, or receiving pre-treatment features, such as subject features, therapeutic cell composition features, and input composition features. The model can also be tested using test data to measure the predictive accuracy of the trained model. It will be recognized that in training a random survival forest model, time and censoring components, such as time to event, are associated with clinical response.

[0052] In some embodiments, the target features, therapeutic cell composition features, and input composition features are preprocessed. Data preprocessing, in some aspects, avoids creating models that produce misleading or inaccurate results. In some embodiments, preprocessing prevents outlying values, missing values, impossible data combinations, highly correlated features, and other confounding features from being incorporated into (e.g., learned from) the model. In provided embodiments, the provided preprocessing steps have been found to be particularly advantageous for training data from small data cohorts, such as may be present in data related to clinical trials of therapeutics, such as those involving T cell therapy drugs (e.g., CAR T cells).

[0053] In some cases, for example, during clinical trials of therapeutic drugs, different dosage levels can be used to treat different numbers of subjects.For example, a group of 100 subjects can be administered a certain dosage, while a different group of 50 subjects can be administered a significantly different dosage.In some cases, this can result in an imbalanced data set.In some cases, the difference in sample size can also be a problem for training a model.In some aspects, imbalance can be corrected by including dosage as a feature for training a model.

[0054] In some cases, preprocessing results in the identification of informative features. For example, preprocessing can be used to remove features with little or no variance, highly correlated features, or missing values, or to replace missing values ​​so that the remaining features are informative, discriminative, and independent (e.g., informative features). In some embodiments, machine learning models, such as random forests and random survival forests, are trained on the informative features identified by preprocessing. In some embodiments, machine learning models, such as random forests and random survival forests, are trained using supervised learning on the informative features identified by preprocessing. In some embodiments, the features used as inputs to a model for determining clinical response, such as subject features, therapeutic cell composition features, and input features, are informative features that are the same informative features used to train the model.

[0055] All publications referenced in this application, including patent documents, scientific papers, and databases, are incorporated by reference in their entirety for all purposes, as if each individual publication were individually incorporated by reference. To the extent that a definition set forth herein conflicts or is otherwise inconsistent with a definition set forth in a patent, patent application, published patent application, or other publication incorporated herein by reference, the definition set forth herein shall take precedence over the definition incorporated herein by reference.

[0056] The section headings used herein are for organizational purposes only and should not be construed as limiting the subject matter described.

[0057] I. Methods for identifying features associated with clinical outcomes and determining clinical outcomes The methods provided herein allow for the identification of features, e.g., subject features, therapeutic cell composition features, and input composition features, associated with a clinical response in a subject after treatment with a therapeutic cell composition. In some embodiments, the methods allow for the clinical response in a subject treated with a therapeutic cell composition to be determined based on the features, e.g., subject features, therapeutic cell composition features, and input composition features, prior to treatment with the therapeutic cell composition. Having this type of information at an early stage, e.g., prior to treatment, allows for the development of treatment strategies (e.g., combination treatments, dosing) prior to treating the subject, thereby increasing the likelihood that the subject will have a positive or favorable clinical response (e.g., durable response, progression-free survival).

[0058] Methods provided herein include generating a therapeutic cell composition (e.g., a therapeutic T cell composition) comprising engineered CD3+, CD4+, CD8+, or CD4+ and CD8+ cells, the therapeutic cell composition being made from an input composition comprising CD3+, CD4+, CD8+, or CD4+ and CD8+ T cells. In some embodiments, the input composition comprises two separate compositions, e.g., a CD4+ composition and a CD8+ composition. In some embodiments, the input composition comprises a single composition comprising CD4+ and CD8+ cells. In some embodiments, methods provided herein for generating a therapeutic cell composition include generating both CD4+ and CD8+ engineered cells for the therapeutic cell composition. In some embodiments, the CD4+ and CD8+ cells are separately engineered, e.g., to generate separate therapeutic cell compositions. In some embodiments, the CD4+ and CD8+ cells are separately engineered to generate separate therapeutic cell compositions from separate CD4+ and CD8+ input compositions. In some embodiments, the therapeutic cell composition comprises mixed CD4+ and CD8+ engineered cells. In some embodiments, the separate CD4+ and CD8+ engineered cells of separate therapeutic cell compositions are combined to create a therapeutic cell composition of mixed CD4+ and CD8+ engineered cells. In some embodiments, the therapeutic cell composition of mixed CD4+ and CD8+ engineered cells is created from a single input composition comprising mixed CD4+ and CD8+ cells. Features of the input composition and the therapeutic cell composition may be measured, received, or obtained from the mixed composition or from separate compositions.

[0059] A. Features and Clinical Responses It is anticipated that a subject's clinical response to treatment with a therapeutic cell composition will depend on many factors, including, but not limited to, features of the subject, features of the therapeutic cell composition, and features of the input composition from which the generated therapeutic cell composition is derived. Accordingly, the methods provided herein involve using machine learning models to assess the relationship between features associated with a subject treated with a therapeutic cell composition, features associated with the therapeutic cell composition, and features of the input composition from which the generated therapeutic cell composition is derived, and the clinical response in the subject after treatment with the therapeutic cell composition.

[0060] In some embodiments, subject-related features used in the methods provided herein include subject attributes, such as age and weight, clinical attributes, such as expression of biomarkers and combinations of biomarkers, disease burden (e.g., measures of tumor burden), treatment history, and combinations thereof.

[0061] In some embodiments, characteristics associated with the therapeutic cell composition and the input composition include a cellular phenotype. In some embodiments, the cellular phenotype is determined by assessing the presence or absence of one or more specific molecules, e.g., surface molecules and / or molecules that may accumulate in or be produced by cells or subpopulations of cells within the input composition or therapeutic cell composition. In some embodiments, the cellular phenotype may include cellular activity, e.g., the production of a factor (e.g., cytokine) in response to a stimulus. In some embodiments, the production of a factor (e.g., cytokine) is in response to recombinant receptor-dependent activation. In some embodiments, the recombinant receptor-dependent activity of cells of the therapeutic cell composition is measured by assessing one or more specific molecules (e.g., cytokines) that may accumulate in or be produced by cells or subpopulations of cells within the therapeutic cell composition. In some embodiments, the recombinant receptor-dependent activity is assessed by measuring the cytolytic activity of cells of the therapeutic cell composition.

[0062] In some embodiments, features of an input composition and / or therapeutic cell composition include measuring, detecting, quantifying, or otherwise assessing the phenotype of a cellular composition (e.g., a surface molecule, cytokine, recombinant receptor). In certain embodiments, features of a composition (e.g., an input composition, a therapeutic cell composition) include measuring, detecting, quantifying, or otherwise assessing the presence, absence, degree of expression, or level of a particular molecule (e.g., a surface molecule, cytokine, recombinant receptor). In some embodiments, the percentage, number, ratio, and / or proportion of cells having a certain attribute is measured. In some embodiments, the percentage, number, ratio, and / or proportion of cells having a certain attribute is a therapeutic cell composition feature or an input composition feature that can be used as input to the machine learning algorithms provided herein.

[0063] In some embodiments, the therapeutic cell composition feature or input composition feature is a phenotype, e.g., a cell phenotype. In some embodiments, the therapeutic cell composition feature or input composition feature is a phenotype indicative of cell viability. In some embodiments, the phenotype is indicative of the absence of apoptosis, the absence of early apoptosis, or the absence of late apoptosis. In some embodiments, the phenotype is the absence of a factor indicative of the absence of apoptosis, early apoptosis, or late apoptosis. In some embodiments, the phenotype is indicative of the viability of a subpopulation or subset of T cells in the therapeutic cell composition, e.g., recombinant receptor-expressing T cells (e.g., CAR + T cells), CD8 + T cells or CD4 + It is the phenotype of T cell.In some embodiments, it is the phenotype of non-activated cell and / or the phenotype of the cell that lacks, or the expression of one or more activation markers is reduced or the expression of the cell is low.In some embodiments, it is the phenotype of non-exhausted cell and / or the phenotype of the cell that lacks, or the expression of one or more exhaustion markers is reduced or the expression of the cell is low.

[0064] In some embodiments, the phenotype is the production of one or more cytokines. In some embodiments, for example, when a cytokine is produced and / or secreted by an engineered cell of a therapeutic cell composition in response to engagement of a recombinant receptor expressed by the cell with its antigen, this activity is referred to as a recombinant receptor-dependent activity. In some embodiments, a therapeutic cell composition characteristic is a recombinant receptor-dependent activity.

[0065] In some embodiments, the production of one or more cytokines is measured, detected, and / or quantified by intracellular cytokine staining. In certain embodiments, the phenotype is the absence of cytokine production. In certain embodiments, the phenotype is positive for cytokine production or high-level cytokine production. Intracellular cytokine staining by flow cytometry (ICS) is an extremely suitable technique for studying cytokine production at the single-cell level. It detects cytokine production and accumulation in the endoplasmic reticulum after cell stimulation, allowing for the identification of cell populations that are positive or negative for specific cytokine production, or the separation of high- and low-producing cells based on a threshold. ICS can also be used in combination with other flow cytometry protocols for immunophenotyping using cell surface markers or MHC multimers to access cytokine production in specific subgroups of cells, making it a highly flexible and versatile method. Other single-cell techniques for measuring or detecting cytokine production include, but are not limited to, ELISPOT, limiting dilution, and T cell cloning.

[0066] In certain embodiments, for example, in therapeutic cell compositions, the characteristic quantity comprises a recombinant receptor-dependent activity. In some embodiments, the activity is a recombinant receptor (e.g., CAR)-dependent activity that is or includes the production and / or secretion of a soluble factor. In some specific embodiments, the soluble factor is a cytokine or chemokine.

[0067] Suitable techniques for measuring the production or secretion of soluble factors are known in the art. The production and / or secretion of soluble factors can be measured by determining the concentration or amount of the extracellular amount of the factor or by determining the amount of transcriptional activity of the gene encoding the factor. Suitable techniques include, but are not limited to, immunoassays, aptamer-based assays, histological or cytological assays, mRNA expression level assays, enzyme-linked immunosorbent assays (ELISAs), alphalisa assays, immunoblotting, immunoprecipitation, radioimmunoassays (RIAs), immunostaining, flow cytometry assays, surface plasmon resonance (SPR), chemiluminescence assays, lateral flow immunoassays, inhibition or avidity assays, protein microarrays, high-performance liquid chromatography (HPLC), Meso Scale Discovery (MSD) electrochemiluminescence, and bead-based multiplex immunoassays (MIAs). In some embodiments, suitable techniques may use detectable binding reagents that specifically bind to soluble factors.

[0068] In some embodiments, the phenotype is indicated by the presence or absence or level of expression in cells of one or more specific molecules, such as specific surface markers, e.g., surface proteins, that indicate the phenotype; intracellular markers that indicate the phenotype; or nucleic acids that indicate the phenotype, or other molecules or factors that indicate the phenotype. In some embodiments, the phenotype is or includes the positive or negative expression of one or more specific molecules. In some embodiments, the specific molecules include, but are not limited to, surface markers, e.g., membrane glycoproteins or receptors; markers related to apoptosis or viability; or specific molecules that indicate the state of immune cells, such as markers related to activation, exhaustion, or mature or naive phenotypes. In some embodiments, any known method for evaluating or measuring, counting and / or quantifying cells based on specific molecules can be used to determine the number of cells of the phenotype in a composition (e.g., input composition, therapeutic cell composition).

[0069] In some embodiments, the phenotype is or includes positive or negative expression of one or more specific molecules in a cell. In some embodiments, positive expression is indicated by a detectable amount of a specific molecule in a cell. In certain embodiments, a detectable amount is any detected amount of a specific molecule in a cell. In certain embodiments, a detectable amount is an amount greater than the background, e.g., background staining, signal, etc., in a cell. In certain embodiments, positive expression is an amount of a specific molecule greater than a threshold, e.g., a predetermined threshold. Similarly, in certain embodiments, a cell having negative expression of a specific molecule may be any cell that has not been determined to have positive expression, or a cell that lacks a detectable amount of a specific molecule, or a detectable amount of a specific molecule greater than the background. In some embodiments, if the amount of a specific molecule is less than the threshold, the cell has negative expression of the specific molecule. Those skilled in the art will understand how to define a threshold for defining positive and / or negative expression of a specific molecule as a matter of routine skill, and will understand that the threshold can be defined according to, for example, but not limited to, the assay or detection method, the identity of the specific molecule, the reagent used for detection, and the specific parameters of the instrument.

[0070] Examples of methods that can be used to detect specific molecules and / or analyze cellular phenotypes include, but are not limited to, biochemical analysis; immunochemical analysis; image analysis; cell morphological analysis; molecular analysis, such as PCR, sequencing, high-throughput sequencing, DNA methylation determination; proteomics analysis, such as protein glycosylation and / or phosphorylation pattern determination; genomics analysis; epigenomics analysis (e.g., ChIP-seq or ATAC-seq); transcriptomics analysis (e.g., RNA-seq); and any combination thereof. In some embodiments, the method may include evaluation of immune receptor repertoire, such as T cell receptor (TCR) repertoire. In some aspects, any determination of phenotype may be evaluated by high-throughput, automated, and / or single-cell-based methods. In some aspects, large-scale or genome-wide methods can be used to identify one or more molecular signatures. In some aspects, one or more molecular signatures, such as the expression of specific RNA or protein in cells, can be determined. In some embodiments, the molecular characteristics of phenotype are analyzed by image analysis, PCR (including standard and all variations of PCR), microarray (including but not limited to DNA microarray, MM chip for microRNA, protein microarray, cell microarray, antibody microarray and carbohydrate array), sequencing, biomarker detection, or methods for determining DNA methylation or protein glycosylation patterns.In certain embodiments, the specific molecule is a polypeptide, i.e., a protein.In some embodiments, the specific molecule is a polynucleotide.

[0071] In some embodiments, positive or negative expression of a particular molecule is expressed on positively or negatively selected cells, respectively (markers + ) or expressed at relatively high levels (markers 高) by incubating the cells with one or more antibodies or other binding agents that specifically bind to one or more surface markers. In certain embodiments, positive or negative expression is determined by flow cytometry, immunohistochemistry, or any other suitable method for detecting a particular marker.

[0072] In certain embodiments, the expression of specific molecules is evaluated by using flow cytometry.Flow cytometry is a biophysical technique that is based on laser or impedance, and is used for cell counting, cell sorting, biomarker detection and protein manipulation by suspending cells in a fluid flow and passing the cells through an electronic detection device.This allows simultaneous multiparametric analysis of the physical and chemical properties of up to several thousand particles per second.

[0073] Data generated by a flow cytometer can be plotted in one dimension to generate histograms, in two-dimensional dot plots, or even in three dimensions. Regions on these plots can be sequentially separated based on fluorescence intensity by creating a series of subset extractions called "gates." Specific gating protocols exist for diagnostic and clinical purposes, particularly in the context of immunology. Plots are often performed on a logarithmic scale. Because the emission spectra of various fluorochromes overlap, the signals at the detector must be electronically and computationally compensated. Data accumulated using a flow cytometer can be analyzed using software such as JMP (statistical software), WinMDI, Flow Software, and the web-based Cytobank (Cytobank), Cellcion, FCS Express, FlowJo, FACSDiva, CytoPaint (also known as Paint-A-Gate), VenturiOne, CellQuest Pro, Infinicyt, or Cytospec.

[0074] Flow cytometry is a standard technique in the art, and one of skill in the art would readily understand how to design or tailor a protocol to detect one or more specific molecules and analyze the data to determine the expression of one or more specific molecules in a population of cells. Standard protocols and techniques for flow cytometry can be found in Loyd "Flow Cytometry in Microbiology; Practical Flow Cytometry by Howard M. Shapiro; Flow Cytometry for Biotechnology by Larry A. Sklar; Handbook of Flow Cytometry Methods by J. Paul Robinson, et al., Current Protocols in Cytometry, Wiley-Liss Pub; Flow Cytometry in Clinical Diagnosis, v4, (Carey, McCoy, and Keren, eds), ASCP Press, 2007; Ormerod, MG (ed.) (2000) Flow Cytometry - A practical approach. 3rd edition. Oxford University Press, Oxford, UK; Ormerod, MG (1999) Flow Cytometry. 2nd edition. BIOS Scientific Publishers, Oxford.; and Flow Cytometry - A basic introduction. Michael G. Ormerod, 2008.

[0075] In some embodiments, cells are sorted by phenotype for subsequent analysis.In some embodiments, cells of different phenotypes in the same cell composition are sorted by fluorescence-activated cell sorting (FACS).FACS is a special type of flow cytometry that allows heterogeneous mixtures of cells to be sorted into two or more containers one cell at a time based on the specific light scattering and fluorescent properties of each cell.FACS is a useful scientific instrument because it provides rapid, objective and quantitative recording of the fluorescent signals from individual cells, and physical separation of specific cells of interest.

[0076] In some embodiments, the input composition features or therapeutic composition features may include any one or more of the parameters or activities associated with features of a cellular composition, e.g., an input cell composition or a therapeutic T cell composition (e.g., CAR-T cells), respectively, described in WO 2019 / 032929, WO 2018 / 223101, WO 2019 / 089848, WO 2020 / 113194, WO 2019 / 090003, WO 2020 / 092848, WO 2019 / 113559, and WO 2018 / 157171, which are incorporated by reference herein in their entireties. In some embodiments, the subject features may include any one or more of subject features or characteristics or subject-related features or characteristics (e.g., subject attributes or clinical attributes for subjects in clinical trials involving administration of a therapeutic T cell composition) described in WO 2019 / 032929, WO 2018 / 223101, WO 2019 / 089848, WO 2020 / 113194, WO 2019 / 090003, WO 2020 / 092848, WO 2019 / 113559, and WO 2018 / 157171, which are incorporated by reference herein in their entireties. In some embodiments, the clinical response to the therapeutic cell composition (e.g., CAR-T cells) may include any one or more of the clinical responses to the therapeutic cell composition (e.g., CAR-T cells) described in International Publication Nos. WO 2019 / 032929, WO 2018 / 223101, WO 2019 / 089848, WO 2020 / 113194, WO 2019 / 090003, WO 2020 / 092848, WO 2019 / 113559, and WO 2018 / 157171, which are incorporated by reference herein in their entireties. Any one or more of such features may be used as data for determining (e.g., predicting) any one or more clinical responses according to the provided methods.

[0077] Non-limiting examples of target features, input composition features, and therapeutic cellular composition features used as data in the provided methods for determining (e.g., predicting) one or more non-limiting clinical responses are described in the subsections below.

[0078] 1. Target features Various features associated with a subject treated with a therapeutic cell composition are contemplated for use by the methods provided herein, e.g., machine learning methods. A subject treated with a therapeutic cell composition may also be referred to herein as a patient.

[0079] In some embodiments, the subject feature includes subject attributes, such as age and weight. In some embodiments, the subject feature is the weight of the subject, e.g., body weight. In some embodiments, the subject weight is the weight of the subject at the time the therapeutic cell composition is administered. In certain embodiments, weight is measured in lb or kg. In some embodiments, the subject feature is age, e.g., the age of the subject at the start of administration of the therapeutic cell composition. Other exemplary subject features include height, ethnicity, race, sex, gender, and body mass index.

[0080] In some embodiments, the features associated with a subject are clinical attributes. Exemplary clinical attributes include, but are not limited to, biomarkers and biomarker combinations, disease diagnosis, disease burden, disease duration, disease severity (e.g., disease grade), and treatment history.

[0081] In some embodiments, clinical attributes (e.g., subject features) associated with a subject include the amount of a prior therapeutic agent, e.g., one or more therapeutic agents, prior to the initiation of administration of a therapeutic T cell composition. In some embodiments, the prior therapeutic agent is administered to treat the same disease and / or condition as the therapeutic cell composition.

[0082] In certain embodiments, the clinical attribute is a platelet count.

[0083] In some embodiments, the clinical attribute is the most recent disease diagnosis. In some embodiments, the clinical attribute is the diagnosis the subject has received.

[0084] In certain embodiments, the clinical attribute is having leukemia. In some embodiments, the target feature is having B-cell leukemia. In certain embodiments, the leukemia is acute lymphoblastic leukemia (ALL), non-Hodgkin's lymphoma (NHL), chronic lymphocytic leukemia (CLL), diffuse large B-cell lymphoma (DLBCL), or acute myeloid leukemia (AML). In certain embodiments, the clinical feature is having acute lymphocytic leukemia (ALL). In some embodiments, the clinical attribute is having lymphoma. In some embodiments, the clinical attribute is having a particular grade of lymphoma. In some embodiments, the clinical attribute is having DLBCL. In some embodiments, the clinical attribute is having follicular lymphoma. In some embodiments, the clinical attribute is having DLBCL transformed from follicular lymphoma. In some embodiments, the clinical attribute is the cellular origin of the DLBCL. For example, in some embodiments, the cell of origin is an activated B cell, a non-germinal center B cell, or a germinal center B cell-like cell. In some embodiments, the clinical attribute is whether the disease is de novo or otherwise. For example, in some embodiments, the clinical attribute is de novo DLBCL or non-de novo DLBCL.

[0085] In some embodiments, the clinical attribute is a genetic phenotype, e.g., the identification of a mutation in a gene known to be correlated with or associated with a disease or condition. In some embodiments, the clinical attribute is whether a gene, e.g., a gene correlated with or associated with a disease or condition, has one or more mutations, e.g., a deletion, insertion, substitution, rearrangement, translocation. In some embodiments, the clinical attribute is the number of mutated genes (e.g., hits). In some embodiments, the clinical attribute is a genetic double hit. For example, in some lymphomas, two genes, e.g., MYC and BCL2, may be mutated. In some embodiments, the clinical attribute is a genetic triple hit. For example, in some lymphomas, three genes, e.g., MYC, BCL6, and BCL2, may be mutated. In some embodiments, the clinical attribute is a genetic double hit or triple hit. In some embodiments, the clinical attribute is a genetic double expressor. For example, in some lymphomas, a dual expressor or double expressor indicates immunohistochemical detection of overexpression of MYC and BCL2. In some embodiments, double expressors indicate genes that are overexpressed, for example, compared to baseline.

[0086] In some aspects, the clinical attribute is whether the subject has relapsed or refractory disease.In some aspects, the clinical attribute is whether the subject has relapsed or become refractory to one or more previous therapeutic agents.In some embodiments, the clinical attribute is whether the subject has relapsed or become refractory after chemotherapy treatment.

[0087] It is contemplated that a subject treated with a therapeutic cell composition may have undergone previous treatments in an attempt to treat a disease or condition. Accordingly, in some embodiments, a clinical attribute is the number of previous lines of therapy a subject has received prior to treatment with a therapeutic cell composition. In some embodiments, a clinical attribute is the number of previous lines of systemic therapy a subject has received prior to treatment with a therapeutic cell composition. In some embodiments, a clinical attribute is whether a subject has undergone an allogeneic hematopoietic stem cell transplant prior to treatment with the therapeutic cell composition. In some embodiments, a clinical attribute is whether a subject has undergone an autologous hematopoietic stem cell transplant prior to treatment with the therapeutic cell composition. In some embodiments, a clinical attribute is best overall response to a previous treatment.

[0088] In some embodiments, the clinical attribute is disease stage.

[0089] In some embodiments, the clinical attribute is disease burden. In certain embodiments, the clinical attribute is high disease burden, e.g., high disease burden before initiation of administration of a therapeutic T cell composition. In certain embodiments, the clinical attribute is high disease burden immediately prior to initiation of administration of a therapeutic cell composition, or within 1 week, 2 weeks, 3 weeks, 4 weeks, 5 weeks, 6 weeks, 7 weeks, 8 weeks, 9 weeks, 1 month, 2 months, 3 months, 4 months, 5 months, 6 months, or more than 6 months before initiation of administration of a therapeutic cell composition. In some embodiments, disease burden is measured by lesion count. In some embodiments, high disease burden is measured based on the percentage of bone marrow blasts. In certain embodiments, the feature of interest is high disease burden, e.g., two-way product sum (SPD) or lactate dehydrogenase (LDH) levels.

[0090] In some embodiments, the clinical attribute is lesion count. In some embodiments, the clinical attribute is SPD. In some embodiments, the clinical attribute is LDH level. In some embodiments, the clinical attribute is fold change in SPD. In some embodiments, the clinical attribute is fold change in LDH level. In some embodiments, the fold change is determined between the time of initial screening and the time point at which lymphocyte depletion therapy is performed prior to administration of the therapeutic cell composition. In some embodiments, the fold change is determined between the time of initial screening and the time point at which the therapeutic cell composition is administered. In some embodiments, the clinical attributes SPD, LDH, lesion count, and fold change, difference, or other quantitative value are used to assess disease burden.

[0091] In some embodiments, the clinical attribute is disease burden, e.g., as indicated by tumor burden. In some embodiments, the clinical attribute is high tumor burden, e.g., high disease burden before initiation of administration of a therapeutic cell composition. In some embodiments, tumor burden is determined by one or more volumetric measurements of the tumor. In some embodiments, the volumetric measurements are measurements of the one or more lesions, e.g., tumor size, tumor diameter, tumor volume, tumor mass, tumor cell burden or tumor-enhancing area, tumor-associated edema, tumor-associated necrosis, and / or the number or extent of metastases. Also, "bulky lesions" can be used to refer to large tumors in the breast. In some embodiments, the volumetric measurements of the tumor are two-dimensional measurements. For example, in some embodiments, the area of ​​one or more lesions is calculated as the product of the longest diameter and the longest perpendicular diameter of all measurable tumors. In some embodiments, the volumetric measurements of the tumor are one-dimensional measurements. In some embodiments, the size of a measurable lesion is assessed as the longest diameter. In some embodiments, tumor size is assessed as the longest diameter. In some embodiments, tumor size is assessed as the perpendicular diameter. In some embodiments, the sum of two-dimensional products (SPD), tumor longest diameter (LD), sum of tumor longest diameters (SLD), necrosis, tumor volume, necrotic volume, necrosis-tumor ratio (NTR), peritumoral edema (PTE), and edema-tumor ratio (ETR) are measured. Exemplary methods for measuring and assessing tumor burden include those described, for example, in Carceller et al., Pediatr Blood Cancer. (2016) 63(8):1400-1406 and Eisenhauer et al., Eur J Cancer. (2009) 45(2):228-247. In some embodiments, the volumetric measurement is the sum of two-dimensional products (SPD), measured by taking the sum of the products of the largest perpendicular diameters of all measurable tumors. In some aspects, tumors or lesions are measured in one dimension by the longest diameter (LD) and / or by taking the sum of the tumor longest diameters (SLD) of all measurable lesions.In some embodiments, the tumor volumetric measurement is a volumetric quantification of tumor necrosis, such as necrotic volume and / or necrosis-tumor ratio (NTR), see Monsky et al., Anticancer Res. (2012) 32(11):4951-4961. In some aspects, the tumor volumetric measurement is a volumetric quantification of tumor-associated edema, such as peritumoral edema (PTE) and / or edema-tumor ratio (ETR). In some embodiments, the measurement can be performed using a subject imaging technique, such as computed tomography (CT), positron emission tomography (PET), and / or magnetic resonance imaging (MRI).

[0092] In some embodiments, tumor volumetric measurements are taken during screening sessions, such as routine evaluations or blood sampling to confirm and / or identify the subject's condition or disease. In some embodiments, tumor burden measurements, such as volumetric measurements (e.g., SPD), are taken before lymphocyte-depleting chemotherapy (LDC). For example, in some embodiments, tumor burden measurements, such as volumetric measurements (e.g., SPD), are taken or evaluated within 1 month, 2 weeks, or 1 week before LDC, for example, within 7 days, 6 days, 5 days, 4 days, 3 days, 2 days, or 1 day before LDC. In certain embodiments, tumor burden measurements, such as volumetric measurements (e.g., SPD), are taken before infusion of T cell therapy into tumor-bearing subjects.

[0093] In some embodiments, the target feature is a categorical cutoff value of tumor burden measurement, for example, volumetric measurement, whether above or below a threshold level.For example, the feature is a categorical cutoff value of tumor burden measurement (for example, volumetric measurement), for example, measured before infusion of T cell therapy, for example, before LDC, and the feature is whether the subject has tumor burden measurement that is at or below a threshold level, or has tumor burden measurement that is greater than a threshold level.In certain embodiments, the tumor burden measurement is SPD, and the threshold level of SPD is 30 cm 2 , 40 cm 2 , 50 cm2 , 60 cm 2 , 70 cm 2 , 80 cm 2 Or 90cm 2 or about 30 cm 2 , 40 cm 2 , 50 cm 2 , 60 cm 2 , 70 cm 2 , 80 cm 2 Or 90cm 2 For example, the feature is a categorical cutoff value of SPD measured before the infusion of a T cell therapy drug, for example, before LDC, and the feature is 2 Or 50 cm 2 Less than or equal to 50 cm 2 The question is whether they have super SPD.

[0094] In some embodiments, a factor indicative of tumor burden is assessed at two time points, and a fold change in the factor indicative of disease burden between the two time points is determined. In some embodiments, the two time points include a first time point and a second time point, and the fold change is the ratio of the factor indicative of disease burden at the first time point to the factor indicative of disease burden at the second time point. In some embodiments, tumor volumetric measurements are obtained at two time points before administration of a treatment, e.g., cell therapy. In some embodiments, tumor volumetric measurements are obtained at a screening session, e.g., during a routine evaluation or blood draw to confirm and / or identify the subject's condition or disease. In certain embodiments, one or more tumor volumetric measurements are measured or measured in a subject who is, will be, or is a candidate for administration of a T cell therapy. In certain embodiments, measurements are obtained before treatment, e.g., cell therapy, or before administration of a therapeutic agent, e.g., a cell therapy. In some embodiments, both of the two time points are no more than one or two months prior to receiving cell therapy. In some embodiments, the two time points are separated by one or more, two, three, four, or five weeks. In some embodiments, the two time points are 3 weeks or more apart. In some embodiments, the two time points are not more than 4 weeks apart, not more than 5 weeks or 6 weeks apart. In some embodiments, the second time point is more than 1, 2, 3, 4, 5, 6, or 7 days prior to administration of the cell therapy.

[0095] In some embodiments, the clinical attribute is disease burden as determined by extranodal disease classification. For example, the clinical attribute can be whether a disease, e.g., lymphoma, has spread to organs outside the lymphatic system. In some embodiments, the clinical attribute is the number of extranodal sites affected.

[0096] In some embodiments, the clinical attribute is whether the subject has a central nervous system (CNS) disease at the time the therapeutic cell composition is administered. In some embodiments, the subject does not have a CNS disease at the time the therapeutic cell composition is administered. In some embodiments, the CNS disease is primary CNS lymphoma (PCNSL). In some embodiments, PCNSL infiltrates the central nervous system (CNS) without the presence of systemic lymphoma. In some embodiments, PCNSL is localized to the brain, spine, cerebrospinal fluid (CSF), and eyes. In some embodiments, PCNSL is diffuse large B-cell lymphoma (DLBCL). In some embodiments, PCNSL is Burkitt's lymphoma, low-grade lymphoma, or T-cell lymphoma. In some embodiments, PCNSL includes neurological signs. In some embodiments, neurological signs include focal neurological deficits, changes in mental status and behavior, symptoms of elevated intracranial pressure, and / or seizures. In some embodiments, exemplary features associated with the disease or condition include those described in Grommes et al. (J. Clin Oncol 2017;35(21):2410-18).

[0097] In some embodiments, the CNS disease is secondary central nervous system lymphoma (SCNSL). In some embodiments, the SCNSL is in patients with systemic lymphoma. In some embodiments, the SCNSL is referred to as metastatic lymphoma. In some embodiments, the SCNSL is DLBCL. In some embodiments, the SCNSL is an aggressive lymphoma that can infiltrate the brain, meninges, spinal cord, and eyes. In some embodiments, the SCNSL includes leptomeningeal spread. In some embodiments, the SCNSL includes disease of the brain parenchyma. In some embodiments, exemplary features associated with the disease or condition include those described in Malikova et al. (Neurophychiatric Disease and Treatment 2018;14:733-40). In some embodiments, the secondary CNS lymphoma invades the brain parenchyma and / or leptomeninges.

[0098] In some embodiments, the clinical attribute is a comorbidity. For example, in some cases, the comorbidity is creatinine clearance (CrCl) before the subject undergoes lymphocyte-depleting chemotherapy before administration of the therapeutic cell composition. In some embodiments, the comorbidity is left ventricular ejection fraction (LVEF).

[0099] In some embodiments, the clinical attribute is Eastern Cooperative Oncology Group (ECOG) performance status. In some embodiments, the subject's ECOG status is defined as: Grade 0 - fully active, able to continue to perform all pre-disease activities without limitation; Grade 1 - ambulatory with limited physical activity but able to perform light or sedentary tasks; Grade 2 - ambulatory and able to care for themselves but unable to perform any work activities; active and mobile more than 50% of waking hours; Grade 3 - limited self-care; more than 50% of waking hours confined to a bed or chair; Grade 4 - completely immobile; unable to continue to care for themselves at all; entirely confined to a bed or chair; or Grade 5 - death.

[0100] In some embodiments, the clinical attribute is the International Prognostic Index score (IPI), e.g., an IPI score for lymphoma. In some embodiments, a subject's IPI score is defined as: low risk (0-1 point) - 73% 5-year survival rate; low-intermediate risk (2 points) - 51% 5-year survival rate; high-intermediate risk (3 points) - 43% 5-year survival rate; or high risk (4-5 points) - 26% 5-year survival rate.

[0101] In some embodiments, the clinical attribute is the subject's body temperature. In some embodiments, the clinical attribute is a blood oxygen level. In some embodiments, the clinical attribute is an albumin level. In some embodiments, the clinical attribute is an alkaline phosphatase level. In some embodiments, the clinical attribute is a basophil count. In some embodiments, the clinical attribute is an absolute basophil count. In some embodiments, the clinical attribute is direct bilirubin. In some embodiments, the clinical attribute is total bilirubin. In some embodiments, the clinical attribute is a lymphocyte count or an absolute lymphocyte count. In some embodiments, the lymphocyte count is the count prior to leukapheresis of the subject to obtain cells for generating a therapeutic cell composition. In some embodiments, the clinical attribute is a blood urea nitrogen level. In some embodiments, the clinical attribute is a calcium level. In some embodiments, the clinical attribute is a carbon dioxide level. In some embodiments, the clinical attribute is a chloride level. In some embodiments, the clinical attribute is a creatinine level. In some embodiments, the clinical attribute is an eosinophil count or an absolute eosinophil count. In some embodiments, the clinical attribute is a glucose level. In some embodiments, the clinical attribute is a hematocrit level. In some embodiments, the clinical attribute is a hemoglobin level. In some embodiments, the clinical attribute is a magnesium level. In some embodiments, the clinical attribute is a monocyte count or absolute monocyte count. In some embodiments, the clinical attribute is a neutrophil count or absolute neutrophil count. In some embodiments, the clinical attribute is a platelet count. In some embodiments, the clinical attribute is a potassium level. In some embodiments, the clinical attribute is a total protein level. In some embodiments, the clinical attribute is a red blood cell count. In some embodiments, the clinical attribute is a white blood cell count. In some embodiments, the clinical attribute is a uric acid level. In some embodiments, the clinical attribute is a sodium level. In some embodiments, the clinical attribute is a triglyceride level. In some embodiments, the clinical attribute is an aspartate amine transferase level.In some cases, aspartate aminotransferase levels can be measured by a serum glutamic oxaloacetic transaminase test. In some embodiments, the clinical attribute is alanine aminotransferase levels. In some cases, aspartate aminotransferase levels can be measured by a serum glutamic pyruvic transaminase test. Any suitable method for detecting the described levels or numbers is contemplated.

[0102] In some embodiments, the clinical attribute is the level, amount, and / or concentration of an inflammatory marker. In some embodiments, the inflammatory marker is or includes the level or presence of C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR), albumin, ferritin, β2-microglobulin (β2-M), or lactate dehydrogenase (LDH) is detected and evaluated. In some embodiments, the inflammatory marker is evaluated using an immunoassay. For example, enzyme-linked immunosorbent assay (ELISA), enzyme-linked immunosorbent assay (EIA), radioimmunoassay (RIA), surface plasmon resonance (SPR), Western blot, lateral flow assay, immunohistochemistry, protein array, or immuno-PCR (iPCR) can be used to detect the inflammatory marker. In some embodiments, the presence, level, amount, and / or concentration of the inflammatory marker indicates tumor burden, e.g., high tumor burden. In some cases, the assay or evaluation of the inflammatory marker is by flow cytometry. In some cases, the reagent is a soluble protein that binds to the inflammatory marker. In some examples, the reagent is a protein that binds to C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), albumin, ferritin, β2 microglobulin (β2-M), or lactate dehydrogenase (LDH).

[0103] In some embodiments, the clinical attribute is a biomarker. In some embodiments, the biomarker is an inflammatory marker, such as C-reactive protein (CRP). In some embodiments, CRP is assessed using an in vitro enzyme-linked immunosorbent assay to obtain a quantitative measurement of human CRP from a sample, such as serum, plasma, or blood. In some examples, CRP is detected using a human enzyme-linked immunosorbent assay (ELISA). In some embodiments, the biomarker is an inflammatory marker, such as erythrocyte sedimentation rate (ESR). In some embodiments, ESR is assessed by measuring the distance (in millimeters per hour) that red blood cells fall after separation from plasma in an upright pipette or tube. In some embodiments, the biomarker is or includes albumin. In some aspects, albumin is assessed using a colorimetric test or an in vitro enzyme-linked immunosorbent assay. In some examples, albumin is detected using a human enzyme-linked immunosorbent assay (ELISA). In some embodiments, the biomarker is an inflammatory marker, such as ferritin or β2 microglobulin. In some embodiments, ferritin or β2 microglobulin is evaluated using immunoassay or detected using ELISA. In some aspects, the biomarker is an inflammatory marker, such as lactate dehydrogenase (LDH), and LDH is evaluated using colorimetric test or in vitro enzyme-linked immunosorbent assay. In some embodiments, the clinical attribute is the level, concentration and / or amount of LDH.

[0104] In some embodiments, the level, concentration and / or number of LDH is a surrogate for disease burden, eg, tumor or cancer.

[0105] In some embodiments, the clinical attribute is receiving bridging chemotherapy prior to initiation of administration of the therapeutic T cell composition. In some embodiments, the bridging chemotherapy is a systemic treatment. In some embodiments, the clinical attribute is receiving bridging chemotherapy and radiation therapy prior to initiation of administration of the therapeutic T cell composition. The treating physician can determine whether bridging therapy is required, for example, for disease control, during the manufacture of the provided composition or cells.

[0106] In certain embodiments, the clinical attribute is preconditioning with, for example, lymphocyte depletion therapy prior to the initiation of administration of a therapeutic T cell composition. In some embodiments, the lymphocyte depletion therapy is or includes the administration of a chemotherapy drug. In certain embodiments, the target feature is preconditioning with fludarabine and / or cyclophosphamide prior to the initiation of administration of a therapeutic cell composition. In certain embodiments, the target feature is preconditioning with cyclophosphamide prior to the initiation of administration of a therapeutic T cell composition. In some embodiments, the target feature is preconditioning with fludarabine and cyclophosphamide prior to the initiation of administration of a therapeutic cell composition.

[0107] In some embodiments, the clinical attribute is the level, amount, or concentration of a cytokine in a blood, serum, or plasma sample prior to the start of administration of the therapeutic T cell composition. In some embodiments, the cytokine is an interleukin, such as interleukin-15 (IL-15).

[0108] In some embodiments, the subject feature is the dose group of the study in which the subject is treated. In some cases, this feature can be particularly useful in some clinical trials where different dose levels are used to account for differences in dosing, for example, when assessing clinical response according to the methods provided herein.

[0109] In some embodiments, the subject features include any one or more subject features described herein, e.g., clinical attributes and subject attributes.

[0110] In some embodiments, the subject features include dose group, bridging chemotherapy, bridging chemotherapy and radiation therapy, bridging chemotherapy systemic treatment, cell origin (e.g., ABC (activated B cell-like or non-GCB) or GCB (germinal center B cell-like), relapse or refractory after chemotherapy, type of diagnosis, disease cohort (e.g., DLBCL), disease burden, relapsed or refractory disease, disease origin (e.g., de novo DLBCL or other DLBCL), gender, route of administration of the therapeutic cell composition (e.g., infusion), fold change in LDH, height, lesion count, oxygen saturation, body temperature (°C), maximum diameter of tumor before treatment with the therapeutic cell composition, fold change in SPD, SPD value before lymphodepleting chemotherapy (e.g., SPD<=50 cm^2 or >50 cm^2), or(with categorical threshold for group with 1000 or more cm^2), BMI, weight, sex, ethnicity, race, age, IPI score, ECOG score, disease stage, disease burden based on LDH before lymphocyte-depleting chemotherapy, disease burden based on SPD before lymphocyte-depleting chemotherapy, subject having active CNS disease at the time of treatment, disease burden based on extranodal disease classification, number of extranodal sites, disease burden based on bulky disease classification, medical history, number of previous lines of therapy, number of previous lines of systemic therapy, prior allogeneic hematopoietic stem cell transplant (allo -HSCT), prior autologous hematopoietic stem cell transplant (auto-HSCT), chemotherapy-refractory or chemotherapy-sensitive disease type, bridging anticancer treatment for disease control, number of days from date of leukapheresis to first infusion, number of months from diagnosis to treatment with therapeutic cell composition, comorbidities (e.g., creatinine clearance (CrCl) before lymphodepletion, left ventricular ejection fraction (LVEF) at screening), baseline C-reactive protein (CRP), lymphocyte count before leukapheresis (10^9 / L), gene double expressor, gene double hit, gene triple hit, gene double or triple hit, gene double or triple hit or double expressor, albumin level, alkaline phosphatase level, basophil count, absolute basophil count, direct bilirubin, total bilirubin, blood urea nitrogen level, calcium level, carbon dioxide level, chloride level, creatinine level, eosinophil count, absolute eosinophil count, glucose level, hematocrit level, hemoglobin level, LDH level, lesion count, lymphocyte count, absolute lymphocyte count, magnesium level, absolute monocyte count, monocyte count, absolute neutrophil count, neutrophil count, phosphate level, platelet count, potassium level, total protein, red blood cell count, aspartate aminotransferase level, alanine aminotransferase level, sodium level, two-way product sum, triglycerides, tumor maximum diameter, tumor perpendicular diameter, uric acid level, and white blood cell count.

[0111] In some embodiments, the features of interest include any one or more of the features of interest shown in Table E4 below.

[0112] In some embodiments, the target feature is determined during an initial screening, e.g., screening prior to leukapheresis to generate an input composition for making a therapeutic cell composition. In some embodiments, the target feature is determined prior to administration of a lymphocyte-depleting therapy drug prior to administration of the therapeutic cell composition. In some embodiments, the target feature is determined at the time the therapeutic cell composition is administered. In some cases, e.g., when the target feature is or comprises a change in a target feature, the target feature can be measured at two or more time points, e.g., during initial screening, prior to administration of lymphocyte-depleting therapy, and prior to administration of the therapeutic cell composition, and the difference or change in the target feature, e.g., percent change, fold change, etc., is determined.

[0113] 2. Input composition features In some embodiments, the input composition contains cells isolated from a sample (e.g., a biological sample), cells obtained from a subject, such as a subject in need of cell therapy or a subject to whom cell therapy will be administered, or cells derived from a subject. Methods for isolating cells from a sample (e.g., a biological sample) are described, for example, in Section II-A. In some aspects, the subject is a human, such as a patient in need of a specific therapeutic intervention, such as adoptive cell therapy, from which cells have been isolated, treated, and / or manipulated. Thus, in some embodiments, the cells are primary cells, e.g., primary human cells. In some embodiments, the input composition contains CD4+ and CD8+ T cells. In some embodiments, the input composition contains CD4+ or CD8+ T cells.

[0114] In some embodiments, the input composition features comprise a cell phenotype. In some embodiments, the phenotype is the number of total T cells. In some embodiments, the phenotype is the number of total CD3 +In some embodiments, the phenotype is or includes the identity of a T cell subtype. Various populations or subtypes of T cells include effector T cells, helper T cells, memory T cells, regulatory T cells, naive T cells, CD4 + cells and CD8 + This includes, but is not limited to, T cell.In some embodiments, T cell subtypes can be identified by detecting the presence or absence of specific molecules.In some specific embodiments, specific molecules are surface markers that can be used to identify T cell subtypes.

[0115] In some embodiments, the phenotype is positive or high expression of one or more specific molecules that are surface markers, e.g., CD3, CD4, CD8, CD28, CD62L, CCR7, CD27, CD127, CD4, CD8, CD45RA and / or CD45RO. In certain embodiments, the phenotype is, for example, positive or high expression of one or more surface markers, e.g., CD3 + , CD4 + , CD8 + , CD28 + , CD62L + , CCR7 + , CD27 + , CD127 + , CD4 + , CD8 + , CD45RA + and / or CD45RO +The phenotype is a surface marker of T cells, or a surface marker of a subpopulation or subset of T cells, based on positive surface marker expression of one or more specific molecules that are surface markers, such as CC chemokine receptor type 7 (CCR7), cluster of differentiation 27 (CD27), cluster of differentiation 28 (CD28), and cluster of differentiation 45 RA (CD45RA). In certain embodiments, phenotypic markers include CCR7, CD27, CD28, CD44, CD45RA, CD62L, and L-selectin. In some embodiments, the phenotype is negative expression or absence of expression of one or more specific molecules that are surface markers, such as CD3, CD4, CD8, CD28, CD62L, CCR7, CD27, CD127, CD4, CD8, CD45RA, and / or CD45RO. In certain embodiments, the phenotype is, for example, negative expression or absence of expression of one or more surface markers, such as CD3 - , CD4 - , CD8-, CD28 - , CD62L - , CCR7 - , CD27 - , CD127 - , CD4 - , CD8 - , CD45RA - and / or CD45RO - The phenotype is a surface marker of T cells, or a surface marker of a subpopulation or subset of T cells, based on the absence of surface marker expression. In some embodiments, the phenotype is the negative expression or absence of expression of one or more specific molecules that are surface markers, such as CC chemokine receptor type 7 (CCR7), cluster of differentiation 27 (CD27), cluster of differentiation 28 (CD28), and cluster of differentiation 45 RA (CD45RA). In some specific embodiments, the phenotype markers include CCR7, CD27, CD28, CD44, CD45RA, CD62L, and L-selectin.

[0116] In certain embodiments, the phenotype is or includes positive or negative expression of CD27, CCR7 and / or CD45RA. + In some embodiments, the phenotype is CD27 + In some embodiments, the phenotype is CCR7 - In some embodiments, the phenotype is CD27 - In some embodiments, the phenotype is CCR7 + / CD27 + In some embodiments, the phenotype is CCR7 - / CD27 + In some embodiments, the phenotype is CCR7 + In some embodiments, the phenotype is CCR7 - / CD27 - In some embodiments, the phenotype is CD45RA - In some embodiments, the phenotype is CD45RA + In some embodiments, the phenotype is CCR7 + / CD45RA - In some embodiments, the phenotype is CD27 + / CD45RA - In some embodiments, the phenotype is CD27 + / CD45RA + In some embodiments, the phenotype is CD27 - / CD45RA + In some embodiments, the phenotype is CD27 - / CD45RA - In some embodiments, the phenotype is CCR7 + / CD27 + / CD45RA - In some embodiments, the phenotype is CCR7 + / CD27 + / CD45RA+.

[0117] In some embodiments, the phenotype is viability.In some specific embodiments, the phenotype is the positive expression of markers that indicate that cells undergo normal functional cellular processes and / or do not undergo necrosis or programmed cell death, or are not in the process of necrosis or programmed cell death.In some embodiments, viability can be assessed by cellular redox potential, cell membrane integrity, or mitochondrial activity or function.In some embodiments, viability is the absence of specific molecules associated with cell death, or the absence of cell death signs in assays.

[0118] In some embodiments, the phenotype is or includes cell viability. In some specific embodiments, cell viability can be detected, measured, and / or evaluated by several means commonly used in the art. Non-limiting examples of such viability assays include, but are not limited to, dye uptake assays (e.g., calcein AM assay), XTT cell viability assays, and dye exclusion assays (e.g., trypan blue dye exclusion assay, eosin dye exclusion assay, or propidium dye exclusion assay). Viability assays are useful for determining the number or percentage (e.g., frequency) of viable cells in cell dose, cell composition, and / or cell sample. In certain embodiments, the phenotype includes cell viability along with other characteristics, such as surface markers, molecules.

[0119] In some particular embodiments, the phenotype is cell viability, viable CD3 + , survival CD4 + , survival CD8 + , survival CD4 + / CCR7 + , survival CD8 + / CD27 + , survival CD4 + / CD27 + , survival CD8 + / CCR7 + / CD27 + , survival CD4 + / CCR7 + / CD27+ , survival CD8 + / CCR7 + / CD45RA - or viable CD4 + / CCR7 + / CD45RA - or a combination thereof.

[0120] In certain embodiments, phenotype is or comprises the absence of apoptosis and / or the sign that cell is undergoing apoptosis process.Apoptosis is a programmed cell death process that includes a series of stereotypical morphological and biochemical events that lead to characteristic cell changes and cell death.These changes include blebbing, cell shrinkage, nuclear fragmentation, chromatin condensation, chromosomal DNA fragmentation and global mRNA degradation.Apoptosis is a well-characterized process, and the specific molecules associated with various stages are well known in the art.

[0121] In some embodiments, the phenotype is the absence of the early stages of apoptosis and / or the absence of indicators and / or specific molecules associated with the early stages of apoptosis. In the early stages of apoptosis, changes in the cell membrane and mitochondrial membrane are evident. Biochemical changes are also evident in the cell cytoplasm and nucleus. For example, the early stages of apoptosis can be indicated by the activation of specific caspases, such as 2, 8, 9, and 10. In certain embodiments, the phenotype is the absence of the late stages of apoptosis and / or the absence of indicators and / or specific molecules associated with the late stages of apoptosis. The middle to late stages of apoptosis are characterized by further loss of membrane integrity, chromatin condensation, DNA fragmentation, and biochemical events such as the activation of caspases 3, 6, and 7.

[0122] In certain embodiments, the phenotype is negative expression of one or more factors associated with apoptosis, including pro-apoptotic factors known to initiate apoptosis, such as members of the death receptor pathway, activated members of the mitochondrial (intrinsic) pathway, such as Bcl-2 family members, e.g., Bax, Bad, and Bid, and caspases. In some embodiments, the phenotype is negative or low levels of apoptotic markers. In certain embodiments, the phenotype is negative expression of apoptotic markers. In certain embodiments, the phenotype is an indicator, such as the absence of Annexin V molecules, which preferentially bind to cells undergoing apoptosis when incubated with or contacted with a cellular composition. In some embodiments, the phenotype is or includes the expression of one or more markers indicative of an apoptotic state in a cell.

[0123] In some embodiments, phenotype is the negative (or low) expression of certain molecules that are the markers of apoptosis.Various apoptosis markers are known to those skilled in the art, and include but are not limited to the activity of one or more caspases, namely the increase in activated caspase (active caspase, CAS), the increase in PARP cleavage, the activation and / or translocation of Bcl-2 family proteins, cell death pathway members such as Fas and FADD, the existence of nuclear shrinkage (for example, by microscopic monitoring), and the existence of chromosomal DNA fragmentation (for example, the existence of chromosomal DNA ladder), or by apoptosis assays including TUNEL staining and Annexin V staining.

[0124] Caspases are enzymes that cleave proteins after aspartic acid residues, and the term derives from "cysteine-aspartic acid protease." Because caspases are involved in apoptosis, activation of caspases, such as caspase-3, indicates increased or resurrected apoptosis. In some embodiments, activated caspase-3 is referred to herein as 3CAS. In certain embodiments, caspase activation can be detected by methods known to those skilled in the art. In some embodiments, caspase activation can be detected using an antibody that specifically binds to activated caspase (i.e., specifically binds to the cleaved polypeptide). In another example, a fluorochrome inhibitor of caspase activity (FLICA) assay can be used to detect caspase-3 activation by detecting the hydrolysis of acetylAsp-Glu-Val-Asp 7-amido-4-methylcoumarin (Ac-DEVD-AMC) by caspase-3 (i.e., detecting the release of fluorescent 7-amino-4-methylcoumarin (AMC)). FLICA assays can be used to determine caspase activation by detecting the products of substrates processed by multiple caspases (e.g., FAM-VAD-FMK FLICA). Other techniques include the CASPASE-GLO® Caspase Assay (PROMEGA), which uses luminogenic caspase-8 tetrapeptide substrates (Z-LETD-aminoluciferin), caspase-9 tetrapeptide substrates (Z-LEHD-aminoluciferin), caspase-3 / 7 substrates (Z-DEVD-aminoluciferin), caspase-6 substrates (Z-VEID-aminoluciferin), or caspase-2 substrates (Z-VDVAD-aminoluciferin).

[0125] In certain embodiments, the phenotype is or comprises negative expression of activated caspase-1, activated caspase-2, activated caspase-3, activated caspase-7, activated caspase-8, activated caspase-9, activated caspase-10, and / or activated caspase-13 in the cell. -In some embodiments, a proform (zymogen-cleaved) form of a caspase, such as any of those described above, is also a marker indicating the presence of apoptosis. In some embodiments, the phenotype is or includes the absence or negative expression of a proform of a caspase, e.g., the proform of caspase-3.

[0126] In some embodiments, the marker of apoptosis is cleaved poly ADP-ribose polymerase 1 (PARP).PARP is cleaved by caspase during the early stage of apoptosis.Therefore, the detection of cleaved PARP peptide is the marker of apoptosis.In certain embodiments, the phenotype is or comprises the positive or negative expression of cleaved PARP.

[0127] In some embodiments, the apoptosis marker is a reagent that detects apoptosis-related characteristics in cells. In certain embodiments, the reagent is an annexin V molecule. During the early stages of apoptosis, the lipid phosphatidylserine (PS) translocates from the inner leaflet to the outer leaflet of the plasma membrane. PS is usually restricted to the inner membrane in healthy and / or non-apoptotic cells. Annexin V is a protein that preferentially binds to phosphatidylserine (PS) with high affinity. When conjugated to a fluorescent tag or other reporter, annexin V can be used to rapidly detect this early cell surface indicator of apoptosis. In some embodiments, the presence of PS on the outer membrane persists until the later stages of apoptosis. Thus, in some embodiments, annexin V staining is an indicator of both the early and late stages of apoptosis. In certain embodiments, annexin, such as annexin V, is tagged with a detectable label and incubated with, exposed to, and / or contacted with cells of the cell composition to detect cells undergoing apoptosis, for example, by flow cytometry. In some embodiments, a fluorescently tagged annexin, e.g., annexin V, is used to detect, e.g., annexin -V / 7 - Cells are stained for flow cytometry analysis using AAD assay.Another suitable protocol for detecting apoptosis by annexin includes techniques and assays that utilize radiolabeled annexin V.In some specific embodiments, phenotypes are detected by annexin, such as annexin V. - In certain embodiments, the phenotype is or comprises negative staining by annexin V. In certain embodiments, the phenotype is or comprises the absence of PS on the outer plasma membrane. In some specific embodiments, the phenotype is or comprises cells that are not bound by annexin, such as annexin V. In some specific embodiments, cells that lack detectable PS on the outer membrane are not bound by annexin V. - In certain embodiments, the assay, e.g., flow cytometry after incubation with labeled annexin V, detects annexin V. - Cells that are not bound by annexin V - is.

[0128] In certain embodiments, the phenotype is annexin V - , annexin V - CD3 + , annexin V - CD4 + , annexin V - CD8 + , annexin V - CD3 + , annexin V - CD4 + , annexin V - CD8 + , activated caspase 3 - , activated caspase 3 - / CD3 + , activated caspase 3 - / CD4 + , activated caspase 3 - / CD8 + , activated caspase 3 - / CD3 + , activated caspase 3 - / CD4 + , activated caspase 3 - / CD8 +, annexin V - / CD4 + / CCR7 + , annexin V - / CD8 + / CD27 + , Annexin V / CD4 + / CD27 + , annexin V - / CD8 + / CCR7 + / CD27 + , Annexin V / CD4 + / CCR7 + / CD27 + , annexin V - / CD8 + / CCR7 + / CD45RA - or Annexin V- / CD4 + / CCR7 + / CD45RA - ; Activated caspase 3 - / CD4 + / CCR7 + , activated caspase-3 / CD8 + / CD27 + , activated caspase 3 - / CD4 + / CD27 + , activated caspase 3 - / CD8 + / CCR7 + / CD27 + , activated caspase 3 - / CD4 + / CCR7 + / CD27 + , activated caspase 3 - / CD8 + / CCR7 + / CD45RA - or activated caspase 3 - / CD4 + / CCR7 + / CD45RA -or a combination thereof. In some embodiments, the phenotype is 3CAS- / CCR7- / CD27-. In some embodiments, the phenotype is 3CAS- / CCR7- / CD27+. In some embodiments, the phenotype is 3CAS- / CCR7+. In some embodiments, the phenotype is 3CAS- / CCR7+ / CD27-. In some embodiments, the phenotype is 3CAS- / CCR7+ / CD27+. In some embodiments, the phenotype is 3CAS- / CD27+. In some embodiments, the phenotype is 3CAS- / CD28- / CD27-. In some embodiments, the phenotype is 3CAS- / CD28- / CD27+. In some embodiments, the phenotype is 3CAS- / CD28+. In some embodiments, the phenotype is 3CAS- / CD28+ / CD27-, and in some embodiments, the phenotype is 3CAS- / CD28+ / CD27+. In some embodiments, the phenotype is 3CAS- / CCR7- / CD45RA-. In some embodiments, the phenotype is 3CAS- / CCR7- / CD45RA+. In some embodiments, the phenotype is 3CAS- / CCR7+ / CD45RA-. In some embodiments, the phenotype is 3CAS- / CCR7+ / CD45RA+. In some embodiments, the phenotype is further CD4+. In some embodiments, the phenotype is further CD8+.

[0129] In certain embodiments, it is contemplated that cells positive for expression of markers of apoptosis have undergone programmed cell death, exhibit reduced or no immune function, and have reduced, if any, ability to undergo activation, expansion, and / or bind antigen to initiate, carry out, or contribute to an immune response or activity. In certain embodiments, the phenotype is defined by negative expression of activated caspases and / or negative staining with Annexin V.

[0130] In certain embodiments, the phenotype is or comprises activated caspase 3 (caspase 3, 3CAS) and / or annexin V.

[0131] Among the phenotypes are the expression or surface expression of one or more markers generally associated with one or more subtypes or subpopulations of T cells, or their phenotypes. T cell subtypes and subpopulations include CD4 + and / or CD8 + These may include T cells and their subtypes, including naive T (T N ) cells, naive-like cells, effector T cells (T EFF ), memory T cells and their subtypes, such as stem cell memory T (T SCM ), Central Memory T(T CM ), Effector Memory T(T EM ), T EMRA These may include terminally differentiated effector memory T cells, tumor infiltrating lymphocytes (TILs), immature T cells, mature T cells, helper T cells, cytotoxic T cells, mucosal-associated invariant T (MAIT) cells, natural and adaptive regulatory T (Treg) cells, helper T cells, such as TH1 cells, TH2 cells, TH3 cells, TH17 cells, TH9 cells, TH22 cells, follicular helper T cells, alpha / beta T cells, and delta / gamma T cells.

[0132] In some embodiments, the input composition feature is the clonality of the cells of the input composition. In some embodiments, assessing the clonality of a population of T cells is assessing the clonal diversity of the population of T cells. In some embodiments, the T cells are polyclonal or multiclonal. The clonality, e.g., polyclonality, of the input composition of T cells is a measure of the breadth of the population's response to a given antigen. In some aspects, the input composition can be assessed by measuring the number of distinct epitopes recognized by antigen-specific cells. This can be done using standard techniques for generating and cloning antigen-specific T cells in vitro. In some embodiments, the T cells are polyclonal (or multiclonal), with no single clonotype population dominating the population of naive-like T cells.

[0133] In the context of a population of T cells, e.g., an input composition, in some aspects, a polyclonal signature indicates a population of T cells with multiple, broad antigen specificities. In some embodiments, polyclonality refers to a population of T cells that exhibits high diversity in their TCR repertoire. In some cases, the diversity of the TCR repertoire is due in some respects to V(D)J recombination events induced by self and foreign antigen selection events. In some embodiments, a diverse or polyclonal population of T cells is one in which analysis indicates the presence of multiple, various, or distinct TCR transcripts or products present in the population. In some embodiments, a population of T cells that exhibits high or relatively high clonality is one in which the diversity of the TCR repertoire is low. In some embodiments, T cells are oligoclonal when analysis indicates the presence of several, e.g., two or three, TCR transcripts or products in the population of T cells. In some embodiments, monoclonality indicates a population of T cells that is low in diversity. In some embodiments, T cells are monoclonal when analysis indicates the presence of a single TCR transcript or product in the population of T cells.

[0134] In some instances, the clonality of cells in input composition, such as T cells, is determined by clonal sequencing, for example, next-generation sequencing or spectratyping.In some aspects, next-generation sequencing method can be used with genomic DNA or cDNA from T cells to evaluate the sequence encoding TCR repertoire, for example, complementarity determining region 3 (CDR3).In some embodiments, whole transcriptome sequencing by RNA-seq can be used.In some embodiments, single-cell sequencing method can be used.

[0135] In some embodiments, clonality, e.g., polyclonality, can be assessed or determined by spectratyping (a measure of the hypervariable region repertoire of TCR Vβ, Vα, Vγ, or Vδ chains). Spectratyping distinguishes rearranged gene variable regions of a particular size from non-rearranged sequences. Therefore, it is understood that a single peak may represent a population of T cells expressing any one of a limited number of rearranged TCR gene variable regions (Vβ, Vα, Vγ, or Vδ) containing any one of the four possible nucleotides (adenine (a), guanine (g), cytosine (c), or thymine (t)) or a combination of these four nucleotides in the junction region. A population of T cells is considered polyclonal if the Vβ spectratyping profile of a given TCR Vβ, Vα, Vγ, or Vδ family has multiple peaks, typically five or more major peaks, most often with a Gaussian distribution. Polyclonality can also be defined by the generation and characterization of antigen-specific clones for an antigen of interest. In the context of a population of T cells, e.g., input composition, monoclonality refers to a population of T cells with a single specificity as defined by spectratyping (a measure of the hypervariable region repertoire of the TCR Vβ, Vα, Vγ, or Vδ chain). A population of T cells is considered monoclonal (or monospecific) if the Vβ, Vα, Vγ, and / or Vδ spectratype profile for a given TCR Vβ, Vα, Vγ, and / or Vδ family has a single dominant peak.

[0136] In some embodiments, the method for assessing clonality may include various features of the methods described in International Publication Nos. 2012 / 048341, 2014 / 144495, 2017 / 053902, 2016044227, 2016176322, and 2012048340, each of which is incorporated by reference in its entirety. In some embodiments, such methods may be used to obtain sequence information about a target polynucleotide of interest within a cell, such as a TCR. The target gene may be obtained from genomic DNA or mRNA of cells from a cell sample or cell population. The cell sample or cell population may include immune cells. For example, for a target TCR molecule, a gene encoding a TCR chain may be obtained from genomic DNA or mRNA of an immune cell or T cell. In some embodiments, the starting material is RNA from a T cell, which is composed of a gene encoding a TCR chain.

[0137] In some embodiments, the Shannon index is applied as a threshold to filter clones for clonality ("Shannon-adjusted clonality"), see Chaara et al. (2018) Front Immunol 9:1038). In some embodiments, the input composition feature is the clonality of CD4+ cells of the input composition. In some embodiments, the input composition feature is the clonality of CD8+ cells of the input composition.

[0138] In some aspects, the phenotype includes expression or markers or functions, e.g., antigen-specific functions such as cytokine secretion associated with a less differentiated cell subset or a more differentiated subset. In some embodiments, the phenotype is a phenotype associated with a less differentiated subset, e.g., CCR7 + , CD27 +and interleukin-2 (IL-2) production. In some aspects, less differentiated cells, e.g., central memory cells, live longer and become exhausted less rapidly, thereby enhancing persistence and endurance. In some embodiments, the phenotype is one or more of a phenotype associated with a well-differentiated subset, e.g., interferon-gamma (IFN-γ) or IL-13 production. In some aspects, well-differentiated subsets may also be associated with senescence and effector function.

[0139] In some embodiments, the phenotype is or comprises the phenotype of memory T cells or memory T cell subsets exposed to their cognate antigen. In some embodiments, the phenotype is or comprises the phenotype of memory T cells (or one or more markers associated therewith), e.g., T CM cell, T EM Cells, or T EMRA cell, T SCM In certain embodiments, the phenotype is or comprises the expression of one or more specific molecules that are markers of memory cells and / or memory T cells or subtypes thereof. In some aspects, the phenotype is or comprises the expression of one or more specific molecules that are markers of memory cells and / or memory T cells or subtypes thereof. CM Exemplary phenotypes associated with cells include CD45RA - , CD62L + , CCR7 + , CD27+, CD28+ and CD95 + In some aspects, one or more of T EM Exemplary phenotypes associated with cells include CD45RA - , CD62L - , CCR7 - , CD27-, CD28- and CD95- + may include one or more of:

[0140] In certain embodiments, the phenotype is or comprises the expression of one or more particular molecules that are markers of naive T cells.

[0141] In some embodiments, the phenotype is or comprises memory T cells or naive T cells. In some specific embodiments, the phenotype is the positive or negative expression of one or more specific molecules that are memory markers. In some embodiments, the memory marker is a specific molecule that can be used to define a memory T cell population.

[0142] In some embodiments, the phenotype is or includes one or more markers associated with non-memory T cells or subtypes thereof. In some aspects, the phenotype is or includes a phenotype or marker associated with naive cells. In some aspects, exemplary phenotypes associated with naive T cells may include one or more of CCR7+, CD45RA+, CD27+, and CD28+. In some embodiments, the phenotype is CCR7 + / CD27 + / CD28 + / CD45RA + In certain embodiments, the phenotype is CCR7 + / CD45RA + In certain embodiments, the phenotype is or comprises CCR7 + In certain embodiments, the phenotype is or comprises CD27+ / CD28+. In some embodiments, the phenotype is or comprises a central memory T cell phenotype. In certain embodiments, the phenotype is CCR7 + / CD27 + / CD28 + / CD45RA - In some embodiments, the phenotype is or comprises CCR7 - / CD27 + / CD28 + / CD45RA - In some embodiments, the phenotype is or comprises CCR7 + / CD27+ In some embodiments, the phenotype is or comprises CD27 + / CD28 + In certain embodiments, the phenotype is or comprises T EMRA Cells or T SCM In certain embodiments, the phenotype is or comprises a cellular phenotype. + In certain embodiments, the phenotype is or comprises CCR7 - / CD27 - / CD28 - / CD45RA + In some embodiments, the phenotype is or comprises CD27 + / CD28 + , CD27 - / CD28 + , CD27 + / CD28 - or CD27 - / CD28 - In some embodiments, the phenotype is or comprises one of: CCR7 + / CD27 + / CD45RA + In certain embodiments, the phenotype is CCR7 + / CD45RA + In certain embodiments, the phenotype is or comprises CD27- / CD28-. In certain embodiments, the phenotype is CCR7 + / CD27 + / CD45RA - In some embodiments, the phenotype is or comprises CCR7 - / CD27 + / CD45RA - In certain embodiments, the phenotype is or comprises CD45RA + In some embodiments, the phenotype is or comprises CCR7 - / CD27 - / CD45RA +In some embodiments, the phenotype is or comprises CCR7 + / CD27 + / CD28 + / CD45RA - ;CCR7 - / CD27 + / CD28 + / CD45RA - ;CCR7 - / CD27 - / CD28 - / CD45RA + ;CD27 + / CD28 + ;CD27 - / CD28 + ;CD27 + / CD28 - or CD27 - / CD28 - In certain embodiments, the phenotype is or comprises CCR7 + / CD27 + / CD45RA - ;CCR7 - / CD27 + / CD45RA - ;CCR7 - / CD27 - / CD28 - / CD45RA + ;CD27 + ;CD27 - ;CD27 + / CD28 - or CD27 - / CD28 - is or contains

[0143] In some embodiments, the phenotype is or includes one or more markers associated with naive-like T cells. In some embodiments, naive-like T cells may include cells of various differentiation states and may be characterized by positive or high expression (e.g., surface or intracellular expression) of certain cell markers, and / or negative or low expression (e.g., surface or intracellular expression) of other cell markers. In some aspects, naive-like T cells are characterized by positive or high expression of CCR7, CD45RA, CD28, and / or CD27. In some aspects, naive-like T cells are characterized by negative expression of CD25, CD45RO, CD56, CD62L, and / or KLRG1. In some aspects, naive-like T cells are characterized by low expression of CD95. In some specific embodiments, naive-like T cells or T cells that are surface-positive for the markers expressed on naive-like T cells are CCR7+CD45RA+, and the cells are CD27+ or CD27-. In some specific embodiments, naive-like T cells or T cells that are surface-positive for the markers expressed on naive-like T cells are CD27+ / CCR7+, and the cells are CD45RA+ or CD45RA-. In some specific embodiments, naive-like T cells or T cells that are surface-positive for the markers expressed on naive-like T cells are CD62L-CCR7+.

[0144] In certain embodiments, the phenotype is or comprises a T cell phenotype that is negative for a marker of apoptosis. In certain embodiments, the phenotype is or comprises a naive cell that is negative for a marker of apoptosis. In some embodiments, the marker of apoptosis is activated caspase 3 (3CAS). In some embodiments, the marker of apoptosis is positive staining with Annexin V. In certain embodiments, the phenotype is CD27 + / CD28 + , CD27 - / CD28 + , CD27 + / CD28 -, CD27 - / CD28 - or a combination thereof.

[0145] In certain embodiments, the phenotype is activated caspase 3 - / CD27 + / CD28 + , activated caspase 3 - / CD27 - / CD28 + , activated caspase 3 - / CD27 + / CD28 - , activated caspase 3 - / CD27 - / CD28 - or a combination thereof. In certain embodiments, the phenotype is or comprises Annexin V - / CD27 + / CD28 + , annexin V - / CD27 - / CD28 + , annexin V - / CD27 + / CD28 - , annexin V - / CD27 - / CD28 - or a combination thereof. In certain embodiments, the phenotype is or comprises CD27 + , CD27 - , CD27 + , CD27 - or a combination thereof. In some embodiments, the phenotype is or comprises CD27 + , CD27 - , CD27 + , CD27 - In certain embodiments, the phenotype is or comprises activated caspase 3 or a combination thereof. - / CD27 + , activated caspase 3 - / CD27 - , activated caspase 3 - / CD27 + , activated caspase 3 - / CD27 - or a combination thereof. In certain embodiments, the phenotype is or comprises Annexin V - / CD27 + , annexin V - / CD27 - , annexin V - / CD27 + , annexin V - / CD27 - or a combination thereof.

[0146] In certain embodiments, the phenotype is CCR7 + / CD28 + , CCR7 - / CD28 + , CCR7 + / CD28 - , CCR7 - / CD28 - or a combination thereof. In some embodiments, the phenotype is or comprises CCR7 + / CD28 + , CCR7 - / CD28 + , CCR7 + / CD28 - , CCR7 - / CD28 - In certain embodiments, the phenotype is or comprises activated caspase 3 or a combination thereof. - / CCR7 + / CD28 + , activated caspase 3 - / CCR7 - / CD28 + , activated caspase 3 - / CCR7 + / CD28 - , activated caspase 3 - / CCR7 - / CD28 - or a combination thereof. In certain embodiments, the phenotype is or comprises Annexin V- / CCR7 + / CD28 + , annexin V - / CCR7 - / CD28 + , annexin V - / CCR7 + / CD28 - , annexin V - / CCR7 - / CD28 - or a combination thereof. In certain embodiments, the phenotype is or comprises CCR7 + , CCR7 - , CCR7 + , CCR7 - or a combination thereof. In some embodiments, the phenotype is or comprises CCR7 + , CCR7 - , CCR7 + , CCR7 - In certain embodiments, the phenotype is or comprises activated caspase 3 or a combination thereof. - / CCR7 + , activated caspase 3 - / CCR7 - , activated caspase 3 - / CCR7 + , activated caspase 3 - / CCR7 - or a combination thereof. In certain embodiments, the phenotype is or comprises Annexin V - / CCR7 + , annexin V - / CCR7 - , annexin V - / CCR7 + , annexin V - / CCR7 - or a combination thereof.

[0147] In some embodiments, the input composition features include any one or more or all of the input composition features described herein, e.g., phenotypes. In some embodiments, input composition features comprise one or more of CAS3- / CCR7- / CD27-, CAS3- / CCR7- / CD27+, CAS3- / CCR7+, CAS3- / CCR7+ / CD27-, CAS3- / CCR7+ / CD27+, CAS3- / CD27+, CAS3- / CD28- / CD27-, CAS3- / CD28- / CD27+, CAS3- / CD28+, CAS3- / CD28+ / CD27-, CAS3- / CD28+ / CD27+, CAS3- / CCR7- / CD45RA-, CAS3- / CCR7-, CD45RA+, CAS3- / CCR7+ / CD45RA-, CAS3- / CCR7+ / CD45RA+, CAS+, CAS+ / CD3+, and clonality. In some embodiments, the input composition features are CAS3- / CCR7- / CD27- / CD4+, CAS3- / CCR7- / CD27+ / CD4+, CAS3- / CCR7+ / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CAS3- / CCR7+ / CD27+ / CD4+, CAS3- / CD27+ / CD4+, CAS3- / CD28- / CD27- / CD4+, CAS3- / CD28- / CD27+ / CD4+, CAS3- / CD28+ / and one or more of the following clonalities: CD4+, CAS3- / CD28+ / CD27- / CD4+, CAS3- / CD28+ / CD27+ / CD4+, CAS3- / CCR7- / CD45RA- / CD4+, CAS3- / CCR7- / CD4+, CD45RA+ / CD4+, CAS3- / CCR7+ / CD45RA- / CD4+, CAS3- / CCR7+ / CD45RA+ / CD4+, CAS+ / CD4+, CAS+ / CD3+ / CD4+, and CD4+.In some embodiments, the input composition features are CAS3- / CCR7- / CD27- / CD8+, CAS3- / CCR7- / CD27+ / CD8+, CAS3- / CCR7+ / CD8+, CAS3- / CCR7+ / CD27- / CD8+, CAS3- / CCR7+ / CD27+ / CD8+, CAS3- / CD27+ / CD8+, CAS3- / CD28- / CD27- / CD8+, CAS3- / CD28+ / and one or more of the following clonalities: CD8+, CAS3- / CD28+ / CD27- / CD8+, CAS3- / CD28+ / CD27+ / CD8+, CAS3- / CCR7- / CD85RA- / CD8+, CAS3- / CCR7- / CD8+, CD85RA+ / CD8+, CAS3- / CCR7+ / CD85RA- / CD8+, CAS3- / CCR7+ / CD85RA+ / CD8+, CAS+ / CD8+, CAS+ / CD3+ / CD8+, and CD8+.In some embodiments, the input composition features are CAS3- / CCR7- / CD27- / CD4+, CAS3- / CCR7- / CD27+ / CD4+, CAS3- / CCR7+ / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CAS3- / CCR7+ / CD27+ / CD4+, CAS3- / CD27+ / CD4+, CAS3- / CD28- / CD27- / CD4+, CAS3- / CD28- / CD27+ / CD4+, CAS S3- / CD28+ / CD4+, CAS3- / CD28+ / CD27- / CD4+, CAS3- / CD28+ / CD27+ / CD4+, CAS3- / CCR7- / CD45RA- / CD4+, CAS3- / CCR7- / C D4+, CD45RA+ / CD4+, CAS3- / CCR7+ / CD45RA- / CD4+, CAS3- / CCR7+ / CD45RA+ / CD4+, CAS+ / CD4+, CAS+ / CD3+ / CD4+, CD4+ CAS3- / CCR7- / CD27- / CD8+, CAS3- / CCR7- / CD27+ / CD8+, CAS3- / CCR7+ / CD8+, CAS3- / CCR7+ / CD27- / CD8+, CAS3- / CCR 7+ / CD27+ / CD8+, CAS3- / CD27+ / CD8+, CAS3- / CD28- / CD27- / CD8+, CAS3- / CD28- / CD27+ / CD8+, CAS3- / CD28+ / CD8+, CAS3- / CD28+ / CD27- / CD8+, CAS3- / CD28+ / CD27+ / CD8+, CAS3- / CCR7- / CD85RA- / CD8+, CAS3- / CCR7- / CD8+, CD85RA+ / CD8+, CAS3- / CCR7+ / CD85RA- / CD8+, CAS3- / CCR7+ / CD85RA+ / CD8+, CAS+ / CD8+, CAS+ / CD3+ / CD8+ and CD8+ clonality.

[0148] In some embodiments, the input composition feature comprises any one or more of the input composition features set forth in Table E4 below. In some of any of the above embodiments, the percentage, number, and / or proportion of cells having the phenotype as described above is determined, measured, obtained, detected, observed, and / or identified. In some embodiments, the number of cells of the phenotype as described above is the total number of cells of the phenotype in the input composition. In some embodiments, the number of cells of the phenotype as described above can be expressed as a frequency, ratio, and / or percentage of cells of the phenotype present in the input composition. In some embodiments, the input composition feature is a frequency, ratio, and / or percentage of cells having a phenotype as described herein.

[0149] 3. Therapeutic Cell Composition Features In some embodiments, the therapeutic cell composition is generated (e.g., as described herein) from an input composition, e.g., as described above. In some embodiments, the therapeutic cell composition is a therapeutic T cell composition. In some embodiments, the therapeutic cell composition comprises engineered CD4+ T cells. In some embodiments, the therapeutic cell composition comprises engineered CD8+ T cells. In some embodiments, the therapeutic cell composition comprises engineered CD4+ and CD8+ T cells. In some embodiments, the engineered T cells of the therapeutic cell composition, e.g., CD4+ and / or CD8+ engineered T cells, express a recombinant receptor, e.g., a recombinant T cell receptor (TCR) or chimeric antigen receptor (CAR). In some embodiments, the recombinant receptor, e.g., a TCR or CAR, binds to an antigen associated with a disease or pathology. For example, the antigen associated with a disease or pathology can be an antigen expressed on cells or tissues of the disease or pathology. In some embodiments, the recombinant receptor specifically binds to an antigen associated with a disease or pathology, or expressed in cells in the environment of a lesion associated with a disease or pathology. In some embodiments, the antigen is associated with and / or involved in the pathogenesis of a disease, condition, or disorder, e.g., causes, exacerbates, or otherwise participates in such disease, condition, or disorder. Exemplary diseases and conditions can include diseases or conditions associated with cellular malignancy or transformation (e.g., cancer), autoimmune or inflammatory diseases, or infectious diseases caused, for example, by bacterial, viral, or other pathogens.

[0150] In some embodiments, the therapeutic cell composition, e.g., the therapeutic T cell composition, is for treating a disease or condition.

[0151] In some embodiments, the therapeutic cell composition characteristic comprises a cell phenotype. In some embodiments, the phenotype is the number of total T cells. In some embodiments, the phenotype is the number of total CD3 +The number of T cells. In certain embodiments, the phenotype includes cells expressing a recombinant receptor or CAR. In some embodiments, the recombinant receptor or CAR binds to an antigen associated with a disease or pathology. In some embodiments, the phenotype includes one or more different subtypes of T cells. In some embodiments, one or more different subtypes further express a recombinant receptor or CAR. In some embodiments, the phenotype is or includes the identity of a T cell subtype. Different populations or subtypes of T cells include effector T cells, helper T cells, memory T cells, effector memory T cells, regulatory T cells, naive T cells, naive-like T cells, CD4 + cells and CD8 + This includes, but is not limited to, T cell.In some specific embodiments, T cell subtypes can be identified by detecting the presence or absence of specific molecules.In some specific embodiments, specific molecules are surface markers that can be used to identify T cell subtypes.

[0152] In some embodiments, the phenotype is positive or high expression of one or more specific molecules that are surface markers, e.g., CD3, CD4, CD8, CD28, CD62L, CCR7, CD27, CD127, CD4, CD8, CD45RA and / or CD45RO. In certain embodiments, the phenotype is, for example, positive or high expression of one or more surface markers, e.g., CD3 + , CD4 + , CD8 + , CD28 + , CD62L + , CCR7 + , CD27 + , CD127 + , CD4 + , CD8 + , CD45RA + and / or CD45RO +The phenotype is a surface marker of T cells, or a surface marker of a subpopulation or subset of T cells, based on positive surface marker expression of one or more specific molecules that are surface markers, such as CC chemokine receptor type 7 (CCR7), cluster of differentiation 27 (CD27), cluster of differentiation 28 (CD28), and cluster of differentiation 45 RA (CD45RA). In certain embodiments, phenotypic markers include CCR7, CD27, CD28, CD44, CD45RA, CD62L, and L-selectin. In some embodiments, the phenotype is negative expression or absence of expression of one or more specific molecules that are surface markers, such as CD3, CD4, CD8, CD28, CD62L, CCR7, CD27, CD127, CD45RA, and / or CD45RO. In certain embodiments, the phenotype is, for example, negative expression or absence of expression of one or more surface markers, such as CD3 - , CD4 - , CD8-, CD28 - , CD62L - , CCR7 - , CD27 - , CD127 - , CD4 - , CD8 - , CD45RA - and / or CD45RO - The phenotype is a surface marker of T cells, or a surface marker of a subpopulation or subset of T cells, based on the absence of surface marker expression. In some embodiments, the phenotype is the negative expression or absence of expression of one or more specific molecules that are surface markers, such as CC chemokine receptor type 7 (CCR7), cluster of differentiation 27 (CD27), cluster of differentiation 28 (CD28), and cluster of differentiation 45 RA (CD45RA). In some specific embodiments, the phenotype markers include CCR7, CD27, CD28, CD44, CD45RA, CD62L, and L-selectin.

[0153] In certain embodiments, the phenotype is or includes positive or negative expression of CD27, CCR7 and / or CD45RA. + In some embodiments, the phenotype is CD27 + In some embodiments, the phenotype is CCR7 - In some embodiments, the phenotype is CD27 - In some embodiments, the phenotype is CCR7 + / CD27 + In some embodiments, the phenotype is CCR7 - / CD27 + In some embodiments, the phenotype is CCR7 + In some embodiments, the phenotype is CCR7 - / CD27 - In some embodiments, the phenotype is CD45RA - In some embodiments, the phenotype is CD45RA + In some embodiments, the phenotype is CCR7 + / CD45RA - In some embodiments, the phenotype is CD27 + / CD45RA + In some embodiments, the phenotype is CD27 - / CD45RA + In some embodiments, the phenotype is CD27 - / CD45RA - In some embodiments, the phenotype is CD27 + / CD45RA - In some embodiments, the phenotype is CCR7 + / CD27 + / CD45RA - In some embodiments, the phenotype is CCR7 + / CD27 + / CD45RA+.

[0154] In certain embodiments, the surface marker indicates the expression of a recombinant receptor, such as a CAR. In certain embodiments, the surface marker is the expression of a recombinant receptor, such as a CAR, which in some aspects can be determined using an antibody such as an anti-idiotype antibody. In some embodiments, the surface marker indicating the expression of a recombinant receptor is a surrogate marker. In certain embodiments, such a surrogate marker is a surface protein that has been modified to have little or no activity. In certain embodiments, the surrogate marker is encoded on the same polynucleotide that encodes the recombinant receptor. In some embodiments, the nucleic acid sequence encoding the recombinant receptor is operably linked to a nucleic acid sequence encoding a marker, or a nucleic acid encoding a self-cleaving peptide or a peptide that causes ribosome skipping, such as a 2A sequence, for example, T2A (e.g., SEQ ID NOs:1 and 4), P2A (e.g., SEQ ID NOs:5 and 6), E2A (e.g., SEQ ID NO:7) or F2A (e.g., SEQ ID NO:8), optionally separated by an internal ribosome entry site (IRES). Exogenous marker genes may be utilized in conjunction with engineered cells to allow for cell detection or selection in some circumstances, and in some circumstances also to promote cell suicide.

[0155] Exemplary surrogate markers may include truncated cell surface polypeptides, such as truncated human epidermal growth factor receptor 2 (tHER2), truncated epidermal growth factor receptor (EGFRt, exemplary EGFRt sequences shown in SEQ ID NO: 2 or 3), or prostate-specific membrane antigen (PSMA) or modified forms thereof. EGFRt may contain an epitope recognized by the antibody cetuximab (Erbitux®) or other therapeutic anti-EGFR antibodies or binding molecules, which can be used to identify or select cells engineered with EGFRt constructs and recombinant receptors, such as chimeric antigen receptors (CARs), and / or to eliminate or separate cells expressing the receptor. See U.S. Patent No. 8,802,374 and Liu et al., Nature Biotech. 2016 April;34(4):430-434). In some aspects, markers, e.g., surrogate markers, include all or a portion (e.g., truncated) of CD34, NGFR, or epidermal growth factor receptor (e.g., tEGFR). In some embodiments, the nucleic acid encoding the marker is operably linked to a polynucleotide encoding a cleavable linker sequence, such as T2A. For example, the marker, and optionally the linker sequence, can be any of those disclosed in International Publication No. 2014031687. For example, the marker can be a truncated EGFR (tEGFR, EGFRt) optionally linked to a linker sequence, such as a T2A cleavable linker sequence. Exemplary polypeptides of truncated EGFR (e.g., tEGFR, EGFRt) include the amino acid sequence set forth in SEQ ID NO: 2 or 3, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more sequence identity to SEQ ID NO: 2 or 3. In some embodiments, the phenotype is EGFRt+.

[0156] In some embodiments, the marker is or comprises a fluorescent protein, such as green fluorescent protein (GFP), enhanced green fluorescent protein (EGFP), e.g., superfold GFP, red fluorescent protein (RFP), e.g., tdTomato, mCherry, mStrawberry, AsRed2, DsRed or DsRed2, cyan fluorescent protein (CFP), blue-green fluorescent protein (BFP), enhanced blue fluorescent protein (EBFP) and yellow fluorescent protein (YFP), as well as species variants of fluorescent proteins, monomeric variants, and variants thereof, including codon-optimized and / or highly sensitive variants. In some embodiments, the marker is or comprises an enzyme, such as luciferase, the lacZ gene from Escherichia coli (E. coli), alkaline phosphatase, secreted embryonic alkaline phosphatase (SEAP), or chloramphenicol acetyltransferase (CAT). Exemplary luminescent reporter genes include luciferase (luc), β-galactosidase, chloramphenicol acetyltransferase (CAT), β-glucuronidase (GUS) or variants thereof.

[0157] In certain embodiments, the phenotype comprises expression, e.g., surface expression, of one or more of the surface markers CD3, CD4, CD8, and / or a recombinant receptor (e.g., a CAR), or a surrogate marker thereof that indicates or correlates with expression of the recombinant receptor (e.g., a CAR). In some embodiments, the surrogate marker is EGFRt.

[0158] In certain embodiments, the phenotype is identified by the expression of one or more specific molecules that are surface markers.In some specific embodiments, the phenotype is or includes the positive or negative expression of CD3, CD4, CD8 and / or recombinant receptors, such as CAR.In some specific embodiments, the recombinant receptor is CAR.In certain embodiments, the phenotype is CD3 + / CAR + , CD4 + / CAR +and / or CD8 + / CAR + Includes.

[0159] In certain embodiments, the phenotype is or comprises positive or negative expression of CD27, CCR7 and / or CD45RA, and / or a recombinant receptor, such as a CAR. + / CAR + In some embodiments, the phenotype is CD27 + / CAR + In some embodiments, the phenotype is CCR7 + / CD27 + / CAR + In some embodiments, the phenotype is CD45RA - / CAR + In some embodiments, the phenotype is CCR7 + / CD45RA - / CAR + In some embodiments, the phenotype is CD27 + / CD45RA - / CAR + In some embodiments, the phenotype is CCR7 + / CD27 + / CD45RA - / CAR +In some embodiments, the phenotype is CCR7- / CD27- / CAR+. In some embodiments, the phenotype is CCR7- / CD27+ / CAR+. In some embodiments, the phenotype is CCR7+ / CD27- / CAR+. In some embodiments, the phenotype is CD28- / CD27- / CAR+. In some embodiments, the phenotype is CD28- / CD27+ / CAR+. In some embodiments, the phenotype is CD28+ / CAR+. In some embodiments, the phenotype is CD28+ / CD27- / CAR+. In some embodiments, the phenotype is CD28+ / CD27+ / CAR+. In some embodiments, the phenotype is CCR7- / CD45RA- / CAR+. In some embodiments, the phenotype is CCR7- / CD45RA+ / CAR+. In some embodiments, the phenotype is CCR7+ / CD45RA- / CAR+. In some embodiments, the phenotype is CCR7+ / CD45RA+ / CAR+. In some embodiments, the phenotype is further CD4+, In some embodiments, the phenotype is further CD8+.

[0160] In some embodiments, the phenotype is viability.In some specific embodiments, the phenotype is the positive expression of markers that indicate that cells are undergoing normal functional cellular processes and / or are not undergoing necrosis or programmed cell death, or are not undergoing necrosis or programmed cell death.In some embodiments, viability can be assessed by cellular redox potential, cell membrane integrity, or mitochondrial activity or function.In some embodiments, viability is the absence of specific molecules associated with cell death, or the absence of signs of cell death in assays.In some embodiments, the phenotype is viable cell concentration.

[0161] In some embodiments, the phenotype is or includes cell viability. In certain embodiments, cell viability can be detected, measured, and / or assessed by several means routine in the art. Non-limiting examples of such viability assays include, but are not limited to, dye uptake assays (e.g., calcein AM assay), XTT cell viability assays, and dye exclusion assays (e.g., trypan blue dye exclusion assay, eosin dye exclusion assay, or propidium dye exclusion assay). Viability assays are useful for determining the number or percentage (e.g., frequency) of viable cells in a cell dose, cell composition, and / or cell sample. In certain embodiments, the phenotype includes cell viability along with other characteristics, such as recombinant receptor expression. In some embodiments, the phenotype is or includes soluble CD137 (sCD137, 4-IBB). In some embodiments, sCD137 indicates activation-induced cell death. In some embodiments, sCD137 is detected in the supernatant.

[0162] In some particular embodiments, the phenotype is cell viability, viable CD3 + , survival CD4 + , survival CD8 + , survival CD3 + / CAR + , survival CD4 + / CAR + , survival CD8 + / CAR + , survival CD4 + / CCR7 + / CAR + , survival CD8 + / CD27 + / CAR + , survival CD4 + / CD27 + / CAR + , survival CD8 + / CCR7 + / CD27 + / CAR + , survival CD4 + / CCR7 + / CD27 + / CAR+ , survival CD8 + / CCR7 + / CD45RA - / CAR + or viable CD4 + / CCR7 + / CD45RA - or a combination thereof.

[0163] In certain embodiments, phenotype is or comprises the absence of apoptosis and / or the sign that cell is undergoing apoptosis process.Apoptosis is a programmed cell death process that includes a series of stereotypical morphological and biochemical events that lead to characteristic cell changes and cell death.These changes include blebbing, cell shrinkage, nuclear fragmentation, chromatin condensation, chromosomal DNA fragmentation and global mRNA degradation.Apoptosis is a well-characterized process, and the specific molecules associated with various stages are well known in the art.

[0164] In some embodiments, the phenotype is the absence of the early stages of apoptosis and / or the absence of indicators and / or specific molecules associated with the early stages of apoptosis. In the early stages of apoptosis, changes in the cell membrane and mitochondrial membrane are evident. Biochemical changes are also evident in the cell cytoplasm and nucleus. For example, the early stages of apoptosis can be indicated by the activation of specific caspases, such as 2, 8, 9, and 10. In certain embodiments, the phenotype is the absence of the late stages of apoptosis and / or the absence of indicators and / or specific molecules associated with the late stages of apoptosis. The middle to late stages of apoptosis are characterized by further loss of membrane integrity, chromatin condensation, DNA fragmentation, and biochemical events such as the activation of caspases 3, 6, and 7.

[0165] In certain embodiments, the phenotype is negative expression of one or more factors associated with apoptosis, including pro-apoptotic factors known to initiate apoptosis, such as members of the death receptor pathway, activated members of the mitochondrial (intrinsic) pathway, such as Bcl-2 family members, e.g., Bax, Bad, and Bid, and caspases. In some embodiments, the phenotype is negative or low levels of apoptotic markers. In certain embodiments, the phenotype is negative expression of apoptotic markers. In certain embodiments, the phenotype is the absence of an indicator, such as Annexin V molecules, that preferentially binds to cells undergoing apoptosis when incubated with or contacted with a cell composition. In some embodiments, the phenotype is or includes the expression of one or more markers indicative of the apoptotic state in cells.

[0166] In some embodiments, phenotype is the negative (or low) expression of certain molecules that are the markers of apoptosis.Various apoptosis markers are known to those skilled in the art, and include but are not limited to the activity of one or more caspases, that is, the increase of activated caspases (active caspases), the increase of PARP cleavage, the activation and / or translocation of Bcl-2 family proteins, cell death pathway members, for example, Fas and FADD, the existence of nuclear shrinkage (for example, by microscopic monitoring), and the existence of chromosomal DNA fragmentation (for example, the existence of chromosomal DNA ladder), or by apoptosis assays, including TUNEL staining and Annexin V staining.

[0167] Caspases are enzymes that cleave proteins after aspartic acid residues, and the term derives from "cysteine-aspartic acid protease." Because caspases are involved in apoptosis, activation of caspases, such as caspase-3, indicates increased or resurrected apoptosis. In certain embodiments, caspase activation can be detected by methods known to those skilled in the art. In some embodiments, caspase activation can be detected using an antibody that specifically binds to activated caspases (i.e., specifically binds to the cleaved polypeptide). In another example, a fluorochrome inhibitor of caspase activity (FLICA) assay can be used to detect caspase-3 activation by detecting the hydrolysis of acetylAsp-Glu-Val-Asp 7-amido-4-methylcoumarin (Ac-DEVD-AMC) by caspase-3 (i.e., detecting the release of fluorescent 7-amino-4-methylcoumarin (AMC)). FLICA assays can be used to determine caspase activation by detecting the products of substrates processed by multiple caspases (e.g., FAM-VAD-FMK FLICA). Other techniques include the CASPASE-GLO® Caspase Assay (PROMEGA), which uses luminogenic caspase-8 tetrapeptide substrates (Z-LETD-aminoluciferin), caspase-9 tetrapeptide substrates (Z-LEHD-aminoluciferin), caspase-3 / 7 substrates (Z-DEVD-aminoluciferin), caspase-6 substrates (Z-VEID-aminoluciferin), or caspase-2 substrates (Z-VDVAD-aminoluciferin).

[0168] In certain embodiments, the phenotype is or comprises negative expression of activated caspase-1, activated caspase-2, activated caspase-3, activated caspase-7, activated caspase-8, activated caspase-9, activated caspase-10, and / or activated caspase-13 in the cell. -In some embodiments, a proform (zymogen-cleaved) form of a caspase, such as any of those described above, is also a marker indicating the presence of apoptosis. In some embodiments, the phenotype is or includes the absence or negative expression of a proform of a caspase, e.g., the proform of caspase-3.

[0169] In some embodiments, the marker of apoptosis is cleaved poly ADP-ribose polymerase 1 (PARP).PARP is cleaved by caspase during the early stage of apoptosis.Therefore, the detection of cleaved PARP peptide is the marker of apoptosis.In certain embodiments, the phenotype is or comprises the positive or negative expression of cleaved PARP.

[0170] In some embodiments, the apoptosis marker is a reagent that detects apoptosis-related characteristics in cells. In certain embodiments, the reagent is an annexin V molecule. During the early stages of apoptosis, the lipid phosphatidylserine (PS) translocates from the inner leaflet to the outer leaflet of the plasma membrane. PS is usually restricted to the inner membrane in healthy and / or non-apoptotic cells. Annexin V is a protein that preferentially binds to phosphatidylserine (PS) with high affinity. When conjugated to a fluorescent tag or other reporter, annexin V can be used to rapidly detect this early cell surface indicator of apoptosis. In some embodiments, the presence of PS on the outer membrane persists until the later stages of apoptosis. Thus, in some embodiments, annexin V staining is an indicator of both the early and late stages of apoptosis. In certain embodiments, annexin, such as annexin V, is tagged with a detectable label and incubated with, exposed to, and / or contacted with cells of the cell composition to detect cells undergoing apoptosis, for example, by flow cytometry. In some embodiments, a fluorescently tagged annexin, e.g., annexin V, is used to detect, e.g., annexin -V / 7 - Cells are stained for flow cytometry analysis using AAD assay.Another suitable protocol for detecting apoptosis by annexin includes techniques and assays that utilize radiolabeled annexin V.In some specific embodiments, phenotypes are detected by annexin, such as annexin V. - In certain embodiments, the phenotype is or comprises negative staining with annexin V. In certain embodiments, the phenotype is or comprises the absence of PS on the outer plasma membrane. In some specific embodiments, the phenotype is or comprises cells that are not bound by annexin, such as annexin V. In some specific embodiments, cells that lack detectable PS on the outer membrane are not bound by annexin V. - In certain embodiments, the assay, e.g., flow cytometry after incubation with labeled annexin V, detects annexin V. - Cells that are not bound by annexin V - is.

[0171] In certain embodiments, the phenotype is annexin V - , annexin V - CD3 + , annexin V - CD4 + , annexin V - CD8 + , annexin V - CD3 + / CAR + , annexin V - CD4 + / CAR + , annexin V - CD8 + / CAR + , activated caspase 3 - , activated caspase 3 - / CD3 + , activated caspase 3 - / CD4 + , activated caspase 3 - / CD8 + , activated caspase 3 - / CD3 + / CAR + , activated caspase 3- / CD4 + / CAR + , activated caspase 3 - / CD8 + / CAR + , annexin V - / CD4 + / CCR7 + / CAR + , annexin V - / CD8 + / CD27 + / CAR + , Annexin V / CD4 + / CD27 + / CAR + , annexin V - / CD8 + / CCR7 + / CD27 + / CAR + , Annexin V / CD4 + / CCR7 + / CD27 + / CAR + , annexin V - / CD8 + / CCR7 + / CD45RA - / CAR + or Annexin V- / CD4 + / CCR7 + / CD45RA - ; Activated caspase 3 - / CD4 + / CCR7 + / CAR + , activated caspase-3 / CD8 + / CD27 + / CAR + , activated caspase 3 - / CD4 + / CD27 + / CAR + , activated caspase 3 - / CD8 + / CCR7 + / CD27 + / CAR + , activated caspase 3 - / CD4 + / CCR7 + / CD27+ / CAR + , activated caspase 3 - / CD8 + / CCR7 + / CD45RA - / CAR + or activated caspase 3 - / CD4 + / CCR7 + / CD45RA - or a combination thereof. In some embodiments, the phenotype is 3CAS- / CCR7- / CD27- / CAR+. In some embodiments, the phenotype is 3CAS- / CCR7- / CD27+ / CAR+. In some embodiments, the phenotype is 3CAS- / CCR7+ / CAR+. In some embodiments, the phenotype is 3CAS- / CCR7+ / CD27- / CAR+. In some embodiments, the phenotype is 3CAS- / CCR7+ / CD27+ / CAR+. In some embodiments, the phenotype is 3CAS- / CD27+ / CAR+. In some embodiments, the phenotype is 3CAS- / CD28- / CD27- / CAR+. In some embodiments, the phenotype is 3CAS- / CD28- / CD27+ / CAR+. In some embodiments, the phenotype is 3CAS- / CD28+ / CAR+. In some embodiments, the phenotype is 3CAS- / CD28+ / CD27- / CAR+, and in some embodiments, the phenotype is 3CAS- / CD28+ / CD27+ / CAR+. In some embodiments, the phenotype is 3CAS- / CCR7- / CD45RA- / CAR+. In some embodiments, the phenotype is 3CAS- / CCR7- / CD45RA+ / CAR+. In some embodiments, the phenotype is 3CAS- / CCR7+ / CD45RA- / CAR+. In some embodiments, the phenotype is 3CAS- / CCR7+ / CD45RA+ / CAR+. In some embodiments, the phenotype is further CD4+. In some embodiments, the phenotype is further CD8+.

[0172] In certain embodiments, it is contemplated that cells positive for expression of markers of apoptosis have undergone programmed cell death, exhibit reduced or no immune function, and have reduced, if any, ability to undergo activation, expansion, and / or bind antigen to initiate, carry out, or contribute to an immune response or activity. In certain embodiments, the phenotype is defined by negative expression of activated caspases and / or negative staining with Annexin V.

[0173] In certain embodiments, the phenotype is activated caspase 3 - (3CAS-, caspase 3 - ) and / or Annexin V - is or contains

[0174] Among the phenotypes are the expression or surface expression of one or more markers commonly associated with one or more subtypes or subpopulations of T cells, or their phenotypes. T cell subtypes and subpopulations include CD4 + and / or CD8 + These may include T cells and their subtypes, including naive T (T N ) cells, naive-like T cells, effector T cells (T EFF ), memory T cells and their subtypes, such as stem cell memory T (T SCM ), Central Memory T(T CM ), Effector Memory T(T EM ), T EMRA These may include terminally differentiated effector memory T cells, tumor infiltrating lymphocytes (TILs), immature T cells, mature T cells, helper T cells, cytotoxic T cells, mucosal-associated invariant T (MAIT) cells, natural and adaptive regulatory T (Treg) cells, helper T cells, such as TH1 cells, TH2 cells, TH3 cells, TH17 cells, TH9 cells, TH22 cells, follicular helper T cells, alpha / beta T cells, and delta / gamma T cells.

[0175] In some aspects, the phenotype includes expression or markers or functions, e.g., antigen-specific functions such as cytokine secretion associated with a less differentiated cell subset or a more differentiated subset. In some embodiments, the phenotype is a phenotype associated with a less differentiated subset, e.g., CCR7 + , CD27 + and interleukin-2 (IL-2) production. In some aspects, the less differentiated subset may also be associated with therapeutic efficacy, self-renewal, survival function, or graft-versus-host disease. In some embodiments, the phenotype is one or more of a phenotype associated with a well-differentiated subset, such as interferon-gamma (IFN-γ) or IL-13 production. In some aspects, the well-differentiated subset may also be associated with senescence and effector function.

[0176] In some embodiments, the phenotype is or comprises the phenotype of memory T cells or memory T cell subsets exposed to their cognate antigen. In some embodiments, the phenotype is or comprises the phenotype of memory T cells (or one or more markers associated therewith), e.g., T CM cell, T EM Cells, or T EMRA cell, T SCM In certain embodiments, the phenotype is or comprises the expression of one or more specific molecules that are markers of memory cells and / or memory T cells or subtypes thereof. In some aspects, the phenotype is or comprises the expression of one or more specific molecules that are markers of memory cells and / or memory T cells or subtypes thereof. CM Exemplary phenotypes associated with cells include CD45RA - , CD62L + , CCR7 + , CD27+, CD28+ and CD95 + In some aspects, one or more of T EM Exemplary phenotypes associated with cells include CD45RA - , CD62L - , CCR7 -, CD27-, CD28- and CD95- + may include one or more of:

[0177] In certain embodiments, the phenotype is or comprises the expression of one or more particular molecules that are markers of naive T cells.

[0178] In some embodiments, the phenotype is or comprises memory T cells or naive T cells. In some specific embodiments, the phenotype is the positive or negative expression of one or more specific molecules that are memory markers. In some embodiments, the memory marker is a specific molecule that can be used to define a memory T cell population.

[0179] In some embodiments, the phenotype is or includes one or more markers associated with naive-like T cells. In some specific embodiments, naive-like T cells may include cells in various differentiation states and may be characterized by positive or high expression (e.g., surface or intracellular expression) of certain cell markers, and / or negative or low expression (e.g., surface or intracellular expression) of other cell markers. In some aspects, naive-like T cells are characterized by positive or high expression of CCR7, CD45RA, CD28, and / or CD27. In some aspects, naive-like T cells are characterized by negative expression of CD25, CD45RO, CD56, CD62L, and / or KLRG1. In some aspects, naive-like T cells are characterized by low expression of CD95. In some specific embodiments, naive-like T cells or T cells that are surface-positive for the markers expressed on naive-like T cells are CCR7+CD45RA+, and the cells are CD27+ or CD27-. In some specific embodiments, naive-like T cells or T cells that are surface-positive for the markers expressed on naive-like T cells are CD27+CCR7+, and the cells are CD45RA+ or CD45RA-. In some specific embodiments, naive-like T cells or T cells that are surface-positive for the markers expressed on naive-like T cells are CD62L- / CCR7+.

[0180] In some embodiments, the phenotype is or includes one or more markers associated with non-memory T cells or subtypes thereof. In some aspects, the phenotype is or includes a phenotype or marker associated with naive cells. In some aspects, exemplary phenotypes associated with naive T cells may include one or more of CCR7+, CD45RA+, CD27+, and CD28+. In some embodiments, the phenotype is CCR7 + / CD27 + / CD28 + / CD45RA + In certain embodiments, the phenotype is CCR7+ / CD45RA + In certain embodiments, the phenotype is or comprises CCR7 + In certain embodiments, the phenotype is or comprises CD27+ / CD28+. In some embodiments, the phenotype is or comprises a central memory T cell phenotype. In certain embodiments, the phenotype is CCR7 + / CD27 + / CD28 + / CD45RA - In some embodiments, the phenotype is or comprises CCR7 - / CD27 + / CD28 + / CD45RA - In some embodiments, the phenotype is or comprises CCR7 + / CD27 + In some embodiments, the phenotype is or comprises CD27 + / CD28 + In certain embodiments, the phenotype is or comprises T EMRA Cells or T SCM In certain embodiments, the phenotype is or comprises a cellular phenotype. + In certain embodiments, the phenotype is or comprises CCR7 - / CD27 - / CD28 - / CD45RA + In some embodiments, the phenotype is or comprises CD27 + / CD28 + , CD27 - / CD28 + , CD27 + / CD28 - or CD27 - / CD28 - In some embodiments, the phenotype is or comprises one of: CCR7 + / CD27 + / CD45RA+ In certain embodiments, the phenotype is CCR7 + / CD45RA + In certain embodiments, the phenotype is or comprises CD27- / CD28-. In certain embodiments, the phenotype is CCR7 + / CD27 + / CD45RA - In some embodiments, the phenotype is or comprises CCR7 - / CD27 + / CD45RA - In certain embodiments, the phenotype is or comprises CD45RA + In certain embodiments, the phenotype is or comprises CCR7 - / CD27 - / CD45RA + is or contains

[0181] In some embodiments, the phenotype is or comprises any of the aforementioned phenotypic characteristics and further comprises expression of a recombinant receptor, e.g., a phenotype associated with memory T cells or memory subtypes, expressing a CAR, or a phenotype associated with naive cells expressing a CAR. In some specific embodiments, the phenotype is or comprises a phenotype of a central memory T cell or stem central memory T cell expressing a CAR. In certain embodiments, the phenotype is or comprises a phenotype of an effector memory cell expressing a CAR. In some embodiments, the phenotype is or comprises a phenotype of a T cell expressing a CAR. EMRA In certain embodiments, the phenotype is or comprises a phenotype of a cell. + / CCR7 + / CD27 + / CD28 + / CD45RA - ;CAR + / CCR7 - / CD27 + / CD28 + / CD45RA - ;CAR + / CCR7 - / CD27 - / CD28 - / CD45RA + ;CAR + / CD27 + / CD28 + ;CAR + / CD27 - / CD28 + ;CAR + / CD27 + / CD28 - ;or CAR + / CD27 - / CD28 - In certain embodiments, the phenotype is or comprises a CAR + / CCR7 + / CD27 + / CD45RA - ;CAR + / CCR7 - / CD27 + / CD45RA - ;CAR + / CCR7 - / CD27 - / CD28 - / CD45RA + ;CAR + / CD27 + ;CAR + / CD27 - ;CAR + / CD27 + / CD28 - ;or CAR + / CD27 - / CD28 - is or contains

[0182] In some specific embodiments, the phenotype is or comprises a T cell phenotype that is negative for apoptosis markers. In some specific embodiments, the phenotype is or comprises a naive cell that is negative for apoptosis markers. In some embodiments, the apoptosis marker is activated caspase 3 (3CAS). In some embodiments, the apoptosis marker is positive staining with Annexin V.

[0183] In certain embodiments, the phenotype is or comprises a phenotype of a memory T cell or subtype thereof that expresses a CAR and is negative for a marker of apoptosis. In certain embodiments, the phenotype is or comprises a phenotype of a memory T cell or a particular subtype that expresses a CAR and is negative for a marker of apoptosis. In some specific embodiments, the phenotype is or comprises a naive cell that expresses a CAR and is negative for a marker of apoptosis. In some specific embodiments, the phenotype is a central memory T cell or T that expresses a CAR and is negative for a marker of apoptosis. SCM In certain embodiments, the phenotype is or comprises the phenotype of effector memory cells that express CAR and are negative for markers of apoptosis. In some specific embodiments, the phenotype is or comprises the phenotype of effector memory cells that express CAR and are negative for markers of apoptosis. - / CAR + / CCR7 + / CD27 + / CD28 + / CD45RA - ;Annexin V - / CAR + / CCR7 - / CD27 + / CD28 + / CD45RA - ;Annexin V - / CAR + / CCR7 - / CD27 - / CD28 - / CD45RA + ;Annexin V - / CAR + / CD27 + / CD28 + ;Annexin V - / CAR + / CD27 - / CD28 + ;Annexin V - / CAR + / CD27 + / CD28 - or Annexin V - / CAR + / CD27 - / CD28 - In certain embodiments, the phenotype is or comprises activated caspase 3. - / CAR + / CCR7 + / CD27 + / CD28 + / CD45RA - ; Activated caspase 3 - / CAR + / CCR7 - / CD27 + / CD28 + / CD45RA - ; Activated caspase 3 - / CAR + / CCR7 - / CD27 - / CD28 - / CD45RA + ; Activated caspase 3 - / CAR + / CD27 + / CD28 + ; Activated caspase 3 - / CAR + / CD27 - / CD28 + ; Activated caspase 3 - / CAR + / CD27 + / CD28 - or activated caspase 3 - / CAR + / CD27 - / CD28 - In certain embodiments, the phenotype is or comprises Annexin V. - / CAR + / CCR7 + / CD27 + / CD45RA - ;Annexin V - / CAR + / CCR7 - / CD27 + / CD45RA - ;Annexin V- / CAR + / CCR7 - / CD27 - / CD45RA + ;Annexin V - / CAR + / CD27 + / CD28 + ;Annexin V - / CAR + / CD27 - / CD28 + ;Annexin V - / CAR + / CD27 + or Annexin V - / CAR + / CD27 - In certain embodiments, the phenotype is or comprises activated caspase 3. - / CAR + / CCR7 + / CD27 + / CD45RA - ; Activated caspase 3 - / CAR + / CCR7 - / CD27 + / CD45RA - ; Activated caspase 3 - / CAR + / CCR7 - / CD27 - / CD45RA + ; Activated caspase 3 - / CAR + / CD27 + / CD28 + ; Activated caspase 3 - / CAR + / CD27 - / CD28 + ; Activated caspase 3 - / CAR + / CD27 + or activated caspase 3 - / CAR + / CD27 - is or contains

[0184] In certain embodiments, the phenotype is CD27 + / CD28 + , CD27 - / CD28 + , CD27 + / CD28 - , CD27 - / CD28 - In some embodiments, the phenotype is or comprises a CAR + / CD27 + / CD28 + , CAR + / CD27 - / CD28 + , CAR + / CD27 + / CD28 - , CAR + / CD27 - / CD28 - In certain embodiments, the phenotype is or comprises activated caspase 3 or a combination thereof. - / CAR + / CD27 + / CD28 + , activated caspase 3 - / CAR + / CD27 - / CD28 + , activated caspase 3 - / CAR + / CD27 + / CD28 - , activated caspase 3 - / CAR + / CD27 - / CD28 - or a combination thereof. In certain embodiments, the phenotype is or comprises Annexin V - / CAR + / CD27 + / CD28 + , annexin V - / CAR + / CD27 - / CD28 + , annexin V - / CAR + / CD27 + / CD28 - , annexin V - / CAR + / CD27 - / CD28 - or a combination thereof. In certain embodiments, the phenotype is or comprises CD27 + , CD27 - , CD27 + , CD27 - In some embodiments, the phenotype is or comprises a CAR + / CD27 + , CAR + / CD27 - , CAR + / CD27 + , CAR + / CD27 - In certain embodiments, the phenotype is or comprises activated caspase 3 or a combination thereof. - / CAR + / CD27 + , activated caspase 3 - / CAR + / CD27 - , activated caspase 3 - / CAR + / CD27 + , activated caspase 3 - / CAR + / CD27 - or a combination thereof. In certain embodiments, the phenotype is or comprises Annexin V - / CAR + / CD27 + , annexin V - / CAR + / CD27 - , annexin V - / CAR + / CD27 + , annexin V - / CAR + / CD27 - or a combination thereof.

[0185] In certain embodiments, the phenotype is CCR7 + / CD28 + , CCR7 - / CD28 + , CCR7 + / CD28 - , CCR7 - / CD28 - In some embodiments, the phenotype is or comprises a CAR + / CCR7 + / CD28 + , CAR + / CCR7 - / CD28 + , CAR + / CCR7 + / CD28 - , CAR + / CCR7 - / CD28 - In certain embodiments, the phenotype is or comprises activated caspase 3 or a combination thereof. - / CAR + / CCR7 + / CD28 + , activated caspase 3 - / CAR + / CCR7 - / CD28 + , activated caspase 3 - / CAR + / CCR7 + / CD28 - , activated caspase 3 - / CAR + / CCR7 - / CD28 - or a combination thereof. In certain embodiments, the phenotype is or comprises Annexin V - / CAR + / CCR7 + / CD28 + , annexin V - / CAR + / CCR7 - / CD28 + , annexin V - / CAR + / CCR7 + / CD28 - , annexin V - / CAR + / CCR7 - / CD28 - or a combination thereof. In certain embodiments, the phenotype is or comprises CCR7 + , CCR7 - , CCR7 + , CCR7 - In some embodiments, the phenotype is or comprises a CAR + / CCR7 + , CAR + / CCR7 - , CAR + / CCR7 + , CAR + / CCR7 - In certain embodiments, the phenotype is or comprises activated caspase 3 or a combination thereof. - / CAR + / CCR7 + , activated caspase 3 - / CAR + / CCR7 - , activated caspase 3 - / CAR + / CCR7 + , activated caspase 3 - / CAR + / CCR7 - or a combination thereof. In certain embodiments, the phenotype is or comprises Annexin V - / CAR + / CCR7 + , annexin V - / CAR + / CCR7 - , annexin V - / CAR + / CCR7 + , annexin V - / CAR + / CCR7 - or a combination thereof.

[0186] In some embodiments, the phenotype is assessed by response to a stimulus, e.g., a stimulus that triggers, induces, stimulates, or prolongs immune cell function. In certain embodiments, the cells are incubated under stimulatory conditions or in the presence of a stimulatory agent, and the phenotype is or includes a response to the stimulus. In certain embodiments, the phenotype is or includes the production or secretion of a soluble factor in response to one or more stimuli. In some embodiments, the phenotype is or includes the absence, production, or secretion of a soluble factor in response to one or more stimuli. In certain embodiments, the soluble factor is a cytokine. In some embodiments, the cytokine is IL-2. In some embodiments, the cytokine is TNFα. In some embodiments, the cytokine is IL-17. In some embodiments, the cytokine is IL-10. In some embodiments, the cytokine is IFN-γ. In some embodiments, the cytokine is IL-13. In some embodiments, the cytokine is IL-5. In some embodiments, the cytokine is GMCSF. In some embodiments, the cells do not produce cytokines (cyto-). In some embodiments, the cell phenotype is cytokine-negative (Cyto-).

[0187] The conditions used to stimulate cells may include one or more of the following: specific medium, temperature, oxygen content, carbon dioxide content, time, agents such as nutrients, amino acids, antibiotics, ions and / or stimulatory factors, such as cytokines, chemokines, antigens, binding partners, fusion proteins, recombinant soluble receptors, and any other agents designed to activate cells. In some embodiments, cells are stimulated, and the phenotype is determined by whether soluble factors, such as cytokines or chemokines, are produced or secreted. In some embodiments, the stimulation is non-specific, i.e., not antigen-specific. In some embodiments, the stimulation includes PMA and ionomycin. In some embodiments, the cells are incubated in the presence of stimulatory conditions or stimulatory agents for about 1 hour, about 2 hours, about 3 hours, about 4 hours, about 5 hours, about 6 hours, about 7 hours, about 8 hours, about 9 hours, about 10 hours, about 11 hours, about 12 hours, about 18 hours, about 24 hours, about 48 hours, or for a period of 1 hour to 4 hours, 1 hour to 12 hours, 12 hours to 24 hours, or for more than 24 hours.

[0188] In some embodiments, the therapeutic cell composition characteristic includes a recombinant receptor-dependent activity. For example, in some embodiments, cells of the therapeutic cell composition are stimulated with an agent that is an antigen or epitope specific to the recombinant receptor, or an antibody or fragment thereof that binds to and / or recognizes the recombinant receptor, or a combination thereof. In certain embodiments, the recombinant receptor-dependent activity, e.g., CAR-dependent activity, is intended to be an activity that occurs in cells expressing the recombinant receptor that does not and / or cannot occur in cells that do not express the recombinant receptor. In some embodiments, the recombinant receptor-dependent activity is an activity that depends on the activity or presence of the recombinant receptor. The recombinant receptor-dependent activity can be any cellular process that is directly or indirectly affected by the expression and / or presence of the recombinant receptor, or by changes in the activity of the recombinant receptor, such as receptor stimulation. In some embodiments, the recombinant receptor-dependent activity can include, but is not limited to, cellular processes such as cell division, DNA replication, transcription, protein synthesis, membrane trafficking, protein translocation and / or secretion, or the recombinant receptor-dependent activity can be an immune cell function, e.g., cytolytic activity. In certain embodiments, recombinant receptor-dependent activity can be measured by changes in CAR receptor confirmation, phosphorylation of intracellular signaling molecules, protein degradation, protein transcription, translation, translocation, and / or production and secretion of factors such as proteins, or growth factors, cytokines.

[0189] In some embodiments, the recombinant receptor is a CAR, and the agent is an antigen or epitope specific to the CAR, or an antibody or fragment thereof that binds to and / or recognizes the CAR, or a combination thereof. In certain embodiments, the cells are stimulated by incubating the cells in the presence of target cells that have the surface expression of the antigen recognized by the CAR. In certain embodiments, the recombinant receptor is a CAR, and the agent is an antibody or an active fragment, variant, or portion thereof that binds to the CAR. In certain embodiments, the antibody or an active fragment, variant, or portion thereof that binds to the CAR is an anti-idiotype (anti-ID) antibody. In certain embodiments, the recombinant receptor-specific agent is a cell that expresses an antigen on its surface, e.g., a target cell. In some embodiments, the recombinant receptor-dependent activity is stimulated by an antigen or an epitope thereof that is bound and / or recognized by the recombinant receptor (e.g., associated with the recombinant receptor).

[0190] In some embodiments, the stimulatory conditions or agents include one or more agents, e.g., ligands, that can stimulate or activate the intracellular signaling domain of the TCR complex. In some aspects, the agents turn on or initiate the TCR / CD3 intracellular signaling cascade in the T cell. Such agents may include antibodies, such as those specific for TCR components and / or costimulatory receptors, e.g., anti-CD3, anti-CD28, and / or one or more cytokines, e.g., bound to a solid support such as beads. In some embodiments, the one or more agents are PMA and ionomycin.

[0191] In certain embodiments, recombinant receptor-dependent activity, for example, CAR-dependent activity, is the measurement of a factor, for example, amount or concentration, or change in amount or concentration, after stimulation of cell composition.In certain embodiments, the factor can be a protein, a phosphorylated protein, a cleaved protein, a translocated protein, a protein in active confirmation, a polynucleotide, an RNA polynucleotide, an mRNA and / or an shRNA.In certain embodiments, the measurement can include, but is not limited to, kinase activity, protease activity, phosphatase activity, cAMP production, ATP metabolism, translocation, for example, increase or decrease in the nuclear localization of protein, increase in transcription activity, increase in translation activity, production and / or secretion of soluble factors, cellular uptake, ubiquitination and / or increase in protein degradation.In certain embodiments, the factor is a secreted soluble factor, for example, hormone, growth factor, chemokine and / or cytokine.

[0192] In some embodiments, recombinant receptor-dependent activity, for example, CAR-dependent activity, is a response to stimulation.In some specific embodiments, cells are incubated under stimulating conditions or in the presence of stimulating agents, and activity is or comprises at least one aspect of the response to stimulation.Responses can include, but are not limited to, intracellular signaling events, such as increased activity of receptor molecules, increased kinase activity of one or more kinases, increased transcription of one or more genes, increased protein synthesis of one or more proteins and / or increased intracellular signaling molecules, for example, increased kinase activity of proteins.In some embodiments, response or activity is related to immune activity, and can include, but are not limited to, the production and / or moiety of soluble factors, such as cytokines, increased antibody production, and / or increased cytolytic activity.

[0193] In certain embodiments, the response of cell composition to stimulation is evaluated by measuring, detecting or quantifying the response to stimulation, i.e., at least one activity that is initiated, triggered, supported, extended, and / or caused by stimulation.In some specific embodiments, cells are stimulated, and the response to stimulation is the activity that is specific to the cells that express recombinant receptor.In some specific embodiments, the activity is recombinant receptor-specific activity, and the activity occurs in the cells that express recombinant receptor, but does not occur or occurs only minimally in the cells that do not express the receptor.In certain embodiments, the recombinant receptor is CAR.In some embodiments, the activity is CAR-dependent activity.

[0194] The conditions used to stimulate cells, such as immune cells or T cells, can include one or more of the following: specific medium, temperature, oxygen content, carbon dioxide content, time, agents, such as nutrients, amino acids, antibiotics, ions and / or stimulatory factors, such as cytokines, chemokines, antigens, binding partners, fusion proteins, recombinant soluble receptors, and any other agents designed to activate cells. In some embodiments, cells are stimulated, and activity is determined by whether soluble factors, such as cytokines or chemokines, are produced or secreted.

[0195] In some embodiments, the activity is specific to cells expressing the recombinant receptor. In some embodiments, the activity specific to cells expressing the recombinant receptor does not occur in cells lacking expression of the recombinant receptor. In some specific embodiments, the recombinant receptor is a CAR, and the activity is a CAR-dependent activity. In certain embodiments, the activity is not present in cells lacking expression of the recombinant receptor under the same conditions as the activity is present in cells expressing the recombinant receptor. In some specific embodiments, the CAR-dependent activity is about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, about 97%, about 98%, about 99% or about 99% lower than the CAR-dependent activity in CAR cells under the same conditions.

[0196] In some embodiments, the activity is specific to cells expressing a recombinant receptor, such as a CAR, and the activity is brought about by stimulation with an agent specific to the cells expressing the recombinant receptor or under stimulatory conditions. In some embodiments, the recombinant receptor is a CAR, and CAR-specific stimulation stimulates, induces, initiates, and / or prolongs activity in CAR+ cells, but does not stimulate, induce, initiate, and / or prolong activity in CAR- cells. In some embodiments, CAR-dependent activity is about 10%, about 20%, about 30%, about 40%, about 50%, about 60%, about 70%, about 75%, about 80%, about 85%, about 90%, about 95%, about 97%, about 98%, about 99%, or about 99% lower in CAR- cells than in CAR+ cells after stimulation with a CAR-specific stimulation.

[0197] In some embodiments, activity is measured in a cell composition containing cells expressing a recombinant receptor, such as a CAR, and the measurement is compared to one or more controls. In certain embodiments, the control is a similar or identical composition of cells that has not been stimulated. For example, in some embodiments, activity is measured in a cell composition after or during incubation with an agent, and the resulting measurement is compared to a control measurement of activity from a similar or identical cell composition that has not been incubated with the agent. In some embodiments, the activity is a recombinant receptor-dependent activity, and both the cell composition and the control cell composition contain cells expressing a recombinant receptor. In some embodiments, the activity is a recombinant receptor-dependent activity, and the control is obtained from a similar cell composition that does not contain cells expressing a recombinant receptor, such as CAR+ cells. Thus, in some embodiments, a cell composition containing recombinant receptor-expressing cells and a control cell composition that does not contain recombinant receptor-expressing cells are contacted with a recombinant receptor-expressing specific agent. In certain embodiments, the control is a measurement from the same cell composition expressing a recombinant receptor obtained before any stimulation. In certain embodiments, a control measurement is obtained to determine background signal, and the control measurement is subtracted from the activity measurement. In some embodiments, the measured value of activity in a cell composition is divided by the control measured value to obtain a value that is a ratio of activity to the control level.

[0198] In certain embodiments, the activity is or comprises the production and / or secretion of a soluble factor. In some embodiments, the activity is a recombinant receptor (e.g., CAR)-dependent activity that is or comprises the production and / or secretion of a soluble factor. In some specific embodiments, the soluble factor is a cytokine or chemokine.

[0199] In certain embodiments, the measurement of soluble factor is measured by ELISA (enzyme-linked immunosorbent assay).ELISA is a plate-based assay technology designed to detect and quantify substances such as peptides, cytokines, antibodies and hormones.In ELISA, soluble factor must be immobilized on a solid surface and then complexed with an antibody linked to an enzyme.Detection is achieved by evaluating the activity of the conjugated enzyme through incubation with a substrate to generate a detectable signal.In some embodiments, CAR-dependent activity is measured by ELISA assay.

[0200] In some embodiments, recombinant receptor-dependent activity is the secretion or production of soluble factors.In some specific embodiments, the production or secretion is stimulated by recombinant receptor-specific agents, such as CAR+ specific agents, in a cell composition containing recombinant receptor-expressing cells, for example, CAR-expressing cells.In some embodiments, the recombinant receptor-specific agent is an antigen specific to the recombinant receptor or its epitope; a cell expressing the antigen, for example, a target cell; or an antibody or its part or variant that binds to and / or recognizes the recombinant receptor; or a combination thereof.In some specific embodiments, the recombinant receptor-specific agent is a recombinant protein that comprises the antigen or its epitope that is bound or recognized by the recombinant receptor.

[0201] In some specific embodiments, the production and / or secretion of recombinant receptor-dependent soluble factors is measured by incubating a cell composition containing recombinant receptor-expressing cells, such as CAR, with recombinant receptor-specific acting substances, such as CAR+ specific acting substances.In some specific embodiments, the soluble factors are cytokines or chemokines.In some embodiments, the cells of the cell composition containing recombinant receptor-expressing cells are incubated in the presence of recombinant receptor-specific acting substances for a certain amount of time, and the production and / or secretion of soluble factors is measured at one or more time points during incubation. In some embodiments, the cells are incubated with a CAR-specific agent for up to 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 7 hours, 8 hours, 9 hours, 10 hours, 11 hours, 12 hours, 18 hours, 19 hours, 20 hours, 21 hours, 22 hours, 23 hours, 24 hours, 48 ​​hours, or about 1 hour, about 2 hours, about 3 hours, about 4 hours, about 5 hours, about 6 hours, about 7 hours, about 8 hours, about 9 hours, about 10 hours, about 11 hours, about 12 hours, about 18 hours, about 19 hours, about 20 hours, about 21 hours, about 22 hours, about 23 hours, about 24 hours, about 48 hours, or for a period of 1 hour to 4 hours, 1 hour to 12 hours, 12 hours to 24 hours inclusive, or for more than 24 hours, and the amount of a soluble factor, e.g., a cytokine, is detected.

[0202] In some embodiments, the recombinant receptor-specific agent is a target cell that expresses an antigen recognized by the recombinant receptor. In some embodiments, the recombinant receptor is a CAR, and the cells of the cell composition are incubated with the target cells at a ratio of total cells, CAR+ cells, CAR+ / CD8+ cells, or Annexin- / CAR+ / CD8+ cells to target cells of the cell composition of about 10:1, about 5:1, about 4:1, about 3:1, about 2:1, about 1:1, about 1:2, about 1:3, about 1:4, about 1:5, about 1:6, about 1:7, about 1:8, about 1:9, or about 1:10, or a range between any of the foregoing, for example, 10:1 to 1:1, 3:1 to 1:3, or 1:1 to 1:10 (inclusive).

[0203] In some specific embodiments, the measured value of recombinant receptor-dependent activity, for example, CAR+ specific activity, is the amount or concentration, or relative amount or concentration, of soluble factors in the T cell composition during incubation or at the end of incubation.In certain embodiments, the measured value is subtracted by a control measurement value or normalized to the control measurement value.In some embodiments, the control measurement value is the measurement value obtained from the same cell composition before incubation.In certain embodiments, the control measurement value is the measurement value obtained from the same control cell composition that is not incubated with recombinant receptor-specific stimulating agent.In some specific embodiments, the control measurement value is the measurement value obtained from the cell composition that does not contain recombinant receptor-positive cells, at the same time point during incubation with recombinant receptor-specific stimulating agent.

[0204] In some embodiments, the measured value is the normalized ratio of the amount or concentration compared to the control.In certain embodiments, the measured value is the amount or concentration of soluble factor per time amount, for example, per minute or per hour.In some embodiments, the measured value is per cell, or per set or reference number of cells, for example, per 100 cells, per 10 cells. 3 Per 10 cells 4 Per 10 cells 5 Per 10 cells 6In some embodiments, the measured value is the amount or concentration of soluble factor per recombinant receptor-expressing cell, CAR+ cell, CAR+ / CD8+ cell, Annexin- / CAR+ / CD8+ cell, 3CAS- / CAR+ / CD8+ cell, CAR+ / CD4+ cell, Annexin- / CAR+ / CD4+ cell or 3CAS- / CAR+ / CD4+ cell of the cell composition. In some specific embodiments, the measurement is the amount or concentration of soluble factor per recombinant receptor-expressing cell, CAR+ cell, CAR+ / CD8+ cell, annexin- / CAR+ / CD8+ cell, 3CAS- / CAR+ / CD8+ cell, CAR+ / CD4+ cell, annexin- / CAR+ / CD4+ cell or 3CAS- / CAR+ / CD4+ cell of the cell composition per time period (e.g., per minute or per hour). In some embodiments, the measurement is the amount or concentration of soluble factor per recombinant receptor- or CAR+-specific agent per time period. In some embodiments, the measurement is the amount or concentration of soluble factor per cell, or per set or reference number of cells, or per amount or concentration of CAR+-specific agent. In certain cases, the measurement is the amount or concentration of soluble factor per time period, per recombinant receptor- or CAR+-specific agent per amount or concentration, per cell, or per reference number of cells. In some embodiments, the measure is the amount or concentration of soluble factor per recombinant receptor-expressing cell, CAR+ cell, CAR+ / CD8+ cell, annexin- / CAR+ / CD8+ cell, 3CAS- / CAR+ / CD8+ cell, CAR+ / CD4+ cell, annexin- / CAR+ / CD4+ cell, or 3CAS- / CAR+ / CD4+ cell of the cell composition per amount or concentration of recombinant receptor- or CAR+-specific agent.In some particular embodiments, the measurement is the amount or concentration of soluble factor per amount of time, per amount or concentration of recombinant receptor or CAR+ specific agent, per amount of CAR+ cells, CAR+ / CD8+ cells, annexin- / CAR+ / CD8+ cells, 3CAS- / CAR+ / CD8+ cells, CAR+ / CD4+ cells, annexin- / CAR+ / CD4+ cells, or 3CAS- / CAR+ / CD4+ cells of the cell composition.

[0205] In certain embodiments, recombinant receptor or CAR-dependent activity is the production or secretion of two or more soluble factors. In some specific embodiments, recombinant receptor or CAR-dependent activity is the production or secretion of 2, 3, 4, 5, 6, 7, 8, 9, 10, or more than 10 soluble factors. In some embodiments, measurements of 2, 3, 4, 5, 6, 7, 8, 9, 10, or more than 10 soluble factors are combined into an arithmetic or geometric mean. In some measurements, measurements of recombinant receptor activity are the secretion of a complex of 2, 3, 4, 5, 6, 7, 8, 9, 10, or more than 10 soluble factors.

[0206] In certain embodiments, the soluble factor is a cytokine. In certain embodiments, the recombinant receptor-dependent activity is or includes the production or secretion of cytokines in response to one or more stimuli. Cytokines are a large group of small signaling molecules that function widely in cell communication. Cytokines are most often related to various immunoregulatory molecules, including interleukins, chemokines, and interferons. Alternatively, cytokines can be characterized by their structure, which is classified into four families: the IL-2 subfamily and the IFN subfamily; and four alpha-helical families, including cysteine ​​knot cytokines, which include members of the IL-1 family, IL-17 family, IL-10 family, and transforming growth factor beta family. Cytokine production and / or secretion contribute to the immune response and are involved in various processes, including the induction of antiviral proteins and the induction of T cell proliferation. Cytokines are not preformed factors, but are rapidly produced and secreted in response to cell activation. Cytokine production or secretion can be measured, detected, and / or quantified by any suitable technique known in the art. In some embodiments, the recombinant receptor-dependent activity is the production or secretion of one or more soluble factors, including interleukins, interferons, and chemokines. In particular embodiments, the recombinant receptor-dependent activity is the production or secretion of one or more of an IL-2 family member, an IFN subfamily member, an IL-1 family member, an IL-10 member, an IL-17 family member, a cysteine ​​knot cytokine, and / or a member of the transforming growth factor beta family.

[0207] In some specific embodiments, the phenotype is the production of one or more cytokines. In some embodiments, the production of two or more cytokines from the same cell may indicate the multifunctional characteristics of such cells. In certain embodiments, the production of one or more cytokines is measured, detected, and / or quantified by intracellular cytokine staining. Intracellular cytokine staining (ICS) by flow cytometry is a highly suitable technique for studying cytokine production at the single-cell level. It detects the production and accumulation of cytokines in the endoplasmic reticulum after cell stimulation, allowing the identification of cell populations that are positive or negative for the production of a specific cytokine, or the separation of high-producing cells and low-producing cells based on a threshold value. In some embodiments, as described above, stimulation can be performed using non-specific stimulation, for example, not antigen-specific stimulation. For example, PMA / ionomycin can be used for non-specific cell stimulation. In some embodiments, stimulation can be performed by an agent that is an antigen or epitope specific to a recombinant receptor (e.g., CAR), or an antibody or fragment thereof that binds to and / or recognizes the recombinant receptor, or a combination thereof. ICS can also be used in combination with other flow cytometry protocols for immunophenotyping using cell surface markers or MHC multimers to access cytokine production in specific subpopulations of cells, making it an extremely flexible and versatile method. Other single-cell techniques for measuring or detecting cytokine production include, but are not limited to, ELISPOT, limiting dilution, and T cell cloning.

[0208] In some embodiments, the phenotype is, for example, the production of cytokines after stimulating the recombinant receptor with an antigen specific to and / or recognized by the recombinant receptor. In certain embodiments, the phenotype is, for example, the lack of cytokine production after stimulating the recombinant receptor with an antigen specific to and / or recognized by the recombinant receptor. In certain embodiments, the phenotype is positive for cytokine production or high-level cytokine production. In some specific embodiments, the phenotype is negative for cytokine production or low-level cytokine production. Cytokines may include, but are not limited to, interleukin-1 (IL-1), IL-1β, IL-2, sIL-2Ra, IL-3, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12, IL-13, IL-17, IL-27, IL-33, IL-35, TNF, tumor necrosis factor alpha (TNFA), CXCL2, CCL2, CCL3, CCL5, CCL17, CCL24, PGD2, LTB4, interferon gamma (IFNG), granulocyte-macrophage colony-stimulating factor (GMCSF), macrophage inflammatory proteins MIP1α, MIP1β, Flt-3L, fractalkine, and / or IL-5. In some embodiments, the phenotype comprises the production of cytokines, e.g., cytokines associated with a particular cell type, e.g., cytokines associated with Th1, Th2, Th17, and / or Treg subtypes. In some embodiments, exemplary Th1-associated cytokines include IL-2, IFN-γ, and transforming growth factor beta (TGF-β), and in some circumstances, exemplary Th1-associated cytokines are involved in cellular immune responses. In some embodiments, exemplary Th2-associated cytokines include IL-4, IL-5, IL-6, IL-10, and IL-13, and in some circumstances, exemplary Th2-associated cytokines are associated with humoral immunity and anti-inflammatory properties.In some embodiments, exemplary Th17-associated cytokines include IL-17A and IL-17F, and in some circumstances, exemplary Th17-associated cytokines are involved in recruiting neutrophils and macrophages, for example, during inflammatory responses.

[0209] In certain embodiments, the recombinant receptor-dependent activity is the production and / or secretion of one or more of IL-1, IL-1β, IL-2, sIL-2Ra, IL-3, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12, IL-13, IL-27, IL-33, IL-35, TNF, TNFalpha, CXCL2, CCL2, CCL3, CCL5, CCL17, CCL24, PGD2, LTB4, interferon gamma (IFN-γ), granulocyte-macrophage colony-stimulating factor (GM-CSF), macrophage inflammatory protein (MIP)-1a, MIP-1b, Flt-3L, fractalkine, and / or IL-5. In some specific embodiments, the recombinant receptor-dependent activity is the production or secretion of Th17 cytokines. In some embodiments, the Th17 cytokine is GMCSF. In some embodiments, the recombinant receptor-dependent activity comprises the production or secretion of a Th2 cytokine, wherein the Th2 cytokine is IL-4, IL-5, IL-10, or IL-13.

[0210] In certain embodiments, the recombinant receptor-dependent activity is the production or secretion of proinflammatory cytokines, which play a role in initiating the inflammatory response, regulating host defense against pathogens, and mediating the innate immune response. Inflammatory cytokines include, but are not limited to, interleukin (IL), interleukin-1-beta (IL-1), interleukin-3 (IL-3), interleukin-5 (IL-5), interleukin-6 (IL-6), interleukin-13 (IL-13), tumor necrosis factor (TNF), C-X-C-chemokine ligand 2 (CXCL2), CC-chemokine ligand 2 (CCL2), CC-chemokine ligand 3 (CCL3), CC-chemokine ligand 5 (CCL5), CC-chemokine ligand 17 (CCL17), CC-chemokine ligand 24 (CCL24), prostaglandin D2 (PGD2), and leukotriene B4 (LTB4), and IL-33. In some embodiments, the CAR-dependent activity is the production and / or secretion of interleukin and / or TNF family members. In certain embodiments, the CAR-dependent activity is the production and / or secretion of IL-1, IL-6, IL-8 and IL-18, TNF-alpha, or combinations thereof.

[0211] In certain embodiments, the recombinant receptor-dependent activity is the secretion of IL-2, IFN-gamma, TNF-alpha, or a combination thereof.

[0212] In some embodiments, the phenotype (e.g., recombinant receptor-dependent activity) is or includes cytokine production. In certain embodiments, the phenotype is or includes production of multiple cytokines (e.g., multifunctional). In certain embodiments, the recombinant receptor-dependent activity is or includes lack of production of one or more cytokines. In certain embodiments, the phenotype is or includes production or lack of one or more of IL-2, IL-5, IL-10, IL-13, IL-17, IFNG, or TNFA. In certain embodiments, the recombinant receptor-dependent activity is or includes production or lack of one or more of IL-2, IL-13, IFNG, or TNFA. In some embodiments, the recombinant receptor-dependent activity is the presence of cytokine production and / or the presence of high levels of cytokine production. In some embodiments, the phenotype is low, reduced, or absent cytokine production.

[0213] In some embodiments, the phenotype is or includes the internal (intracellular) production of cytokines, for example, when assessed in the presence of a stimulatory agent or under stimulatory conditions in which secretion is prevented or inhibited. In some embodiments, the stimulatory agent is a non-specific stimulatory agent, for example, a stimulatory agent that does not bind to an antigen-binding domain, such as a recombinant receptor (e.g., CAR). In some embodiments, the stimulatory agent is PMA / ionomycin, which can act as a non-specific stimulatory agent. In some embodiments, the stimulatory agent is a specific stimulatory agent, for example, a stimulatory agent that is an antigen or epitope specific to a recombinant receptor (e.g., CAR), or a stimulatory agent that is an antibody or fragment thereof that binds to and / or recognizes the recombinant receptor, or a combination thereof. In certain embodiments, the phenotype is or includes the lack or absence of internal production of cytokines. In some specific embodiments, the phenotype is or includes the internal amount of one or more cytokines in the case of production of multiple cytokines assessed using an ICS assay. In certain embodiments, the phenotype is or comprises the internal amount of one or more of IL-2, IL-5, IL-13, IFNG or TNFA assessed using an ICS assay. In some embodiments, the phenotype is or comprises the low internal amount or lack of detectable amount of one or more cytokines assessed using an ICS assay. In certain embodiments, the phenotype is or comprises the low internal amount or lack of detectable amount of IL-2, IL-5, IL-13, IFNG or TNFA assessed using an ICS assay. In some embodiments, the phenotype comprises the evaluation of multiple cytokines, for example, by a multiplexed assay or an assay to evaluate multifunctionality (see, for example, Xue et al., (2017) Journal for ImmunoTherapy of Cancer 5:85).In some embodiments, the lack of cytokine expression is inversely correlated with or associated with cellular activity and / or function and / or response and progression-free survival duration. In some embodiments, cells with reduced, minimal, or no cytokine production, as assessed according to any known method or method described herein, are reduced in the cell composition (e.g., output composition, therapeutic cell composition).

[0214] In certain embodiments, it is contemplated that the phenotype may include cytokine production, or the absence or low production of cytokines. This may depend on several factors, including, but not limited to, the identity of the cytokine, the assay performed to detect the cytokine, and the stimulatory agent or conditions used in conjunction with the assay. For example, in some embodiments, the phenotype is or includes the absence or low level of IL-13 production as indicated by ICS, while in some embodiments, the phenotype is or includes the production of IFN-gamma as indicated by ICS.

[0215] In some embodiments, the phenotype is determined by one or more cytokines and CD3 + , CD4 + , CD8 + , CD3 + / CAR + , CD4 + / CAR + , CD8 + / CAR + , annexin V - , annexin V - CD3 + , annexin V - CD4 + , annexin V - CD8 + , annexin V - CD3 + / CAR + , annexin V - CD4 + / CAR+ , annexin V - CD8 + / CAR + , activated caspase 3 - , activated caspase 3 - / CD3 + , activated caspase 3 - / CD4 + , activated caspase 3 - / CD8 + , activated caspase 3 - / CD3 + / CAR + , activated caspase 3 - / CD4 + / CAR + or activated caspase 3 - / CD8 + / CAR + In certain embodiments, the phenotype is or comprises the production of any one of, or a combination thereof. + / CAR + and / or CD8 + / CAR + In some embodiments, the phenotype is or comprises the production of one or more cytokines in the CD4 + / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 in CD4 cells. + / CAR + In some embodiments, the phenotype is or comprises the production of TNF-alpha in CD4 cells. + / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 and TNF-alpha in CD4 cells. + / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 and IFN-gamma in CD8 cells. + / CAR +In some embodiments, the phenotype is or comprises the production of TNF-alpha in CD8 cells. + / CAR + In some embodiments, the phenotype is or comprises the production of IFN-gamma and TNF-alpha in the cell. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 in the cell. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises the production of TNF-alpha in the cell. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 and TNF-alpha in the cell. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 and IFN-gamma in the cell. - / CD8 + / CAR + In some embodiments, the phenotype is or comprises the production of TNF-alpha in the cell. - / CD8 + / CAR + In some embodiments, the phenotype is or comprises the production of IFN-gamma and TNF-alpha in the cell. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises the production of TNF-alpha in the cell. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 and TNF-alpha in the cell. - / CD4+ / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 and IFN-gamma in the cell. - / CD8 + / CAR + In some embodiments, the phenotype is or comprises the production of TNF-alpha in the cell. - / CD8 + / CAR + The phenotypes described in this paragraph are positively correlated with durable response and progression-free survival. Thus, in some embodiments, cells containing these phenotypes are maximized or increased in cell compositions (e.g., output compositions, therapeutic cell compositions).

[0216] In some embodiments, the phenotype is or comprises a lack of production of one or more cytokines. In certain embodiments, the phenotype comprises a lack of production of one or more cytokines and CD3 + , CD4 + , CD8 + , CD3 + / CAR + , CD4 + / CAR + , CD8 + / CAR + , annexin V - , annexin V - CD3 + , annexin V - CD4 + , annexin V - CD8 + , annexin V - CD3 + / CAR + , annexin V - CD4 + / CAR + , annexin V - CD8 + / CAR + , activated caspase 3 - , activated caspase 3- / CD3 + , activated caspase 3 - / CD4 + , activated caspase 3 - / CD8 + , activated caspase 3 - / CD3 + / CAR + , activated caspase 3 - / CD4 + / CAR + or activated caspase 3 - / CD8 + / CAR + In some embodiments, the phenotype is or comprises a lack of production of any one or a combination thereof. In some embodiments, the one or more cytokines are IL-2, IFN-gamma, and / or TNF-alpha. In some embodiments, the phenotype is or comprises a lack of production of activated caspase 3 - / CD4 + / CAR + In some embodiments, the phenotype is or comprises a lack of production of IL-2 in the cell. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises a lack of production of TNF-alpha in the cell. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises a lack of production of IL-2 and TNF-alpha in the cells. - / CD4 + / CAR + In some embodiments, the phenotype is or comprises a lack of production of IL-2 and IFN-gamma in the cells. - / CD8 + / CAR + In some embodiments, the phenotype is or comprises a lack of production of TNF-alpha in the cell. - / CD8 + / CAR +is or comprises a lack of production of INF-gamma and TNF-alpha in cells. In some embodiments, the phenotype described in this paragraph negatively correlates with durable response and progression-free survival.

[0217] In certain embodiments, the phenotype is or includes the presence or absence of internal amounts of one or more of IL-2, IL-13, IFN-gamma, or TNF-alpha, as assessed by an ICS assay and one or more specific markers for a subset or cells of a particular cell type. In some embodiments, the phenotype is or includes IL-2, IL-13, IFN-gamma, or TNF-alpha, as well as CD4 + / CAR + and / or CD8 + / CAR + In certain embodiments, the phenotype is or comprises the production or lack thereof of one or more of IL-2 and CD4 + / CAR + and / or CD8 + / CAR + In some embodiments, the phenotype is or comprises the production of IL-2 and CD4 + / CAR + and / or CD8 + / CAR + In some embodiments, the phenotype is or comprises a lack or underproduction of IL-13 and CD4 + / CAR + and / or CD8 + / CAR + In some embodiments, the phenotype is or comprises the production of IL-13 as well as CD4 + / CAR + and / or CD8 + / CAR + In certain embodiments, the phenotype is or comprises the production of IL - 13 and CD4 + / CAR +and / or CD8 + / CAR + In some embodiments, the phenotype is or comprises a lack or underproduction of IFN-gamma as well as CD4 + / CAR + and / or CD8 + / CAR + In certain embodiments, the phenotype is or comprises the production of TNF-alpha as well as CD4 + / CAR + and / or CD8 + / CAR + In certain embodiments, the phenotype is or comprises the production of TNF-alpha as well as CD4 + / CAR + and / or CD8 + / CAR + is or involves the absence or underproduction of

[0218] Any one or more of the phenotypes, alone or in combination, can be assessed or determined according to the methods provided. In some embodiments, the phenotype is CD3 + , CD3 + / CAR + , CD4 + / CAR + , CD8 + / CAR + Or a combination thereof.

[0219] In some particular embodiments, the phenotype is CD3 + In certain embodiments, the phenotype is or comprises CD3 + / CAR + In some embodiments, the phenotype is or comprises CD8 + / CAR + In certain embodiments, the phenotype is or comprises CD4+ / CAR+.

[0220] In certain embodiments, the phenotype is annexin - / CD3 + / CAR + In some embodiments, the phenotype is or comprises annexin - / CD4 + / CAR + In certain embodiments, the phenotype is or comprises annexin - / CD8 + / CAR.

[0221] In certain embodiments, the phenotype is intracellular IL-2 and CD4 + / CAR + In certain embodiments, the phenotype is characterized by the absence or low amounts of intracellular IL-13 and CD4 + / CAR + In some embodiments, the phenotype is a lack or low amount of IL-13 and CD8 + / CAR + In certain embodiments, the phenotype is the absence or low amount of intracellular expression of TNF-alpha CD4 + / CAR + Absence or low amount of

[0222] In some particular embodiments, the phenotype is CD8 + / CAR + In certain embodiments, the phenotype is or comprises annexin - / CD8 + / CAR + is or contains

[0223] In some embodiments, the phenotype optionally includes an indication of production of one or a combination of cytokines non-specific to the antigen or recombinant receptor and / or polyclonally produced, wherein the one or more cytokines are IL-2, IL-13, IL-17, IFN-gamma, or TNF-alpha. In some embodiments, the indication of production is measured in an assay comprising incubating a sample of the T cell composition with a polyclonal agent, an antigen-specific agent, or a recombinant receptor, optionally an agent that binds to a CAR, optionally an intracellular cytokine staining assay. In some embodiments, the agent is or comprises PMA and ionomycin, or is or comprises a T cell receptor or T cell receptor complex agonist. In some embodiments, the phenotype includes a naive phenotype or a memory phenotype, optionally wherein the memory phenotype comprises a T effector memory phenotype, a T central memory phenotype, or a T effector memory phenotype expressing CD45RA (Temra).

[0224] In some embodiments, recombinant receptor-dependent (e.g., CAR) activity is a measure of the production or accumulation of inflammatory cytokines, optionally one or a combination of TNF-alpha, IFN-gamma, and IL-2. In some embodiments, recombinant receptor-dependent (e.g., CAR) activity is a measure of the production or accumulation of TNF-alpha, IFN-gamma, and a combination of IL-2 and IL-17. In some embodiments, recombinant receptor-dependent (e.g., CAR) activity is a measure of the production or accumulation of IFN-gamma and IL-2. In some embodiments, recombinant receptor-dependent (e.g., CAR) activity is a measure of the production or accumulation of IFN-gamma, TNFA, and IL-2. In some embodiments, recombinant receptor-dependent (e.g., CAR) activity is a measure of the production or accumulation of IFN-gamma and TNFA.

[0225] In some embodiments, recombinant receptor activity is recombinant receptor-specific killing (e.g., cytolytic behavior). In some embodiments, the cytolytic activity of engineered CD8+ cells is assessed (e.g., quantified). In some embodiments, recombinant receptor-dependent cytolytic activity is assessed by exposing, incubating, and / or contacting cells expressing the recombinant receptor, or a cell composition containing cells expressing the recombinant receptor, with target cells expressing the antigen and / or epitope bound and / or recognized by the recombinant receptor. Cytolytic activity can be measured by directly or indirectly measuring target cell number over time. For example, target cells can be incubated with a detectable marker before incubation with recombinant receptor-expressing cells, such marker being detectable after the target cells are lysed, or a detectable marker detectable in live target cells. These readouts provide direct or indirect target cell number and / or target cell death and can be measured at different time points during the assay. A decrease in target cell number and / or an increase in target cell death indicates cytolytic activity of the cells. Suitable methods for performing cytolytic assays are known in the art and include, but are not limited to, chromium-51 release assays, non-radioactive chromium assays, and flow cytometry assays using fluorescent dyes such as carboxyfluorescein succinimidyl ester (CFSE), PKH-2, and PKH-26. In some cases, cytolytic activity is also referred to herein as lysis of cells.

[0226] In some specific embodiments, recombinant receptor (e.g., CAR) dependent cytolytic activity is measured by incubating a cell composition comprising cells expressing recombinant receptor with target cells expressing the antigen or epitope thereof that is bound or recognized by recombinant receptor.In some specific embodiments, the recombinant receptor is CAR.

[0227] In some embodiments, the activity measurement is compared with a control.In some specific embodiments, the control is a culture of target cells that is not incubated with the cell composition.In some embodiments, the control is a measurement from a control cell composition that does not contain CAR+ cells that are incubated with the target cells at the same ratio.

[0228] In certain embodiments, the measurement value of the cytolytic activity assay is the number of viable target cells during incubation or at the end of incubation. In certain embodiments, the measurement value is the amount of target cell death marker, such as chromium-51, released during incubation. In some embodiments, the measurement value is the amount of target cell death determined by subtracting the amount of co-incubated target cells at a given time point from the amount of control target cells incubated alone. In some embodiments, the measurement value is the percentage of target cells remaining at a given time point compared to the initial amount of target cells. In certain embodiments, the measurement value is the amount of cells killed over a certain amount of time. In certain embodiments, the measurement value is the amount of cells killed per cell, or the amount of cells killed per set number or reference of cells, for example, but not limited to, per 100 cells of the composition, per 10 cells. 3 Per 10 cells 4 Per 10 cells 5 Per 10 cells 6 Per 10 cells 7 Per 10 cells 8 Per 10 cells 9 per piece, or 10 cells 10The measured value is the amount of target cells killed per cell. In certain embodiments, the measured value is the amount of cells killed per each CAR+ cell, CAR+ / CD8+ cell or Annexin- / CAR+ / CD8+ cell of the cell composition, or a reference or set number thereof. In some specific embodiments, the measured value is the amount of cells killed per cell of the cell composition over a certain amount of time. In certain embodiments, the measured value is the amount of cells killed per CAR+ cell, CAR+ / CD8+ cell or Annexin- / CAR+ / CD8+ cell of the cell composition over a certain amount of time.

[0229] In some embodiments, the cell phenotype comprises assessing the genomic integration of a transgene sequence, such as a transgene sequence encoding a recombinant receptor, e.g., a CAR. In some embodiments, the cell phenotype is the integrated copy number, e.g., vector copy number, which is the number of copies of the transgene sequence integrated into the chromosomal or genomic DNA of the cell. In some embodiments, the vector copy number can be expressed as an average or mean copy number. In some aspects, the vector copy number of a particular integrated transgene comprises the number of integrants (including the transgene sequence) per cell. In some embodiments, the vector copy number of a particular integrated transgene comprises the number of integrants (including the transgene sequence) per diploid genome. In some aspects, the vector copy number of a transgene sequence is expressed as the number of integrated transgene sequences per cell. In some aspects, the vector copy number of a transgene sequence is expressed as the number of integrated transgene sequences per diploid genome. In some embodiments, the copy number is the average or mean copy number per diploid genome or per cell in a population of cells.

[0230] In some embodiments, the therapeutic cell composition characteristic is the clonality of the cells of the therapeutic cell composition. In some embodiments, assessing the clonality of a population of T cells is assessing the clonal diversity of the population of T cells. In some embodiments, the T cells are polyclonal or multiclonal. The clonality, e.g., polyclonality, of the therapeutic cell composition of T cells is a measure of the breadth of the population's response to a given antigen. In some aspects, the therapeutic cell composition can be assessed by measuring the number of distinct epitopes recognized by antigen-specific cells. This can be done using standard techniques for generating and cloning antigen-specific T cells in vitro. In some embodiments, the T cells are polyclonal (or multiclonal), with no single clonotype population dominating the population of naive-like T cells.

[0231] In the context of a population of T cells, e.g., a therapeutic cell composition, in some aspects, a polyclonal signature refers to a population of T cells with multiple, broad antigen specificities. In some embodiments, polyclonality refers to a population of T cells that exhibits high diversity in their TCR repertoire. In some cases, the diversity of the TCR repertoire is due in some respects to V(D)J recombination events induced by self and foreign antigen selection events. In some embodiments, a diverse or polyclonal population of T cells is one in which analysis indicates the presence of multiple, various, or distinct TCR transcripts or products present in the population. In some embodiments, a population of T cells that exhibits high or relatively high clonality is one in which the diversity of the TCR repertoire is low. In some embodiments, T cells are oligoclonal when analysis indicates the presence of several, e.g., two or three, TCR transcripts or products in the population of T cells. In some embodiments, monoclonality refers to a population of T cells that is low in diversity. In some embodiments, T cells are monoclonal when analysis indicates the presence of a single TCR transcript or product in the population of T cells.

[0232] In some instances, the clonality of cells, such as T cells, in therapeutic cell compositions is determined by clonal sequencing, for example, next-generation sequencing or spectratyping.In some aspects, next-generation sequencing method can be used with genomic DNA or cDNA from T cells to evaluate the sequence encoding TCR repertoire, for example, complementarity determining region 3 (CDR3).In some embodiments, whole transcriptome sequencing by RNA-seq can be used.In some embodiments, single-cell sequencing method can be used.

[0233] In some embodiments, clonality, e.g., polyclonality, can be assessed or determined by spectratyping (a measure of the hypervariable region repertoire of TCR Vβ, Vα, Vγ, or Vδ chains). Spectratyping distinguishes rearranged gene variable regions of a particular size from non-rearranged sequences. Therefore, it is understood that a single peak may represent a population of T cells expressing any one of a limited number of rearranged TCR gene variable regions (Vβ, Vα, Vγ, or Vδ) containing any one of the four possible nucleotides (adenine (a), guanine (g), cytosine (c), or thymine (t)) or a combination of these four nucleotides in the junction region. A population of T cells is considered polyclonal if the Vβ spectratyping profile of a given TCR Vβ, Vα, Vγ, or Vδ family has multiple peaks, typically five or more major peaks, most often with a Gaussian distribution. Polyclonality can also be defined by the generation and characterization of antigen-specific clones for an antigen of interest. In the context of a population of T cells, e.g., a therapeutic cell composition, monoclonality refers to a population of T cells with a single specificity as defined by spectratyping (a measure of the hypervariable region repertoire of the TCR Vβ, Vα, Vγ, or Vδ chain). A population of T cells is considered monoclonal (or monospecific) if the Vβ, Vα, Vγ, and / or Vδ spectratyping profile for a given TCR Vβ, Vα, Vγ, and / or Vδ family has a single dominant peak.

[0234] In some embodiments, the method for assessing clonality may include various features of the methods described in International Publication Nos. 2012 / 048341, 2014 / 144495, 2017 / 053902, 2016044227, 2016176322, and 2012048340, each of which is incorporated by reference in its entirety. In some embodiments, such methods may be used to obtain sequence information about a target polynucleotide of interest within a cell, such as a TCR. The target gene may be obtained from genomic DNA or mRNA of cells from a cell sample or cell population. The cell sample or cell population may include immune cells. For example, for a target TCR molecule, a gene encoding a TCR chain may be obtained from genomic DNA or mRNA of an immune cell or T cell. In some embodiments, the starting material is RNA from a T cell, which is composed of a gene encoding a TCR chain.

[0235] In some embodiments, the Shannon index is applied as a threshold to filter clones for clonality ("Shannon-adjusted clonality"), see Chaara et al. (2018) Front Immunol 9:1038). In some embodiments, the therapeutic cell composition feature is the clonality of CD4+ cells of the therapeutic cell composition. In some embodiments, the therapeutic cell composition feature is the clonality of CD8+ cells of the therapeutic cell composition.

[0236] In some embodiments, the therapeutic cell composition characteristic is a dose. In some embodiments, the dose is a single dose of CD4+ and CD8+ engineered cells. In some embodiments, the single dose comprises separately administering to the subject CD4+ engineered cells and CD8+ engineered cells. In some embodiments, the single dose comprises administering to the subject 25×10 6 CD8+CAR+ T cells and 25 × 10 6 In some embodiments, the single dose comprises administering 50 x 10 CD4+ CAR+ T cells to the subject. 6CD8+CAR+ T cells and 50 × 10 6 In some embodiments, the single dose comprises administering 75 x 10 CD4+ CAR+ T cells to the subject. 6 CD8+CAR+ T cells and 75 × 10 6 The method includes administering CD4+ CAR+ T cells separately to a subject.

[0237] In some embodiments, the therapeutic cell composition characteristics include any one or more or all of the therapeutic cell composition characteristics described herein, e.g., phenotype and recombinant receptor-dependent activity. In some embodiments, the therapeutic cell composition characteristics are CAS3- / CCR7- / CD27-, CAS3- / CCR7- / CD27+, CAS3- / CCR7+, CAS3- / CCR7+ / CD27-, CAS3- / CCR7+ / CD27+, CAS3- / CD27+, CAS3- / CD28+, CAS3- / CD28+ / CD27-, CAS3- / CD28+ / CD27+, CAS3- / CCR7- / CD45RA-, CAS3- / CCR7- / CD45RA+, CAS3- / CCR7+ / CD45RA-, CAS3- / CCR7+ / CD45RA+, CAS+ / CD3+ / CAR+, CD3+ / CAR+, CD3+, CAR+, clonality, EGFRt+, cytokine-, IFNG+, IFNg+ / IL2, IFNg+ / IL17+ / TNFa+, IFNg + / IL2+ / IL17+ / TNFa+, IFNg+ / IL2+ / TNFa+, CAR+ / IFNg+, IFNg+ / TNFa+, CAR+ / IL2+, IL2+ / TNFa+, cytolysis, CAR+ / TNFa+, viable cell concentration, vector copy number (VCN), EGFRt+ vector copy number, viability, GMCSF+ / CD19+, IFNG+ / CD19+, IL10+ / CD1 9+, IL13+ / CD19+, IL2+ / CD19+, IL4+ / CD19+, IL5+ / CD19+, IL6+ / CD19+, MIP1A+ / CD19+, MIP1B+ / CD19+, sCD137+ / CD19+, TNFa+ / CD19+, dose, dose level, percent viable administered cells, total non-viable administered cells, total viable administered cells, total dose.

[0238] In some embodiments, the therapeutic cell composition characteristics are CAS3- / CCR7- / CD27- / CD4+, CAS3- / CCR7- / CD27+ / CD4+, CAS3- / CCR7+ / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CAS3- / CCR7+ / CD27+ / CD4+, CAS3- / CD27+ / CD4+, CAS3- / CD28+ / CD4+, CAS3- / CD28+ / CD27- / CD4+, CAS3- / CD28+ / CD27+ / CD4+, CAS3- / CCR7- / CD45RA- / CD4+, CAS 3- / CCR7- / CD45RA+ / CD4+, CAS3- / CCR7+ / CD45RA- / CD4+, CAS3- / CCR7+ / CD45RA+ / CD4+, CAS+ / CD3+ / CAR+ / CD4+, CD3+ / CAR+ / CD4+, CD3+ / CD4+, CAR+ / CD4+, clonality of CD4+ cells, EGFRt+ / CD4+, cytokine- / CD4+, IFNG+ / CD4+, IFNg+ / IL2 / CD4+, IFNg+ / IL17+ / TNFa+ / CD4+, IFNg+ / IL2+ / IL17+ / TNFa+ / CD4+, IFNg+ / IL2+ / TNFa+ / CD4+, CAR+ / IFNg+ / CD4+, IFNg+ / TNFa+ / CD4+, CAR+ / IL2+ / CD4+, IL2+ / TNFa+ / CD4+, CD4+ cytolysis, CAR+ / TNFa+ / CD4+, viable cell concentration of CD4+ cells, vector copy number in CD4+ cells, EGFRt+ vector copy number in CD4+ cells, viability of CD4+ cells, GMCSF+ / CD19+ / CD4+, IFNG+ / CD19+ / CD4+, IL10+ / CD19+ / CD4+, IL13+ / CD19+ / CD4+ D4+, IL2+ / CD19+ / CD4+, IL4+ / CD19+ / CD4+, IL5+ / CD19+ / CD4+, IL6+ / CD19+ / CD4+, MIP1A+ / CD19+ / CD4+, MIP1B+ / CD19+ / CD4+, sCD137+ / CD19+ / CD4+, TNFa+ / CD19+ / CD4+, dose of CD4+ cells, dose level of CD4+ cells, percent viable administered CD4+ cells, total non-viable administered CD4+ cells, total viable administered CD4+ cells, total dose of CD4+ cells.

[0239] In some embodiments, the therapeutic cell composition characteristics are CAS3- / CCR7- / CD27- / CD8+, CAS3- / CCR7- / CD27+ / CD8+, CAS3- / CCR7+ / CD8+, CAS3- / CCR7+ / CD27- / CD8+, CAS3- / CCR7+ / CD27+ / CD8+, CAS3- / CD27+ / CD8+, CAS3- / CD28+ / CD8+, CAS3- / CD28+ / CD27- / CD8+, CAS3- / CD28+ / CD27+ / CD8+, CAS3- / CCR7- / CD45RA- / CD8+, CAS 3- / CCR7- / CD45RA+ / CD8+, CAS3- / CCR7+ / CD45RA- / CD8+, CAS3- / CCR7+ / CD45RA+ / CD8+, CAS+ / CD3+ / CAR+ / CD8+, CD3+ / CAR+ / CD8+, CD3+ / CD8+, CAR+ / CD8+, CD8+ clonality, EGFRt+ / CD8+, cytokine- / CD8+, IFNG+ / CD8+, IFNg+ / IL2 / CD8+, IFNg+ / IL17+ / TNFa+ / CD8+, IFNg+ / IL2+ / IL17+ / TNFa+ / CD8+, IF Ng+ / IL2+ / TNFa+ / CD8+, CAR+ / IFNg+ / CD8+, IFNg+ / TNFa+ / CD8+, CAR+ / IL2+ / CD8+, IL2+ / TNFa+ / CD8+, cytolysis by CD8+ cells, CAR+ / TNFa+ / CD8+, viable cell concentration of CD8+ cells, vector copy number in CD8+ cells, EGFRt+ vector copy number in CD8+ cells, viability of CD8+ cells, GMCSF+ / CD19+ / CD8+, IFNG+ / CD19+ / CD8+, IL10+ / CD19+ / CD8+, IL13+ / CD19+ / CD8+, IL2+ / CD19+ / CD8+, IL4+ / CD19+ / CD8+, IL5+ / CD19+ / CD8+, IL6+ / CD19+ / CD8+, MIP1A+ / CD19+ / CD8+, MIP1B+ / CD19+ / CD8+, sCD137+ / CD19+ / CD8+, TNFa+ / CD19+ / CD8+, dose of CD8+ cells, CD8+ dose level, percent viable administered CD8+ cells, total non-viable administered CD8+ cells, total viable administered CD8+ cells, total dose of CD8+ cells.

[0240] In some embodiments, the therapeutic cell composition characteristics are CAS3- / CCR7- / CD27-, CAS3- / CCR7- / CD27+, CAS3- / CCR7+, CAS3- / CCR7+ / CD27-, CAS3- / CCR7+ / CD27+, CAS3- / CD27+, CAS3- / CD28+, CAS3- / CD28+ / CD27-, CAS3- / CD28+ / CD27+, CAS3- / CCR7- / CD45RA-, CAS3- / CCR7- / CD45RA+, CAS3- / CCR7+ / CD45RA-, CAS3- / CCR7+ / CD45RA+, CAS+ / CD3+ / CAR+, CD3+ / CAR+, CD3+, CAR+, clonality, EGFRt+, cytokine-, IFNG+ , IFNg+ / IL2, IFNg+ / IL17+ / TNFa+, IFNg+ / IL2+ / IL17+ / TNFa+, IFNg+ / IL2+ / TNFa+, CAR+ / IFNg+, IFNg+ / TNFa+, CAR+ / IL2+, IL2+ / TNFa+, cytolysis, CAR+ / TNFa+, viable cell concentration, vector copy number, EGFRt+ vector copy number, viability, GMCSF+ / , IFNG+ / , IL10+ / , IL13+ / , IL2+ / , IL4+ / , IL5+ / , IL6+ / , MIP1A+ / , MIP1B+ / , sCD137+ / , TNFa+ / , dose, dose level, percent viable administered cells, total non-viable administered cells, total viable administered cells, total dose.

[0241] In some embodiments, the therapeutic cell composition characteristics are CAS3- / CCR7- / CD27- / CD4+, CAS3- / CCR7- / CD27+ / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CAS3- / CCR7+ / CD27+ / CD4+, CAS3- / CD27+ / CD4+, CAS3- / CD28+ / CD4+, CAS3- / CD28+ / CD27- / CD4+, CAS3- / CD28+ / CD27+ / CD4+, CAS3- / CCR7 - / CD45RA- / CD4+, CAS3- / CCR7- / CD45RA+ / CD4+, CAS3- / CCR7+ / CD45RA- / CD4+, CAS3- / CCR7+ / CD45RA+ / CD4+, CAS+ / CD3+ / CAR+ / CD4+, CD 3+ / CAR+ / CD4+, CD3+ / CD4+, CAR+ / CD4+, CD4+ cell clonality, EGFRt+ / CD4+, cytokine- / CD4+, IFNG+ / CD4+, IFNg+ / IL2 / CD4+, IFNg+ / IL17+ / TNF a+ / CD4+, IFNg+ / IL2+ / IL17+ / TNFa+ / CD4+, IFNg+ / IL2+ / TNFa+ / CD4+, CAR+ / IFNg+ / CD4+, IFNg+ / TNFa+ / CD4+, CAR+ / IL2+ / CD4+, IL2+ / TNFa+ / CD4+, CD4+-mediated cytolysis, CAR+ / TNFa+ / CD4+, viable cell concentration in CD4+ cells, vector copy number in CD4+ cells, EGFRt+ vector copy number in CD4+, CD4+ viability, GMCSF+ / CD4+, IF The present invention relates to a method for determining the level of CD4+ cells administered, the method...

Claims

Claim 1: A processor-implemented method for predicting a clinical response in a subject having a disease or condition before the subject is treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with the disease or condition in the subject, comprising: wherein said therapeutic cell composition is for treating said disease or condition, wherein said therapeutic cell composition is administered to said subject, wherein said therapeutic cell composition is made from an input composition; wherein the input composition comprises T cells selected from a sample from a subject, wherein the T cells are used to generate a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR); The method comprises: (a) Below: (i) subject features determined from the subject, the subject features comprising one or more of subject attributes and clinical attributes, wherein the subject attributes comprise one or more of age, weight, height, ethnicity, race, sex, and body mass index, and the clinical attributes comprise one or more of biomarkers, disease diagnosis, disease burden, disease duration, disease grade, and treatment history; (ii) input composition features determined from the input composition, the input composition features comprising a cell phenotype, wherein the cell phenotype comprises an identification of one or more of effector T cells, helper T cells, memory T cells, regulatory T cells, naive T cells, CD4+ T cells, and CD8+ T cells; and (iii) a therapeutic cell composition feature determined from the therapeutic cell composition, the therapeutic cell composition feature comprising one or more of a cell phenotype, a recombinant receptor-dependent activity, and a dose; receiving a feature including: (b) applying the features as inputs to a machine learning model trained to predict the subject's clinical response to treatment with the therapeutic cell composition based on the features prior to treatment of the subject with the therapeutic cell composition, wherein the clinical response is progression-free survival (PFS); overall response rate (ORR); objective response (OR); complete response (CR); safety: neurological event (NE) grade ≧1, NE event grade ≧3, cytokine release syndrome (CRS) grade ≧1, CRS grade ≧3; and pharmacokinetic endpoints: log10 area under the curve (AUC), log10 maximum concentration (C max ), and time to peak concentration (T max ). The method comprising:

2. 10. The method of claim 1, wherein the machine learning model is a supervised machine learning model.

3. 3. The method of claim 1 or 2, wherein the machine learning model is a random forest model.

4. 3. The method of claim 1 or 2, wherein the machine learning model is a random survival forest model.

5. The machine learning model is trained using supervised training, wherein the supervised training is based on features received from a plurality of subjects having a disease or condition, wherein each of the plurality of subjects has been treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with the disease or condition; wherein the therapeutic cell composition from each of the plurality of subjects is engineered to express a CAR and is generated from an input composition from each subject before the subject is treated; wherein the input composition comprises T cells selected from a sample from a subject to be treated, wherein the T cells are used to generate a therapeutic cell composition comprising T cells comprising a CAR, and wherein the supervised training comprises: (a) Below: (i) Object features obtained from each of a plurality of objects; (ii) input composition features derived from the input compositions derived from each of the plurality of subjects; and (iii) Therapeutic cell composition feature quantities obtained from therapeutic cell compositions derived from each of a plurality of subjects receiving a feature including: (b) preprocessing the features to identify informative features, wherein the informative features comprise a subset of the features and include one or more target features, one or more input composition features, and one or more therapeutic cell composition features; and (c) obtaining a clinical response from each of the plurality of subjects following treatment with one of the plurality of therapeutic compositions; and (d) applying the informative features from the plurality of subjects and the obtained clinical responses as inputs to train a machine learning model.

5. The method of any one of claims 1 to 4, comprising:

6. A processor-implemented method for developing a machine learning model, comprising: wherein the machine learning model is for predicting the clinical response of a subject having a disease or condition before the subject is treated with one of a plurality of therapeutic cell compositions, each comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with the disease or condition in the subject; wherein the training is based on features received from a plurality of subjects having a disease or condition, wherein each of the plurality of subjects has been treated with a therapeutic cell composition comprising T cells comprising a chimeric antigen receptor (CAR) that binds to an antigen associated with the disease or condition; wherein a therapeutic cell composition derived from each of a plurality of subjects is generated from an input composition derived from each subject to be treated and is engineered to express a CAR; wherein the input composition comprises T cells selected from a sample from a subject, wherein the T cells are used to generate a therapeutic cell composition comprising T cells comprising a CAR; wherein the method comprises: (a) Below: (i) subject features determined from each of a plurality of subjects, the subject features comprising one or more of subject attributes and clinical attributes, wherein the subject attributes comprise one or more of age, weight, height, ethnicity, race, sex, and body mass index, and the clinical attributes comprise one or more of biomarkers, disease diagnosis, disease burden, disease duration, disease grade, and treatment history; (ii) input composition features determined from the input compositions from each of a plurality of subjects, the input composition features comprising a cellular phenotype, wherein the cellular phenotype comprises an identification of one or more of effector T cells, helper T cells, memory T cells, regulatory T cells, naive T cells, CD4+ T cells, and CD8+ T cells; and (iii) a therapeutic cell composition feature determined from the therapeutic cell composition from each of the plurality of subjects, the therapeutic cell composition feature comprising one or more of a cell phenotype, a recombinant receptor-dependent activity, and a dose. receiving a feature including: (b) obtaining a clinical response from each of the plurality of subjects after treatment with the therapeutic cell composition, wherein the clinical response is progression-free survival (PFS); overall response rate (ORR); objective response (OR); complete response (CR); safety: neurological event (NE) grade ≧1, NE event grade ≧3, cytokine release syndrome (CRS) grade ≧1, CRS grade ≧3; and pharmacokinetic endpoints: log 10 area under the curve (AUC), log 10 maximum concentration (C max ), and time to peak concentration (T max ); and (c) applying the features and the obtained clinical responses from the plurality of subjects as inputs to train the machine learning model. The method comprising:

7. 7. The method of claim 6, wherein the method comprises identifying features associated with the obtained clinical response from a trained machine learning model.

8. 8. The method of claim 6 or 7, wherein the machine learning model is a supervised machine learning model.

9. 9. The method of any one of claims 6 to 8, wherein the machine learning model is a random forest model.

10. The method of any one of claims 6 to 8, wherein the machine learning model is a random survival forest model.

11. The method of any one of claims 6 to 10, wherein the features are informative features identified by preprocessing.

12. the informative features are a subset of the features; and one or more target features, one or more input composition features, and one or more therapeutic cell composition features; and applying the informative features from the plurality of subjects and the obtained clinical responses as inputs to train a machine learning model.

12. The method of claim 11.

13. 13. The method of any one of claims 6 to 12, wherein the features and obtained clinical responses from multiple subjects are applied as inputs to train machine learning using supervised learning.

14. 14. The method of any one of claims 6-13, wherein the clinical response is or comprises a complete response (CR), a partial response (PR), a durable response, progression-free survival (PFS), an objective response (OR), a pharmacokinetic response that is or is greater than the target pharmacokinetic response, no toxic response or a mild toxic response, a toxic response, a pharmacokinetic response that is lower than the target pharmacokinetic response, or no CR, PR, durable response, or OR.

15. 15. The method of any one of claims 11 to 14, wherein identifying informative features associated with the clinical response comprises determining an importance measure for each of the informative features.

16. 16. The method of claim 15, wherein the importance measure comprises a permutation importance measure, an average minimum depth, and / or a total number of trees of the trained machine learning model, where the root node is split by informative features.

17. 17. The method of claim 15 or 16, wherein the informative features associated with the clinical response are the first 10, 9, 8, 7, 6, 5, 4, 3, 2 or 1 informative features identified by ranking the importance scale values ​​of each of the informative features, and the importance scale is the same for each informative feature.

18. The preprocessing for identifying informative features is a) removing target features, input composition features, and therapeutic cell composition features having data loss greater than, greater than about, or equal to 50%; b) removing subject features, input composition features, and therapeutic cell composition features that have (i) zero variance, (ii) more than, about, or 95% of data values ​​equal to a single value, and / or (iii) fewer than 0.1n unique values, where n is the number of subjects in the plurality of subjects; c) inputting missing data for the target features, input composition features, and therapeutic cell composition features by multivariate imputation using chained equations; and d) identifying a covariate cluster including a set of target features, input composition features, therapeutic cell composition features, and combinations thereof having correlation coefficients with absolute values ​​greater than, greater than about, or equal to 0.5, and iteratively selecting target features, input composition features, and therapeutic cell composition features from the covariate cluster, wherein the selected target features, input composition features, and therapeutic cell composition features have the smallest average absolute correlations with all remaining target features, input composition features, and therapeutic cell composition features.

18. The method of any one of claims 11 to 17, comprising one or more of:

19. 19. The method of any one of claims 11 to 18, wherein preprocessing to identify informative features comprises or is removing target features, input composition features, and therapeutic cell composition features that have greater than 50%, about 50%, or 50% data loss.

20. 20. The method of any one of claims 11-19, wherein the preprocessing to identify informative features comprises or is a step of removing target features, input composition features, and therapeutic cell composition features that have (i) zero variance, or (ii) more than 95%, about 95%, or 95% of data values ​​equal to a single value and fewer than 0.1n unique values, where n = the number of samples.

21. 21. The method of any one of claims 11 to 20, wherein the preprocessing for identifying informative features comprises or is a step of imputing missing data for target features, input composition features, and therapeutic cell composition features by multivariate imputation using chained equations.

22. 22. The method of any one of claims 11 to 21, wherein the preprocessing for identifying informative features comprises or is a step of identifying a covariate cluster including a set of target features, input composition features, therapeutic cell composition features, and combinations thereof, having correlation coefficients with absolute values ​​greater than, about, or equal to 0.5, and iteratively selecting target features, input composition features, and therapeutic cell composition features from the covariate cluster, wherein the selected target features, input composition features, and therapeutic cell composition features have the smallest average absolute correlations with all remaining target features, input composition features, and therapeutic cell composition features.

23. 23. The method of any one of claims 5-22, wherein said plurality of subjects is 10, about 10, or more than 10 subjects and less than 500 subjects.

24. The features of interest were: treatment group, bridging chemotherapy, bridging chemotherapy and radiotherapy, bridging chemotherapy systemic treatment, cell origin, relapse or refractory after chemotherapy, type of diagnosis, disease cohort, disease burden, relapsed or refractory disease, disease origin, gender, route of administration of therapeutic cell composition, fold change in LDH, height, lesion count, oxygen saturation, body temperature (°C), maximum diameter of tumor before treatment with therapeutic cell composition, fold change in SPD, SPD value before lymphocyte-depleting chemotherapy, BMI, weight, sex, ethnicity, race, age, IPI score, ECOG score, disease stage, disease burden based on LDH before lymphocyte-depleting chemotherapy, S before lymphocyte-depleting chemotherapy. Disease burden based on PD, subject having active CNS disease at the time of treatment, disease burden based on extranodal disease classification, number of extranodal sites, disease burden based on bulky disease classification, medical history, number of prior lines of therapy, number of prior lines of systemic therapy, prior allogeneic hematopoietic stem cell transplant (allo-HSCT), prior autologous hematopoietic stem cell transplant (auto-HSCT), chemotherapy-refractory or chemotherapy-sensitive disease type, bridging anticancer therapy for disease control, days from date of leukapheresis to first infusion, months from diagnosis to treatment with therapeutic cell composition, baseline C-reactive protein (CRP), lymphocyte count before leukapheresis (10 9 / L), gene double expressor, gene double hit, gene triple hit, gene double hit or triple hit, gene double hit or triple hit or double expressor, albumin level, alkaline phosphatase level, basophil count, absolute basophil count, direct bilirubin, total bilirubin, blood urea nitrogen level, calcium level, carbon dioxide level, chloride level, creatinine level, eosinophil count, absolute eosinophil count, glucose level, hematocrit level, hemoglobin level 24. The method of any one of claims 1-23, wherein the measured values ​​comprise one or more of: leukocyte count, lymphocyte count, absolute lymphocyte count, magnesium level, absolute monocyte count, monocyte count, absolute neutrophil count, neutrophil count, phosphate level, platelet count, potassium level, total protein, red blood cell count, aspartate aminotransferase level, alanine aminotransferase level, sodium level, two-way product sum, triglycerides, tumor maximum diameter, tumor perpendicular diameter, uric acid level, and white blood cell count.

25. Input composition features are CAS3- / CCR7- / CD27- / CD4+, CAS3- / CCR7- / CD27+ / CD4+, CAS3- / CCR7+ / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CA S3- / CCR7+ / CD27+ / CD4+, CAS3- / CD27+ / CD4+, CAS3- / CD28- / CD27- / CD4+, CAS3- / CD28- / CD27+ / CD4+, CAS3- / CD28+ / CD4+ , CAS3- / CD28+ / CD27- / CD4+, CAS3- / CD28+ / CD27+ / CD4+, CAS3- / CCR7- / CD45RA- / CD4+, CAS3- / CCR7- / CD4+, CD45RA+ / CD4 +, CAS3- / CCR7+ / CD45RA- / CD4+, CAS3- / CCR7+ / CD45RA+ / CD4+, CAS+ / CD4+, CAS+ / CD3+ / CD4+, CD4+ clonality, CAS3- / CCR7- / CD27- / CD8+, CAS3- / CCR7- / CD27+ / CD8+, CAS3- / CCR7+ / CD8+, CAS3- / CCR7+ / CD27- / CD8+, CAS3- / CCR7+ / CD27+ / CD8+, CAS 3- / CD27+ / CD8+, CAS3- / CD28- / CD27- / CD8+, CAS3- / CD28- / CD27+ / CD8+, CAS3- / CD28+ / CD8+, CAS3- / CD28+ / CD27- / CD8+, 25. The method of any one of claims 1 to 24, comprising one or more of the following clonalities: CAS3- / CD28+ / CD27+ / CD8+, CAS3- / CCR7- / CD85RA- / CD8+, CAS3- / CCR7- / CD8+, CD85RA+ / CD8+, CAS3- / CCR7+ / CD85RA- / CD8+, CAS3- / CCR7+ / CD85RA+ / CD8+, CAS+ / CD8+, CAS+ / CD3+ / CD8+ and CD8+.

26. 26. The method of any one of claims 1 to 25, wherein the cell phenotype comprises identification of effector T cells, helper T cells, memory T cells, regulatory T cells, naive T cells, CD4+ T cells, and CD8+ T cells.

27. Therapeutic cell composition features are CAS3- / CCR7- / CD27- / CD8+, CAS3- / CCR7- / CD27+ / CD8+, CAS3- / CCR7+ / CD8+, CAS3- / CCR7+ / CD27- / CD8+, CAS3- / CCR7+ / CD27+ / CD8+, CAS 3- / CD27+ / CD8+, CAS3- / CD28+ / CD8+, CAS3- / CD28+ / CD27- / CD8+, CAS3- / CD28+ / CD27+ / CD8+, CAS3- / CCR7- / CD45RA- / CD8+, CAS3- / CCR7- / CD45RA+ / CD 8+, CAS3- / CCR7+ / CD45RA- / CD8+, CAS3- / CCR7+ / CD45RA+ / CD8+, CAS+ / CD3+ / CAR+ / CD8+, CD3+ / CAR+ / CD8+, CD3+ / CD8+, CAR+ / CD8+, CD8+ cell clonality, EGFRt+ / CD8+, cytokine- / CD8+, IFNG+ / CD8+, IFNg+ / IL2 / CD8+, IFNg+ / IL17+ / TNFa+ / CD8+, IFNg+ / IL2+ / IL17+ / TNFa+ / CD8+, IFNg+ / IL2+ / TNFa+ / CD8+, CAR + / IFNg+ / CD8+, IFNg+ / TNFa+ / CD8+, CAR+ / IL2+ / CD8+, IL2+ / TNFa+ / CD8+, CD8+ cytolysis, CAR+ / TNFa+ / CD8+, viable cell concentration of CD8+ cells, vector copy number of CD8+ cells, EGFRt+ vector copy number of CD8+, viability of CD8+, GMCSF+ / CD8+, IFNG+ / CD8+, IL10+ / CD8+, IL13+ / CD8+, IL2+ / CD8+, IL4+ / CD8+, IL5+ / CD8+, IL6+ / CD8+, MIP1A+ / CD8+, MI P1B+ / CD8+, sCD137+ / CD8+, TNFα+ / CD8+, dose of CD8+ cells, dose level of CD8+ cells, percent viable administered CD8+ cells, total non-viable administered CD8+ cells, total viable administered CD8+ cells, total dose of CD8+ cells, CAS3- / CCR7- / CD27- / CD4+, CAS3- / CCR7- / CD27+ / CD4+, CAS3- / CCR7+ / CD27- / CD4+, CAS3- / CCR7+ / CD27+ / CD4+, CAS3- / CD27+ / CD4+,CAS3- / CD28+ / CD4+, CAS3- / CD28+ / CD27- / CD4+, CAS3- / CD28+ / CD27+ / CD4+, CAS3- / CCR7- / CD45RA- / CD4+, CAS3- / C CR7- / CD45RA+ / CD4+, CAS3- / CCR7+ / CD45RA- / CD4+, CAS3- / CCR7+ / CD45RA+ / CD4+, CAS+ / CD3+ / CAR+ / CD4+, CD3+ / CAR + / CD4+, CD3+ / CD4+, CAR+ / CD4+, CD4+ cell clonality, EGFRt+ / CD4+, cytokine- / CD4+, IFNG+ / CD4+, IFNg+ / IL2 / CD4+, IFNg+ / IL17+ / TNFa+ / CD4+, IFNg+ / IL2+ / IL17+ / TNFa+ / CD4+, IFNg+ / IL2+ / TNFa+ / CD4+, CAR+ / IFNg+ / CD4+, IFNg+ / TNFa+ / CD4+, CAR+ / IL2+ / CD4+, IL2+ / TNFa+ / CD4+, CD4+ cytolysis, CAR+ / TNFa+ / CD4+, viable cell concentration in CD4+ cells, vector copy number in CD4+ cells, EGFRt+ vector copy number in CD4+, CD4+ viability, GMCSF+ / CD4+, IFNG+ / CD4+, IL10+ / CD4+, IL13+ / CD4+, IL2+ / CD4+, IL4+ / CD4+, I 27. The method of any one of claims 1 to 26, comprising one or more of L5+ / CD4+, IL6+ / CD4+, MIP1A+ / CD4+, MIP1B+ / CD4+, sCD137+ / CD4+, TNFa+ / CD4+, dose of CD4+ cells, dose level of CD4+ cells, percent viable administered CD4+ cells, total non-viable administered CD4+ cells, total viable administered CD4+ cells, and total dose of CD4+ cells.

28. 28. The method of any one of claims 1 to 27, wherein the sample comprises a whole blood sample, a buffy coat sample, a peripheral blood mononuclear cell (PBMC) sample, an unfractionated T cell sample, a lymphocyte sample, a leukocyte sample, an apheresis product, or a leukapheresis product.

29. The method of any one of claims 1 to 28, wherein the T cells comprise CD4+ and / or CD8+ T cells.

Images