T cell quantitative system pharmacology model

Abstract

A method may include, by at least one data processor and based on a set of cytokinetic parameters corresponding to a plurality of T-cell phenotypes and the plurality of T-cell phenotypes, identifying a plurality of T-cell phenotypes in peripheral tissue and lymph node compartments (i.e., T-cell phenotypes, T-cell phenotypes, T-cell phenotypes, T-cell phenotypes, T-cell phenotypes, T-cell phenotypes, and T-cell phenotypes). The distribution of the plurality of T cell phenotypes over time is determined based on a transport rate between the plurality of T cell phenotypes (e.g., a healthy tissue compartment), a blood compartment, a tumor drainage lymph node compartment, and a tumor compartment. The plurality of T cell phenotypes may include stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells. Related methods and articles are also disclosed.

Description

[0001] Cross-references

[0002] This application claims priority to U.S. Provisional Patent Application No. 63 / 415,649, filed on October 12, 2022, the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0003] The present disclosure relates generally to quantitative systems pharmacology models, and more particularly to T cell quantitative systems pharmacology models. Background Art

[0004] T cell receptor (TCR) (engineered T cell therapy) is an emerging cancer treatment strategy that shows evidence of anti-tumor activity in both solid tumors and hematological cancers. Despite initial promises, there are still many challenges in understanding and characterizing unique cell dynamics (including the transport (trafficking), proliferation, apoptosis and persistence of TCR engineered T cells and other T cells and T cell therapies after infusion into patients). Because TCR engineered T cell therapy includes live and phenotypically diverse T cells, the pharmacological properties of these therapies are different from molecular therapies, and the pharmacokinetics of these therapies and the resulting pharmacodynamics are monitored and predicted, which also brings different challenges. Further, it is not clear how the cell phenotype and composition of the infusion product can affect the cellular pharmacokinetics and subsequent efficacy and safety. Summary of the invention

[0005] Methods, systems, and articles of manufacture, including computer program products, for quantitative systems pharmacology models of T cells are provided. In one aspect, a method is provided. The method may include: generating a model, wherein the model simulates the distribution over time of each of a plurality of T cell phenotypes in a plurality of physiological compartments after delivery of a T cell target (TCT) product to a patient; generating a plurality of digital twins by the model, wherein each of the plurality of digital twins represents the distribution over time of the plurality of T cell phenotypes in the plurality of physiological compartments associated with a corresponding patient, wherein the corresponding patient has a set of patient features; receiving a set of features associated with a target patient; matching the target patient features with the patient features associated with the digital twins to select a subset of digital twins from the plurality of digital twins that is similar to the target patient; using the selected subset of digital twins for the target patient to predict a target patient response to multiple hypothetical deliveries of the TCT product using different doses and T cell phenotype compositions, and generating a treatment plan for the target patient based on the predicted response provided by the subset of digital twins, wherein the treatment plan is predicted to provide a level of T cell persistence over time in at least one of the plurality of physiological compartments of the target patient that is above a predetermined threshold, wherein the treatment plan includes a dose of treatment and a composition of the plurality of T cell phenotypes.

[0006] In some variations, the method further comprises identifying a source of variability across patients by visualizing the parameter space for a second subset of the plurality of digital twins; and modifying the model to account for the source of variability across patients.

[0007] In some variations, the method further comprises generating, by the model, a plurality of simulations across different T cell phenotype compositions and dose levels; and visualizing the distribution over time of each of the plurality of T cell phenotypes associated with the plurality of simulations, wherein the visualization indicates an effect of the T cell phenotype composition on the level of T cell persistence over time.

[0008] In some variations, the set of patient characteristics includes patient biometrics, patient medical history, and baseline biomarker data.

[0009] In some variations, each of the plurality of digital twins further represents a response to a dose of a composition of the TCT product within a plurality of compartments.

[0010] In some variations, each of the multiple digital twins includes a set of cell kinetic parameters that includes at least one of the number, proliferation rate, transport rate, apoptosis rate, and differentiation rate of the multiple T cell phenotypes within at least one of the multiple physiological compartments of the corresponding patient over a period of time.

[0011] In some variations, the composition of the TCT product includes an initial number of each of the plurality of T cell phenotypes.

[0012] In some variations, the plurality of T cell phenotypes includes at least two of stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells.

[0013] In some variations, the plurality of physiological compartments includes a peripheral tissue and lymph node compartment, a blood compartment, a tumor draining lymph node compartment, and a tumor compartment.

[0014] Methods, systems and products for quantitative system pharmacology models of T cells are provided, including computer program products. In one aspect, a method is provided. The method may include determining a set of cell dynamics parameters after delivery of a T cell target (TCT) product by at least one data processor, and the cell dynamics parameter set corresponds to a plurality of T cell phenotypes in the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment), blood compartment, tumor draining lymph node compartment and tumor compartment of the patient. The plurality of T cell phenotypes include stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells and endogenous T cells. The method may include determining a first transport rate between the peripheral tissue and lymph node compartment and the blood compartment of the plurality of T cell phenotypes of the patient by the at least one data processor. The method may include determining a set of cell dynamics parameters corresponding to the plurality of T cell phenotypes in the blood compartment by the at least one data processor. The method may include determining a second transport rate between the blood compartment and the tumor draining lymph node compartment of the plurality of T cell phenotypes of the patient by the at least one data processor. The method may include determining, by the at least one data processor, a set of cell dynamic parameters corresponding to the multiple T cell phenotypes within the tumor draining lymph node compartment. The method may include determining, by the at least one data processor, a third transport rate of the effector memory T cells and the effector T cells from the blood compartment of the patient to the tumor compartment. The method may include determining, by the at least one data processor, a set of cell dynamic parameters corresponding to the effector memory T cells and the effector T cells within the tumor compartment. The method may include determining, by the at least one data processor and based on the set of cell dynamic parameters, the first transport rate, the second transport rate, and the third transport rate, each of the multiple T cell phenotypes in the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment), the blood compartment, the tumor draining lymph node compartment, and the tumor compartment over time.

[0015] In some variations, one or more features disclosed herein including the following features may optionally be included in any feasible combination. In some variations, the method includes determining, by the at least one data processor, a differentiation rate parameter corresponding to a differentiation rate of the effector memory T cells to the effector T cells within the tumor compartment. Determining the distribution is further based on the differentiation rate parameter.

[0016] In some variations, the method includes determining, by the at least one data processor, a differentiation rate parameter corresponding to the differentiation of the stem cell-like memory T cell to the central memory T cell, the central memory T cell to the effector memory T cell, and the effector memory T cell to the effector T cell in the tumor-draining lymph node compartment. Determining the distribution is further based on the differentiation rate parameter.

[0017] In some variations, the method comprises determining a T cell therapy for treating a tumor, wherein the T cell therapy comprises a dose of the composition of the plurality of T cell phenotypes of the TCT product.

[0018] In some variations, the T cell therapy is at least one of a T cell receptor (TCR) engineered T cell therapy, an autologous T cell therapy, an allogeneic T cell therapy, an iPSC-derived T cell therapy, and a CAR T cell therapy.

[0019] In some variations, the TCT product includes doses of the composition of the plurality of T cell phenotypes.

[0020] In some variations, the tumor compartment comprises a tumor site of a tumor.

[0021] In some variations, the profile includes: a first profile corresponding to a dose of the composition of the TCT product administered to a first patient and a second profile corresponding to a dose of the composition of the TCT product administered to a second patient. The method includes: determining, by the at least one data processor and based at least on the first profile and the second profile, a third patient's response to a T cell therapy comprising a dose of the composition of the TCT product.

[0022] In some variations, the method includes determining a treatment plan for the third patient based at least on the third patient's response to the T cell therapy.

[0023] In some variations, the dose of the composition of the TCT product administered to the first patient is administered to the first patient after a lymphodepletion regimen is administered to the first patient. The dose of the composition of the TCT product administered to the second patient is administered to the second patient after another lymphodepletion regimen is administered.

[0024] In some variations, the set of cytokinetic parameters includes at least one of number, proliferation rate, apoptosis rate, and differentiation rate of the plurality of T cell phenotypes.

[0025] In one aspect, a method includes determining, by at least one data processor, a first patient profile representing a first response to a first dose of a first composition of a T cell target (TCT) product within multiple physiological compartments of a first patient. The method includes determining, by the at least one data processor, a second patient profile representing a second response to a second dose of a second composition of the TCT product within the multiple physiological compartments of a second patient. The method includes generating, by the at least one data processor and based at least on the first patient profile and the second patient profile, an output indicating a pharmacokinetic-pharmacodynamic (PKPD) relationship for the TCT product. The method includes determining, by the at least one data processor and based at least on the output, a third patient's response to a third dose of a T cell therapy comprising a third composition of the TCT product. The method includes determining a treatment plan for the third patient based at least on the third patient's response to the third dose of the T cell therapy comprising the third composition of the TCT product.

[0026] In some variations, the first patient data includes a first set of cell kinetic parameters, the first set of cell kinetic parameters including at least one of a first number, a first proliferation rate, a first transport rate, a first apoptosis rate, and a first differentiation rate of a plurality of T cell phenotypes in at least one of the plurality of physiological compartments of the first patient over a period of time. The second patient data includes a second set of cell kinetic parameters, the second set of cell kinetic parameters including at least one of a second number, a second proliferation rate, a second transport rate, a second apoptosis rate, and a second differentiation rate of the plurality of T cell phenotypes in at least one of the plurality of physiological compartments of the second patient over the period of time.

[0027] In some variations, the first composition includes a first quantity of each of a plurality of T cell phenotypes. The second composition includes a second quantity of each of the plurality of T cell phenotypes.

[0028] In some variations, the first patient data and the second patient data each include responses corresponding to each of the plurality of T cell phenotypes.

[0029] In some variations, the plurality of T cell phenotypes includes at least two of stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells.

[0030] In some variations, the plurality of physiological compartments includes a peripheral tissue and lymph node compartment, a blood compartment, a tumor draining lymph node compartment, and a tumor compartment.

[0031] In some variations, the T cell therapy is at least one of a T cell receptor (TCR) engineered T cell therapy, an autologous T cell therapy, an allogeneic T cell therapy, an iPSC-derived T cell therapy, and a CAR T cell therapy.

[0032] In some variations, a linear increase from the first dose to the second dose causes a non-linear change between the first response and the second response.

[0033] In some variations, the first composition is different from the second composition. The first dosage is different from the second dosage.

[0034] In some variations, determining the response to the T cell therapy in the third patient comprises: simulating multiple responses to multiple doses of multiple compositions of the TCT product in multiple simulated patients based at least on the output. The response in the third patient is one of multiple simulated responses.

[0035] In some variations, the first patient data is determined after a first lymphodepleting regimen is administered to the first patient. The second patient data is determined after a second lymphodepleting regimen is administered to the second patient.

[0036] In some variations, the response in the third patient is determined by varying at least one of the first dose, the first composition, the second dose, and the second composition.

[0037] In some variations, simulating the plurality of responses is based on applying a lymphodepletion regimen to the plurality of simulated patients.

[0038] On the one hand, a system is provided. The system may include at least one processor and at least one memory. The at least one memory may store instructions that cause operations when executed by the at least one processor. The operation may include: determining a set of cell dynamics parameters after delivery of a T cell target (TCT) product, the set of cell dynamics parameters corresponding to a plurality of T cell phenotypes in the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) of the patient. The plurality of T cell phenotypes include stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells. The operation may include determining a first transport rate between the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) and the blood compartment of the patient of the plurality of T cell phenotypes. The operation may include determining a set of cell dynamics parameters corresponding to the plurality of T cell phenotypes in the blood compartment. The operation may include determining a second transport rate between the blood compartment and the tumor-draining lymph node compartment of the plurality of T cell phenotypes in the patient. The operation may include determining a set of cell dynamics parameters corresponding to the plurality of T cell phenotypes in the tumor-draining lymph node compartment. The operation may include determining a third transport rate of the effector memory T cells and the effector T cells from the blood compartment to the tumor compartment of the patient. The operation may include determining a set of cell kinetic parameters corresponding to the effector memory T cells and the effector T cells in the tumor compartment. The operation may include determining, by at least one data processor and based on the set of cell kinetic parameters, the first transport rate, the second transport rate, and the third transport rate, the distribution over time of each of the multiple T cell phenotypes in the peripheral tissue and lymph node compartment, the blood compartment, the tumor draining lymph node compartment, and the tumor compartment.

[0039] On the other hand, a system is provided. The system may include at least one processor and at least one memory. The at least one memory may store instructions that cause operations when executed by the at least one processor. The operation may include determining a first patient profile by at least one data processor, the first patient profile representing a first response to a first dose of a first composition of a T cell target (TCT) product in multiple physiological compartments of a first patient. The operation includes determining a second patient profile by the at least one data processor, the second patient profile representing a second response to a second dose of a second composition of the TCT product in multiple physiological compartments of a second patient. The operation includes generating an output indicating a pharmacokinetic-pharmacodynamic (PKPD) relationship for the TCT product by the at least one data processor and at least based on the first patient profile and the second patient profile. The operation includes determining a third patient's response to a third dose of a T cell therapy comprising a third composition of the TCT product by the at least one data processor and at least based on the output. The method includes determining a treatment plan for the third patient based at least on the third patient's response to the third dose of the T cell therapy comprising the third composition of the TCT product.

[0040] On the other hand, a computer program product including a non-transitory computer-readable storage medium is provided. The non-transitory computer-readable storage medium may include a program code that causes an operation when executed by at least one processor. The operation may include: determining a set of cell dynamics parameters after delivering a T cell target (TCT) product, the set of cell dynamics parameters corresponding to a plurality of T cell phenotypes in the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) of the patient. The plurality of T cell phenotypes include stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells. The operation may include determining a first transport rate between the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) and the blood compartment of the patient of a plurality of T cell phenotypes. The operation may include determining a set of cell dynamics parameters corresponding to the plurality of T cell phenotypes in the blood compartment. The operation may include determining a second transport rate between the plurality of T cell phenotypes in the blood compartment and the tumor-draining lymph node compartment of the patient. The operation may include determining a set of cell dynamics parameters corresponding to the plurality of T cell phenotypes in the tumor-draining lymph node compartment. The operation may include determining a third transport rate of the effector memory T cells and the effector T cells from the blood compartment to the tumor compartment of the patient. The operation may include determining a set of cell kinetic parameters corresponding to the effector memory T cells and the effector T cells in the tumor compartment. The operation may include determining, by at least one data processor and based on the set of cell kinetic parameters, the first transport rate, the second transport rate, and the third transport rate, the distribution over time of each of the multiple T cell phenotypes in the peripheral tissue and lymph node compartment, the blood compartment, the tumor draining lymph node compartment, and the tumor compartment.

[0041] On the other hand, a computer program product including a non-transitory computer-readable storage medium is provided. The non-transitory computer-readable storage medium may include a program code that causes an operation when executed by at least one processor. The operation may include determining a first patient data by at least one data processor, the first patient data representing a first response to a first dose of a first composition of a T cell target (TCT) product in multiple physiological compartments of a first patient. The operation includes determining a second patient data by the at least one data processor, the second patient data representing a second response to a second dose of a second composition of the TCT product in multiple physiological compartments of a second patient. The operation includes generating an output indicating a pharmacokinetic-pharmacodynamic (PKPD) relationship for the TCT product by the at least one data processor and at least based on the first patient data and the second patient data. The operation includes determining a third patient's response to a third dose of a T cell therapy including a third composition of the TCT product by the at least one data processor and at least based on the output. The method includes determining a treatment plan for the third patient based at least on the third patient's response to the third dose of the T cell therapy including the third composition of the TCT product.

[0042] Specific implementations of the current subject matter may include methods consistent with the description provided herein and articles including tangibly embodied machine-readable media, which are operable to cause one or more machines (e.g., computers, etc.) to cause operations that implement one or more of the features described. Similarly, a computer system that may include one or more processors and one or more memories coupled to the one or more processors is also described. A memory that may include a non-transitory computer-readable or machine-readable storage medium may include, encode, store, etc., one or more programs that cause one or more processors to perform one or more of the operations described herein. A computer-implemented method consistent with one or more implementations of the current subject matter may be implemented by one or more processors present in a single computing system or multiple computing systems. Such multiple computing systems may be connected and may exchange data and / or commands or other instructions, etc., via one or more connections, including, for example, via a network (e.g., the Internet, a wireless wide area network, a local area network, a wide area network, a wired network, etc.) via a direct connection between one or more computing systems in the multiple computing systems, etc.

[0043] The details of one or more variations of the subject matter described herein are set forth in the accompanying drawings and the following description. Other features and advantages of the subject matter described herein will become apparent with reference to the description and drawings, as well as the claims. Although certain features of the subject matter disclosed herein are described for illustrative purposes related to the quantitative system pharmacology model of T cells, it should be readily understood that such features are not intended to be limiting. The claims following this disclosure are intended to define the scope of the subject matter protected. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings, which are incorporated into and constitute a part of this specification, illustrate certain aspects of the subject matter disclosed herein and, together with the description, help explain some of the principles associated with the disclosed implementations.

[0045] Figure 1 An exemplary T cell phenotyping system consistent with implementations of the current subject matter is depicted;

[0046] Figure 2 An exemplary architecture of a QSP model consistent with implementations of the current subject matter is depicted;

[0047] Figure 3 depicts an exemplary graph showing cell counts over time consistent with implementations of the current subject matter;

[0048] Figure 4 depicts an exemplary diagram showing a set of cell dynamics parameters for T cell phenotypes consistent with implementations of the current subject matter, e.g., a single simulation illustrating a model that captures the known dynamics of T cell therapy across time;

[0049] Figure 5 An exemplary comparison of cellular pharmacokinetics in blood consistent with implementations of the current subject matter is depicted;

[0050] Fig. 6A and 6B An exemplary comparison of cellular pharmacokinetics in blood across various dose groups and patients consistent with implementations of the current subject matter is depicted;

[0051] Figure 7 depicts exemplary graphs showing phenotypic composition of T cell-based products consistent with implementations of the current subject matter and changes in cellular pharmacokinetics across dose levels;

[0052] Figure 8 depicts exemplary graphs showing phenotypic composition of T cell-based products consistent with implementations of the current subject matter and changes in cellular pharmacokinetics across dose levels;

[0053] Fig. 9depicts a flow chart showing an example of a process consistent with an implementation of the current subject matter for generating a distribution of T cell phenotypes;

[0054] Fig.10 depicts a flow chart showing an example of a process consistent with implementations of the current subject matter for determining patient response to T cell therapy;

[0055] Fig.11 A block diagram illustrating an example of a computing system consistent with implementations of the current subject matter is depicted.

[0056] Fig.12 Figure 2. Example parameter space ridge plots and principal component analysis of digital twins revealing sources of variability between patients and across dose groups (A to O).

[0057] Fig.13 Depicts exemplary biological variability (patient-specific) impacts on cellular kinetics of TCR-engineered T cell therapies, resulting in persister or non-persister outcomes.

[0058] Fig.14 Depicted are exemplary graphs showing that dose composition affects cellular kinetics of TCR-engineered T cells.

[0059] Fig.15 Depicts an example predictive simulation of a digital twin showing agreement with observed data in a patient with available clinical data.

[0060] Fig.16 A flow chart showing an example of a process 1600 for generating a personalized treatment plan using a model and a digital twin, consistent with an implementation of the current subject matter, is depicted.

[0061] Wherever practical, like labels are used to refer to the same or similar items in the drawings. DETAILED DESCRIPTION

[0062] As noted, there are many challenges in understanding and characterizing the unique cell dynamics of T cell therapy, such as when T cell therapy is not a conventional therapy, because these therapies include live and phenotypically diverse T cells, the pharmacological properties of these therapies are different from molecular therapies, and the pharmacokinetics and resulting pharmacodynamics of these therapies are monitored and predicted, which also brings different challenges. In such cases, conventional pharmacokinetic-pharmacodynamic (PKPD) methods may not be sufficient. Further, it is not clear how the cell phenotype and composition of the infusion product can affect cell pharmacokinetics (e.g., cell dynamics) and subsequent efficacy and safety. For example, there is currently a lack of sufficient animal models and preclinical data to understand the conversion of T cell therapy (such as engineered TCR T cell therapy). In addition, the cell dynamics of such T cell therapies may be significantly affected by the administered dose, the variability of the composition, and the variability of patient characteristics and / or composition.

[0063] To address these challenges, the framework of the quantitative systems pharmacology (QSP) model provided herein (consistent with the implementation of the current subject matter) takes into account the variability of the dose and / or composition of T cell therapy and the variability of the patient to accurately predict the distribution of multiple T cell phenotypes in various physiological compartments and / or patient responses over time. For example, the QSP model described herein can track the expression of TCR engineered T cells (such as stem cell-like memory T cells (T scm ), central memory T cells (T cm ), effector memory T cells (T em ), effector T cells (T eff ) and / or endogenous T cells) across physiological compartments (including tissue and lymph node compartments, blood compartments, tumor draining lymph node compartments, and tumor compartments). Each subset undergoes steady-state proliferation, antigen-driven proliferation and differentiation, apoptosis, margination, and transport between and / or within physiological compartments. In addition, as described herein, the architecture of the QSP model captures effector memory T cells and effector T cell populations that infiltrate the tumor compartment, where effector T cells can kill tumor cells of tumors within the tumor compartment. The architecture of the QSP model described herein can also be used to treat T cell death by combining lymphocyte depletion modulation therapy and endogenous T cells (T endo ) to capture the potential competition between immune cells after infusion of TCR engineered T cells. As described herein, the QSP model is calibrated to the Phase I clinical trial data of TCR engineered T cells targeting E7 in patients with epithelial cancer. The QSP model described herein may be applicable to other cancer antigens and tumor types. Therefore, the QSP model described herein accurately predicts the effects of different doses and T cell phenotype ratios (e.g., composition) of various T cell therapies.

[0064] Figure 1 To depict a block diagram of an example system 100, the example system includes a client-server architecture and network configured to perform various methods described herein. A platform (e.g., machines and software, possibly interoperating via a series of network connections, protocols, application-level interfaces, etc.) provides server-side functionality to one or more client nodes 102 and / or 106 in the form of a server platform 120 via a communication network 114 (e.g., the Internet or other type of wide area network (WAN) such as a wireless network or a dedicated network with additional security suitable for the tasks performed by the user).

[0065] The client node (e.g., client node 102 and / or client node 106) can be, for example, a user device (e.g., a mobile electronic device, a fixed electronic device, etc.). The client node can be associated with a user and / or accessible to the user. In another example, the client node can be a computing device (e.g., a server) accessible to and / or associated with an individual or entity. The node can include a network module (e.g., a network adapter) configured to transmit and / or receive data. Via the nodes in the computer network, multiple users and / or servers can communicate and exchange data. In some embodiments, the client node can facilitate the transmission of patient data to the platform 120 for further processing.

[0066] like Figure 1 As shown, the client node 102 hosts the web extension 104, thereby allowing the user to access the functionality provided by the server platform 120, such as receiving visualizations of one or more treatment plans in the treatment plan from the server platform 120. The web extension 104 can be compatible with any web browser application used by the user of the client node. Further, Figure 1 Another client node 106 is shown, for example, hosting a mobile application 108, thereby allowing a user to access functionality provided by the server platform 120, such as receiving visualizations of one or more of the treatment plans from the server platform 120. Delivery of the visualizations may be via wired or wireless communication modes.

[0067] In at least some examples, server platform 120 may be one or more computing devices or systems, storage devices, and other components (including Figure 1 These modules may include, for example, a model generation engine 122, a digital twin generation engine 124, a matching engine 126, a treatment plan generation engine 128, a data access module 142, an analysis engine 110, and a data storage device 150. Each of these modules is described in more detail below.

[0068] The model generation engine 122 can generate a model that can capture the biomedical mechanism of the TCT product after being injected into the patient's body. In some embodiments, the model generation engine 122 can collaborate with the analysis engine 110 to generate a model. The model generation operation performed by the analysis engine 110 and the model generation engine 122 is further described in detail elsewhere herein. The digital twin generation engine 124 can facilitate the generation of a digital twin set. In some embodiments, the digital twin generation engine 124 can collaborate with the analysis engine 110 to generate a digital twin set. In some embodiments, the digital twin can be designed in such a way that each patient from a clinical trial is matched with a specific digital twin subset by, for example, a matching engine 126. For example, using model 1202, the digital twin generation engine 124 can generate hundreds, thousands or more digital twins. In some embodiments, each of the multiple digital twins can represent the distribution of multiple T cell phenotypes in multiple physiological compartments associated with corresponding patients over time, wherein the corresponding patient can be associated with a patient feature set. The matching engine 126 can receive a target patient feature set from one or more of the client nodes 102 or 106, and can match the target patient features with patient features associated with the digital twins to select a subset of digital twins similar to the target patient from the plurality of digital twins. The treatment plan generation engine 128 can collaborate with other engines / modules of the server platform 120 to generate a treatment plan for one or more target patients. The treatment plan generation operations performed by the treatment plan generation engine 128 are further described in detail elsewhere herein.

[0069] The data access module 142 may facilitate access to the data storage 150 of the server platform 120 by any of the remaining modules / engines 110, 122, 124, 126, and 128 of the server platform 120. In one example, one or more of the data access modules 142 may be a database access module, or any type of data access module capable of storing data to and / or retrieving data from the data storage 150 as needed by the particular module 110, 122, 124, 126, and 128 employing the data access module 142 to access the data storage 150. Examples of the data storage 150 include, but are not limited to, one or more data storage components, such as magnetic disk drives, optical disk drives, solid state disk (SSD) drives, and other forms of non-volatile and volatile memory components.

[0070] The data storage device 150 can store input clinical data and / or one or more determinations / models / digital twins made and / or generated by the remaining modules / engines 110, 122, 124, 126, and 128 of the server platform 120. The data storage device 150 may include a graph database, a time series database, a relational database, or a combination of these to efficiently manage and organize large amounts of data related to T cell therapy. The data covers various aspects of patient response to TCT products, including a comprehensive collection of patient data / digital twins representing the distribution of T cell phenotypes over time in different physiological compartments.

[0071] In some embodiments, a graph database can be utilized because it has the inherent ability to efficiently model and manage the complex relationships within complex data sets. For example, as discussed elsewhere herein, T cell therapy research may involve a complex network of relationships between various data entities, including different T cell phenotypes, the distribution and / or concentration of different T cell phenotypes in different physiological compartments, and cell dynamics parameters. A graph database may be able to represent this interconnected data structure. In some embodiments, a graph database can be utilized by: using nodes and edges to represent entities and their relationships, and thus modeling the interconnectivity of different components in the field of T cell therapy. For example, a node can represent a T cell phenotype, a compartment, and a parameter, and an edge can depict the relationship and interaction between them. Additionally or alternatively, a graph database can provide flexibility in querying and traversing relationships. For example, a graph database can visualize the following: how different T cell phenotypes affect each other, how different T cell phenotypes move between compartments, and how changes in cell dynamics parameters affect distribution over time. This can provide a deep understanding of the dynamics of T cell responses.

[0072] In some embodiments, a time series database may be used to handle data that evolves over time, for example, the distribution of multiple T cell phenotypes over time in multiple physiological compartments. A time series database can capture the temporal and dynamic aspects of patient responses, enabling visibility into changes in T cell phenotypes within physiological compartments as therapy progresses. For example, a time series database can be used to store and query time-stamped data points, which can facilitate tracking of cell dynamics parameters over time. In some embodiments, a relational database can be used to store structured clinical data and metadata related to T cell therapy, facilitating the linking of patient data with specific treatment regimens, laboratory results, and patient demographics. A relational database can improve data integrity and enable direct retrieval of specific pieces of information.

[0073] Figure 2An example architecture 200 of a QSP model 1202 consistent with the implementation of the current theme is schematically depicted. As noted, consistent with the implementation of the current theme, the QSP model 1202 describes the in vivo dynamics of lymphocyte (e.g., T cell) proliferation after treatment with a T cell target (TCT) product. In other words, the architecture 200 of the QSP model 1202 can represent the pharmacokinetic-pharmacodynamic relationship of a TCT product. TCT products can additionally and / or alternatively increase the anti-tumor activity in both solid tumors and hematological cancers. For example, TCT products can be delivered to patients as part of a T cell therapy for treating tumors, cancers, etc. T cell therapy can be at least one of T cell receptor (TCR) engineered T cell therapy, autologous T cell therapy, allogeneic T cell therapy, induced pluripotent stem cells ("iPSC") derived T cell therapy, and chimeric antigen receptor ("CAR") T cell therapy.

[0074] As described herein, a TCT product can include multiple T cell phenotypes that proliferate in various physiological compartments of a patient. For example, a TCT product can include a dose of a composition of multiple T cell phenotypes. Figure 2 , various T cell phenotypes including stem cell-like memory T cells (T scm )210, Central memory T cells (T cm )212, effector memory T cells (T em )214, effector T cells (T eff ) 216 and / or endogenous T cells 219. Specific T cell phenotypes among the multiple T cell phenotypes included in the architecture 200 are selected to improve the accuracy of predictions generated based on the QSP model 1202, to more accurately track T cell phenotype behaviors within and between multiple physiological compartments, and to more accurately determine the distribution of T cell phenotypes. Therefore, in some implementations, the multiple T cell phenotypes included in the QSP model 1202 are only stem cell-like memory T cells 210, central memory T cells 212, effector memory T cells 214, effector T cells 216 and / or endogenous T cells 219 and / or all of these cells. In other implementations, the multiple T cell phenotypes included in the QSP model 1202 are only stem cell-like memory T cells 210, central memory T cells 212, effector memory T cells 214, and effector T cells 216 and / or all of these cells.

[0075] In some embodiments, the TCT product includes a live drug. In other words, the TCT product includes live and phenotypically diverse T cells. Therefore, the actual composition and / or proportion of each of the multiple T cell phenotypes that make up the TCT product can vary from patient to patient and / or between doses delivered to the same patient. The TCT product delivered to each patient can have a dose that includes a certain number of T cells. The dose of the TCT product can vary based on the patient. In addition and / or alternatively, the dose of the TCT product can be delivered to the patient multiple times as part of a treatment regimen. Each dose delivered to the patient can be the same and / or vary. In other words, each dose can have a dose level that includes the same number of T cells or a different number of T cells. For example, a dose can include approximately 10 9 T cells, 10x 10 9 T cells, 100x 10 9 Thus, there may be a ten-fold difference in the number of T cells included in each dose level corresponding to each dose.

[0076] The TCT product can be delivered to the patient after a lymphodepletion regimen has been administered to the patient. This can help prolong the persistence of the various T cell phenotypes of the TCT product delivered to the patient and improve the effectiveness of the delivered T cell therapy.

[0077] refer to Figure 2 , the architecture 200 of the QSP model 1202 includes multiple physiological compartments. For example, the architecture 200 includes a peripheral tissue and lymph node compartment (i.e., a healthy tissue compartment) 202, a blood compartment 204, a tumor draining lymph node compartment 208, and a tumor compartment 206. The tumor compartment 206 may include a tumor site of a tumor that is configured to be treated by a TCT product. The specific physiological compartments included in the architecture 200 are selected to improve the accuracy of the predictions generated based on the QSP model 1202, to more accurately track T cell behavior within and between multiple physiological compartments, and to more accurately determine the distribution of multiple T cell phenotypes. Therefore, in some implementations, the multiple physiological compartments included in the QSP model 1202 are only the peripheral tissue and lymph node compartment (i.e., a healthy tissue compartment) 202, the blood compartment 204, the tumor draining lymph node compartment 208, and the tumor compartment 206 and / or all of these compartments.

[0078] When the TCT product is delivered to the patient, a variety of T cell phenotypes are transported to each of a plurality of physiological compartments. For example, in some embodiments, at least some or all of a plurality of physiological compartments include a variety of T cell phenotypes. For example, peripheral tissue and lymph node compartments (i.e., healthy tissue compartments) 202, blood compartments 204, tumor draining lymph node compartments 208, and tumor compartments 206 may include stem cell-like memory T cells 210, central memory T cells 212, effector memory T cells 214, effector T cells 216, and / or endogenous T cells 219.

[0079] In some embodiments, each T cell phenotype in a plurality of T cell phenotypes is associated with a cell kinetic parameter set, which describes the behavior of the T cell phenotype after the TCT product is delivered to the patient. The cell kinetic parameter set includes one or more cell kinetic parameters. For example, the cell kinetic parameter set includes at least one of the number, proliferation (e.g., amplification) rate, apoptosis rate, and differentiation rate of a plurality of T cell phenotypes. The value of each cell kinetic parameter in one or more cell kinetic parameters corresponding to each T cell phenotype in a plurality of T cell phenotypes in each physiological compartment can be different from each other.

[0080] As an example, Figure 3 A graph 300 showing T cell counts over time consistent with implementations of the current subject matter is shown. As shown in graph 300, after a TCT product is delivered to a patient, the number of T cells decreases before rapidly expanding. After reaching a maximum number of T cells, the number of T cells decreases slowly over time. As another example, Figure 4 Depicts an exemplary diagram 400 showing the influence of the cell kinetic parameter set for T cell phenotype consistent with the implementation of the current theme over time. In particular, Figure 400 shows the concentration of effector T cells over time after TCT products are delivered to patients. For example, Figure 400 shows T cell behavior, including transport (e.g., edge set), proliferation (e.g., amplification), proliferation peak (e.g., amplification peak), apoptosis and redistribution and persistence of effector T cells in physiological compartments after TCT products are delivered to patients. The cell kinetic parameter set for T cell phenotype affects the behavior of the corresponding T cell phenotype over time, as shown in Figure 400. Although Figure 400 shows the T cell behavior corresponding to effector T cells over time, any T cell phenotype described herein may also experience T cell behavior, including transport (e.g., edge set), proliferation (e.g., amplification), proliferation peak (e.g., amplification peak), apoptosis and redistribution and persistence that can be represented by a cell kinetic parameter set (e.g., a cell kinetic parameter set over time).

[0081] As an example, referring to Figure 400, after delivery of the TCT product, the effector T cells undergo agglomeration 402, followed by rapid proliferation 404, as the concentration of the effector T cells increases before reaching an expansion peak 406. In some implementations, the proliferation rate, the expansion peak 406, etc. can indicate the efficacy of the T cell therapy. After reaching the expansion peak 406, the effector T cells undergo apoptosis and redistribution 408, as the effector T cells are transported between physiological compartments and undergo apoptosis. Finally, the effector T cells undergo persistence 410. Persistence 410 (e.g., the length of persistence, etc.) can additionally and / or alternatively indicate the efficacy of the T cell therapy.

[0082] As noted, the cell dynamics parameter set may include differentiation of multiple T cell phenotypes. In some implementations, such as Figure 2 As shown, after the effector memory T cells 214 have migrated into the tumor compartment 206, at 240, the architecture 200 includes the differentiation of the effector memory T cells 214 into effector T cells 216. The differentiation of the effector memory T cells 214 into effector T cells 216 within the tumor compartment can increase the number of effector T cells used to infiltrate the tumor 218 at 242. Further, the architecture 200 can include the differentiation of multiple T cell phenotypes within the tumor draining lymph node compartment 208. Within the tumor draining lymph node compartment 208, the architecture 200 includes the antigen-driven differentiation of the stem cell-like memory T cells 210 into the central memory T cells 212 at 244, the differentiation of the central memory T cells 212 into the effector memory T cells 214 at 246, and the differentiation of the effector memory T cells 214 into the effector T cells 216 at 248. The activation of various T cell phenotypes within the tumor draining lymph node compartment 208 can be triggered by antigen presenting cells presenting antigens from the tumor thereon. The antigen presenting cells activate various T cell phenotypes within the tumor draining lymph node compartment 208. This allows effector T cells 216 to recognize the tumor 218 and infiltrate the tumor 218.

[0083] Return to reference Figure 2, the framework 200 of the QSP model 1202 can include transport of multiple T cell phenotypes between each of multiple physiological compartments. For example, the framework 200 can include transport of multiple T cell phenotypes between the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 and the blood compartment 204, between the blood compartment 204 and the tumor draining lymph node compartment 208, and between the blood compartment 204 and the tumor compartment 206 (e.g., at transport rates or edge rates). In other words, the architecture 200 can include transport (e.g., at a transport rate or an aggregation rate) of multiple T cell phenotypes from a peripheral tissue and lymph node compartment (i.e., a healthy tissue compartment) 202 to a blood compartment 204, from a blood compartment 204 to a peripheral tissue and lymph node compartment (i.e., a healthy tissue compartment) 202, from a blood compartment 204 to a tumor-draining lymph node compartment 208, from a tumor-draining lymph node compartment to a blood compartment 204, and / or from a blood compartment 204 to a tumor compartment 206.

[0084] In particular, if Figure 2 As shown, the architecture 200 includes the transport of stem cell-like memory T cells 210 between the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 and the blood compartment 204 at 220 and the transport of stem cell-like memory T cells 210 between the blood compartment 204 and the tumor-draining lymph node compartment 208 at 222. The architecture 200 may additionally and / or alternatively include the transport of central memory T cells 212 between the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 and the blood compartment 204 at 224 and the transport of central memory T cells 214 between the blood compartment 204 and the tumor-draining lymph node compartment 208 at 226. The architecture 200 may additionally and / or alternatively include transport of effector memory T cells 214 between the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 and the blood compartment 204 at 228, transport of effector memory T cells 214 between the blood compartment 204 and the tumor draining lymph node compartment 208 at 230, and transport of effector memory T cells 214 between the blood compartment 204 and the tumor compartment 206. The architecture 200 may additionally and / or alternatively include transport of effector T cells 216 between the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 and the blood compartment 204, transport of effector T cells 216 from the blood compartment 204 to the tumor compartment 206, and transport of effector T cells 216 from the tumor draining lymph node compartment 208 to the blood compartment 204 at 232. Figure 2 As shown, the architecture 200 may additionally and / or include transport of endogenous T cells 219 between the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 and the blood compartment 204 and transport of endogenous T cells 219 between the blood compartment 204 and the tumor draining lymph node compartment 208.

[0085] Reference again Figure 2 , architecture 200 includes proliferation (e.g., proliferation rate, such as antigen-driven proliferation rate and / or steady-state proliferation rate) within each of physiological compartments (e.g., peripheral tissue and lymph node compartments (i.e., healthy tissue compartments) 202, blood compartments 204, tumor compartments 206, and tumor-draining lymph node compartments 208) for each of a plurality of T cell phenotypes.

[0086] For example, within the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202, the architecture 200 includes a proliferation rate (e.g., steady-state proliferation rate) 250 corresponding to the stem-like memory T cells 210, a proliferation rate (e.g., steady-state proliferation rate) 252 corresponding to the central memory T cells 212, a proliferation rate (e.g., steady-state proliferation rate) 254 corresponding to the effector memory T cells 214, and a proliferation rate (e.g., steady-state proliferation rate) 256 corresponding to the effector T cells 216. Within the blood compartment, the architecture 200 includes a proliferation rate (e.g., steady-state proliferation rate) 258 corresponding to the stem-like memory T cells 210, a proliferation rate (e.g., steady-state proliferation rate) 260 corresponding to the central memory T cells 212, a proliferation rate (e.g., steady-state proliferation rate) 262 corresponding to the effector memory T cells 214, and a proliferation rate (e.g., steady-state proliferation rate) 264 corresponding to the effector T cells 216. Within the tumor draining lymph node compartment 208, the architecture 200 includes a proliferation rate (e.g., antigen-driven proliferation rate) 266 corresponding to the stem-like memory T cells 210, a proliferation rate (e.g., antigen-driven proliferation rate) 268 corresponding to the central memory T cells 212, a proliferation rate (e.g., antigen-driven proliferation rate) 270 corresponding to the effector memory T cells 214, and a proliferation rate (e.g., steady-state proliferation rate) 272 corresponding to the effector T cells 216. Within the tumor compartment 206, the architecture 200 includes a proliferation rate (e.g., antigen-driven proliferation rate) 274 corresponding to the effector memory T cells 214, and a proliferation rate (e.g., steady-state proliferation rate) 276 corresponding to the stem-like memory T cells 216.

[0087] Reference again Figure 2, the architecture 200 includes apoptosis (e.g., apoptosis rate) in each of the physiological compartments (e.g., the peripheral tissue and lymph node compartment (i.e., the healthy tissue compartment) 202, the blood compartment 204, the tumor compartment 206, and the tumor draining lymph node compartment 208) for each of the multiple T cell phenotypes. For example, within the peripheral tissue and lymph node compartment (i.e., the healthy tissue compartment) 202, the architecture 200 includes an apoptosis rate 278 corresponding to the stem cell-like memory T cells 210, an apoptosis rate 280 corresponding to the central memory T cells 212, an apoptosis rate 282 corresponding to the effector memory T cells 214, and an apoptosis rate 284 corresponding to the effector T cells 216. Within the blood compartment, the architecture 200 includes an apoptosis rate 286 corresponding to the stem cell-like memory T cells 210, an apoptosis rate 288 corresponding to the central memory T cells 212, an apoptosis rate 290 corresponding to the effector memory T cells 214, and an apoptosis rate 292 corresponding to the effector T cells 216. Within the tumor draining lymph node compartment 208, the architecture 200 includes an apoptosis rate 294 corresponding to the stem-like memory T cells 210, an apoptosis rate 296 corresponding to the central memory T cells 212, an apoptosis rate 298 corresponding to the effector memory T cells 214, and an apoptosis rate 299 corresponding to the effector T cells 216. Within the tumor compartment 206, the architecture 200 includes an apoptosis rate 297 corresponding to the effector memory T cells 214 and an apoptosis rate 295 corresponding to the stem-like memory T cells 216.

[0088] As noted above, the QSP model 1202 can be calibrated on clinical data. In some implementations, cell kinetic parameters such as number, proliferation rate, apoptosis rate, differentiation rate, transport rate, etc. as described herein can be determined based at least on clinical data. The calibrated QSP model 1202 is able to reproduce the results of three different dose levels (approximately 10 9 , about 10 10 , about 10 11The QSP model 1202 successfully captures the initial edge set or transport, subsequent proliferation, apoptosis and final persistence and other cell dynamics parameters for each of the T cell phenotypes observed across time for all three dose levels. After calibration, the parameter space is sampled to create a virtual queue that captures the variability of clinical data across each dose level (described in more detail below). In addition, the architecture 200 can be used to determine the relationship between various cell dynamics parameters, such as the persistence of T cells across time and the variability of cell proliferation rate, differentiation rate and transport rate of T cells. Finally, the architecture 200 can provide the effects of the variability of the dose level and phenotypic composition of the initial TCT product on the cell dynamics, distribution and persistence of T cell therapy across time.

[0089] In some implementations, the model generation engine 122 may determine the behavior of a plurality of T cell phenotypes within each of a plurality of physiological compartments (e.g., peripheral tissue and lymph node compartments (i.e., healthy tissue compartments) 202, blood compartments 204, tumor compartments 206, and tumor-draining lymph node compartments 208) at various time points, at various doses (e.g., dose levels) of the TCT product, and / or at various compositions of the TCT product. In some implementations, the model generation engine 122 may determine the distribution of corresponding T cell phenotypes of the plurality of T cell phenotypes in each of the physiological compartments over time, such as based at least on the determined behavior of the plurality of T cell phenotypes within each of the plurality of physiological compartments at various time points and between the plurality of physiological compartments at various time points. In some implementations, the model generation engine 122 can determine the persistence over time of corresponding T cell phenotypes in each of the multiple T cell phenotypes in each of the multiple physiological compartments, such as based at least on the determined behavior of the multiple T cell phenotypes at various time points within each of the multiple physiological compartments and between the multiple physiological compartments at various time points and / or based on the determined distribution.

[0090] For example, the analysis engine 110 or the model generation engine 122 may determine a distribution or patient profile (including a set of cell kinetic parameters, patient responses, etc.) for each patient based on the determined behavior of various T cell phenotypes within and between physiological compartments. Figure 5 An example comparison of cellular pharmacokinetics in blood consistent with implementations of the current subject matter is depicted. In particular, Figure 5A comparison of the behavior of various T cell phenotypes over time (captured at various time points) across three doses and various patients is shown. As noted herein, while the composition of the TCT product delivered to each patient remains constant, the constant composition of the TCT product can have some degree of variation (e.g., within a threshold range), at least because the TCT product can include a live drug. Further, changes in patient type or patient composition may also affect the variability of the results (e.g., Figure 5 However, by varying the dose and incorporating a constant composition, the effect of the dose level on cellular pharmacokinetics can be captured and determined by the analysis engine 110 or the model generation engine 122, such as based on the architecture 200. The architecture 200 can also take into account such changes in the "constant" composition of the TCT product and / or the makeup of the patient.

[0091] For example, in Figure 5 , the first row of graphs corresponds to a low dose of a TCT product delivered to a first patient group comprising two patients. The low dose comprises approximately 10 9 The second row of graphs corresponds to the median dose of the TCT product delivered to the second patient group of three patients. The median dose includes approximately 10 x 10 9 The third row of graphs corresponds to a high dose of the TCT product delivered to a third patient group of five patients. The high dose includes approximately 100 x 10 9 The composition of each T cell. Therefore, various dose levels (low dose, medium dose, high dose, etc.) correspond to the number of T cells and / or T cell concentration in the TCT product delivered to the patient. Each line in each figure corresponds to a different patient. As noted, the variability in the behavior of each of the T cells in response to the delivery of the TCT product at each dose level for each patient is at least partially due to the variability of the constant composition of the TCT product and the specific composition of the patient. Figure 5 , the first column corresponds to the behavior of stem cell-like memory T cells 210 , the second column corresponds to the behavior of central memory T cells 212 , the third column corresponds to the behavior of effector memory T cells 214 , and the fourth column corresponds to the behavior of effector T cells 216 .

[0092] The distribution at each dose level for each patient in the patient can allow the determination of the key drivers of the cell population value across time, which can then be used by the analysis engine 110 or the model generation engine 122 and / or as part of the architecture 200 to accurately determine the dosage and / or composition of the TCT product used to treat a specific patient as a part of a T cell therapy (such as T cell receptor (TCR) engineered T cell therapy, autologous T cell therapy, allogeneic T cell therapy, iPSC-derived T cell therapy, CART cell therapy, or an engineered T cell therapy with TCR targeting various tumor antigens). For example, the distribution and / or patient data determined herein can allow the determination of the effect of the variability of one or more of the cell kinetic parameter sets and / or transport described herein on the effect of effector T cell counts. This can then allow the determination of the efficacy of a specific treatment. In some implementations, cell kinetic parameters (such as proliferation and transport rates of stem cell-like memory T cells 210 and / or transport, proliferation, and differentiation rates of effector memory T cells 214) can be key drivers for accurately determining and capturing patient data (such as those with high levels of cell proliferation or decline). Therefore, based at least on the architecture 200, the analysis engine 110 or the model generation engine 122 can predict the blood and tissue distribution and pharmacokinetics of T cells (e.g., multiple T cell phenotypes) after administration of T cell therapy in patients with solid tumors. The analysis engine 110 or the model generation engine 122 can also utilize the architecture 200 to capture the sensed pharmacokinetics and efficacy of T cells (e.g., multiple T cell phenotypes) in preclinical animal species (such as mice) to assist in the translation of the T cell therapy described herein.

[0093] In some embodiments, as described elsewhere herein Figures 2 to 5As described, a QSP model 1202 can be developed to simulate the dynamic data of various T cell phenotypes in the context of cell therapy treatment. In some embodiments, the QSP model 1202 can be carefully designed to be consistent with the relevant theoretical basis of biological processes. In this way, the QSP model 1202 can encapsulate the basic biomedical mechanisms that control T cell biology (covering both the initial T cells in the patient and the TCR engineered T cells introduced as part of the therapy) in physiologically meaningful compartments. These compartments include key areas within the patient's body, including central blood compartments, healthy tissues, tumor compartments, and tumor-draining lymph node compartments. The model 1202 can integrate the basic biological mechanisms inherent to T cell behavior, such as antigen-induced proliferation and differentiation, cell migration and transport, and key processes of homeostatic proliferation and apoptosis. The QSP model 1202 can be formulated using a system of ordinary differential equations (ODEs), which serves as a mathematical basis for describing and predicting the in vivo dynamics of different T cell phenotypes. In some embodiments, in addition to endogenous T cells (T cells) naturally present in the patient's system throughout the treatment regimen, endo In addition to T cells, these phenotypes can also include stem cell-like memory T cells (T cells) derived from TCR-engineered T cell therapies. scm ), central memory T cells (T cm ), effector memory T cells (Tem) and effector T cells (T eff ).

[0094] In some embodiments, the digital twin generation engine 124 can generate digital twins with the help of the QSP model 1202. For example, by utilizing the QSP model 1202, the digital twin generation engine 124 can generate a patient-specific digital twin that replicates the circulating T cell dynamics recorded in a clinical trial involving TCR-engineered T cells against E7 in patients with metastatic HPV-associated epithelial cancer. In some embodiments, analysis of key factors affecting cell dynamics and differences between these digital twins can provide the following conclusions: 1. Stem cell-like memory T cells (T scm ) can be a key determinant of T cell expansion and persistence; 2. scm The associated differences may play an important role in accounting for the observed changes in cellular dynamics between patients. The details of the analysis using digital twins are further described elsewhere herein. By using digital twins to conduct virtual clinical trials via computational modeling, it is expected to increase T in the administered product. scmThe presence of will enhance the durability of engineered T cells, potentially enabling the use of lower dose levels. Additionally or alternatively, the present disclosure validates the QSP model 1202, digital twins, and T cells by estimating the kinetics of two patients with pancreatic cancer receiving T cell therapy targeting KRAS G12D. scm Enriched meaning is related to the broader applicability of the in-depth understanding.

[0095] In some embodiments, the QSP model 1202 characterizes the cellular dynamics of T cells (covering five different T cell phenotypes) in the bloodstream, various tissues, tumor-draining lymph nodes, and tumor environments. In some embodiments, the model is calibrated using clinical data from TCR-engineered T cells targeting E7 to generate a "reference virtual patient". Subsequently, in some embodiments, using the same data set, the digital twin generation engine 124 can facilitate the generation of a digital twin collection, where each patient from a clinical trial is matched to a specific digital twin subset by, for example, a matching engine 126. For example, using model 1202, the digital twin generation engine 124 can generate hundreds, thousands, or more digital twins. In some embodiments, each of the multiple digital twins can represent the distribution of multiple T cell phenotypes in multiple physiological compartments associated with a corresponding patient over time, where the corresponding patient can be associated with a patient feature set. In some embodiments, the matching engine 126 can receive a target patient feature set, and can match the target patient features with the patient features associated with the digital twin to select a digital twin subset similar to the target patient from the multiple digital twins. In some embodiments, the matching engine 126 can calculate a matching score for each digital twin in the set for the target patient. The matching engine 126 can then select the top N digital twins with the highest scores. This can facilitate selecting a subset of digital twins from the set of digital twins that are similar to the target patient among the available digital twins.

[0096] In some embodiments, a patient feature set (including patient biometrics, patient medical history, and baseline biomarker data) can be used as a basis for calculating a matching score. This matching score is used to assess and determine the similarity between the target patient and the digital twin set. The matching score calculation can include different methods, such as biometric-driven scoring, medical history-driven scoring, biomarker data-driven scoring, integrated scoring that takes into account all features, weighted scoring based on feature correlation, dynamic adjustment of scoring criteria, and threshold-based scoring. These examples provide a general method to facilitate the selection of a digital twin that is closely aligned with the characteristics of the target patient across a range of clinical scenarios and can enable predictive simulations of patient responses.

[0097] In some embodiments, for each target patient, the matching engine 126 can select several digital twin virtual patients, for example, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, etc. In some embodiments, each of these selected digital twins can be carefully crafted to represent a unique combination of potential biological parameter values, with the goal of closely replicating the observed patient dynamics when receiving the corresponding clinically administered dose amount and composition. In view of the inherent uncertainty of biological parameters, the use of multiple digital twins for each patient can provide many benefits, for example, this can enable explicit consideration of alternative parameterizations of the underlying biology (which can be consistent with the observed data). Additionally or alternatively, this approach can accommodate scenarios where divergent predictions may occur under untested regimens, different dosing regimens, or changes in the cellular composition of the administered product.

[0098] These digital twins have been shown to be useful, for example, for making predictions about alternative dosing strategies through the treatment plan generation engine 128. In some embodiments, predictions related to the response of the target patient to a large number of hypothetical deliveries of TCT products are made using a subset of digital twins selected for the target patient. These hypothetical deliveries cover a range of different dose quantities and T cell phenotype compositions. Through this approach, a comprehensive and in-depth understanding of potential responses and outcomes can be generated, which helps to make informed decisions about treatment strategies. In addition, in some embodiments, parameter analysis has been performed to obtain an in-depth understanding of biological mechanisms that may differ between individual patients or drive sustained cell persistence over time. In some embodiments, simulations involving alternative dosing strategies have been performed on digital twins to achieve predictions related to the effects of dose composition and dose quantity on cell dynamics. In some embodiments, predictions generated by digital twins have been verified by unique TCR engineered T cell therapy scenarios (specifically targeting KRAS G12D in patients diagnosed with pancreatic cancer).

[0099] Fig. 6A and 6B Depicted are example comparisons of cellular pharmacokinetics in blood across various dose groups and patients consistent with implementations of the current subject matter. In particular, Fig. 6A and 6B Individualized virtual patients are shown that characterize the variability across dose level groups, TCT treatment components, and patients. Fig. 6A and 6BIn, each row corresponds to a different patient, and the graph in each column corresponds to a specific T cell phenotype. For example, the graph in the first column corresponds to stem cell-like memory T cells 210, the graph in the second column corresponds to central memory T cells 212, the graph in the third column corresponds to effector memory T cells 214, the graph in the fourth column corresponds to effector T cells 216, and the graph in the fifth column corresponds to endogenous T cells. Each line in each figure corresponds to a different dose (e.g., the number and / or concentration of T cells) of the TCT product delivered to each patient and a composition (e.g., a specific ratio or proportion of various T cell phenotypes (such as stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells)) of the TCT product delivered to each patient.

[0100] Based on the architecture 200 of the QSP model 1202, the analysis engine 110 or the model generation engine 122 can determine various distributions (e.g., graphs representing patient data corresponding to each of a plurality of T cell phenotypes), such as Fig. 6A and 6B As shown. Based on the determined distribution (e.g., patient data), the analysis engine 110 or the model generation engine 122 can predict the dynamics of the T cell subsets or T cell phenotype composition for the TCT product and the effect of each composition or subset on the anti-tumor efficacy over an extended period of time. Therefore, the architecture 200 can allow virtual clinical trials and / or experiments to be conducted with the help of digital twins. The architecture 200 can additionally and / or alternatively allow the analysis engine 110 or the model generation engine 122 to determine the effect of changes in lymphocyte clearance efficiency and / or regimen on anti-tumor efficacy, the T cell phenotype composition of the TCT product, and / or the dosage and regimen of T cell therapy administration based on the architecture 200.

[0101] Additionally and / or alternatively, based at least on the architecture 200 and / or the determined patient profile for at least one, two or more patients, the analysis engine 110 or the treatment plan generation engine 128 can predict individual patient responses (or ranges thereof) to a given TCT product by simulations based on cell kinetic parameters determined by individual patient baseline biomarker data (data collected via genomic, transcriptomic and / or flow cytometric methods). For example, the analysis engine 110 can determine the response to a T cell therapy comprising a given TCT product in a target patient by at least: simulating a plurality of simulated patients (e.g., a distribution) based at least on the architecture 200 and / or the determined patient profile (e.g., a distribution) Fig. 6A and 6B The various responses to various doses of the composition (or multiple compositions) of the TCT product (as shown). The various responses can be simulated by at least changing the dosage, composition, etc. of the TCT product.

[0102] The reaction in the target patient can be one of a variety of simulated reactions. For example, it is possible (for example, by analysis engine 110 or matching engine 126) to select at least one of the simulated patient data (for example, distribution, reaction, etc.) corresponding to the patient with similar characteristics, composition, diagnosis, tumor, cancer, etc. to the target patient. Based on the selected simulated patient data, it is possible to determine and / or administer effective T cell therapy (including the dosage and / or composition of TCT products) to the target patient to treat tumors and / or cancer. This allows the ideal dose and / or composition of the TCT products for treating tumors and / or cancers of the target patient to be determined. In some implementations, the simulated patient data can be used to assess both safety (such as about cytokine release syndrome (CRS)) and efficacy of specific T cell therapies (including specific doses and / or compositions of TCT products and regimens).

[0103] Furthermore, if Fig. 6A and 6B As shown, the digital twin generation engine 124 can capture the different cell dynamics exhibited by each patient in response to different dose amounts and compositions of TCR engineered T cells ( Fig. 6A and 6B , Patient ID 2 and 11 were excluded from the dataset due to data unavailability). Digital twins similar to the reference patients replicated the differences observed between different patient cohorts. Notably, the digital twins can reproduce the limited cell expansion in patients belonging to the low-dose group (e.g., ID 1 and 3) and the more significant expansion of Tscm and Teff cell populations in patients within the high-dose cohort (e.g., ID 7, 8, 9, 10, 12). In addition, the digital twins can capture intra-group variability. For example, within the medium-dose group, each simulation simulates changes in cell number and kinetic data between patient 5 and patients 4 and 6. Initial composition differences can be identified as a source of variability for predicting patient responses.

[0104] like Fig. 6A and 6B As shown, digital twins form a virtual population representing 10 patients. This virtual population can be used to predict cell dynamics for these 10 patients under different TCR engineered T cell therapy regimens and compositions, enabling detailed dissection of the biological mechanisms that explain the observed and predicted dynamics and inter-patient variability. Each column in the data provided shows experimental measurements of blood, covering T scm 、T cm 、T em or T eff TCR engineered T cells and endogenous T cells (Tendo). Each row presents the cell measurement results for an individual patient in cells / milliliter (cells / mL). Fig. 6A and6B As shown, patients 1 and 3 received 10^9 cells, patients 4, 5, and 6 received 10^10 cells, and all other patients received 10^11 cells. The data points correspond to experimental measurements over time, while the curves represent the 10 digital twins that exhibit the closest match to the experimental data for each corresponding patient. Fig. 6A and 6B The digital twin shown captures changes in cellular dynamics between patients undergoing treatment with TCR-engineered T cells. Notably, the digital twin reproduces the multiphasic cellular dynamics observed in the blood following administration of TCR-engineered T cells targeting HPV-16E7.

[0105] Fig.12 Example parameter space ridge plots and principal component analysis of digital twins are plotted, revealing sources of variability between patients and across dose groups (A to O). Fig.12 As shown, the ridgeline plot can visualize the distribution of 14 parameters (1 subplot per parameter) across the digital twin of each clinical patient (i.e., target patient). Fig.12 As shown, for each subgraph (i.e., Fig.12 A to Fig.12 O), with individual patient IDs listed along the y-axis and the range of potential parameter values ​​listed along the x-axis. Fig.12 P shows the variability of the principal components (PC1 vs. PC2) across parameter space. In some embodiments, Fig.12 Each dot in P can represent a separate digital twin, for a total of 100 digital twins.

[0106] Digital twins, characterized by different inferred parameterizations of the underlying biological processes that govern cellular dynamics, facilitate this analysis. Fig.12 The ridge plots shown in A to 12O and the principal component analysis (PCA) ( Fig.12 P) is used to visualize the parameter space for the digital twin associated with each patient and facilitate cross-patient comparisons.

[0107] like Fig.12 As shown, several parameters exhibited consistent distributions across patients in the ridge plot ( Fig.12 A to Fig.12 O) For example, antigen-driven T em proliferation( Fig.12 C) and T em Conversion Fig.12 N)-related parameters showed similar distributions, suggesting that these processes did not contribute significantly to the inter-patient differences in cellular dynamics. scmThe rate constant for the control of antigen-driven proliferation of cells (i.e., kprolif_atg_scm subgraph - Fig.12 A) shows significant variability both within and between dose groups. Within the medium dose group, the digital twin for patient 5 exhibited higher T scm The proliferation rate constant was similar to that observed in the high-dose cohort. In contrast, the digital twins for patients 4 and 6 were characterized by a lower T scm The proliferation rate constant was similar to that observed in the low-dose cohort. This observation is consistent with the finding that patient 5 showed higher cell counts compared with patients 4 and 6, suggesting that antigen-driven T scm The propensity to proliferate plays a crucial role in the inter-patient variability of T cell numbers.

[0108] Alternatively or additionally, certain parameters and processes, including T scm , T cm and T em ( Fig.12 F, 12G, 12H) transport rate constants and T cm ( Fig.12 B) and endogenous cells ( Fig.12 E) showed significant inter-patient variability. These findings may highlight the importance of these biological processes as key determinants of cellular dynamics and heterogeneity. Alternatively or additionally, this analysis revealed significant inter-cohort variability between specific parameters. This is reflected in the PCA plot ( Fig.12 P) is further illustrated, where the minimum overlap is observed along the first principal component (PC1) between the low-dose group and the high-dose group. In PCA, as confirmed by contribution analysis, parameters controlling T cell proliferation and transport become contributors to the variability across different dose cohorts. These trends indicate that there is a complex interaction between dose level and early memory T cell proliferation, indicating that the impact exceeds those explicitly incorporated into the model.

[0109] In some embodiments, Fig.12 As shown, identification of sources of variability across patients can be achieved by visualizing the parameter space for subsets of multiple digital twins. This analysis allows for a comprehensive assessment of potential factors that lead to differential kinetic data between patients. In addition, model 1202 can be adapted and / or modified to account for these identified sources of variability across patients.

[0110] Fig.13 Depicts example biological variability (patient-specific) impacts on cell dynamics of TCR engineered T cell therapies, resulting in persister or non-persistent outcomes. Fig.13As shown, the digital twin set generated by the digital twin generation engine 124 can be used as 10 10 The dose of TCR engineered T cells and 1% T scm , 3% T cm 48% T em and 48% T eff Cell dose composition (composed of Fig.13 A shows) to simulate. Fig.13 B to 13E show the cell dynamics over time for each T cell phenotype. In some embodiments, the bars depicted in the figures may represent the 25th and 75th percentiles of each subset, while the solid lines represent the median values ​​for each T cell phenotype. Additionally or alternatively, Fig.13 G presents a bar graph showing partial rank correlation coefficient (PRCC) scores. Each parameter is represented along the y-axis and its relationship with the outcome (persistent vs. non-persistent) is quantified. Empty bars may represent parameters that are statistically insignificant in this context.

[0111] like Fig.13 As shown in Figure 2, the stratification of digital twins within virtual clinical trials provides a deep understanding of the key determinants that influence T cell persistence. Fig.13 As shown, the foregoing analysis can provide a visualization of the biological variability inherent across the digital twin population. In some embodiments, the digital twin can enable an assessment of how biological variability between patients affects T cell persistence over time. Fig.13 As shown in the described example, a virtual clinical trial was simulated in which all digital twins received a medium dose consisting of 10^10 TCR-engineered T cells and a 1% T cell population. scm , 3% T cm 48% T em and 48% T eff The composition of cells Fig.13 A), reflecting representative dose compositions from the E7 clinical trial. Simulated dynamics (medians and interquartile ranges) across virtual populations for various T cell phenotypes are depicted in Fig.13 B to 13E. At this fixed dosage amount and composition, T scm and T eff cell( Fig.13 B. Fig.13 E) Both exhibit an initial distribution and expansion, followed by a subsequent decline across all digital twins. In contrast, T cm and T em cell( Fig.13 C. Fig.13D) is predicted to undergo less expansion. Notably, the predicted responses spanned nearly an order of magnitude for each T cell phenotype, highlighting the expected inter-patient variability under consistent dose composition.

[0112] Given the correlation between cell expansion within the initial months and treatment efficacy in hematological malignancies, a stratification of digital twins based on the frequency of TCR-engineered T cells in the blood at day 50 was performed. Those digital twins (exhibiting a dominant presence of TCR-engineered T cells in the blood, indicating persistence over time) were classified as “persisters” under this treatment regimen, while the rest (showing a decline over time) were classified as “non-persisters” ( Fig.13 F).

[0113] like Fig.13 As shown, a global sensitivity analysis can be performed to delve into the underlying biological processes driving the persister and non-persistent outcomes. scm Transport and proliferation, T endo Proliferation and T cm The rate constants of proliferation and transport were identified as key determinants of persister or non-persister outcome ( Fig.13 G shows the partial rank correlation coefficient values). It is worth noting that these parameters were also identified in the previous analysis of inter-individual variability. Thus, there may be nonlinear relationships between the parameters. For example, due to the overall T scm Expansion depends on both transport to tumor-draining lymph nodes and proliferation within the tumor-draining lymph nodes, thus having a lower T scm Digital twins of transport rate constants may require higher T scm The persistence of T cells can be demonstrated by measuring the proliferation rate constant and vice versa. In the bivariate plot, the differences in these parameters between persisters and non-persisters become more obvious. In some embodiments, these parameters become the key drivers of inter-patient variability in cell persistence, which may indicate the need to optimize the treatment regimen, dosage and / or composition to effectively regulate these parameters and achieve the desired level of T cell persistence.

[0114] Fig.14 Depicted is an example graph showing the impact of dose composition on cellular kinetics of TCR engineered T cells. The figure presents two different sets of digital twins: one matching patient 1 (i.e., combined with Fig.14 A) and another matched patient 6 (i.e., combined Fig.14B). In some embodiments, these digital twins can be re-simulated with alternative dose compositions and different dose amounts. Each row in the figure can correspond to a different dose composition or dose amount that underwent simulation. Each column can show a representation of the dose composition or cell dynamics of a TCR engineered T cell subset. For ease of comparison, the cell dynamics of the digital twin treated with the original dose composition are re-plotted in the corresponding curve as a baseline for the cell dynamics generated for the alternative dosing simulation.

[0115] In some embodiments, in addition to the number of doses and inherent biological patient variability, the dose composition may have an impact on the expansion and persistence of T cells. To illustrate the meaning of alternative dose compositions, different compositions (such as 100 mg / kg / day) for representative patients obtained from both the low dose cohort (patient 1) and the medium dose cohort (patient 6) can be used. Fig. 6A and 6B Specifically, based on the following assumptions, the simulation is performed using T scm Cells (including 85% T scm , 5% T cm , 5% T em and 5% T eff ) Enriched composition: T scm The enrichment may promote more pronounced expansion and prolonged persistence. The kinetics of cells treated with the original dose composition were compared with those treated with T at two different dose levels. scm The enriched doses were then compared to the predicted cellular kinetics after treatment, as Fig.14 shown.

[0116] Predicting cellular dynamics supports the concept that T scm Treatment with induced enhanced overall expansion and persistence of TCR-engineered T cells. Fig.14 As shown, compared with the infusion, mainly including T em and T eff The original dose of cells was 10^10 cells, and the dose of the original dose was 10 times that of the original dose (10^9 cells) to simulate the infusion of T scm The enriched product is expected to produce comparable overall cell expansion and persistence. For patient 1, simulated infusion of T cells at the original dose level of 10^9 cells was performed. scm The enriched material predicted that the increase in T cell counts in the bloodstream was mainly composed of T em and T eff The simulation (indicating a higher T in the infused product) was approximately 100 times the original administered material. scm The composition corresponds to the greater abundance of T cells in these patients) is consistent with earlier findings that T scmThe key role of parameters in shaping cell dynamics and persistence is related. It is worth noting that these insights go beyond the scope of what can be obtained from pure correlation analysis of the E7 clinical trial. Therefore, in some embodiments, the treatment plan generation engine 128 can utilize the QSP model 1202 and the digital twin to optimize the dose composition and quantity for clinical patients (i.e., target patients), with the ultimate goal of enhancing T cell persistence in clinical applications.

[0117] In some embodiments, after using the QSP model and digital twins to explore optimized dose compositions and quantities, these simulations and findings can be used in real-world clinical scenarios. For example, the system can predict the response of a target patient to multiple hypothetical deliveries of a TCT product, each of which can be characterized by a different combination of T cell dose and composition. To predict this target patient response, a subset of digital twins with characteristics similar to those of the target patient can be selected. These selected digital twins provide a predictive framework to estimate how a patient's T cell dynamics will respond to various treatment scenarios, thereby enabling a comprehensive assessment of potential outcomes.

[0118] In some embodiments, the treatment plan generation engine 128 can utilize the predicted target patient response provided by the digital twin subset. In this way, the treatment plan generation engine 128 can formulate an individualized treatment plan for the target patient, taking into account the predicted cell dynamics and persistence levels over time in various physiological compartments. The treatment plan is designed, for example, by the treatment plan generation engine 128 to optimize the T cell persistence in the target patient (which can always exceed a predetermined threshold across multiple physiological compartments). In some embodiments, the treatment plan specifies the specific dose and composition of the TCT product for the patient's therapy, so that it is consistent with the expected goal of maintaining a T cell content above the defined threshold. Using the in-depth understanding derived from the QSP model 1202 and the digital twin, the application establishes a framework for data-driven personalized decision-making in the field of TCR engineered T cell therapy. It introduces the ability to fine-tune the treatment method on a per-patient basis, focusing on optimizing both the composition and quantity of therapeutic T cell doses. The purpose is to achieve and maintain the level of targeted T cell persistence, ultimately improving the accuracy and efficacy of clinical interventions in this professional field.

[0119] Fig.15 Depicts an example predictive simulation of a digital twin showing consistency with observed data in patients with available clinical data. In some embodiments, clinical data related to patients with pancreatic cancer treated with TCR engineered T cells targeting KRAS G12D can be utilized. Fig.15 As shown, depicting the Fig.15 Patient A 1 and Fig.15B The percentage of TCR-engineered T cells relative to total T cells in the blood of patient 2 over time. Fig.15 As shown, 100 digital twins (represented by lighter color curves) can be re-simulated with the dose amounts and compositions specified in the clinical trial for KRAS G12D. Fig.15 The darker colored curve may represent a single digital twin that most accurately reproduces the observed data for each patient, as determined by the root mean square error (RMSE).

[0120] like Fig.15 As shown, the predictive ability of digital twins in estimating T cell data for patients with TCR engineered T cell therapy targeting KRAS G12D is demonstrated. In some embodiments, in order to assess the applicability of the QSP model structure and digital twins to other TCR engineered T cell therapies and medical indications, a 100-member digital twin virtual population is used for simulation, using the same dose and composition parameters used in clinical studies involving TCR engineered T cells targeting KRAS G12D in patients with metastatic pancreatic cancer. The KRAS G12D clinical data set selected for this evaluation can be based on its public availability, thereby providing an in-depth understanding of the dose composition, dose volume, and dynamics of the therapy across different patients. In the context of the KRAS G12D clinical trial, both patients were administered a moderate dose of approximately 1010 cells, mainly including Tem cells, although patient No. 2 received a larger proportion of early memory cells. In this way, changes in the simulation curve can reflect the inherent inter-patient and intra-patient variability encapsulated in the digital twin. Clinical patient trajectories were aligned with the range of predictions made by the digital twins, with the best matching digital twin in each case fused with the densest cluster of simulated data. Fig.15 Model 1202 and the digital twin were additionally validated as the figure demonstrated their ability to accurately predict cellular dynamics of TCR-engineered T cells in patients with pancreatic cancer undergoing therapy targeting KRAS G12D. This agreement between predicted and observed data underscores the robustness and generalizability of the model and digital twin, extending their utility beyond the specific therapy and patient cohort studied, and demonstrating their potential applicability to predict T cell behavior in diverse TCR-engineered therapies and medical contexts.

[0121] Figure 7Depicted are example graphs showing variations in cellular pharmacokinetics based on the phenotypic composition of a T cell product and dose levels consistent with implementations of the current subject matter. As noted, architecture 200 allows for determination of the effects of variability in dose levels and phenotypic composition of a TCT product on cellular pharmacokinetics of multiple T cell phenotypes and the distribution and persistence of T cell therapy across time. Figure 7 Include a graph depicting the resimulation of the high-dose patient group, e.g. Figure 5 The third row of Figures depicts examples. These graphs include previously determined behaviors of T cell phenotypes (solid lines) and tested doses of TCT products that include compositions with greater numbers of stem cell-like memory T cells than effector T cells (dashed lines). Figure 7 As shown, TCT products including a majority of stem cell-like memory T cells may be more likely to have a greater number of total T cells remaining in the patient 200 days after infusion compared to simulations of TCT products with a majority of effector T cells.

[0122] Figure 8 Depicted are example graphs showing the phenotypic composition of T cell-based products consistent with implementations of the current subject matter and changes in cellular pharmacokinetics at dose levels. In particular, Figure 8 The figure in shows a set of simulations performed based on the QSP model 1202. Figure 8 The simulations shown included three different dose levels (10 9 cells, 10 10 cells and 10 11 The re-simulation of the cellular pharmacokinetics of patient 6 based on different compositions of TCT products is shown in Fig. 6A Comparison of cellular pharmacokinetics. Fig. 9 A flow chart showing an example of a process 900 consistent with implementations of the current subject matter for determining the distribution of multiple T cell phenotypes over time following delivery of a TCT product is depicted. Fig. 9 , process 900 can be performed by the analysis engine 110 to generate a treatment plan, determine a T cell therapy (including the dose of a TCT product and / or the composition of a TCT product) for treating a tumor, predict a patient response to the dose of a composition of a TCT product, etc. For example, the analysis engine 110 can implement the QSP model 1202 to determine a patient profile including the distribution of T cell phenotypes of a TCT product over time within a patient's physiological compartment, such as for the dose and / or composition of a TCT product. T cell therapies, treatment plans, etc. can be determined based at least on the determined distribution. The analysis engine 110 can capture the cellular dynamics of multiple T cell phenotypes after administration of a T cell therapy for solid tumors based on the architecture 200. Consistent with the implementation of the current subject matter, process 900 refers to Figure 2 Architecture 200 is shown.

[0123] At 902, the analysis engine 110 (e.g., at least one data processor) can determine a set of cell dynamics parameters corresponding to a plurality of T cell phenotypes within the patient's peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) (e.g., peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202). A plurality of T cell phenotypes may include stem cell-like memory T cells (e.g., stem cell-like memory T cells 210), central memory T cells (e.g., central memory T cells 212), effector memory T cells (e.g., effector memory T cells 214), effector T cells (e.g., effector T cells 216), and endogenous T cells. In some implementations, a plurality of T cell phenotypes include stem cell-like memory T cells 210, central memory T cells 212, effector memory T cells 214, and effector T cells 216. A TCT product may include a dosage of a composition of a plurality of T cell phenotypes. For example, a dosage may include the total number of T cells of a TCT product. The composition of a TCT product may include the number of each T cell phenotype in a plurality of T cell phenotypes. In other words, the composition of the TCT product may include a specific ratio (or range of ratios) of each T cell phenotype among the multiple phenotypes that form the TCT product.

[0124] The cell kinetic parameter set includes at least one of the number, proliferation rate, apoptosis rate, and differentiation rate of the multiple T cell phenotypes. The cell kinetic parameter set corresponding to the multiple T cell phenotypes within the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 can be determined at multiple time points (e.g., a first time point, a second time point, a third time point, etc.). The values ​​of the cell kinetic parameters in the cell kinetic parameter set can be different for each of the multiple T cell phenotypes within the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202. Therefore, the analysis engine 110 can track the behavior of the multiple T cell phenotypes within the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 at various time points.

[0125] At 904, the analysis engine 110 (e.g., at least one data processor) can determine a first transport rate of the plurality of T cell phenotypes between the patient's peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202 and a blood compartment (e.g., blood compartment 204). The first transport rate can be related to the plurality of T cell phenotypes in Figure 2 This corresponds to the transfers at lines 220, 224, 228, 232 shown in architecture 200.

[0126] At 906, the analysis engine 110 (e.g., at least one data processor) may determine a set of cell kinetic parameters corresponding to the plurality of T cell phenotypes within the blood compartment 204. The set of cell kinetic parameters includes at least one of a number, a proliferation rate, an apoptosis rate, and a differentiation rate of the plurality of T cell phenotypes within the blood compartment 204. The set of cell kinetic parameters corresponding to the plurality of T cell phenotypes within the blood compartment 204 may be determined at a plurality of time points (e.g., a first time point, a second time point, a third time point, etc.). The values ​​of the cell kinetic parameters in the set of cell kinetic parameters may be different for each of the plurality of T cell phenotypes within the blood compartment 204. Thus, the analysis engine 110 may track the behavior of the plurality of T cell phenotypes within the blood compartment 204 at various time points.

[0127] At 908, the platform 120 (e.g., at least one data processor) can determine a second transport rate of the plurality of T cell phenotypes between the patient's blood compartment 204 and the tumor-draining lymph node compartment (e.g., the tumor-draining lymph node compartment 208). The second transport rate can be related to the plurality of T cell phenotypes in Figure 2 Corresponding to the transfers at lines 222, 226, 230, 234 shown in architecture 200.

[0128] At 910, the analysis engine 110 (e.g., at least one data processor) may determine a set of cell kinetic parameters corresponding to the plurality of T cell phenotypes within the tumor-draining lymph node compartment 208. The set of cell kinetic parameters includes at least one of number, proliferation, apoptosis, and differentiation of the plurality of T cell phenotypes within the tumor-draining lymph node compartment 208. The values ​​of the cell kinetic parameters in the set of cell kinetic parameters may be different for each of the plurality of T cell phenotypes within the tumor-draining lymph node compartment 208. In some implementations, the analysis engine 110 determines a differentiation rate parameter (e.g., in the set of cell kinetic parameters) corresponding to differentiation of the stem cell-like memory T cells 210 to the central memory T cells 212, the central memory T cells 212 to the effector memory T cells 214, and the effector memory T cells 214 to the effector T cells 216 within the tumor-draining lymph node compartment 208. A set of cell dynamics parameters corresponding to the various T cell phenotypes within the tumor draining lymph node compartment 208 may be determined at multiple time points (e.g., a first time point, a second time point, a third time point, etc.). Thus, the analysis engine 110 may track the behavior of the various T cell phenotypes within the tumor draining lymph node compartment 208 at various time points.

[0129] At 912, the analysis engine 110 (e.g., at least one data processor) can determine a third transport rate of the effector memory T cells 214 and the effector T cells 216 from the patient's blood compartment to the tumor compartment (e.g., the tumor compartment 206). The third transport rate can be related to the multiple T cell phenotypes in Figure 2 This corresponds to the transfers at lines 236, 238 shown in architecture 200.

[0130] At 914, the analysis engine 110 (e.g., at least one data processor) may determine a set of cell kinetic parameters corresponding to the effector memory T cells 214 and the effector T cells 216 within the tumor compartment 206. The values ​​of the cell kinetic parameters in the set of cell kinetic parameters may be different for each of the plurality of T cell phenotypes within the tumor compartment 206. The set of cell kinetic parameters includes at least one of the number, proliferation rate, apoptosis rate, and differentiation rate of the effector memory T cells 214 and the effector T cells 216 within the tumor compartment 206. In some implementations, the analysis engine 110 determines a differentiation rate parameter (e.g., in the set of cell kinetic parameters) corresponding to the differentiation rate of the effector memory T cells 214 to the effector T cells 216 within the tumor compartment 206. The set of cell kinetic parameters corresponding to the effector memory T cells 214 and the effector T cells 216 within the tumor compartment 206 may be determined at a plurality of time points (e.g., a first time point, a second time point, a third time point, etc.). Thus, the analysis engine 110 can track the behavior of the effector memory T cells 214 and the effector T cells 216 within the tumor compartment 206 at various time points.

[0131] At 916, the analysis engine 110 (e.g., at least one data processor) may determine, based at least on the set of cell kinetic parameters, the first transport rate, the second transport rate, and the third transport rate, a distribution over time of each of the plurality of T cell phenotypes in the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202, the blood compartment 204, the tumor draining lymph node compartment 208, and the tumor compartment 206. In some implementations, the analysis engine 110 may determine the persistence of the plurality of T cell phenotypes in the peripheral tissue and lymph node compartment (i.e., healthy tissue compartment) 202, the blood compartment 204, the tumor draining lymph node compartment 208, and the tumor compartment 206, such as based at least on the set of cell kinetic parameters, the first transport rate, the second transport rate, and the third transport rate and / or based on the determined distribution.

[0132] The distribution may include a first distribution corresponding to a dose of a composition of a TCT product administered to a first patient and a second distribution corresponding to a dose of a composition of a TCT product administered to a second patient. Based at least on the first distribution corresponding to the first patient and / or the second distribution corresponding to the second patient (and other distributions and patients), the analysis engine 110 may determine the response of another (e.g., a third) patient to a T cell therapy comprising a dose and composition of a TCT product. Based at least on the determined response of the third patient to T cell therapy, the analysis engine 110 may determine a treatment plan for a third patient to treat a tumor and / or cancer of the third patient. In other words, the treatment plan may be determined based at least on a predicted distribution (e.g., a first distribution and / or a second distribution) corresponding to a specific dose and / or composition of a TCT product. T cell therapy may be administered to a patient according to a treatment plan comprising a determined dose and composition of a TCT product. In some implementations, the dose of the composition of the TCT product is administered to the first patient after a lymphocyte clearance regimen is administered to the first patient, and the dose of the composition of the TCT product is administered to the second patient after a lymphocyte clearance regimen is administered to the second patient. The lymphocyte clearance regimen administered to each patient may be the same or different. Therefore, in some implementations, the analysis engine 110 determines a T cell therapy for treating a tumor, the cell therapy comprising a dosage of a composition of multiple T cell phenotypes of a TCT product. For example, in some implementations, the T cell therapy may be determined based at least on a predicted distribution (e.g., a first distribution and / or a second distribution) corresponding to a specific dose and / or composition of a TCT product. As described herein, the T cell therapy may include at least one of a T cell receptor (TCR) engineered T cell therapy, an autologous T cell therapy, an allogeneic T cell therapy, an iPSC-derived T cell therapy, and a CAR T cell therapy.

[0133] Fig.10 A flow chart illustrating an example of a process 1000 for determining a treatment plan for a patient consistent with an implementation of the current subject matter is depicted. Fig.10 , process 1000 can be performed by the analysis engine 110 to generate a treatment plan, determine a T cell therapy including a dose of a TCT product and / or a composition of a TCT product) for treating a tumor, predict a patient response to a dose of a composition of a TCT product, etc. For example, the analysis engine 110 can implement the QSP model 1202 to determine a patient profile including a distribution of a T cell phenotype of a TCT product within a patient's physiological compartment over time, such as for a dose and / or composition of a TCT product. T cell therapy, a treatment plan, etc. can be determined based at least on the determined distribution. Consistent with the implementation of the current subject matter, process 1000 refers to Figure 2Architecture 200 is shown. Further, consistent with implementations of the current subject matter, process 1000 may include one or more steps of process 900, and process 900 may include one or more steps of process 1000.

[0134] At 1002, the analysis engine 110 (e.g., at least one data processor) can determine a first patient profile representing a first response to a first dose of a first composition of a T cell target (TCT) product in a plurality of physiological compartments of a first patient. The first patient profile includes a first set of cell kinetic parameters, the first set of cell kinetic parameters including at least one of a first number, a first proliferation rate, a first transport rate, a first apoptosis rate, and a first differentiation rate of a plurality of T cell phenotypes in at least one of a plurality of physiological compartments of the first patient over a period of time. Thus, the first patient profile can include one or more cell kinetic parameters in a cell kinetic parameter set (e.g., a first set of cell kinetic parameters), as described herein. Further, consistent with implementations of the current subject matter, the plurality of physiological compartments include a peripheral tissue and lymph node compartment (i.e., a healthy tissue compartment) 202, a blood compartment 204, a tumor draining lymph node compartment 208, and a tumor compartment 206.

[0135] The plurality of T cell phenotypes include at least two of stem cell-like memory T cells 210, central memory T cells 212, effector memory T cells 214, effector T cells 216, and endogenous T cells. The first composition of the TCT product may include a first quantity of each of the plurality of T cell phenotypes. The first patient data includes a first response corresponding to each of the plurality of T cell phenotypes.

[0136] At 1004, the analysis engine 110 (e.g., at least one data processor) can determine a second patient profile representing a second response to a second dose of a second composition of a T cell target (TCT) product in a plurality of physiological compartments of a second patient. The second patient profile includes a second set of cytokinetic parameters, the second set of cytokinetic parameters including at least one of a second number, a second proliferation rate, a second transport rate, a second apoptosis rate, and a second differentiation rate of a plurality of T cell phenotypes in at least one of the plurality of physiological compartments of the second patient over a period of time. Thus, the second patient profile can include one or more cytokinetic parameters in a cytokinetic parameter set (e.g., a second set of cytokinetic parameters), as described herein.

[0137] The plurality of T cell phenotypes include at least two of stem cell-like memory T cells 210, central memory T cells 212, effector memory T cells 214, effector T cells 216, and endogenous T cells. The second composition of the TCT product may include a second quantity of each of the plurality of T cell phenotypes. The second patient data includes a second response corresponding to each of the plurality of T cell phenotypes.

[0138] In some implementations, the first composition is different from the second composition, and the first dose is different from the second dose.A linear increase from the first dose to the second dose can cause a nonlinear change between the first response and the second response.

[0139] At 1006, the analysis engine 110 (e.g., at least one data processor) generates an output indicating a pharmacokinetic-pharmacodynamic (PKPD) relationship for the TCT product based on at least the first patient data and the second patient data. The PKPD relationship can be schematically illustrated as the architecture 200 of the QSP model 1202. The PKPD relationship can represent at least the distribution of each of the plurality of T cell phenotypes in the physiological compartments of the patient over time.

[0140] At 1008, analysis engine 110 (e.g., at least one data processor) determines the third patient's response to T cell therapy, which includes the third dose of the third composition of TCT products. As described herein, T cell therapy is at least one of T cell receptor (TCR) engineered T cell therapy, autologous T cell therapy, allogeneic T cell therapy, iPSC-derived T cell therapy and CAR T cell therapy. In some implementations, determining the response to T cell therapy in the third patient includes: at least based on the output, simulating multiple responses of multiple doses of multiple compositions of TCT products in multiple simulated patients. In some implementations, simulating multiple reactions is based on applying lymphocyte clearance schemes to multiple simulated patients. The reaction in the third patient can be one of multiple simulated reactions. This allows determination of the ideal dose and / or composition of TCT products for treating tumors and / or cancers of the third patient. Further, the third patient can be determined by at least changing at least one of the first dose, the first composition, the second dose and the second composition. Such configurations can also allow determination of the ideal dose and / or composition of TCT products for treating tumors and / or cancers of the third patient.

[0141] At 1010, the analysis engine 110 (e.g., at least one data processor) determines a treatment plan for a third patient based at least on the third patient's response to the T cell therapy, the T cell therapy comprising a third dose of a third composition of the TCT product. In other words, the treatment plan can be determined based at least on the predicted PKPD relationship, the first patient data, the second patient data, the predicted response of the third patient, etc. The T cell therapy can be administered to the third patient according to the treatment plan.

[0142] Fig.16 A flow chart showing an example of a process 1600 for generating a personalized treatment plan using a model and a digital twin is depicted, consistent with an implementation of the current subject matter. Fig.16 , process 1600 can be performed by the platform 120 to generate a treatment plan, determine a T cell therapy (including a dose of a TCT product and / or a composition of a TCT product) for treating a tumor, predict a patient response to a dose of a composition of a TCT product, etc. For example, the platform 120 can generate a model at operation 1602, wherein the model can be a QSP model, such as that generated by the model generation engine 122 using Figure 2 1600. In some embodiments, the generated model can simulate the distribution over time of each of a plurality of T cell phenotypes in a plurality of physiological compartments after delivery of a T cell target (TCT) product to a patient. In some embodiments, process 1600 can proceed to operation 1604, where the digital twin generation engine 124 can generate a plurality of digital twins through a model (e.g., through QSP model 1202). In some embodiments, each of a plurality of digital twins can represent the distribution over time of a plurality of T cell phenotypes in a plurality of physiological compartments associated with a corresponding patient, where the corresponding patient has a set of patient features. In some embodiments, patient features can include patient biometrics, patient medical history, and baseline biomarker data. In some embodiments, patient features can include the type of cancer / tumor that the corresponding patient has. Next, process 1600 can proceed to operation 1606, where the matching engine 126 can receive a set of features associated with a target patient. In some embodiments, the matching engine 126 can generate a plurality of digital twins through a model (e.g., through QSP model 1202). In some embodiments, each of a plurality of digital twins can represent the distribution over time of a plurality of T cell phenotypes in a plurality of physiological compartments associated with a corresponding patient, where the corresponding patient has a set of patient features. In some embodiments, the patient features can include patient biometrics, patient medical history, and baseline biomarker data. In some embodiments, the patient features can include the type of cancer / tumor that the corresponding patient has. Next, process 1600 can proceed to operation 1606, where the matching engine 126 can receive a set of features associated with a target patient. In some embodiments, the matching engine 126 can generate a plurality of digital twins through a model (e.g., through QSP model 1202). Figure 1The client node 102 or 106 shown receives the feature set. Next, the matching engine 126 can match the target patient features with the patient features associated with the digital twin at operation 1608 to select a digital twin subset similar to the target patient from multiple digital twins. In some embodiments, the matching engine 126 can calculate a matching score for each digital twin of the target patient in the set. Subsequently, the matching engine 126 can select the top N digital twins with the highest score. This can facilitate the selection of a digital twin subset from a digital twin set similar to the target patient among the available digital twins. Next, the process 1600 can proceed to operation 1610, where the platform 120 can use the selected digital twin subset for the target patient to predict the target patient response to multiple hypothetical deliveries of TCT products using different doses and T cell phenotype compositions. The treatment plan generation engine 128 can then generate a treatment plan for the target patient based on the predicted response provided by the digital twin subset at operation 1612. In some embodiments, a treatment plan may be predicted to provide a level of T cell persistence over time above a predetermined threshold in multiple physiological compartments of a target patient, wherein the treatment plan provides / includes a dose of treatment and a composition of multiple T cell phenotypes. In some embodiments, a treatment plan may be predicted to provide a level of T cell persistence over time above a predetermined threshold in at least one of multiple physiological compartments of a target patient. In some embodiments, a treatment plan may be predicted to provide a level of T cell persistence over time above a predetermined threshold in at least a tumor compartment of a target patient.

[0143] Fig.11 A block diagram illustrating a computing system 1100 consistent with an implementation of the current subject matter is depicted. Figures 1 to 10 , computing system 1100 may be used to implement analysis engine 110, QSP model 1202, and / or any components thereof.

[0144] like Fig.11As shown, the computing system 1100 may include a processor 1110, a memory 1120, a storage device 1130, and an input / output device 1140. The processor 1110, the memory 1120, the storage device 1130, and the input / output device 1140 may be interconnected via a system bus 1150. The computing system 1100 may additionally or alternatively include a graphics processing unit (GPU) such as for image processing and / or an associated memory for the GPU. The GPU and / or the associated memory of the GPU may be interconnected with the processor 1110, the memory 1120, the storage device 1130, and the input / output device 1140 via the system bus 1150. The memory associated with the GPU may store one or more images described herein, and the GPU may process one or more of the images described herein. The GPU may be coupled to the processor 1110 and / or form a part of the processor. The processor 1110 is capable of processing instructions for execution within the computing system 1100. Such executed instructions may implement one or more components, such as the analysis engine 110, the QSP model 1202, and the like. In some specific implementations of the current subject matter, the processor 1110 may be a single-threaded processor. Alternatively, the processor 1110 may be a multi-threaded processor. The processor 1110 is capable of processing instructions stored in the memory 1120 and / or the storage device 1130 to display graphical information for a user interface provided via the input / output device 1140.

[0145] The memory 1120 is a computer-readable medium, such as a volatile or non-volatile computer-readable medium, that stores information within the computing system 1100. For example, the memory 1120 may store a data structure representing a configuration object database. The storage device 1130 is capable of providing persistent storage for the computing system 1100. The storage device 1130 may be a floppy disk device, a hard disk device, an optical disk device, a magnetic tape device, or other suitable persistent storage device. The input / output device 1140 provides input / output operations for the computing system 1100. In some specific implementations of the current subject matter, the input / output device 1140 includes a keyboard and / or a pointing device. In various specific implementations, the input / output device 1140 includes a display unit for displaying a graphical user interface.

[0146] According to some specific implementations of the present subject matter, the input / output device 1140 can provide input / output operations for network devices. For example, the input / output device 1140 can include an Ethernet port or other networking port to communicate with one or more wired and / or wireless networks (e.g., a local area network (LAN), a wide area network (WAN), the Internet).

[0147] In some implementations of the current subject matter, computing system 1100 can be used to execute various interactive computer software applications that can be used to display text in various (e.g., spreadsheet) formats (e.g., Microsoft and / or any other type of software) to organize, analyze and / or store data. Alternatively, computing system 1100 can be used to execute any type of software application. These applications can be used to perform various functions, such as planning functions (e.g., generating, managing, editing spreadsheet documents, word processing documents and / or any other objects, etc.), computing functions, communication functions, etc. The application may include various additional functions or may be an independent computing product and / or function. After activation within the application, the function can be used to generate a user interface provided via input / output device 1140. The user interface can be generated by computing system 1100 and presented to the user (e.g., on a computer screen monitor, etc.).

[0148] One or more aspects or features of the subject matter described herein can be implemented in digital electronic circuits, integrated circuits, specially designed ASICs, field programmable gate arrays (FPGAs) computer hardware, firmware, software, and / or combinations thereof. These various aspects or features can be implemented in one or more computer programs, which are executable and / or interpretable on a programmable system, which includes at least one programmable processor (which can be dedicated or general, coupled to receive data and instructions from a storage system and send data and instructions to the storage system), at least one input device, and at least one output device. A programmable system or computing system can include a client and a server. Typically, the client and the server are remotely arranged from each other, and generally interact through a communication network. The relationship between the client and the server is generated by means of computer programs running on respective computers and the client-server relationship between each other.

[0149] These computer programs (which may also be referred to as programs, software, software applications, applications, components or codes) include machine instructions for programmable processors and may be implemented in high-level procedural and / or object-oriented programming languages ​​and / or in assembly / machine languages. As used herein, the term "machine-readable medium" refers to any computer product, device and / or apparatus (such as, for example, a disk, an optical disk, a memory and a programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor. A machine-readable medium may store such machine instructions non-temporarily (such as, for example, a non-temporary solid-state memory or a magnetic hard drive or any equivalent storage medium). A machine-readable medium may store such machine instructions in a temporary manner (such as, for example, a processor cache or other random access memory associated with one or more physical processor cores) alternatively or additionally.

[0150] To provide interaction with a user, one or more aspects or features of the subject matter described herein may be implemented on a computer having a display device (such as, for example, a cathode ray tube (CRT) or a liquid crystal display (LCD) or a light emitting diode (LED) monitor for displaying information to a user) and a keyboard and a pointing device (such as, for example, a mouse or trackball, through which a user can provide input to the computer). Other types of devices may also be used to provide interaction with a user. For example, the feedback provided to the user may be any form of sensory feedback, such as, for example, visual feedback, auditory feedback, or tactile feedback; input from the user may be received in any form, including sound, voice, or tactile input. Other possible input devices include touch screens or other touch-sensitive devices, such as single-point or multi-point resistive or capacitive tracking pads, voice recognition hardware and software, optical scanners, optical pointers, digital image capture devices and associated interpretation software, and the like.

[0151] Depending on the desired configuration, the subject matter described herein may be embodied in a system, device, method and / or article. The embodiments described in the foregoing description do not represent all embodiments consistent with the subject matter described herein. Instead, they are only some examples consistent with aspects related to the described subject matter. Although some variations have been described in detail above, other modifications or additions are possible. In particular, in addition to those features and / or variations described herein, other features and / or variations may also be provided. For example, the above-mentioned specific implementation may be directed to various combinations and sub-combinations of the disclosed features and / or to combinations and sub-combinations of several further features disclosed above. In addition, the logic flow depicted in the drawings and / or described herein does not necessarily require the specific order or continuous order shown to achieve the desired results. For example, without departing from the scope of the present disclosure, the logic flow may include different and / or additional operations than those shown. Without departing from the scope of the present disclosure, one or more operations of the logic flow may be repeated and / or omitted. Other specific implementations may be within the scope of the following claims.

Claims

1. A method comprising: generating a model, wherein the model simulates the distribution over time of each of a plurality of T cell phenotypes in a plurality of physiological compartments following delivery of a T cell targeted (TCT) product to a patient; generating a plurality of digital twins by the model, wherein each of the plurality of digital twins represents a distribution over time of the plurality of T cell phenotypes in the plurality of physiological compartments associated with a corresponding patient, wherein the corresponding patient has a patient feature set; receiving a set of features associated with a target patient; matching target patient characteristics with patient characteristics associated with digital twins to select a subset of digital twins from the plurality of digital twins that are similar to the target patient; using a selected subset of digital twins for the target patients to predict target patient responses to multiple hypothetical deliveries of the TCT product using different doses and T cell phenotypic compositions; as well as A treatment plan is generated for the target patient based on the predicted responses provided by the subset of digital twins, wherein the treatment plan is predicted to provide a level of T cell persistence over time in at least one of the multiple physiological compartments of the target patient that is above a predetermined threshold, wherein the treatment plan includes a dose of the treatment and a composition of the multiple T cell phenotypes.

2. The method according to claim 1, further comprising: identifying sources of variability across patients by visualizing a parameter space for a second subset of the plurality of digital twins; as well as The model was modified to account for this source of variability across patients.

3. The method according to claim 1, further comprising: generating, by the model, a plurality of simulations across different T cell phenotype compositions and dose levels; and visualizing the distribution over time associated with the plurality of simulations for each of the plurality of T cell phenotypes, wherein the visualization indicates the effect of T cell phenotype composition on T cell persistence levels over time.

4. The method of claim 1, wherein the set of patient characteristics comprises patient biometrics, patient medical history, and baseline biomarker data.

5. The method of claim 1, wherein each of the plurality of digital twins further represents a response to a dose of a composition of the TCT product within a plurality of compartments.

6. A method according to claim 5, wherein each of the multiple digital twins includes a set of cell dynamics parameters, the set of cell dynamics parameters including at least one of the number, proliferation rate, transport rate, apoptosis rate and differentiation rate of the multiple T cell phenotypes in at least one of the multiple physiological compartments of the corresponding patient over a period of time.

7. The method of claim 6, wherein the composition of the TCT product comprises an initial number of each of the plurality of T cell phenotypes.

8. The method of claim 1, wherein the plurality of T cell phenotypes comprises at least two of stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells.

9. The method of claim 1, wherein the plurality of physiological compartments include a peripheral tissue and lymph node compartment, a blood compartment, a tumor draining lymph node compartment, and a tumor compartment.

10. The method of claim 1, wherein generating the model further comprises: determining, by at least one data processor, a set of cell kinetic parameters after delivery of a T cell targeted (TCT) product, the set of cell kinetic parameters corresponding to the plurality of T cell phenotypes within the patient's peripheral tissues and lymph node compartments, wherein the plurality of T cell phenotypes include stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells; determining, by the at least one data processor, a first transport rate of the plurality of T cell phenotypes between the peripheral tissue and lymph node compartment and the blood compartment of the patient; determining, by the at least one data processor, a set of cell kinetic parameters corresponding to the plurality of T cell phenotypes within the blood compartment; determining, by the at least one data processor, a second transport rate of the plurality of T cell phenotypes between the blood compartment and the tumor-draining lymph node compartment of the patient; determining, by the at least one data processor, a set of cell kinetic parameters corresponding to the plurality of T cell phenotypes within the tumor-draining lymph node compartment; determining, by the at least one data processor, a third transport rate of the effector memory T cells and the effector T cells from the blood compartment to the tumor compartment of the patient; determining, by the at least one data processor, a set of cell kinetic parameters corresponding to the effector memory T cells and the effector T cells within the tumor compartment; as well as The distribution over time of each of the plurality of T cell phenotypes in the peripheral tissue and lymph node compartment, the blood compartment, the tumor draining lymph node compartment, and the tumor compartment is determined by the at least one data processor and based on the set of cell kinetic parameters, the first transport rate, the second transport rate, and the third transport rate.

11. The method according to claim 10, further comprising: A differentiation rate parameter corresponding to the differentiation of the effector memory T cells to the effector T cells within the tumor compartment is determined by the at least one data processor; and wherein determining the distribution is further based on the differentiation rate parameter.

12. The method according to claim 10, further comprising: Determining, by the at least one data processor, a differentiation rate parameter corresponding to the differentiation of the stem cell-like memory T cells to the central memory T cells, the central memory T cells to the effector memory T cells, and the effector memory T cells to the effector T cells in the tumor-draining lymph node compartment; and wherein determining the distribution is further based on the differentiation rate parameter.

13. The method according to claim 10, further comprising: Identify T cell therapies for treating tumors, wherein the T cell therapy comprises a dose composed of the multiple T cell phenotypes of the TCT product, and wherein the T cell therapy is at least one of a T cell receptor (TCR) engineered T cell therapy, an autologous T cell therapy, an allogeneic T cell therapy, an iPSC-derived T cell therapy, and a CAR T cell therapy.

14. A method comprising: determining, by at least one data processor, first patient data representing a first response to a first dose of a first composition of a T cell target (TCT) product within a plurality of physiological compartments of a first patient; determining, by the at least one data processor, second patient data representing a second response to a second dose of a second composition of the TCT product within the plurality of physiological compartments of a second patient; generating, by the at least one data processor and based at least on the first patient data and the second patient data, an output indicative of a pharmacokinetic-pharmacodynamic (PKPD) relationship for the TCT product; determining, by the at least one data processor and based at least on the output, a third patient's response to a third dose of T-cell therapy comprising a third composition of the TCT product; and A treatment plan for the third patient is determined based at least on the third patient's response to the third dose of the T cell therapy comprising the third composition of the TCT product.

15. The method of claim 14, wherein the first patient data comprises a first set of cell kinetic parameters comprising at least one of a first number, a first proliferation rate, a first transport rate, a first apoptosis rate, and a first differentiation rate of a plurality of T cell phenotypes in at least one of the plurality of physiological compartments of the first patient over a period of time; and The second patient data includes a second set of cell kinetic parameters, which includes at least one of a second number, a second proliferation rate, a second transport rate, a second apoptosis rate, and a second differentiation rate of the multiple T cell phenotypes in at least one of the multiple physiological compartments of the second patient over the period of time.

16. The method of claim 15, wherein the first composition comprises a first quantity of each of a plurality of T cell phenotypes; and wherein the second composition comprises a second quantity of each of the plurality of T cell phenotypes.

17. The method of claim 16, wherein the first patient data and the second patient data each include responses corresponding to each of the plurality of T cell phenotypes.

18. The method according to any one of claims 15 to 17, wherein the plurality of T cell phenotypes comprises at least two of stem cell-like memory T cells, central memory T cells, effector memory T cells, effector T cells, and endogenous T cells.

19. The method of claim 14, wherein the plurality of physiological compartments include a peripheral tissue and lymph node compartment, a blood compartment, a tumor draining lymph node compartment, and a tumor compartment.

20. The method of claim 14, wherein the T cell therapy is at least one of T cell receptor (TCR) engineered T cell therapy, autologous T cell therapy, allogeneic T cell therapy, iPSC-derived T cell therapy, and CAR T cell therapy.

21. The method of claim 14, wherein a linear increase from the first dose to the second dose causes a non-linear change between the first response and the second response.

22. The method of claim 14, wherein the first composition is different from the second composition; and wherein the first dose is different from the second dose.

23. The method of claim 14, wherein determining the response to the T cell therapy in the third patient comprises: simulating, based at least on the output, a plurality of responses to a plurality of doses of a plurality of compositions of the TCT product in a plurality of simulated patients; And wherein the response in the third patient is one of a plurality of simulated responses.

24. The method of claim 14, wherein the first patient data is determined after a first lymphodepleting regimen is administered to the first patient; and wherein the second patient data is determined after a second lymphodepleting regimen is administered to the second patient.

25. The method of claim 14, wherein the response in the third patient is determined by varying at least one of the first dose, the first composition, the second dose, and the second composition.

26. The method of claim 23, wherein simulating the plurality of responses is based on applying a lymphodepletion regimen to the plurality of simulated patients.

27. A system comprising: at least one data processor; as well as At least one memory storing instructions which, when executed by said at least one data processor, result in operations including the method according to any one of claims 14 to 26.

28. A non-transitory computer readable medium storing instructions which, when executed by at least one data processor, result in operations comprising: the method of any one of claims 14 to 26.

29. A method comprising: maintaining a plurality of patient data at a database, wherein the plurality of patient data indicates patient responses to doses of a composition of a TCT product in the form of distributions of a plurality of T cell phenotypes in a plurality of physiological compartments over time, receiving, by at least one data processor, a baseline biomarker dataset associated with a target patient; determining, by the at least one data processor, a set of cell kinetic parameters based on the baseline biomarker dataset for the target patient; generating, by the at least one data processor, a plurality of simulations associated with the target patient based at least in part on cell kinetic parameters; selecting, by the at least one data processor, a subset of simulations from the plurality of simulations based on a level of similarity between the target patient and a patient; as well as A treatment plan is generated for the target patient based on the set of predicted responses generated by the simulation subset.

Images