Blood sequencing-based method for assessing intratumor heterogeneity and its use in predicting response to immunotherapy

Abstract

The present invention relates to a method and system for evaluating intratumor heterogeneity based on blood sequencing. The present invention also relates to the use of intratumor heterogeneity as a biomarker for predicting the response to tumor immunotherapy, selecting treatment methods for tumor patients, dividing tumor patients into different treatment groups, and predicting treatment effects.

Description

Technical Field

[0001] The present invention belongs to the field of biotechnology, and more specifically, the present invention relates to a method and system for evaluating intratumor heterogeneity based on blood sequencing. The present invention also relates to the use of intratumor heterogeneity as a biomarker for selecting treatment methods suitable for tumor patients, predicting the response to tumor immunotherapy, or for dividing tumor patients into different treatment groups, as well as for predicting treatment effects. Background Art

[0002] Immunotherapy has achieved great success in the clinical treatment of tumors. However, there are still a large number of tumor patients who do not benefit from immunotherapy, and the reasons are not clear. Immunotherapy based on immune checkpoint inhibitors (ICIs) has also made significant progress in the clinical treatment of tumors. However, immunotherapy based on immune checkpoint inhibitors also has problems such as low proportion of effective population and drug resistance. Only about 20% of patients can benefit from ICIs monotherapy.

[0003] In the prior art, tumor mutational burden (TMB) and neoantigens are two major factors in evaluating whether immunotherapy is effective. Those skilled in the art usually believe that tumors with high mutation loads will have more neoantigens and are therefore more likely to produce an immune response. This theory seems logical and intuitive. However, in fact, immunotherapy is still effective for many tumors with low mutation loads. It can be seen that the degree of mutation load cannot be accurately used to predict the efficacy of immunotherapy. Therefore, there is an urgent need to find effective biomarkers to screen the beneficiaries.

[0004] In recent years, intra-tumor heterogeneity (ITH) has attracted much attention. Intra-tumor heterogeneity characterizes the distribution of mutations in the main clone and subclone of the tumor. The higher the degree of intra-tumor heterogeneity, the more subclones there are and the higher the diversity of mutation abundance. Summary of the invention

[0005] The present invention focuses on intratumor heterogeneity as a biomarker.

[0006] The present invention finds that the intratumor heterogeneity biomarkers obtained based on blood sampling and the improved algorithm of the present invention can be effectively used to evaluate whether tumor patients can benefit from immunotherapy.

[0007] The present invention also finds that the intra-tumor heterogeneity biomarkers obtained based on blood sampling and the improved evaluation method of the present invention can be effectively used to select tumor patient subgroups suitable for immunotherapy.

[0008] The method for evaluating intratumor heterogeneity of the present invention includes the following improvements:

[0009] (1) High-throughput sequencing based on blood uses the MSAF concept in blood tests to correct all tumor-specific mutations detected, making the mutation abundance between samples comparable;

[0010] (2) Based on the Shannon diversity formula, the tumor heterogeneity evaluation index was weighted. The weighting corrected the impact of the mutation distribution characteristics of the blood on the evaluation, thereby optimizing the method for determining tumor heterogeneity and verifying the feasibility of blood sampling-based tumor heterogeneity indicators in predicting the efficacy of immunotherapy for different cancer types and selecting patient subgroups suitable for immunotherapy.

[0011] Specifically, the present invention includes the following technical solutions:

[0012] In a first aspect, the present invention provides a method for predicting the prognosis of a tumor patient receiving immunotherapy, the method comprising scoring ITH from a blood sample of the patient, wherein the prognosis of a patient with a low ITH score after immunotherapy is better than that of a patient with a high ITH score.

[0013] In a second aspect, the present invention provides the use of ITH as a predictive marker for determining the prognosis of tumor patients receiving immunotherapy, wherein the determination comprises scoring ITH from a patient's blood sample, and the prognosis of patients with low ITH scores after immunotherapy is determined to be better than that of patients with high ITH scores.

[0014] In a third aspect, the present invention provides the use of ITH as a predictive marker in the preparation of a kit for determining the prognosis of tumor patients receiving immunotherapy, wherein the determination comprises scoring ITH from a patient's blood sample, and the prognosis of patients with low ITH scores after immunotherapy is determined to be better than that of patients with high ITH scores.

[0015] In a fourth aspect, the present invention provides a method for identifying tumor patients who may benefit from immunotherapy, the method comprising scoring ITH from a blood sample of the patient, wherein patients with low ITH scores are identified as patients who may benefit from immunotherapy.

[0016] In a fifth aspect, the present invention provides use of ITH as a predictive marker for identifying tumor patients who may benefit from immunotherapy, wherein the identification comprises scoring bITH from a blood sample of the patient, wherein patients with low bITH scores are identified as patients who may benefit from immunotherapy.

[0017] In a sixth aspect, the present invention provides the use of bITH as a predictive marker in the preparation of a kit for identifying tumor patients who may benefit from immunotherapy, wherein the identification comprises scoring bITH from a patient's blood sample, wherein patients with low bITH scores are identified as patients who may benefit from immunotherapy.

[0018] In a seventh aspect, the present invention provides an improved method for assessing bITH, the method comprising:

[0019] (a) sequencing the patient's cell-free DNA to obtain sequencing information; and using the sequencing information to determine tumor-specific mutations;

[0020] (b) calculating the variant allele frequency (VAF) of each mutation site using the sequencing information and the determined tumor-specific variation;

[0021] (c) Divide the VAF into multiple intervals and calculate the probability of mutation distribution within each VAF distribution interval;

[0022] (d) The diversity of mutation distribution of each sample was evaluated as the bITH metric.

[0023] Specifically, the variant allele frequency (VAF) of each mutation site is calculated using sequencing information and determined tumor-specific mutations, which is the depth of mutated allele sequencing (Dmut) / the total sequencing depth of the site; the calculation formula is as follows (1):

[0024]

[0025] Specifically, the total sequencing depth of the site = the reference allele sequencing depth + the variant allele sequencing depth;

[0026] More specifically, the reference allele sequencing depth is the number of normal sequences without the mutation at the corresponding site in the sequencing results;

[0027] More specifically, the variant allele sequencing depth is the number of variant sequences with the variant at the corresponding site in the sequencing results;

[0028] More specifically, wherein the variation is selected from a single nucleotide variation (SNV) or an indel (insertion or deletion);

[0029] Specifically, the maximum somatic allele frequency (MSAF) of each sample was calculated, and the VAF of each mutation was divided by the MSAF to normalize the VAF of each mutation, thereby obtaining a set of corrected tumor mutation allele frequencies (tumorVAF) distributed between 0 and 1. The tumorVAF parameter represents the degree of development of the mutation in the tumor. For the calculation of tumorVAF, see formula (2).

[0030]

[0031] Specifically, MSAF is the highest variation frequency or mutation abundance in ctDNA sequencing.

[0032] Specifically, according to formula (3), 0-1 is divided into intervals of 0.1, and the number of corrected tumorVAFs falling into each interval is counted to obtain a set of integer arrays with a length of 10.

[0033] This array represents the number of mutations in each tumorVAF distribution interval with a length of 0.1.

[0034]

[0035] Specifically, the array is divided by the total number of mutations, as shown in formula (4), to obtain the mutation distribution probability within each VAF distribution interval with a length of 0.1.

[0036]

[0037] Specifically, according to formula (5), the Shannon Diversity Index (SDI) formula is used to evaluate the diversity of mutation distribution of each sample. The value of SDI ( / MSAF) can characterize the degree of heterogeneity of the tumor. The larger the value, the higher the heterogeneity within the tumor, and vice versa.

[0038]

[0039] Compared with the low ITH population calculated by SDI in the existing technology, the corrected SDI ( / MSAF) indicator can better predict that tumor patients with low intra-tumor heterogeneity have a good immunotherapy effect.

[0040] Furthermore, since the distribution of tumorVAF in the blood test data is very uneven, there will be a distribution trend that the higher the tumorVAF, the lower the proportion of mutations, which will cause most mutations to be concentrated in the area with low tumorVAF. Therefore, when using SDI ( / MSAF) to evaluate heterogeneity, an estimation bias will be introduced. Changes in the number of mutations in low tumorVAF will cause large fluctuations in the score, while there are fewer mutations in the high tumorVAF area and the number changes are also less, making it difficult to change the score. Therefore, the assessment of intratumor heterogeneity is deeply limited by the number of mutations detected. Based on this, the present invention further corrects the algorithm in response to this practical defect.

[0041] Specifically, the present invention gives a weight function to the original SDI ( / MSAF) formula according to the prior distribution characteristics of tumorVAF. The purpose of the weight function is to give higher weights to high tumorVAF regions with less mutation distribution, and lower weights to low tumorVAF regions with more mutation distribution.

[0042] Specifically, suitable weight functions found by the present invention are as follows:

[0043] 1. Count the mean tumorVAF in each distribution interval and get mean tumorVAF as the weight function;

[0044] 2. Use the median of each tumorVAF distribution interval as the weight function; or

[0045] 3. For a sufficient number of data sets, the mutation distribution function of the population can be estimated, and the main clone region and subclone region can be divided (the default value can be 0.25), and the main clone region is given twice the weight.

[0046] More specifically, in an eighth aspect, the present invention provides an improved method for assessing ITH, the method comprising:

[0047] (1) sequencing the patient's cell-free DNA to obtain sequencing information; and using the sequencing information to determine tumor-specific mutations;

[0048] (2) Using sequencing information and identified tumor-specific variants, the variant allele frequency (VAF) of each mutation site is calculated as the depth of mutated allele sequencing (Dmut) / the total sequencing depth of the site; the calculation formula is as follows (1):

[0049]

[0050] Specifically, the total sequencing depth of the site = the reference allele sequencing depth + the variant allele sequencing depth;

[0051] More specifically, the reference allele sequencing depth is the number of normal sequences without the mutation at the corresponding site in the sequencing results;

[0052] More specifically, the variant allele sequencing depth is the number of variant sequences with the variant at the corresponding site in the sequencing results;

[0053] More specifically, wherein the variation is selected from a single nucleotide variation (SNV) or an indel (insertion or deletion);

[0054] (3) Calculate the maximum somatic allele frequency (MSAF) of each sample and divide the VAF of each mutation by the MSAF to normalize the VAF of each mutation and obtain a set of corrected tumor mutation allele frequencies (tumorVAF) distributed between 0 and 1. The tumorVAF parameter represents the degree of development of the mutation in the tumor. For the calculation of tumorVAF, see formula (2).

[0055]

[0056] Specifically, MSAF is the highest variation frequency or mutation abundance in ctDNA sequencing.

[0057] (4) According to Formula 3, 0-1 is divided into intervals of 0.1, and the number of corrected tumorVAFs falling into each interval is counted to obtain an integer array of length 10. This array represents the number of mutations in each tumorVAF distribution interval with a length of 0.1.

[0058]

[0059] (5) Divide the array by the total number of mutations, as shown in Formula 4, to obtain the mutation distribution probability within each VAF distribution interval with a length of 0.1.

[0060]

[0061] (6) According to Formula 6, a higher weight is given to the region with high tumorVAF, and a lower weight is given to the region with low tumorVAF. Thus, the present invention obtains bITH based on blood testing and weighted calculation. bITH is more sensitive to changes in the region with high tumorVAF, thereby overcoming the influence of the number of mutation detections on the heterogeneity assessment and being able to better assess the degree of heterogeneity within the tumor.

[0062]

[0063] Specifically, low bITH is a specific predictor for the selection of immunotherapy and an effective marker for predicting the prognosis of immunotherapy.

[0064] More specifically, the reference bITH value may vary depending on the type of tumor.

[0065] In a ninth aspect, the present invention also provides a system for evaluating intratumor heterogeneity in tumor patients, the system comprising: modules for executing the various steps of the above method, such as: a calculation module, a processing module, a judgment module, a sample acquisition module, and a receiving module.

[0066] In this application, the term "computing module" generally refers to a functional module used for calculation. The computing module can calculate output values ​​or obtain conclusions or results based on input values. For example, the computing module can be mainly used to calculate output values. The computing module can be tangible, such as a processor of an electronic computer, a computer or electronic device with a processor, or a computer network, or a program, command line, or software package stored on an electronic medium.

[0067] In this application, the term "processing module" generally refers to a functional module for data processing. The processing module may be based on processing input values ​​into statistically significant data, for example, it may be a classification of data for input values. The processing module may be tangible, such as an electronic or magnetic medium for storing data, a processor of an electronic computer, a computer or electronic device with a processor, or a computer network, or it may be a program, command line, or software package stored on an electronic medium.

[0068] In this application, the term "judgment module" generally refers to a functional module used to obtain relevant judgment results. In this application, the judgment module can calculate output values ​​or obtain conclusions or results based on input values. For example, the judgment module can be mainly used to obtain conclusions or results. The judgment module can be tangible, such as a processor of an electronic computer, a computer or electronic device with a processor, or a computer network, or a program, command line, or software package stored on an electronic medium.

[0069] In the present application, the term "sample acquisition module" generally refers to a functional module for obtaining the sample of the subject. For example, the sample acquisition module may include reagents and / or instruments required to obtain the sample (e.g., tissue sample, blood sample, saliva, pleural effusion, peritoneal effusion, cerebrospinal fluid, etc.). For example, it may include a blood collection needle, a blood collection tube and / or a blood sample transport box. For example, the device of the present application may not contain or include one or more sample acquisition modules, and may optionally have the function of outputting the measurement value of the sample described in the present application.

[0070] In the present application, the term "receiving module" generally refers to a functional module for obtaining the measured value in the sample. In the present application, the receiving module can input the sample described in the present application (e.g., tissue sample, blood sample, saliva, pleural effusion, peritoneal effusion, cerebrospinal fluid, etc.). In the present application, the receiving module can input the measured value of the sample described in the present application (e.g., tissue sample, blood sample, saliva, pleural effusion, peritoneal effusion, cerebrospinal fluid, etc.). The receiving module can detect the state of the sample. For example, the data receiving module can optionally perform gene sequencing (e.g., second generation gene sequencing) on ​​the sample. For example, the data receiving module can optionally include reagents and / or instruments required for performing the gene sequencing. The data receiving module can optionally detect sequencing depth, sequencing read count or sequencing sequence information.

[0071] In the above first to sixth aspects, the ITH (or bITH) is assessed by the method described in the seventh or eighth aspect.

[0072] In the above aspects, the demarcation threshold between low and high bITH scores is the lower third of the bITH score, and the lowest third of the bITH scores are considered to be the low bITH population.

[0073] In all the above aspects, the immunotherapy is immune checkpoint inhibitor therapy.

[0074] In the above aspects, the immunotherapy is the administration of a therapeutic agent that targets at least one immune checkpoint protein to alter the regulation of an immune response (eg, downregulate or inhibit an immune response).

[0075] In the above aspects, the immune checkpoint protein is selected from one or more of the following groups: programmed cell death 1 (PD-1), programmed cell death ligand 1 (PD-L1), programmed cell death ligand 2 (PD-L2), V-domain Ig inhibitor of T cell activation (VISTA), B7-H2, B7-H3, B7-H4, B7-H6, 2B4, ICOS, HVEM, CD160, gp49B, PIR-B, KIR family receptors, TIM-1, TIM-3, TIM-4, LAG-3, BTLA, SIRPα (CD47), CD48, 2B4 (CD244), B7.1, B7.2, ILT-2, ILT-4, TIGIT, LAG-3, BTLA, IDO, OX40, and A2aR.

[0076] In the above aspects, the immunotherapy is the administration of an agent targeting PD-1 or PD-L1.

[0077] Compared with the prior art, the beneficial technical effects achieved by the present invention are as follows:

[0078] Non-invasiveness: The use of ctDNA testing, an economical, universal and non-invasive method, can further reduce the burden on patients.

[0079] Feasibility: Patients with low bITH tumors screened by the improved algorithm of the present invention can significantly benefit from immunotherapy, and patients with low bITH tumors are more suitable for immunotherapy than chemotherapy.

[0080] Universality: The ITH biological indicator can be applied to the evaluation of treatment or prognosis plans for multiple cancer types. For example, it is not only applicable to lung cancer, but also to multiple other cancer types, especially to low TMB populations and tumors with poor TMB predictive value, such as nasopharyngeal carcinoma and clear renal cell carcinoma. It has universality. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 :In the OAK dataset, the ability of SDI ( / MSAF) and original SDI in predicting disease control rate (DCR) was compared. According to the clinical trial endpoint DCR evaluated by RECIST, the immunotherapy group patients were divided into the disease control group and the disease progression group. It was found that the original SDI was almost indistinguishable from the random classification, with an AUC value of 0.507, which was almost close to random; and the SDI ( / MSAF) after MSAF correction, the AUC value increased to 0.583. Among them, Figure 1 The farther the middle curve is from the 45-degree right angle line, the better the prediction effect. The larger the AUC (area under the curve) value, the better the prediction effect. The prediction effect of SDI ( / MSAF) is better than the original SDI. The value of SDI ( / MSAF) represents the degree of tumor heterogeneity. The smaller the value, the lower the heterogeneity within the tumor, and the better the corresponding immunotherapy prognosis, and vice versa.

[0082] Figure 2 : In the OAK dataset, the cox risk regression model was used to perform KM survival curve analysis on the population treated with atezolizumab (see Figure 2 ). The results showed that the low heterogeneity population corresponding to the low SDI ( / MSAF) calculated by MSAF correction had significantly better immunotherapy effects, whether in terms of progression-free survival (PFS) or overall survival (OS). The experimental results showed that the mPFS of the low heterogeneity population and the high heterogeneity population were 4.1 and 1.7 months, respectively, and the mOS were 19 and 8.4 months, respectively, which proved that SDI ( / MSAF) can better predict the prognostic effect of immunotherapy than the SDI obtained by conventional calculation.

[0083] Figure 3 : In the OAK dataset, patients were divided into a disease control group and a disease progression group based on the clinical trial endpoint DCR (Disease Control Rate) evaluated by RECIST. The prediction of SDI was close to random, with an AUC value of 0.507. The prediction effect of the bITH score was improved, with an AUC value of 0.602.

[0084] Figure 4 : In the OAK dataset, the cox risk regression model was used to perform KM survival curve analysis on the population receiving immunotherapy with atezolizumab, which verified that the progression-free survival of the low bITH population was significantly better than that of the high bITH population (mPFS was 4.2 months and 1.6 months, respectively); and the overall survival rate of the low bITH population was also significantly better than that of the high bITH population (mOS was 18 months and 8.4 months, respectively).

[0085] Figure 5 : In the OAK dataset, the bITH score is used as a more independent evaluation indicator, and low bITH can also be used as a predictor of immunotherapy. Figure 5 As shown in the data, in the low bITH population, the effect of atezolizumab immunotherapy was significantly better than that of docetaxel chemotherapy. In addition, the interaction P values ​​of the intergroup tests for treatment methods and heterogeneity between groups in OS and PFS were significant (P values ​​were 0.038 and 0.030, respectively), indicating that low bITH is a specific predictor for immunotherapy.

[0086] Figure 6 : In the POPLAR cohort data set, it was verified that among the people using atezolizumab, the progression-free survival of the low-bITH population was significantly better than that of the high-bITH population (mPFS was 5.5 months and 1.9 months, respectively). The overall survival rate of the low-bITH population during immunotherapy showed a better trend than that of the high-bITH population (mOS was 15 months and 10 months, respectively).

[0087] Figure 7 : In the POPLAR cohort data set, it was verified that the overall survival rate of atezolizumab in the low bITH population was significantly better than that of docetaxel (mOS was 15 months and 8.5 months, respectively), and the progression-free survival rate was also better than that of docetaxel (mPFS was 5.5 months and 3 months, respectively). It further verified that the low bITH population assessed by the bITH score is more suitable for immunotherapy than chemotherapy. DETAILED DESCRIPTION OF THE INVENTION

[0089] definition

[0090] As used herein, the singular forms "a," "an," and "the" include plural referents unless otherwise indicated.

[0091] As used herein, the term "about" refers to the usual error range for the corresponding value that is easily known to those skilled in the art. The "about" value or parameter mentioned herein includes (and describes) embodiments related to the value or parameter itself. For example, a description of "about X" includes a description of "X". In some embodiments, "about" indicates a value of up to ±10% of the listed value, such as ±1%, ±2%, ±3%, ±4%, ±5%, ±6%, ±7%, ±8%, ±9%, or ±10%.

[0092] Tumor mutation burden (TMB)

[0093] As used herein, the term "tumor mutation load" or "TMB" refers to the level (e.g., number) of changes (e.g., one or more changes, such as one or more somatic changes) of each preselected unit (e.g., per million bases (Mb)) in a predetermined genome (e.g., in a coding region of a predetermined genome) detected from a tumor (e.g., a tumor tissue sample, such as a formalin-fixed and paraffin-embedded (FFPE) tumor sample, a fresh frozen tumor sample). For example, the TMB score can be measured on the basis of the entire genome or exome, or can be measured on the basis of a subset of a genome or exome. In certain embodiments, the TMB score measured on the basis of a subset of a genome or exome can be extrapolated to determine the entire genome or exome mutation load. In certain embodiments, the TMB score refers to the level of somatic mutations accumulated in an individual (e.g., an animal, such as a human being). The TMB score can refer to somatic mutations accumulated in patients with tumors (e.g., melanoma). In certain embodiments, the TMB score refers to mutations accumulated in the entire genome of an individual. In some embodiments, the TMB score refers to the mutations accumulated within a particular sample (eg, a tumor sample, eg, a NSCLC sample) collected from a patient.

[0094] Somatic mutation

[0095] The term "somatic mutation" or "somatic alteration" refers to a genetic alteration that occurs in somatic tissues (e.g., cells other than germ cells). Examples of genetic alterations include, but are not limited to, point mutations (e.g., replacement of a single nucleotide by another nucleotide (e.g., silent mutations, missense mutations, and nonsense mutations)), insertions and deletions (e.g., addition and / or removal of one or more nucleotides (e.g., insertions and deletions (indels))), amplifications, gene duplications, copy number alterations (CNA), rearrangements, and splice site mutations. The presence of a particular mutation may be associated with a disease state (e.g., a tumor, such as NSCLC).

[0096] In certain embodiments, the somatic changes are silent mutations (e.g., synonymous changes). In other embodiments, the somatic changes are non-synonymous single nucleotide variations (SNVs). In other embodiments, the somatic changes are silent mutations (e.g., changes that have no detectable effect on the fitness of the clone). In certain embodiments, the somatic changes are variations of unknown significance (VUS), e.g., changes whose pathogenicity has neither been confirmed nor ruled out. In certain embodiments, the somatic changes have not been identified as being associated with a tumor phenotype.

[0097] In certain embodiments, the somatic changes are not associated with, or are not believed to be associated with, effects on cell division, growth, or survival. In other embodiments, the somatic changes are associated with effects on cell division, growth, or survival.

[0098] In certain embodiments, the number of somatic alterations excludes functional alterations in subgenomic intervals.

[0099] In some embodiments, the functional change is a change that has an effect on cell division, growth or survival (e.g., promotes cell division, growth or survival) compared to a reference sequence (e.g., a wild-type or unmutated sequence). In certain embodiments, the functional change is identified as is by inclusion in a functional change database (e.g., COSMIC database). In other embodiments, the functional change is a change with a known functional state (e.g., occurring as a somatic change known in the COSMIC database). In certain embodiments, the functional change is a change with a possible functional state (e.g., a truncation in a tumor suppressor gene). In certain embodiments, the functional change is a driver mutation (e.g., a change that gives a selective advantage to a clone in its microenvironment, e.g., by increasing cell survival or reproduction).

[0100] In certain embodiments, the number of somatic alterations excludes alterations that are present in the sample below a threshold frequency (eg, below 5%, below 3%, below 1%).

[0101] In certain embodiments, the number of somatic alterations excludes germline mutations in subgenomic intervals.

[0102] In certain embodiments, the germline alteration is a SNP, base substitution, insertion, deletion, indel, or silent mutation (eg, a synonymous mutation).

[0103] The somatic mutations in the present invention refer to SNV and Indel.

[0104] Variant allele frequency

[0105] Variant Allele Frequency (VAF) or Variant Allele Fraction refers to the ratio of the read coverage depth supporting the mutant allele at a certain site in the genome to the total read coverage depth of this site. VAF is expressed as follows:

[0106]

[0107] Shannon's Diversity Index (SDI)

[0108] The Shannon diversity index is a measurement index based on information theory, which is used to estimate the diversity of a community. It is also called the Shannon-Wiener or Shannon-Weaver index.

[0109] A higher SDI value indicates a higher degree of distribution diversity, and a lower SDI value indicates a lower degree of distribution diversity.

[0110] Intratumor heterogeneity (ITH)

[0111] Cancer cells continuously accumulate somatic mutations during division and proliferation, some of which confer stronger fitness advantages to cancer cells and may lead to genetically different tumor cell populations. Specific tumor cell subpopulations carry specific subclonal mutations, which is called intratumor heterogeneity (ITH).

[0112] Unlike mutational load, ITH reflects the genetic differences between cells. A tumor with a high mutational load can have a very low ITH; and a tumor with high heterogeneity can also have a very low mutational load.

[0113] For an individual, the ITH associated with an increased clinical benefit is a detectable level in a biological sample. ITH can be measured by methods known to those skilled in the art. The present invention further improves the method for assessing ITH, which is specifically disclosed in the invention section.

[0114] As used herein, a "low ITH score" refers to an ITH score at or below a reference ITH score, and a "high ITH score" refers to an ITH score above the reference ITH score.

[0115] As used herein, the term "reference ITH score" refers to an ITH score to which another ITH score is compared to make a diagnostic, predictive, prognostic, and / or therapeutic decision. For example, the reference ITH score can be an ITH score in a reference sample, a reference population, and / or a predetermined value.

[0116] Those skilled in the art will appreciate that the numerical value of the reference ITH score may vary depending on the type of tumor, such as lung cancer (e.g., non-small cell lung cancer (NSCLC) or small cell lung cancer), kidney cancer (e.g., renal urothelial carcinoma or renal cell carcinoma (RCC)), bladder cancer (e.g., bladder urothelial (transitional cell) carcinoma (e.g., locally advanced or metastatic urothelial carcinoma, including first-line (1L) or second-line or above (2L+) locally advanced or metastatic urothelial carcinoma)), breast cancer (e.g., human epidermal growth factor receptor 2 (HER2)+ breast cancer or hormone receptor positive (HR+) breast cancer), colorectal cancer (e.g., colon adenocarcinoma), ovarian cancer, pancreatic cancer, gastric cancer, esophageal cancer, mesothelioma, melanoma (e.g., skin melanoma), skin cancer (e.g., squamous cell carcinoma of the skin), head and neck cancer (e.g., head and neck squamous cell carcinoma (HNSCC)), thyroid cancer, sarcoma (e.g., soft tissue sarcoma, fibrosarcoma, myxosarcoma, liposarcoma, osteogenic sarcoma, osteosarcoma, chondrosarcoma, angiosarcoma, endotheliosarcoma, lymphangiosarcoma, lymphangioendotheliosarcoma, leiomyosarcoma, or rhabdomyosarcoma), prostate cancer, glioblastoma, cervical cancer, thymic cancer, leukemia (e.g., acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic myeloid leukemia (CML), chronic eosinophilic leukemia, or chronic lymphocytic leukemia (CLL)), lymphoma (e.g., Hodgkin lymphoma or non-Hodgkin lymphoma Hodgkin lymphoma (NHL), myeloma (e.g., multiple myeloma (MM)), mycosis fungoides, Merkel cell carcinoma, hematologic malignancies, cancer of blood tissue, B-cell cancer, bronchial cancer, stomach cancer, brain cancer or central nervous system cancer, peripheral nervous system cancer, uterine cancer or endometrial cancer, oral cancer or pharyngeal cancer, liver cancer, testicular cancer, bile duct cancer, small intestine cancer or appendix cancer, salivary gland cancer, adrenal gland cancer, adenocarcinoma, inflammatory myofibroblastic tumor, gastrointestinal stromal tumor (GIST), colon cancer, myelodysplastic syndrome (MDS), myeloproliferative disorder (MPD), polycythemia vera, chordoma, synovioma, Ewing's tumor, squamous cell carcinoma, basal cell carcinoma, adenocarcinoma, sweat gland carcinoma, sebaceous gland carcinoma, breast cancer Carcinoma capitis, papillary adenocarcinoma, medullary carcinoma, bronchogenic carcinoma, renal cell carcinoma, hepatoma, cholangiocarcinoma, choriocarcinoma, seminoma, embryonal carcinoma, Wilms' tumor, bladder cancer, epithelial carcinoma, glioma, astrocytoma, medulloblastoma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, oligodendroglioma, meningioma, neuroblastoma, retinoblastoma, follicular lymphoma, diffuse large B-cell lymphoma, mantle cell lymphoma, hepatocellular carcinoma, thyroid cancer, small cell carcinoma, essential thrombocythaemia, idiopathic myeloid metaplasia, hypereosinophilic syndrome, systemic mastocytosis, common eosinophilia, neuroendocrine carcinoma, or carcinoid tumor.

[0117] Immunotherapy

[0118] As used herein, "immunotherapy" is immune checkpoint inhibitor treatment, which refers to a therapeutic agent that targets at least one immune checkpoint protein to alter the regulation of an immune response (e.g., downregulate or inhibit an immune response). Immune checkpoint proteins are known in the art and include but are not limited to programmed cell death 1 (PD-1), programmed cell death ligand 1 (PD-L1), programmed cell death ligand 2 (PD-L2), V domain Ig inhibitor of T cell activation (VISTA), B7-H2, B7-H3, B7-H4, B7-H6, 2B4, ICOS, HVEM, CD160, gp49B, PIR-B, KIR family receptors, TIM-1, TIM-3, TIM-4, LAG-3, BTLA, SIRPα (CD47), CD48, 2B4 (CD244), B7.1, B7.2, ILT-2, ILT-4, TIGIT, LAG-3, BTLA, IDO, OX40, and A2aR. In some cases, immune checkpoint proteins can be expressed on the surface of activated T cells.

[0119] The therapeutic agent of immune checkpoint inhibitor includes but is not limited to targeting one or more of the following therapeutic agents: CTLA-4, PD-1, PD-L1, PD-L2, VISTA, B7-H2, B7-H3, B7-H4, B7-H6, 2B4, ICOS, HVEM, CD160, gp49B, PIR-B, KIR family receptors, TIM-1, TIM-3, TIM-4, LAG-3, BTLA, SIRPα (CD47), CD48, 2B4 (CD244), B7.1, B7.2, ILT-2, ILT-4, TIGIT, LAG-3, BTLA, IDO, OX40 and A2aR. In some cases, immune checkpoint inhibitors enhance or contain the function of one or more targeted immune checkpoint proteins. In some cases, immune checkpoint inhibitors are PD-L1 axis binding antagonists as described herein.

[0120] In certain embodiments, the PD-1 targeting agent described herein includes a PD-1 inhibitor. In some embodiments, the PD-1 inhibitor is selected from PDR001 (Novartis), nivolumab (Bristol-Myers Squibb), pembrolizumab (Merck & Co), pidilizumab (CureTech), MEDI0680 (Medical Immune), REGN2810 (Regeneron), TSR-042 (Tesaro), PF-06801591 (Pfizer), BGB-A317 (BeiGene), BGB-108 (BeiGene), INCSHR1210 (Incyte), or AMP-224 (Amplimmune).

[0121] In certain embodiments, the targeted PD-L1 agents described herein include PD-L1 inhibitors. In some embodiments, the PD-L1 inhibitor is selected from FAZ053 (Novartis), Atezolizumab (Roche Pharmaceuticals), Avelumab (Merck Serono and Pfizer Pharmaceuticals), Durvalumab (AstraZeneca), or BMS-936559 (Bristol-Myers Squibb).

[0122] Progression-free survival

[0123] As used herein, "progression-free survival" or "PFS" refers to the length of time after complete surgical resection of a tumor during or after treatment without any recurrence of the disease, during which no signs or symptoms of the treated disease (e.g., tumor) appear. Recurrence-free survival can include the amount of time a patient has experienced a complete response or partial response and the amount of time a patient has experienced stable disease.

[0124] Overall survival

[0125] As used herein, "overall survival" or "OS" refers to the percentage of individuals in a group who are likely to be alive after a specified duration of time. Specific implementation method:

[0126] In the examples, 585 patients with non-small cell lung cancer in the OAK Phase II trial (NCT02008227) and 200 patients in the POPLAR Phase II trial (NCT01903993) are used as examples to illustrate the present invention. It should be noted that the examples are only for illustrative purposes and cannot be interpreted as limiting the present application in any way.

[0127] Example 1: Verification of SDI( / MSAF) as a better biological predictor of immunotherapy efficacy compared to SDI in the OAK dataset

[0128] This embodiment includes the following steps:

[0129] (1) 585 patients with non-small cell lung cancer from the OAK Phase II trial (NCT02008227) were randomly divided into the atezolizumab immunotherapy group (295 patients) and the chemotherapy group (290 patients) according to the PRISMA-IPD and TRIPOD guidelines. Patients were required to have an average sequencing coverage depth greater than 800X and an MSAF greater than 1%. Patients with EGFR driver mutations and ALK rearrangement mutations were excluded.

[0130] (2) Collect the mutation spectrum information including mutation abundance of the patient, and implement steps 1-6 as described in the invention content to evaluate the SDI / (MSAF) index, i.e., the SDI after MSAF correction; as a comparative experiment, calculate the SDI according to the conventional Shannon formula, i.e., the SDI without MSAF correction.

[0131] (3) Based on the clinical trial endpoint DCR (disease control rate) evaluated by RECIST, the patients in the immunotherapy group were divided into a disease control group and a disease progression group. The ability of SDI ( / MSAF) and the original SDI in predicting DCR was compared (see Figure 1 ), and found that the original SDI was almost indistinguishable from the random classification, with an AUC value of 0.507, which was almost close to random. After the SDI( / MSAF) was corrected by MSAF, the AUC value increased to 0.583. A lower SDI( / MSAF) score corresponds to lower intratumor heterogeneity, which can indicate a better prognosis for immunotherapy.

[0132] (4) The immunotherapy group was divided into tertiles, and the population with SDI ( / MSAF) lower than one-third of the tertile was defined as the low heterogeneity population, and the population with SDI ( / MSAF) higher than two-thirds of the tertile was defined as the high heterogeneity population. The Cox hazard regression model was used to perform KM survival curve analysis on the population receiving immunotherapy with atezolizumab (see Figure 2 ). The results showed that the population with low heterogeneity calculated by the adjusted SDI ( / MSAF) had significantly better immunotherapy effects, whether in terms of progression-free survival (PFS) or overall survival (OS). Comparison showed that the mPFS of the low heterogeneity population and the high heterogeneity population were 4.1 and 1.7 months, respectively, and the mOS were 19 and 8.4 months, respectively, which proved that SDI ( / MSAF) can predict better immunotherapy prognosis.

[0133] Example 2: Verification of bITH as a better biological predictor of immunotherapy efficacy compared to SDI in the OAK dataset

[0134] This embodiment includes the following steps:

[0135] (1) A total of 585 patients with non-small cell lung cancer were enrolled in the OAK Phase II trial (NCT02008227). These patients were randomly divided into an immunotherapy group (atezolizumab) and a chemotherapy group (docetaxel) according to the PRISMA-IPD and TRIPOD guidelines. Patients were required to have an average sequencing coverage depth greater than 800X and an MSAF greater than 1%, and patients with EGFR driver mutations and ALK rearrangement mutations were excluded.

[0136] (2) Collecting the patient's mutation spectrum information including mutation abundance, and calculating the bITH index according to steps 1-7 described in the Summary of the Invention.

[0137] (3) Based on the clinical trial endpoint DCR (disease control rate) evaluated by RECIST, patients were divided into a disease control group and a disease progression group. The prediction of SDI was close to random, with an AUC value of 0.507. The prediction effect of bITH score was improved, with an AUC value of 0.602 (see Figure 3 ).

[0138] (4) The immunotherapy group was divided into two groups, low heterogeneity and high heterogeneity, according to the tertiles of the population. The Cox hazard regression model was used to perform KM survival curve analysis on the population treated with atezolizumab immunotherapy (see Figure 4 ), which verified that the progression-free survival of the low bITH population was significantly better than that of the high bITH population (mPFS was 4.2 months and 1.6 months, respectively). And the overall survival rate of the low bITH population was also significantly better than that of the high bITH population (mOS was 18 months and 8.4 months, respectively).

[0139] (5) Compared with SDI ( / MSAF), the bITH score further increases the weight function, increases the sensitivity of capturing mutation changes, and avoids the impact of the number of mutation detections on the result estimation. As a more independent evaluation indicator, low bITH can also be used as a predictor for the selection of immunotherapy. As shown in the figure, in the low bITH population, the effect of atezolizumab immunotherapy was significantly better than that of docetaxel chemotherapy. In addition, the interaction P values ​​of the inter-group tests of treatment methods and heterogeneity in OS and PFS were significant (P values ​​were 0.038 and 0.030, respectively), indicating that low bITH is an immunotherapy-specific predictor (see Figure 5 ).

[0140] Example 3: Validation of the evaluation effect of bITH in the POPLAR dataset

[0141] (1) 200 patients from the POPLAR Phase II trial (NCT01903993) were obtained. These patients were randomly divided into the atezolizumab immunotherapy group (99 patients) and the docetaxel chemotherapy group (101 patients) according to the PRISMA-IPD and TRIPOD guidelines. Patients were required to have an average sequencing coverage depth greater than 800X and MSAF greater than 1%, and patients with EGFR driver mutations and ALK rearrangement mutations were excluded.

[0142] (2) Collecting the patient's mutation spectrum information including mutation abundance, and calculating the bITH index according to steps 1-7 described in the Summary of the Invention.

[0143] (3) Using the population tertile (cutoff = 0.5) as the dividing point, the patients were divided into two groups: low bITH and high bITH. The Cox risk regression model was used to perform KM survival curve analysis on the population using atezolizumab (see Figure 6 ), which verified that among the people using atezolizumab, the progression-free survival of the low-bITH group was significantly better than that of the high-bITH group (mPFS was 5.5 months and 1.9 months, respectively). The overall survival rate of the low-bITH group in immunotherapy showed a better trend than that of the high-bITH group (mOS was 15 months and 10 months, respectively).

[0144] (4) Using the Cox risk regression model, KM survival curve analysis was performed on the population with low bITH, such as Figure 7 , which verified that the overall survival rate of atezolizumab in the low bITH population was significantly better than that of docetaxel (mOS was 15 months and 8.5 months, respectively), and the progression-free survival rate was also better than that of docetaxel (mPFS was 5.5 months and 3 months, respectively). It further verified that the low bITH population assessed by the bITH score is more suitable for immunotherapy than chemotherapy.

Claims

1. A method for assessing intratumor heterogeneity (bITH), characterized in that The method comprises: a) sequencing the patient's cell-free DNA to obtain sequencing information; and using the sequencing information to determine tumor-specific mutations; b) calculating the variant allele frequency (VAF) of each mutation site using the sequencing information and the determined tumor-specific variation; c) Divide the VAF into multiple intervals and calculate the probability of mutation distribution within each VAF distribution interval; d) Evaluate the diversity of mutation distribution of each sample as the bITH indicator; wherein the variant allele frequency (VAF) of each mutation site is the depth of mutant allele sequencing (Dmut) / the total sequencing depth of the site; the calculation formula is as follows (1): The following steps are also included: b-2) Calculate the maximum system mutation frequency (MSAF) of each sample, and divide the VAF of each mutation by the MSAF to normalize the VAF of each mutation, and obtain a set of corrected tumor mutation allele frequencies (tumorVAF) distributed between 0 and 1. The tumorVAF parameter represents the degree of development of the mutation in the tumor. For the calculation of tumorVAF, see formula (2) c-2) According to the following formula (3), 0-1 is divided into intervals of 0.1, and the number of corrected tumorVAFs falling into each interval is counted to obtain a set of integer arrays of length 10, which represents the number of mutations in each tumorVAF distribution interval with a length of 0.

1. c-3) Divide the array by the total number of mutations, as shown in the following formula (4), to obtain the probability of mutation distribution within each VAF distribution interval with a length of 0.1: d-2) According to Formula 5, the diversity of mutation distribution of each sample is evaluated. Among them, the size of the SDI ( / MSAF) value represents the degree of tumor heterogeneity, and the larger the value, the higher the heterogeneity within the tumor.

2. The method according to claim 1, characterized in that The following steps are also included: d-3) Using the weight function, a relatively low weight is given to the low tumorVAF area, and bITH is obtained by weighted calculation. The larger the bITH value, the higher the intratumor heterogeneity.

3. The method according to claim 2, characterized in that The following steps are also included: (1) Calculate the mean tumorVAF in each distribution interval to obtain the mean tumorVAF as the weight function; or (2) Use the median of each tumorVAF distribution interval as the weight function; or, (3) By estimating the mutation distribution function of the population, the main clone region and the subclone region are divided, and the main clone region is given a double weight.

4. The method according to claim 3, characterized in that By estimating the mutation distribution function of the population, the main clone region and the subclone region are divided. The default value is 0.25, and the main clone region is given twice the weight.

5. The method according to claim 1, characterized in that The following steps are also included: d-4) According to the following formula (6), a higher weight is given to the area with high tumorVAF, and a lower weight is given to the area with low tumorVAF. The bITH is obtained based on blood tests and weighted calculations. The larger the bITH value, the higher the intratumor heterogeneity:

6. A use of bITH as a predictive marker for determining the prognosis of a tumor patient receiving immunotherapy, characterized in that: The method comprises scoring bITH from a blood sample of a patient, wherein the bITH score is calculated by any one of claims 1 to 5, wherein the prognosis of a patient with a low bITH score after immunotherapy is judged to be better than that of a patient with a high bITH score.

7. The use according to claim 6, characterized in that The demarcation threshold between low and high bITH scores is the lower tertile of the bITH score, that is, the lowest one-third of the bITH score is considered to be a low-bITH population, or the highest two-thirds of the bITH score is considered to be a high-bITH population.

8. The use according to claim 6, characterized in that The immunotherapy is immune checkpoint inhibitor therapy.

9. The use according to claim 8, characterized in that The immunotherapy is the administration of a therapeutic agent that targets at least one immune checkpoint protein to alter the regulation of the immune response.

10. The use according to claim 9, characterized in that The regulation of the immune response is to downregulate or inhibit the immune response.

11. The use according to claim 9 or 10, characterized in that: The immune checkpoint protein is selected from one or more of the following groups: programmed cell death 1 (PD-1), programmed cell death ligand 1 (PD-L1), programmed cell death ligand 2 (PD-L2), V-domain Ig inhibitor of T cell activation (VISTA), B7-H2, B7-H3, B7-H4, B7-H6, 2B4, ICOS, HVEM, CD160, gp49B, PIR-B, KIR family receptors, TIM-1, TIM-3, TIM-4, LAG-3, BTLA, SIRPα (CD47), CD48, 2B4 (CD244), B7.1, B7.2, ILT-2, ILT-4, TIGIT, LAG-3, BTLA, IDO, OX40, and A2aR.

12. The use according to claim 8, characterized in that The immunotherapy is to administer an agent that targets PD-1 or PD-L1.

13. A system for evaluating intratumor heterogeneity in a tumor patient, characterized in that: The system comprises modules for executing the steps of the method according to any one of claims 1 to 5.

Images