Cancer Biomarkers for Immune Checkpoint Inhibitors

Abstract

Disclosed herein is a method for identifying and treating cancer patients who will benefit from immune checkpoint inhibitor therapy.By utilizing PD-(L)1 therapy treatment data and CGP and quantitative transcriptome profiling (CGP+qTP) data from the Strata Trial (NCT03061305), an observational clinical trial evaluating the impact of molecular profiling on patients with advanced solid tumors, the inventors have developed and validated an integrated immunotherapy response score (IRS) that predicts the benefit of PD-(L)1 in pan-solid tumors by both real-world progression-free survival (rwPFS) and overall survival (OS) with an analytically and clinically validated CGP+qTP non-regulated test (LDT) that can be applied to small formalin-fixed paraffin-embedded (FFPE) tissue specimens.

Description

[Technical field]

[0001] Related Applications

[0002] This application claims priority to and the benefit of co-pending U.S. Provisional Application No. 63 / 277,158, filed November 8, 2021, and co-pending U.S. Provisional Application No. 63 / 407,606, filed September 16, 2022. The disclosures of both provisional applications are hereby incorporated by reference in their entireties herein. [Background technology]

[0003] 2. Background of the Invention

[0004] Anti-PD-1 and anti-PD-L1 (PD-[L]1) monoclonal antibodies, known as checkpoint inhibitors (CPIs), have transformed cancer care and are approved for use in multiple tumor types and pan-tumor indications (microsatellite instability-high / mismatch repair deficient [MSI-H / dMMR] and tumor mutational burden [TMB] ≥ 10 mutations per megabase [Mut / Mb]). 1~3 Improved biomarkers that can predict the benefit of anti-PD-(L)1 have the potential to extend CPI to additional patient populations outside of currently approved indications and to focus their application to more effective, likely responsive patients when alternative therapies exist. In addition, this focused application would reduce unnecessary activation of the immune system via checkpoint inhibitors in subjects unlikely to respond to therapy, thus reducing the number of adverse events in these subjects, which may include colitis, hepatitis, adrenocorticotropic hormone insufficiency, hypothyroidism, type 1 diabetes, acute kidney injury, and myocarditis. PD-L1 immunohistochemistry (IHC) is necessary for the treatment of many tumor types and serves as a companion diagnostic biomarker, although the antibodies, staining platforms, PD-L1 expressing cells included in the scoring algorithm, and cutoffs vary by tumor type. 4~14In addition, although high TMB predicts CPI response across multiple tumor types, TMB determination approaches vary across studies and tests, with only a few TMB-high (TMB-H) patients benefiting, and a single TMB cutoff may not be optimal across tumor types or CPIs. 15~24 For example, in the KEYNOTE-158 study of nine tumor types that led to pan-solid tumor approval of second-line pembrolizumab (anti-PD-1) in patients with TMB ≥ 10 Mut / Mb by the FoundationOne companion diagnostic (CDx) comprehensive genomic profiling (CGP) device, objective responses were observed in 37%, 13%, and 6% of patients with TMB ≥ 13 Mut / Mb, ≥ 10 and < 13 Mut / Mb, and < 10 Mut / Mb, respectively. 25、26 .

[0005] In addition, although only pembrolizumab is approved for patients with high TMB, multiple retrospective and prospective analyses support the clinical utility of high TMB by comprehensive genomic profiling (CGP) for predicting durable response to other anti-PD-(L)1 monotherapies, including both other PD-1 (e.g., nivolumab) and PD-L1 (e.g., atezolizumab) monoclonal antibodies. 27~32 Notably, in a prospective basket study of patients with second-line or greater solid tumors with high TMB by FoundationOne CDx treated with nivolumab or atezolizumab, an ORR of 28% (n=10 / 36) and 19% (n=17 / 90), respectively, was observed in patients with TMB ≥ 10 Mut / Mb, and an increased ORR of 47% (n=8 / 17) and 38% (n=16 / 42), respectively, was observed in patients with TMB ≥ 16 Mut / Mb. 31、32In addition to potentially identifying patients outside of current indications who may benefit from PD-(L)1 monotherapy, given the increasing number of approved PD-(L)1 combination therapy regimens and the thousands of ongoing combination trials, biomarkers capable of identifying benefit of PD-(L)1 monotherapy are particularly important in tumor types where only combination therapy regimens are approved (or where monotherapy is approved only in later lines), as combination regimens are associated with increased clinical and economic toxicity, and recent meta-analyses have essentially demonstrated a lack of evidence for additive or synergistic benefit between PD-(L)1 therapy and other agents in approved combination regimens. 33 .

[0006] Many translational research studies have demonstrated that PD-L1 expression, TMB (with clonal TMB showing increased predictive power versus TMB methods including all somatic mutations), and other immune-related gene expression markers focusing on the tumor microenvironment (TME) are independent predictors of response. 15、34~47 For example, in bladder cancer, multiple studies have demonstrated the potential for PD-L1 by IHC, TMB, and T cell inflammatory gene expression to predict benefit of PD-(L)1 therapy, and the need to maximize the benefit of PD-(L)1 only increases when considering the number of other approved agents in different therapy classes (chemotherapies, antibody drug conjugates, and small molecule inhibitors) that must be sequenced, whether alone or in combination with chemotherapy. 48~53 Importantly, however, a single integrative, clinically applicable and validated test for treatment selection across solid tumors is lacking. Summary of the Invention [Means for solving the problem]

[0007] Summary of the Invention

[0008] By leveraging PD-(L)1 therapy treatment data and CGP and quantitative transcriptome profiling (CGP+qTP) data from the Strata Trial (NCT03061305), an observational clinical trial evaluating the impact of molecular profiling on patients with advanced solid tumors, we developed and validated an integrated immunotherapy response score (IRS) to predict PD-(L)1 benefit in pan-solid tumors by both real-world progression-free survival (rwPFS) and overall survival (OS) with an analytically and clinically validated CGP+qTP non-regulated test (LDT) applicable to small formalin-fixed paraffin-embedded (FFPE) tissue specimens.

[0009] Some aspects of the present invention are directed to a method of treatment comprising the steps of: (a)(i) measuring expression levels of RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12 obtained from a tumor specimen from a subject; (b) measuring the tumor mutational burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; (c) calculating an Immunotherapy Response Score (IRS) from the expression levels of the RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12 obtained in step (a) and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and (d) administering checkpoint inhibitor therapy to the subject.

[0010] In some embodiments, measuring the expression levels of RNA transcripts comprises measuring the expression levels of RNA transcripts for at least three of PD-1, TOP2A, PD-L1 and ADAM12, and calculating the IRS comprises calculating the IRS from the expression levels of the RNA transcripts for at least three of PD-1, TOP2A, PD-L1 and ADAM12, and the converted TMB measurement. In some embodiments, measuring the expression levels of RNA transcripts comprises measuring the expression levels of RNA transcripts for all of PD-1, TOP2A, PD-L1 and ADAM12, and calculating the IRS comprises calculating the IRS from the expression levels of RNA transcripts for all of PD-1, TOP2A, PD-L1 and ADAM12, and the converted TMB measurement. In some embodiments, measuring the expression levels of RNA transcripts comprises measuring the expression levels of RNA transcripts for at least PD-1 and PD-L1, and calculating the IRS comprises calculating the IRS from the expression levels of RNA transcripts for at least PD-1 and PD-L1, and the converted TMB measurement.

[0011] In some embodiments, measuring the expression levels of the RNA transcripts comprises measuring the expression levels of RNA transcripts of PD-1, PD-L1, and ADAM12, and calculating the IRS comprises calculating the IRS from the expression levels of the RNA transcripts of PD-1, PD-L1, and ADAM12, and the converted TMB measurement. In some embodiments, step (a) further comprises ii) measuring the expression levels of RNA transcripts for at least one reference gene in the biological sample, and iii) normalizing the measured expression levels of at least two, at least three, or all of the measured RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12 to the level of the RNA transcript of the at least one reference gene to provide normalized expression levels of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12. In some embodiments, the expression levels of the RNA transcripts used to calculate the IRS comprise the normalized expression levels of the RNA transcripts.

[0012] In some embodiments, step (a) further comprises the step of: iv) median centering the measured expression levels of at least two, at least three, or all of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12, before or after normalizing the expression levels of the measured RNA transcripts.

[0013] In some embodiments, step (a) further comprises v) log2 transforming the measured expression levels, median centered expression levels, normalized expression levels or median centered normalized expression levels of at least two, at least three, or all of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12, wherein the expression levels utilized to calculate the IRS in step c are the transformed expression levels, transformed median centered expression levels, transformed normalized expression levels, or transformed median centered normalized expression levels.

[0014] In some embodiments, the IRS is calculated as follows: IRS=approximately 0.27×[transformed TMB measurement]+approximately 0.11×[transformed PD-1 level]+approximately 0.06×[transformed PD-L1 level]-approximately 0.06[transformed ADAM12 level]-approximately 0.077×[transformed TOP2A level].

[0015] (wherein the transformed PD-1 level, the transformed PD-L1 level, the transformed ADAM12 level, the transformed TOP2A level are each one of the transformed expression level, the transformed median centered expression level, the transformed normalized expression level, or the transformed median centered normalized expression level).

[0016] In some embodiments, the IRS is calculated as follows: IRS=0.27×[transformed TMB measurement]+0.11×[transformed PD-1 level]+0.06×[transformed PD-L1 level]−0.06[transformed ADAM12 level]−0.077×[transformed TOP2A level].

[0017] (wherein the transformed PD-1 level, the transformed PD-L1 level, the transformed ADAM12 level, the transformed TOP2A level are each one of the transformed expression level, the transformed median centered expression level, the transformed normalized expression level, or the transformed median centered normalized expression level).

[0018] In some embodiments, the IRS is calculated as follows: IRS=0.273758×[transformed TMB measurement]+0.112641×[transformed PD-1 level]+0.061904×[transformed PD-L1 level]−0.057991[transformed ADAM12 level]−0.077011×[transformed TOP2A level].

[0019] (wherein the transformed PD-1 level, the transformed PD-L1 level, the transformed ADAM12 level, the transformed TOP2A level are each one of the transformed expression level, the transformed median centered expression level, the transformed normalized expression level, or the transformed median centered normalized expression level).

[0020] In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.87 or greater. In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.873569 or greater.

[0021] In some embodiments, the one or more reference genes comprise three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC and SLC4A1AP.In some embodiments, the one or more reference genes comprise a combination of CIAO1, EIF2B1 and HMBS.

[0022] In some embodiments, the tumor specimen is a formalin-fixed, paraffin-embedded (FFPE) tumor specimen. In some embodiments, the tumor specimen contains at least 20% tumor content.

[0023] In some embodiments, the tumor specimen is breast cancer, central or peripheral nervous system cancer, cancer of unknown primary, colorectal cancer, endometrial cancer, gastrointestinal stromal tumor, glioma, hepatobiliary cancer, neuroendocrine cancer, ovarian cancer, pancreatic cancer, prostate cancer, salivary gland cancer, sarcoma, or thyroid cancer.

[0024] In some embodiments, the tumor specimen is assessed as having low microsatellite instability or microsatellite stability. In some embodiments, the tumor specimen is assessed as having low tumor mutation burden, where low tumor mutation burden is classified as having less than 10 mutations per megabase (mut / Mb). In some embodiments, the expression level of RNA transcripts is measured using PCR and next generation sequencing.

[0025] In some embodiments, the checkpoint inhibitor is an anti-PD-1 antibody, an anti-CTLA-4 antibody, an anti-PD-L1 antibody, or an anti-PD-L2. In some embodiments, the checkpoint inhibitor is nivolumab, pembrolizumab, atezolizumab, cemiplimab, durvalumab, pidilizumab, atezolimumab, PDR001, BMS-936559, avelumab, ipilimumab, or SHR-1210. In some embodiments, the checkpoint inhibitor therapy is administered as a monotherapy. In some embodiments, the checkpoint inhibitor therapy is administered in combination with one or more other chemotherapeutic agents.

[0026] In some embodiments, the tumor specimen exhibits a TPS score of 1-49%. In some embodiments, the checkpoint inhibitor is administered as part of a first line treatment regimen. In some embodiments, the checkpoint inhibitor is administered as part of a second line or subsequent treatment regimen.

[0027] Some aspects of the disclosure include a method of identifying a subject who would benefit from checkpoint inhibitor therapy, comprising the steps of: (a) measuring expression levels of RNA transcripts for PD-1, TOP2A, PD-L1, and ADAM12, and one or more reference genes in a biological sample obtained from a tumor specimen from the subject, the one or more reference genes comprising three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC, and SLC4A1AP; (b) log2-transform, median-center, and normalize the measured expression levels of the RNA transcripts to the level of the RNA transcript of at least one reference gene to identify positive results for the RNA transcripts. (c) measuring tumor mutation burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; (d) calculating an Immunotherapy Response Score (IRS) from the transformed normalized levels of PD-1, TOP2A, PD-L1 and ADAM12 RNA transcripts and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and (e) identifying a subject as benefiting from checkpoint inhibitor therapy.

[0028] In some embodiments, the tumor specimen is from a cancer that is not approved for the indicated use of checkpoint inhibitor therapy. In some embodiments, the tumor specimen is assessed as having low frequency microsatellite instability or microsatellite stability. In some embodiments, the tumor specimen is assessed as having low tumor mutational burden, where low tumor mutational burden is classified as less than 10 mutations per megabase (mut / Mb). In some embodiments, the tumor specimen exhibits a TPS score of 1-49%.

[0029] Some aspects of the disclosure include a method of identifying a subject that would benefit from checkpoint inhibitor therapy, comprising the steps of: (a) receiving, by a processor, measured expression levels of RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; (b) log2 transforming, median centering, and normalizing, by the processor, the measured expression levels of the RNA transcripts to the level of the RNA transcript of the at least one reference gene to provide transformed normalized levels of the RNA transcripts; (c) receiving, by the processor, measured expression levels of the RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from the subject; The present invention is directed to a method comprising the steps of: receiving a measured tumor mutation burden (TMB) in a body sample; (d) log2 transforming, by a processor, the TMB measurement to provide a transformed TMB measurement; (e) calculating, by the processor, an immunotherapy response score (IRS) from the transformed normalized levels of RNA transcripts of at least two of PD-1, TOP2A, PD-L1, and ADAM12, and the transformed TMB measurement, which positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy; and (f) providing a determination that the subject has a checkpoint inhibitor responsive cancer.

[0030] In some embodiments, the IRS is calculated as follows: IRS=0.27×[transformed TMB measurement]+0.11×[transformed normalized level of PD-1]+0.06×[transformed normalized level of PD-L1]−0.06[transformed normalized level of ADAM12]−0.077×[transformed normalized level of TOP2A].

[0031] In some embodiments, the IRS is calculated as follows: IRS=0.273758×[transformed TMB measurement]+0.112641×[transformed normalized level of PD-1]+0.061904×[transformed normalized level of PD-L1]−0.057991[transformed normalized level of ADAM12]−0.077011×[transformed normalized level of TOP2A].

[0032] In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.87 or greater.In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.873569 or greater.In some embodiments, the one or more reference genes comprise three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC and SLC4A1AP.In some embodiments, the one or more reference genes comprise a combination of CIAO1, EIF2B1 and HMBS.

[0033] Some aspects of the disclosure relate to a method of treating a subject in need of treatment with checkpoint inhibitor therapy, comprising administering checkpoint inhibitor therapy to the subject, wherein the subject in need of treatment is further provided with: (a) measuring expression levels of RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12, and one or more reference genes in a biological sample obtained from a tumor specimen from the subject, wherein the one or more reference genes comprise three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC, and SLC4A1AP; (b) comparing the measured expression levels of the RNA transcripts relative to the level of the RNA transcripts of the at least one reference gene; (c) measuring tumor mutation burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; and (d) calculating an Immunotherapy Response Score (IRS) from the transformed normalized levels of at least two of the RNA transcripts of PD-1, TOP2A, PD-L1 and ADAM12, and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and obtains an IRS having a value indicative of a beneficial response to checkpoint inhibitor therapy.

[0034] In some embodiments, the IRS is calculated as follows: IRS=0.27×[transformed TMB measurement]+0.11×[transformed normalized level of PD-1]+0.06×[transformed normalized level of PD-L1]−0.06[transformed normalized level of ADAM12]−0.077×[transformed normalized level of TOP2A].

[0035] In some embodiments, the IRS is calculated as follows: IRS=0.273758×[transformed TMB measurement]+0.112641×[transformed normalized level of PD-1]+0.061904×[transformed normalized level of PD-L1]−0.057991[transformed normalized level of ADAM12]−0.077011×[transformed normalized level of TOP2A].

[0036] In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.87 or greater. In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.873569 or greater.

[0037] In some embodiments, the one or more reference genes comprise a combination of HMBS, CIAO1 and EIF2B1. In some embodiments, the tumor specimen is from a cancer that is not approved for the indicated use of checkpoint inhibitor therapy. In some embodiments, the tumor specimen is evaluated as having low frequency microsatellite instability or microsatellite stability. In some embodiments, the tumor specimen is evaluated as having low tumor mutation burden, and low tumor mutation burden is classified as less than 10 mutations per megabase (mut / Mb).

[0038] In some embodiments, the checkpoint inhibitor therapy is administered as a monotherapy. In some embodiments, the checkpoint inhibitor therapy is administered in combination with one or more other chemotherapeutic agents. In some embodiments, the tumor specimen exhibits a TPS score of 1-49%.

[0039] In some embodiments, the tumor specimen is a formalin-fixed paraffin-embedded (FFPE) tumor specimen. In some embodiments, the tumor specimen is breast cancer, central or peripheral nervous system cancer, cancer of unknown primary, colorectal cancer, endometrial cancer, gastrointestinal stromal tumor, glioma, hepatobiliary cancer, neuroendocrine cancer, ovarian cancer, pancreatic cancer, prostate cancer, salivary gland cancer, sarcoma, or thyroid cancer.

[0040] In some embodiments, the expression levels of RNA transcripts are measured using PCR and next generation sequencing.

[0041] In some embodiments, the checkpoint inhibitor is an anti-PD-1 antibody, an anti-CTLA-4 antibody, an anti-PD-L1 antibody, or an anti-PD-L2. In some embodiments, the checkpoint inhibitor is nivolumab, pembrolizumab, atezolizumab, cemiplimab, durvalumab, pidilizumab, atezolimumab, PDR001, BMS-936559, avelumab, ipilimumab, or SHR-1210.

[0042] Some aspects of the present invention are directed to a method of treatment of a subject in need of treatment with checkpoint inhibitor therapy comprising the step of administering checkpoint inhibitor therapy to said subject, wherein said subject in need of treatment has been pre-identified by a method comprising the steps of: a. measuring expression levels of RNA transcripts for PD-1 and PD-L2 obtained from a tumor specimen from the subject; b. log2 transforming, median centering, and normalizing the measured expression levels of the RNA transcripts to levels of RNA transcripts of one or more reference genes to provide transformed normalized levels of the RNA transcripts; c. measuring tumor mutational burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; and d. calculating an Immunotherapy Response Score (IRS) from the expression levels or normalized levels of the RNA transcripts of PD-1 and PD-L2 and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and obtaining an IRS having a value indicative of a beneficial response to checkpoint inhibitor therapy.

[0043] In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0044] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM 12. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM 12]).

[0045] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0046] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0047] In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0048] In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0049] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0050] In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0051] In some embodiments of the methods disclosed herein, step a. further comprises measuring an expression level of an RNA transcript for at least one reference gene in the biological sample, step b. comprises normalizing the measured expression levels of the other measured RNA transcripts to the level of the RNA transcript of the at least one reference gene to provide normalized levels of the other RNA transcripts, and step c. comprises calculating an IRS from the normalized levels.

[0052] In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is greater than or equal to 10. In some embodiments, the one or more reference genes include three genes selected from LRP1, MRPL13, TBP, HMBS, ITGB7, MYC, CIAO1, CTCF, EIF2B1, GGNBP2, and SLC4A1AP.

[0053] In some embodiments, the tumor specimen is a formalin-fixed paraffin-embedded (FFPE) tumor specimen. In some embodiments, the tumor specimen is adrenal gland cancer, biliary tract cancer, bladder cancer, brain cancer, breast cancer, cervical cancer, colon cancer, rectal cancer, endometrial cancer, esophageal cancer, head and neck cancer, kidney cancer, liver cancer, non-small cell lung cancer, lung cancer, lymphoma, melanoma, meningeal cancer, non-melanoma skin cancer, ovarian cancer, pancreatic cancer, prostate cancer, sarcoma, small intestine cancer, or gastric cancer.

[0054] In some embodiments, the expression levels of RNA transcripts are measured using PCR and next generation sequencing.

[0055] In some embodiments, the checkpoint inhibitor is an anti-PD-1 antibody, an anti-CTLA-4 antibody, an anti-PD-L1 antibody, or an anti-PD-L2. In some embodiments, the checkpoint inhibitor is nivolumab, pembrolizumab, atezolizumab, durvalumab, pidilizumab, PDR001, BMS-936559, avelumab, or SHR-1210.

[0056] Some aspects of the present disclosure include a method of identifying a subject who would benefit from checkpoint inhibitor therapy, comprising the steps of: a. measuring expression levels of RNA transcripts for PD-1, PD-L2, and optionally one or more of CD4, ADAM12, PD-L1, and VTCN1, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; b. log2-transform, median-center, and normalize the measured expression levels of the RNA transcripts to the level of the RNA transcript of the at least one reference gene to provide a normalized level of the RNA transcript; c. tumor mutation burden in the biological sample. d. measuring normalized levels of RNA transcripts of one or more of PD-1, PD-L2, and optionally CD4, ADAM12, PD-L1, and VTCN1, and the transformed TMB measurement, obtaining an IRS that positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and having a value indicative of a beneficial response to checkpoint inhibitor therapy; and e. identifying subjects as benefiting from checkpoint inhibitor therapy.

[0057] In some embodiments, the tumor specimen is from a cancer that is not approved for the indicated use of checkpoint inhibitor therapy. In some embodiments, the calculated IRS value indicates that the median time to next treatment (TNTT) is 24 months or longer.

[0058] Some aspects of the disclosure include a method of identifying a subject who would benefit from checkpoint inhibitor therapy, comprising the steps of: a. receiving, by a processor, measured expression levels of RNA transcripts for PD-1, PD-L2, and optionally expression levels of RNA transcripts for one or more of CD4, ADAM12, PD-L1, and VTCN1, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; b. log2 transforming, median centering, and normalizing, by the processor, the measured expression levels of the RNA transcripts to the level of the RNA transcript of the at least one reference gene to provide normalized levels of the RNA transcripts; c. receiving, by a processor, a measured tumor mutation burden (TMB) in the biological sample; d. log2 transforming, by a processor, the TMB measurement to provide a transformed TMB measurement; e. calculating, by the processor, an immunotherapy response score (IRS) from normalized levels of RNA transcripts of PD-1, PD-L2, and optionally one or more of CD4, ADAM12, PD-L1, and VTCN1, and the transformed TMB measurement, which positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy; and f. providing a determination that the subject has a checkpoint inhibitor responsive cancer.

[0059] In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0060] In some embodiments, the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0061] In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0062] In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[PD-1 normalized level]+0.112×[PD-L2 normalized level]+−0.128×[CD4 normalized level]).

[0063] In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0064] In some embodiments, the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0065] In some embodiments, the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0066] In some embodiments, the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0067] Some aspects of the present disclosure are directed to a method of treating a subject in need of treatment with checkpoint inhibitor therapy, comprising administering checkpoint inhibitor therapy to the subject, wherein the subject in need of treatment comprises the steps of: a. measuring expression levels of RNA transcripts for PD-1, PD-L2, and optionally one or more of CD4, ADAM12, PD-L1, and VTCN1, and one or more reference genes in a biological sample obtained from a tumor specimen from the subject, wherein the one or more reference genes comprise three genes selected from LRP1, MRPL13, TBP, HMBS, ITGB7, MYC, CIAO1, CTCF, EIF2B1, GGNBP2, SLC4A1AP; b. comparing the measured expression levels of the RNA transcripts with at least one of the following: log2 transforming, median centering, and normalizing to the level of an RNA transcript of a reference gene to provide a normalized level of the RNA transcript; c. measuring tumor mutation burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; d. calculating an Immunotherapy Response Score (IRS) from the normalized levels of the RNA transcripts of one or more of PD-1, PD-L2, and optionally CD4, ADAM12, PD-L1, and VTCN1, and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and obtains an IRS having a value indicative of a beneficial response to checkpoint inhibitor therapy.

[0068] In some embodiments, the IRS is calculated as 10 times the inverse of the hazard ratio for a patient compared to the median hazard ratio using a Cox model.

[0069] In some embodiments, the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0070] In some embodiments, an IRS value of 10 or greater indicates a beneficial response to checkpoint inhibitor therapy.

[0071] Some embodiments of the present disclosure are directed to a method of treatment comprising the steps of: a. measuring expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from a subject; b. log2 transforming, median centering, and normalizing the measured expression levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12 to levels of the RNA transcripts of the at least one reference gene to provide normalized levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12; c. measuring expression levels of the RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 in a biological sample obtained from a tumor specimen from a subject; measuring mutational burden (TMB) and log2 transforming the TMB measurement to provide a transformed TMB measurement; d. calculating an Immunotherapy Response Score (IRS) from normalized levels of PD-1, PD-L2, CD4, and ADAM12 RNA transcripts and the transformed TMB measurement, obtaining an IRS that positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and e. administering checkpoint inhibitor therapy to the subject.

[0072] In some embodiments, the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]). In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is 10 or greater. In some embodiments, the tumor specimen is a formalin-fixed, paraffin-embedded (FFPE) tumor specimen. In some embodiments, the tumor specimen is adrenal gland cancer, biliary tract cancer, bladder cancer, brain cancer, breast cancer, cervical cancer, colon cancer, rectal cancer, endometrial cancer, esophageal cancer, head and neck cancer, kidney cancer, liver cancer, non-small cell lung cancer, lung cancer, lymphoma, melanoma, meningeal cancer, non-melanoma skin cancer, ovarian cancer, pancreatic cancer, prostate cancer, sarcoma, small intestine cancer, or gastric cancer. In some embodiments, the expression level of the RNA transcript is measured using PCR and next generation sequencing. In some embodiments, the checkpoint inhibitor is an anti-PD-1 antibody, an anti-CTLA-4 antibody, an anti-PD-L1 antibody, or an anti-PD-L2. In some embodiments, the checkpoint inhibitor is nivolumab, pembrolizumab, atezolizumab, durvalumab, pidilizumab, PDR001, BMS-936559, avelumab, or SHR-1210.

[0073] All patents, patent applications, and other publications (e.g., scientific articles, books, websites, and databases) mentioned herein are incorporated herein by reference in their entirety. In the event of a conflict between this specification and any of the incorporated references, this specification (including any amendments thereto that may be based on the incorporated references) shall control. Standard art-accepted meanings of terms are used herein unless otherwise indicated. Standard abbreviations for various terms are used herein.

[0074] The above-discussed and many other features and attendant advantages of the present invention will become better understood by reference to the following detailed description of the invention.

[0075] BRIEF DESCRIPTION OF THE DRAWINGS

[0076] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee. [Brief description of the drawings]

[0077] [Figure 1] FIG. 1 provides a flow chart of the process used to obtain the present method.

[0078] [Diagram 2] FIG. 2 shows that expression from the 3′ and 5′ amplicons for PD-L1 or PD-1 was highly correlated and therefore were averaged together.

[0079] [Diagram 3] FIG. 3 shows the partial AIC / BIC for each tested model when trained on all 708 samples in the dataset (i.e., the complete dataset).

[0080] [Figure 4] Figure 4 shows the median partial AIC and log-likelihood for each model when trained on a random 2 / 3 of the dataset x 100 iterations. The median log-likelihood score for the test set (1 / 3 omitted) is also shown.

[0081] [Diagram 5] FIG. 5 shows the process of dividing the IRS scores into three groups: low, medium, and high, with the high group having the highest benefit from ICI treatment.

[0082] [Figure 6] FIG. 6 shows that TTNT of pembrolizumab correlates with overall survival (OS).

[0083] [Figure 7] FIG. 7 shows the covariate correlations for the top 10 biomarkers.

[0084] [Figure 8] FIG. 8 shows survival for pembrolizumab (pembro) and chemotherapy (chemo) high / medium / low IRS groups, demonstrating that the IRS score determined by the methods disclosed herein predicts response to ICI instead of predicting the overall efficacy of any cancer treatment.

[0085] [Figure 9] FIG. 9 shows actual clinical progression free survival by low / intermediate / high IRS groups with pembro or chemo treatment.

[0086] [Figure 10] Figure 10 shows actual clinical progression-free survival for patients with NSCLC and other cancers by pembro low / intermediate / high IRS group.

[0087] [Figure 11] FIG. 11 shows the IRG rates (i.e., IRS group) for on-label (melanoma, lung-NSCLC, lung-other, head and neck, lymphoma, bladder, esophagus, bile duct, stomach, cervix, liver, kidney, melanoma) and off-label cancers.

[0088] [Figure 12] FIG. 12 shows IRG rates across the Strata Trial cohort by cancer type.

[0089] [Figure 13] FIG. 13 shows Pembro monotherapy versus pembro in combination with chemotherapy for the IRS group.

[0090] [Figure 14] Figure 14 shows the actual clinical progression-free survival for the IRS group divided into TMB-high (TMB-H) and TMB-low (TMB-L) groups. TMB-high is defined as 10 or more mutations per megabase. TMB-low is less than 10 mutations per megabase.

[0091] [Figure 15] FIG. 15 shows the rates of immunotherapy responders versus TMB-H / L.

[0092] [Figure 16] FIG. 16 shows the actual clinical progression-free survival of patients treated with Pembro with IRG and the tumor content of the samples.

[0093] [Figure 17] FIG. 17 shows the actual clinical progression free survival of patients treated with Chemo with IRG and the tumor content of the samples.

[0094] [Figure 18] Figure 18 shows a brute force search of the 2 / 3-dataset for the best two covariates to add to the minimalist model (PD-1, PD-L2, and TMB). The model with the lowest AIC was selected to derive the training / testing statistics for that cut.

[0095] [Figure 19] FIG. 19 shows the results of a small model search illustrating the claimed methods using ADAM12+CD4+PD-1+PD-L2+TMB.

[0096] [Figure 20] Figure 20 shows the results of backward selection starting with 21 markers and TMB. A multivariate fit was performed to drop and remove the least significant markers. Partial AIC was compared before and after dropping.

[0097] [Figure 21]FIG. 21 shows the best model via brute force training on the full dataset.

[0098] [Figure 22] Figure 22 shows the best model with added PD-L1 via brute force training on the full dataset.

[0099] [Diagram 23] Figure 23 shows a minimalist model using only PD-1, PD-L2 and TMB trained on the full dataset.

[0100] [Figure 24] FIG. 24 shows the large-scale model derived from backward selection.

[0101] [Diagram 25] Figure 25 shows coefficients from 2 / 3 cross-validation iterations of the "brute force" model without and with PD-L1, the yellow line is the final model coefficients from the model without PD-L1.

[0102] [Figure 26] Figure 26 shows coefficients from 2 / 3 cross-validation iterations. Small ("minimalist") and large ("backward selection") models. The yellow line is the brute-force ("bru") model coefficients from the no PD-L1 model.

[0103] [Figure 27] FIG. 27 shows model cross-validation, where specimens are collected after the pembro start date.

[0104] [Figure 28] Figure 28 shows the surrogate biomarker equation: TMB+PD-L2+PD-1+ADAM12+CD4+PD-L1.

[0105] [Figure 29] FIG. 29 shows the alternative biomarker equation of TMB+PD-L2+PD-1.

[0106] [Diagram 30] FIG. 30 shows the alternative biomarker equation TMB+PD-L2+PD-1+ADAM12.

[0107] [Diagram 31] FIG. 31 shows the alternative biomarker equation: TMB+PD-L2+PD-1+PD-L1.

[0108] [Diagram 32] FIG. 32 shows the surrogate biomarker equation: TMB+PD-L2+PD-1+CD4.

[0109] [Diagram 33] FIG. 33 shows the surrogate biomarker equation TMB+PD-L2+PD-1+VTCN1.

[0110] [Diagram 34] FIG. 34 shows the surrogate biomarker equation: TMB+PD-L2+PD-1+CD4+ADAM12+VTCN1.

[0111] [Figure 35a]Figure 35a-c represent the development of an integrated immunotherapy response score (IRS) model to stratify the benefit of PD-(L)1 therapy in patients with advanced solid tumors. Figure 35a) represents real-world clinical treatment, where molecular profiling data from formalin-fixed paraffin-embedded (FFPE) tumor tissue from patients enrolled in the StrataTrial (NCT03061305) are collected in the Strata Clinical Molecular Database (SCMD). Molecular data from both DNA (yellow) and RNA (blue) include both comprehensive genomic profiling (CGP) with both DNA and RNA components, and parallel quantitative transcriptional profiling (qTP) composed of RNA from analytical and clinically validated trials. To develop an integrated predictor of benefit of PD-(L)1 therapy, a cohort of 648 patients (from 26 tumor types) was identified with available molecular information, treated with a systemic therapy line treatment containing pembrolizumab (pembro; PD-1). Using five-fold cross-validated Lasso-penalized Cox proportional hazards modeling, we developed an IRS model to predict real-world progression-free survival (rwPFS; by time to next therapy), which includes tumor mutation burden (TMB; from CGP) and expression of PD-1, PD-L1, ADAM12 and TOP2A (from qTP). The locked IRS model and the thresholds that assign patients to IRS-low [L] or IRS-high [H; increased benefit] were then applied to an independent validation cohort of 248 patients (from 24 tumor types) treated with non-pembrolizumab PD-[L]1 systemic monotherapy. Pie charts for the development and validation cohorts show the tumor type distribution for the 11 most common tumor types and other tumor types. Figure 35b) depicts IRS stratification of rwPFS of pembrolizumab in the development cohort. The rwPFS of pembrolizumab in the development cohort stratified by IRS group is shown by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​(adjusted by variables shown in Figure 35c) for IRS-H vs. IRS-L. The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown.Figure 35c shows that the IRS is robust to potential confounders in the development cohort. A forest plot of the variables included in the adjusted Cox proportional hazards model was used to assess the ability of the IRS to stratify rwPFS for pembrolizumab. Adjusted hazard ratios with 95% confidence intervals (CI) are shown for each variable, with statistically significant variables in bold. [Fig. 35b-c] Same as above.

[0112] [Fig. 36a-b]Figures 36a-f represent real-world progression-free survival (rwPFS) and overall survival (OS) of PD-[L]1 monotherapy by immunotherapy response score (IRS) status. Figure 36a represents rwPFS for pembrolizumab (pembro; PD-1 therapy)-treated patients in the discovery cohort. The rwPFS of pembrolizumab monotherapy in the development cohort stratified by IRS group is shown by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​for IRS-high [H] vs. IRS-low [L] groups. The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. Figure 36b represents the Kaplan-Meier analysis as in 36a, except for OS. Figures 36c-d represent Kaplan-Meier analyses as in Figures 36a-b, except evaluating rwPFS (36c) and OS (36d) in an independent validation cohort of patients treated with non-pembrolizumab PD-(L)1 monotherapy. Figure 36e represents forest plots of adjusted HRs with 95% CIs for IRS and tumor mutation burden (TMB; TMB-high [H] is ≧10 mutations per megabase) in otherwise equivalent models adjusted separately for IRS and TMB (H vs. L for each) in both cohorts for rwPFS and OS. Venn diagrams show the number (n) and overlap of IRS-H (blue) and TMB-H (red) populations in both cohorts. Figure 36f represents the overlap of IRS-H and TMB-H populations for 24,463 patients with informative IRS and TMB status (regardless of treatment status) in the Strata Clinical Molecular Database (SCMD). [Fig. 36c-d] Same as above. [Fig. 36e-f] Same as above.

[0113] [Figure 37]Figures 37a-c represent validation of the predictive nature of the Immunotherapy Response Score (IRS) biomarker. To establish the predictive nature of the IRS model, we evaluated the internal control in the pembrolizumab monotherapy cohort, consisting of 146 patients who received a prior line of systemic therapy prior to pembrolizumab monotherapy. Figure 37a represents that for each patient, rwPFS was determined for the line of systemic therapy immediately preceding pembrolizumab (yellow) and the line of pembrolizumab monotherapy (purple), and then the rwPFS for each group was stratified by IRS status. Figure 37b represents a Kaplan-Meier analysis of the rwPFS of pembrolizumab monotherapy (purple) versus the rwPFS of prior systemic therapy (yellow) in the IRS-low [L] subset of patients (log-rank p-values ​​are shown). The number of patients (n), events, and median rwPFS (with 95% confidence interval [CI]) for each group are shown. Figure 37c represents a Kaplan-Meier analysis of rwPFS of pembrolizumab monotherapy (purple) versus prior systemic therapy (yellow) in the IRS-H subset of patients (log-rank p-values ​​are shown). The likelihood ratio test (LRT) p-value for the interaction between pembrolizumab versus previous line of treatment and IRS status (IRS-L versus IRS-high [H]) is also shown.

[0114] [Figure 38]Figures 38a-c depict the Immunotherapy Response Score (IRS) for predicting benefit of pembrolizumab monotherapy versus combination chemotherapy in first-line NSCLC. Using propensity score matching (see Methods), we identified matched cohorts of patients with NSCLC treated with first-line systemic pembrolizumab (pembro) monotherapy (n=77) or pembrolizumab + chemotherapy (chemo) combination therapy (n=77) who did not differ significantly in age, sex, TMB status, PD-L1 expression by quantitative transcriptome profiling (qTP; expression biomarker component of IRS), or IRS status; PD-L1 immunohistochemistry (IHC) was only available for 24 / 154 samples in the matched cohort (see Figure 13 for validation of PD-L1 by qTP versus PD-L1 IHC). Figure 38a represents a Kaplan-Meier analysis of rwPFS of pembrolizumab monotherapy (orange) versus pembrolizumab + chemotherapy combination therapy (yellow) in the IRS-low [L] subset of patients (log-rank p-values ​​are shown). The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. Figure 38b represents a Kaplan-Meier analysis of rwPFS of pembrolizumab monotherapy (orange) versus pembrolizumab + chemotherapy combination therapy (yellow) in the IRS-high [H] subset of patients (log-rank p-values ​​are shown). The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. Figure 38c depicts the distribution of IRS status in a separate cohort of NSCLC tumor samples with PD-L1 IHC stratified by clinically relevant tumor proportion score (TPS) bins (Figure S8).

[0115] [Fig. 39a-b]Figures 39a-d represent the distribution of pan-solid tumors by immunotherapy response score (IRS) group. Figure 39a represents that IRS groups were determined for all 24,463 patients in the Strata Clinical Molecular Database (SCMD) with informative tumor mutation burden (TMB) and gene expression data required to create an IRS. The distribution of IRS groups (low [L; light blue] vs. high [H; dark blue]) is shown by box plots (numbers indicated percentages); Figure 39b represents the stratification of 24,463 patients by tumor type for approved and unapproved PD-(L)1 monotherapy; Figure 39c represents the decomposition of Figure 39b by individual tumor type; Figure 39d represents the decomposition of Figure 39b by IRS and TMB (high [H] vs. low [L]; TMB-H for ≧10 mutations per megabase). Results may not sum to 100% or may not be equal in subanalyses due to rounding. Tumor type abbreviations: NSCLC (non-small cell lung cancer), RCC (renal cell carcinoma), NMSC (non-melanoma skin cancer), SCLC (small cell lung cancer), CNS and PNS (central and peripheral nervous system), CUP (cancer of unknown primary), CRC (colorectal cancer), GIST (gastrointestinal stromal tumor). [Figure 39c-d] Same as above.

[0116] [Figure 40-1]Figure 40 represents the study landscape from the Strata Clinical Molecular Database (SCMD) used to develop and validate the Immunotherapy Response Score (IRS). The calibre of patients from the Strata Trial (NCT03061305) used to develop and validate the IRS is shown. Included populations are indicated by grey boxes. Because patients may contribute to multiple analyses (e.g., a subject treated with a first-line angiogenesis inhibitor and second-line pembrolizumab may be eligible for both the "non-IO first-line analysis" and the "discovery cohort" as long as they meet both inclusion / exclusion criteria [including samples collected before both lines of therapy]), the number of common patients is indicated by green arrows at the highest branching point. The entire SCMD population is shown in bold yellow. There were no common patients between the discovery and validation cohorts (bold blue). Analyses in groups are indicated by figure numbers. *For the "Clone samples by IRS" group, common subjects are not shown due to the complexity of the figure (maximum common subjects by "Non-IO first line" n=26). [Figure 40-2] Same as above.

[0117] [Diagram 41] Figure 41 represents the allocation of therapy lines from real-world treatment data. For all Strata Trial (NCT03061305) subjects with treatment data (treatment start and stop dates), a standardized allocation of adjuvant / systemic therapy lines was made, taking into account potential overlaps of adjuvant / systemic therapy, monotherapy / combination therapy, treatment start / stop dates, and repeated lines of therapy (whether monotherapy or combination). Examples of real-world progression-free survival measurements by assigned treatment line and time to next therapy (TTNT; from start date of therapy to start date of next therapy) are shown for patients with metastatic renal cell carcinoma.

[0118] [Diagram 42]Figure 42 represents the time to next therapy (TTNT) of patients in the Strata Clinical Molecular Database (SCMD) by line of therapy. Actual clinical progression-free survival by TTNT for first-line (line 1, blue line), second-line (line 2, green line) or third or subsequent line (line 3+, orange line) therapy for 9,899 patients in SCMD with treatment data from at least one systemic anti-neoplastic agent. The number of patients (n), events, and median rwPFS (with 95% confidence interval [CI]) for each group are shown along with the overall log-rank p-value.

[0119] [Fig. 43a-b]Figures 43a-c represent non-small cell lung cancer (NSCLC) analysis of the Strata Clinical Molecular Database (SCMD). Figure 43a, Actual clinical progression-free survival (rwPFS, by time to next therapy) in SCMD of patients with first-line NSCLC when treated with first-generation (gen) targeted EGFR, ALK, ROS1 or MET tyrosine kinase inhibitors ([TKI]; erlotinib, gefitinib, or crizotinib, n=37) versus later-generation inhibitors (n=120, green), is shown by Kaplan-Meier analysis with adjusted hazard ratios (HR) and p-values ​​shown for later-generation versus first-generation inhibitors. Number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. Figure 43b, rwPFS in SCMD patients with first-line NSCLC when treated with first-line oncogene NCCN preferred targeted monotherapy TKIs based on whether treatment was administered before (n=57, blue line; treatment decision made from orthogonal testing) or after (n=72, green line; treatment decision made using StrataNGS) receiving StrataNGS CGP test results by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​shown for treatment before vs. after receiving CGP results. Figure 43c, rwPFS in SCMD of patients with first-line NSCLC treated with biomarker-matched first-line cancer genes NCCN preferred targeted monotherapy TKIs after receiving StrataNGS results based on whether the sample 1) passed both StrataNGS sample input criteria and associated sequencing QC metrics (n=48) or 2) did not meet sample input or failed sequencing QC metrics but reported therapy-matched biomarkers (n=17) by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and fail (2) vs. pass (1) p-values ​​shown for QC metrics. [Figure 43c] Same as above.

[0120] [Diagram 44] Figure 44 depicts the correlation of real-world pembrolizumab progression-free survival (rwPFS) and overall survival (OS). Correlation of rwPFS and OS with time to next therapy (TTNT) for pembrolizumab for patients in the discovery cohort with two or more lines of systemic therapy. Colored boxes indicate patients discussed in the supplemental results.

[0121] [Fig. 45a-b] Figures 45a-d represent real-world progression-free survival (rwPFS) and overall survival (OS) of PD-(L)1 monotherapy by tumor mutation burden (TMB) status. Figure 45a represents rwPFS (by time to next therapy) of pembrolizumab monotherapy (PD-1) in the discovery cohort stratified by TMB group (≥10 mutations per megabase TMB-high [H] vs. TMB-low [L] by StrataNGS testing) and shown by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​for TMB-H vs. -L groups. Number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. As in Figure 45a, except for Figure 45b, OS. Figures 45c-d, As in Figures 45a-b, except evaluating an independent validation cohort of patients treated with non-pembrolizumab PD-(L)1 monotherapy. [Fig. 45c-d] Same as above.

[0122] [Figure 46a]Figure 46a-f represent housekeeping gene selection and validation, precision vs. qRT-PCR, and replicate amplicon correlation for the quantitative expression component of an integrated comprehensive genomic profiling and quantitative transcriptional profiling (CGP+qTP) pre-approved test used to report the Immunotherapy Response Score (IRS). The IRS is reported from an integrated CGP+qTP test that combines comprehensive genomic profiling (CGP) from the analytically and clinically validated StrataNGS test with parallel quantitative transcriptional profiling (qTP) by multiplex RT-PCR-based next generation sequencing. Figure 46a represents an initial pre-clinical version of the qTP panel containing six "positive control" genes across two RNA primer pools previously used in the RNA fusion component (OPA positive) of the Oncomine Focus / Precision assay. To assess the suitability of these markers as pan-cancer housekeeping genes for quantitative expression profiling, we performed a multi-part evaluation of pan-cancer, pan-normal tissue stability transcriptome profiles. The mean expression levels (transcripts per million [TPM]) and coefficient of variation (CV) are shown for OPA-positive genes, additional candidate housekeeping genes from TCGA evaluation (TCGA stable), and the commonly used housekeeping gene GAPDH from >20,000 tumor, normal, and cancer cell line samples. Eight bolded genes were included in the gene expression panel used to develop the CGP+qTP test and IRS. Figure 46b shows a letter-value plot of normalized expression for the three final housekeeping genes used in the qTP panel (CIAO1, EIF2B1, and HMBS), and the remaining five candidates from the CGP+qTP clinical test over a consecutive 4-month period of samples with reportable quantitative expression using the current test version (n=3,417; regardless of tumor content). Figure 46c shows that the clinical accuracy of the qTP component was first determined by determining target gene expression consistent with hydrolysis probe-based qRT-PCR through expression validation on 24 FFPE tumor samples. The expression of each included target gene amplicon (n=32) is indicated by color.Concordance correlation coefficients for panel-wide validation and only four IRS expression biomarkers (PD-L1, PD-1, ADAM12 and TOP2A) are shown. Equivalence lines are shown. Figures 46d-f show that two separate PD-L1, PD-1 and ADAM12 amplicons are present in the current qTP panel (only one of the two ADAM12 amplicons was also present in all previous panels used to develop and validate the IRS). Since multiplex PCR-based qTP allows unambiguous read assignment to each target gene amplicon, we determined the correlation coefficients of replicate amplicons across 24,463 Strata Trial samples used to assess IRS distribution (n=7,911 samples in panels with both ADAM12 amplicons). Scatter plots are overlaid on density heatmaps. Equivalence lines are shown. [Figure 46b] Same as above. [Figure 46c] Same as above. [Fig. 46d-f] Same as above.

[0123] [Fig. 47a-b]Figures 47a-d depict the accuracy vs. clinical immunohistochemistry and reproducibility for the quantitative expression components of an integrated comprehensive genomic profiling and quantitative transcriptional profiling (CGP+qTP) non-approved test used to report the Immunotherapy Response Score (IRS). The IRS is reported from an integrated CGP+qTP test that combines comprehensive genomic profiling (CGP) from the analytically and clinically validated StrataNGS test with parallel quantitative transcriptional profiling (qTP) by multiplex RT-PCR-based next generation sequencing. Figure 47a depicts the accuracy of the PD-L1 qTP component of the IRS was validated against clinical IHC using a cohort of 276 non-small cell lung cancer (NSCLC) formalin-fixed paraffin-embedded (FFPE) tumor samples with reportable qTP (including tumor content [TC] ≥ 20%) in the accompanying pathology report and PD-L1 IHC expression by the 22C3 clone (using tumor proportion score [TPS]). Box plots of qTP PD-L1 expression stratified by TPS bins (0%, 1-49%, and ≥50%) are shown together with results from the Kruskal-Wallis [KW] test (by Jonkheel-Tapstra [JT] trend test [increasing median from 0%, 1-49%, and ≥50%]). Only 24 of these samples were derived from 154 patients in a propensity-matched first-line NSCLC treatment analysis, which precludes direct evaluation of IRS versus PD-L1 IHC to predict pembrolizumab benefit (see Figure 4). However, IRS status was generated for all 276 NSCLC samples with PD-L1 IHC, and the percentage of IRS-H samples by TPS bin is shown in dark blue. Figure 47b shows that the accuracy of the TOP2A qTP component of the IRS was validated against clinical IHC using a cohort of 956 FFPE tumor tissue samples (36 tumor types) with reportable qTP (including TC > 20%) with a proliferation index (percentage of Ki67 positive tumor cells) in the accompanying pathology report.Pearson correlation coefficients of qTP TOP2A expression vs. clinical proliferation index from scatter plots are shown with 95% confidence intervals [CI] and p-values, with a line of best fit shown by dots and dashed lines overlaid with density heatmap. Figure 47c shows that panel-wide qTP reproducibility across operators, lots, and instruments was established using separate replicate nucleic acid aliquots isolated from FFPE tumor samples. 27 unique samples were assessed by two operators on different days, using different laboratory-prepared instruments, different laboratory-prepared reagent lots, and different template and sequencing lots and instruments. For each sample, the maximum and minimum nRPM for each target gene across all replicates was plotted (individual target gene amplicons are indicated by color) and the concordance correlation coefficient was determined. As in Figure 47c, except that Figure 47d, IRS reproducibility was determined by plotting the maximum and minimum IRS across all replicates for each sample, and the concordance correlation coefficient was determined. Qualitative agreement of IRS status (high vs. low) from maximum and minimum IRS scores across all iterations was also determined. [Fig. 47c-d] Same as above.

[0124] [Figure 48] Figure 48 depicts Cox proportional hazards regression with Lasso penalization for immunotherapy response score (IRS) development. To develop a benefit prediction model of integrated PD-1 / PD-L1 blockade, we performed Cox proportional hazards regression with Lasso penalization with 5-fold cross-validation in the pembrolizumab (PD-1 therapy) discovery cohort of 648 patients to perform feature selection from tumor mutation burden (TMB; log2) and 23 candidate immune and proliferation expression biomarkers associated with TTNT of pembrolizumab. The Lasso penalty condition was selected as the value that maximized the concordance index (upper panel; gray line) via 5-fold cross-validation with coefficients shown for TMB and 23 candidate expression biomarkers versus alpha (α) (lower panel), resulting in a 5-condition model including TMB, PD-1, PD-L1, ADAM12, and TOP2A.

[0125] [Figure 49] Figures 49a-b depict pembrolizumab overall survival (OS) by immunotherapy response score (IRS) status. Figure 49a depicts pembrolizumab OS in the discovery cohort stratified by IRS group as shown by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​(adjusted by variables shown in b) for IRS-H vs.-L. The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. Figure 49b depicts the ability of IRS to stratify pembrolizumab OS was assessed using a forest plot of variables included in the adjusted Cox proportional hazards model. Adjusted hazard ratios with 95% confidence intervals (CIs) are shown for each variable. Statistically significant variables are in bold.

[0126] [Fig. 50a-b] Figures 50a-d represent real-world clinical progression-free survival (rwPFS) and overall survival (OS) by immunotherapy response score (IRS) status in the validation cohort stratified by PD-1 vs. PD-L1 therapy. Figure 50a represents rwPFS (by time to next therapy) for the monotherapy PD-L1-treated subset of the validation cohort stratified by IRS group, as shown by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​for the IRS-H vs. -L group. The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. Figure 50b, as in Figure 50a, except for assessing OS. Figures 50c and d, as in Figures 50a and b, except for assessing TTNT (Figure 50c) and OS (Figure 50d) for the monotherapy PD-1-treated subset of the validation cohort. In addition to adjusted HRs and p-values, log-rank p-values ​​are also shown. [Fig. 50c-d] Same as above.

[0127] [Fig. 51a-b]Figures 51a-d represent real-world clinical progression-free survival (rwPFS) and overall survival (OS) of PD-(L)1 monotherapy by Immunotherapy Response Score (IRS) status and Tumor Mutational Burden (TMB). Figure 51a depicts rwPFS of pembrolizumab monotherapy in the discovery cohort stratified by IRS (IRS-high [-H] vs. -low [L]) and TMB (TMB-H [>= 10 mutations per megabase] vs. TMB-L as shown by Kaplan-Meier analysis. Benjamini-Hochberg (BH) adjusted p-values ​​for pairwise log-rank tests between the IRS-H / TMB-H and IRS-H / TMB-L groups are shown. Number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for the analyzed groups are shown. Figure 51b, as in Figure 51a, except for OS. Figures 51c and d, as in Figures 51a and b, except evaluating rwPFS (Figure 51c) and OS (Figure 51d) in an independent validation cohort of patients treated with non-pembrolizumab PD-(L)1 monotherapy. [Fig. 51c-d] Same as above.

[0128] [Fig. 52a-b]Figures 52a-d show that CDKN2A deep deletion status does not add to the immunotherapy response score (IRS) for predicting real-world progression-free survival (rwPFS) or overall survival (OS) of PD-(L)1 monotherapy. Figure 52a shows the rwPFS of pembrolizumab monotherapy in a subset (n=310) of the discovery cohort evaluable for CDKN2A deep deletion (equivalent to 2-copy deletion if homozygous / diploid) status (≧40% tumor content and evaluable copy number alteration) stratified by IRS group, as shown by Kaplan-Meier analysis with adjusted hazard ratios (HR) and p-values ​​for IRS-high vs. -L groups. The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. Forest plots show adjusted hazard ratios with 95% confidence intervals (CIs) for IRS (IRS-H vs. IRS-L) and CDKN2A deep deletion status (CDKN2A deep deletion present vs. CDKN2A deep deletion absent) in the same adjusted model. Figure 52b, as in Figure 52a, but assessing OS. Figures 52c and d, as in Figures 52a and b, but assessing TTNT (Figure 52c) and OS (Figure 52d) in a subset (n=199) of the independent validation cohort evaluable for CDKN2A deep deletions treated with non-pembrolizumab PD-(L)1 monotherapy. [Fig. 52c-d] Same as above.

[0129] [Figure 53]Figure 53a-b shows that the immunotherapy response score (IRS) is robust to the timing of pre-PD-(L)1 sample collection. Figure 53a shows the Pearson correlation of IRS from clonal tumor specimens from the same patient with different collection dates and without checkpoint inhibitor therapy (n=104 patients) during the sample collection dates tested. Figure 53b shows the rwPFS of PD-(L)1 stratified by IRS group in 181 patients who had their samples collected after the start of PD-(L)1 therapy but were otherwise included in the discovery or validation cohorts, as shown by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​for IRS-H vs. IRS-L. The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown.

[0130] [Fig. 54a-b]Figures 54a-e show that the Immunotherapy Response Score (IRS) is robust to variable tumor content. Figure 54a shows that the continuous tumor content condition was included in the adjusted Cox proportional hazards (CPH) model for real-world progression-free survival (rwPFS; by time to next therapy) of pembrolizumab in the entire discovery cohort (including age, sex, most common tumor type [NSCLC] vs. other, type of therapy [monotherapy / combination], and line of therapy). Adjusted hazard ratios with 95% confidence intervals (CI) are shown for each variable, with statistically significant variables in bold. Figures 54b-d show Kaplan-Meier analysis of rwPFS of pembrolizumab binned by tumor content (20-35%, 40-70%, and >70%) and stratified by IRS group. The number of patients (n), events, and median rwPFS (with 95% confidence interval [CI]) for each group are shown. Figure 54e shows the rwPFS of PD-(L)1 stratified by IRS group in 64 patients who were otherwise included in the discovery or validation cohort except for those whose tested samples had <20% tumor content, as shown by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​for IRS-H vs. -L groups. The number of patients (n), events, and median rwPFS (with 95% confidence interval [CI]) for each group are shown. [Fig. 54c-d] Same as above. [Figure 54e] Same as above.

[0131] [Fig. 55a-b]Figures 55a-d present additional analyses supporting the predictive nature of the Immunotherapy Response Score (IRS) biomarker. Figures 55a and b: To establish the predictive nature of the IRS model, we evaluated an internal control cohort for the pembrolizumab monotherapy cohort, consisting of 146 patients who received a previous line of systemic therapy prior to monotherapy pembrolizumab therapy. For each patient, real-world progression-free survival (rwPFS) was determined for the previous line of systemic therapy prior to pembrolizumab and for the pembrolizumab monotherapy line, with rwPFS stratified by IRS status (see Figures 37a-c). Here, we show a Kaplan-Meier analysis of rwPFS of pembrolizumab monotherapy (purple) versus rwPFS of prior systemic therapy (yellow) in a subset of non-high-frequency microsatellite instability (MSI-H) tumors in tumor types approved for non-PD-(L)1 monotherapy, along with log-rank p-values ​​between pembrolizumab and prior therapy in (Fig. 55b) IRS-high [H] and (Fig. 55a) IRS-low [L] populations. The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. To confirm the predictive nature of the IRS model, Fig. 55c, we determined rwPFS in 3,184 patients in SCMD treated with systemic first-line non-immunotherapy (IO) who otherwise met the criteria for the discovery and validation cohorts. Kaplan-Meier analysis of rwPFS of non-IO systemic first-line stratified by IRS status with adjusted hazard ratios (HRs) and p-values ​​for IRS-H vs. IRS-L groups. Figure 55d shows rwPFS of ipilimumab + nivolumab (ipi+nivo) in 70 patients who received combination ipilimumab (CTLA4) + nivolumab (PD-1) therapy but were otherwise eligible for the validation cohort, stratified by IRS status, as shown by Kaplan-Meier analysis with adjusted hazard ratios (HRs) and p-values ​​for IRS-H vs. -L groups. [Fig. 55c-d] Same as above.

[0132] [Fig. 56a-c]Figure 56a-g represent the clinical utility of integrated comprehensive genomic profiling and quantitative transcriptional profiling (CGP+qTP) outside of immunotherapy treatment decision making. IRS are reported from an integrated CGP+qTP test, which combines comprehensive genomic profiling (CGP) from the analytically and clinically validated StrataNGS test with parallel quantitative transcriptional profiling (qTP) by multiplex RT-PCR-based next generation sequencing. Figure 56a represents the accuracy of ESR1 (estrogen receptor; ER) by qTP validated against clinical IHC using a cohort of 300 breast cancer formalin-fixed paraffin-embedded (FFPE) tumor samples with reportable qTP (including tumor content [TC] ≥ 20%) and ER IHC expression (by tumor cell positivity %) in the accompanying pathology report. The entire cohort was used for accuracy; however, the cohort was randomly split into equivalent training (n=150) and validation (n=150) cohorts to establish clinical validity before accuracy assessment (see FIG. 56d). Correlation coefficients of qTP ER expression vs. clinical ER tumor cell positivity (log2) from scatter plots are shown with 95% confidence intervals [CI] and p-values, with a line of best fit indicated by dots and dashed lines overlaid on the density heatmap. FIG. 56b depicts the accuracy of PGR (progesterone receptor; PR) by qTP validated against clinical IHC using a cohort of 291 breast cancer formalin-fixed paraffin-embedded (FFPE) tumor samples with reportable qTP (including tumor content [TC]≧20%) and PR IHC expression (by tumor cell positivity) in the accompanying pathology report. The entire cohort was used for accuracy; however, the cohort was randomly split into equivalent training (n=145) and validation (n=146) cohorts to establish clinical validity prior to accuracy assessment (see FIG. 56e). Correlation coefficients of qTP PR expression vs. clinical PR tumor cell positivity % (log2) from scatter plots are shown with 95% confidence intervals [CI] and p-values, with a line of best fit indicated by dots and dashed lines overlaid on the density heatmap.Figure 56c depicts the accuracy of HER2 (ERBB2) by qTP validated against clinical IHC using a cohort of 545 breast cancer formalin-fixed paraffin-embedded (FFPE) tumor samples with reportable qTP (including tumor content [TC] > 20%) and HER2 IHC expression (0, 1+, 2+ or 3+) in the accompanying pathology report. The entire cohort was used for accuracy, however, the cohort was randomly split into equivalent training (n = 273) and validation (n = 272) cohorts to establish clinical validity before accuracy assessment (see Figure 56f). Box plots of qTP HER2 expression stratified by clinical IHC category are shown along with Kruskal-Wallis (KW) test p-values ​​and Jonkheel-Tapstra [JT] trend test p-values ​​(increasing median from 0 to 1+ to 2+ to 3+) in the accuracy cohort. Figure 56d represents that clinical validity for ER status by qTP was established by setting thresholds for qTP ER negative (<12.75; green dashed line) and positive (>14.5; red dashed line) in a training cohort (n=150) of breast cancer FFPE tissue samples with clinical ER status (by tumor cell positivity %) in the accompanying pathology report (see Figure 56a) based on clinical IHC-defined categories of ER negative (Neg.; 0%), low (1-10%) and positive (Pos; >10%). Expressions between the negative and positive thresholds were defined as indeterminate qTP ER (light grey). The desired sensitivity (sens; percent positive agreement [PPA]) and specificity (spec; percent negative agreement [NPA]) for qTP ER negative / positive status (vs. IHC negative and positive), respectively, were prespecified as >95%. The locked threshold was then applied to the validation cohort (n=150) and box plots of qTP ER expression by clinical IHC category are shown, along with PPA and NPA values ​​and 95% confidence intervals (CI). In this validation cohort, the qTP ER indeterminate category (n=8 of 150 validation samples) accurately identified 6 / 7 clinical IHC ER-low samples, supporting the clinical utility of this category.Fig. 56e, clinical validity for PR status by qTP was established by setting a threshold for qTP PR negativity (<12.3; red dashed line) in a training cohort of breast cancer FFPE tissue samples (n=145) with clinical PR status (by tumor cell positivity %) in the accompanying pathology report (see b). Although PR does not have a group reporting "low" clinical IHC, three clinical IHC-defined categories of PR negative (Neg.; 0%), low (1-10%) and positive (Pos; >10%) were used in the training cohort to facilitate appropriate balancing of PPA and NPA in threshold setting. Since the potential clinical significance of a false-positive PR status, i.e., inappropriately considering ER-negative / HER2-negative breast cancer as hormone receptor positive (vs. triple negative), is more influential than a false-negative PR status (which is unclear if ER-negative / PR-positive breast cancer is biologically plausible), the threshold was set to favor NPA, with a predefined acceptable NPA (vs. PR 0% IHC) set at >95%. The locked threshold was then applied to the validation cohort (n=146), and box plots of qTP PR expression by IHC category are shown, along with PPA and NPA values, and 95% confidence intervals (CI). f) Clinical validity for HER2 status by qTP was established by setting the threshold in a training cohort (n=273) of breast cancer FFPE tissue samples with clinical HER2 status (by clinically recognized categories of 0, 1+, 2+ or 3+) in the accompanying pathology report (see c).Given that, similar to ER, the clinical utility of HER2 IHC 2+ reflects the uncertain validity of FISH / ISH (and StrataNGS provides ERBB2 copy status) and 0 vs. 1+ expression in retrospective samples clinically scored prior to FDA approval of trastuzumab deruxtecan in HER2 1+ and 2+ (FISH / ISH negative) breast cancer, we set thresholds for qTP HER2 low (<18.0; green dashed line) and high (>19.2; red dashed line), with expression between these thresholds reported as qTP HER2 indeterminate (light grey); thresholds were set by balancing the desired maximum sensitivity vs. IHC 3+ with the observation that the majority of IHC 3+ tumors with the lowest qTP HER2 expression also lacked ERBB2 amplification in the training cohort. Therefore, the desired NPA and PPA for qTP HER2 low / high status (vs. IHC 0-1+ and 3+) were predefined as NPA>95% and PPA>70%; no performance metrics were predefined for IHC 2+ samples. The locked thresholds were then applied to the validation cohort (n=272 [including 51 IHC 2+ not formally assessed]) and box plots of qTP HER2 expression by clinical IHC category, stratified by ERBB2 copy number status (red=amplified, green=no amplification [wild type; wt], grey=non-evaluable copy number status), are shown along with PPA and NPA values ​​and 95% confidence intervals (CIs). Notably, of these three false-negative samples in the validation cohort (IHC 3+ but qTP HER2 low), there are two missing ERBB2 amplifications by StrataNGS testing. In addition, more than 50% (n=7) of the qTP HER2 indeterminate category (n=12 of 272 total validated samples) were IHC 2+, supporting the clinical utility of this category and its deferral to amplification status.Figure 56g, The above analyses support the clinical utility of ER and PR (collectively hormone receptor [HR]), as well as HER2 status by qTP, as the clinical utility of these biomarkers has already been established, but as an additional demonstration of the clinical utility of integrating qTP results with CGP results, we determined the impact of qTP HR status on the association of PIK3CA mutation treatment in patients with breast cancer (standard of care [SOC] PIK3CA mutations associated with the FDA-approved alpelisib + fulvestrant regimen in patients with hormone receptor positive / HER2 negative breast cancer only [green box]) from consecutively tested pan-solid FFPE tumor samples (n=3,904) over a 4-month period submitted for clinical CGP testing. Of the 3,904 samples, 288 samples were breast cancer and met the necessary qTP and CGP QC metrics (including the requirement of a final ≥20% tumor content) to assess HR status, PIK3CA mutations, and ERBB2 copy number status as shown in the sample layout diagram, of which 31% (n=90) harbored SOC PIK3CA mutations associated with alpelisib therapy. Of the 90 PIK3CA-mutated samples, 2 (2%) were correctly identified as not associated with alpelisib therapy by CGP testing alone (based on the presence of ERBB2 amplification; pink box), whereas only 11 (12%) were correctly identified as not associated with alpelisib therapy by integration of qTP findings (based on HR-negative status; dark red box). [Fig. 56d-f] Same as above. [Fig. 56g] Same as above.

[0133] [Fig. 57a-b]Figures 57a-c represent exploratory analyses defining the Immunotherapy Response Score (IRS) very-low subset. In post-hoc exploratory analyses of combined discovery (n=648; pembrolizumab [pembro]-treated) and validation (n=248; non-pembrolizumab PD-[L]1-treated) cohorts, we identified a threshold (<0.41) that subdivided the IRS-low (-L) group into intermediate (IRS-L[I]) and very-low (IRS-L[U]) subsets. Figure 57a shows the real clinical progression-free survival (rwPFS) of PD-(L)1 in the combined cohort stratified by IRS-high [H], IRS-L(I), and IRS-L(U) groups, as shown by Kaplan-Meier analysis with Benjamini-Hochberg (BH) adjusted p-values ​​for pairwise log-rank tests between IRS-L(I) vs. IRS-L(U) groups, and adjusted hazard ratios (HRs) and p-values ​​for IRS-L(I) vs. IRS-L(U) groups shown. The Cox proportional hazards model was adjusted for age, sex, most common tumor type (NSCLC vs. other), line of therapy, type of therapy (monotherapy vs. combination therapy), and IRS (-H, -L[I], and -L[U]). The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for the analyzed groups are shown. Fig. 57b, As in Fig. 57a, except for overall survival (OS). Fig. 57c, This three-group IRS classification was applied to all 24,463 patients in the Strata Clinical Molecular Database (SCMD) with valid tumor mutation burden (TMB) and gene expression data. The distribution of IRS groups is shown by box plots (numbers indicate percentages). Hierarchy and decomposition of approved tumor types for PD-(L)1 monotherapy is shown. Tumor type abbreviations: NSCLC (non-small cell lung cancer), RCC (renal cell carcinoma), NMSC (non-melanoma skin cancer), SCLC (small cell lung cancer). [Fig. 57c] Same as above.

[0134] [Fig. 58a-b]Figure 58a-d represents confirmation of the predictive nature of the Immunotherapy Response Score (IRS) biomarker when the very-low subset is defined. To establish the predictive nature of the IRS model, we evaluated the internal comparator in the pembrolizumab monotherapy cohort consisting of 146 patients who received a previous line of systemic therapy before pembrolizumab (pembro) monotherapy. Here, we subdivided the IRS-low (-L) group into intermediate (IRS-L[I]) and very-low (IRS-L[U]) subsets as defined in Figure 57a-c. Figure 58a represents that for each patient, rwPFS was determined for the line of systemic therapy immediately preceding pembrolizumab and the line of pembrolizumab monotherapy, and rwPFS was stratified by IRS status. Figure 58a represents a Kaplan-Meier analysis of rwPFS of previous systemic therapy in the IRS-high [H], IRS-L(I), and IRS-L(U) groups (overall log-rank p-values ​​shown). The number of patients (n), events, and median rwPFS (with 95% confidence intervals [CI]) for each group are shown. Figures 58b-d represent a Kaplan-Meier analysis of rwPFS of pembrolizumab monotherapy (purple) versus previous systemic therapy (yellow) in the (Figure 58b) IRS-L(U), (Figure 58c) IRS-L(I), and (Figure 58d) IRS-H groups of patients (log-rank p-values ​​shown). [Fig. 58c-d] Same as above. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0135] Detailed Description of the Invention

[0136] Immune checkpoint inhibitors are FDA approved and provide clinical benefit across a wide range of tumor types. However, in most indication tumor types, only a small number of patients benefit, and additional patients benefit outside of indication tumor types. Therefore, improved diagnostic tools are needed to select patients for immunotherapy treatment. Utilizing real-world pembrolizumab outcome data combined with DNA mutation and RNA expression data from clinical NGS testing for 610 diverse solid tumor patients, we demonstrated that TMB, PD-L1 and PD-L2 are independent predictors of treatment benefit, and that multivariate immunotherapy response score (IRS) predicted the benefit of pembrolizumab compared to chemotherapy across solid tumors. IRS scores were characterized across nearly 20,000 advanced solid tumors, showing that a proportion of patients in the high IRS group predicted the observed tumor type response rate of pembrolizumab. In another embodiment, utilizing PD-(L)1 therapy treatment data and CGP and quantitative transcriptome profiling (CGP+qTP) data from the Strata Trial (NCT03061305), enabled the development and cross-validation of an integrated immunotherapy response score (IRS) predicting PD-(L)1 benefit in pan-solid tumors by both real-world progression-free survival (rwPFS) and overall survival (OS) with an analytically and clinically validated CGP+qTP non-regulated test (LDT) applicable to small formalin-fixed paraffin-embedded (FFPE) tissue specimens. The IRS diagnostic algorithm disclosed herein significantly improves patient selection for immunotherapy.

[0137] Some aspects of the present invention are directed to a method of treatment comprising the steps of: (a)(i) measuring expression levels of RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12 obtained from a tumor specimen from a subject; (b) measuring the tumor mutational burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; (c) calculating an Immunotherapy Response Score (IRS) from the expression levels of the RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12 obtained in step (a) and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and (d) administering checkpoint inhibitor therapy to the subject.

[0138] In some embodiments, measuring the expression levels of RNA transcripts comprises measuring the expression levels of RNA transcripts for at least three of PD-1, TOP2A, PD-L1 and ADAM12, and calculating the IRS comprises calculating the IRS from the expression levels of the RNA transcripts for at least three of PD-1, TOP2A, PD-L1 and ADAM12, and the converted TMB measurement. In some embodiments, measuring the expression levels of RNA transcripts comprises measuring the expression levels of RNA transcripts for all of PD-1, TOP2A, PD-L1 and ADAM12, and calculating the IRS comprises calculating the IRS from the expression levels of RNA transcripts for all of PD-1, TOP2A, PD-L1 and ADAM12, and the converted TMB measurement. In some embodiments, measuring the expression levels of RNA transcripts comprises measuring the expression levels of RNA transcripts for at least PD-1 and PD-L1, and calculating the IRS comprises calculating the IRS from the expression levels of RNA transcripts for at least PD-1 and PD-L1, and the converted TMB measurement.

[0139] In some embodiments, measuring the expression level of the RNA transcript comprises measuring the expression level of the RNA transcript of PD-1, PD-L1, and ADAM12, and calculating the IRS comprises calculating the IRS from the expression levels of the RNA transcript of PD-1, PD-L1, and ADAM12, and the converted TMB measurement. In some embodiments, the expression level of both PD-1 and PD-L1 positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, a high conversion level of TMB positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the expression level of both PD-1 and PD-L1 positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, as does a high conversion level of TMB. In some embodiments, the expression level of ADAM12 negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, PD-L and PD-L1 expression levels, as well as high TMB transformation levels, both positively correlate with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, while ADAM12 expression levels negatively correlate with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy.

[0140] In some embodiments, step (a) further comprises: ii) measuring the expression level of an RNA transcript for at least one reference gene in the biological sample; and iii) normalizing the measured expression levels of at least two, at least three, or all of the measured RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12 to the level of the RNA transcript of the at least one reference gene to provide normalized expression levels of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12. In some embodiments, the one or more reference genes comprise 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or more housekeeping genes. In some embodiments, the reference genes are selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC, and SLC4A1AP. In some embodiments, one or more reference genes comprise three or more of CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC and SLC4A1AP.In some embodiments, one or more reference genes comprise a combination of CIAO1, EIF2B1 and HMBS.Thus, in some embodiments, the expression level of RNA transcript used to calculate IRS comprises the normalized expression level of RNA transcript.

[0141] In some embodiments, the normalized expression levels of both PD-1 and PD-L1, as well as high conversion levels of TMB, are positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the normalized expression levels of both ADAM12 and TOP2A are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the normalized expression levels of PD-L and PD-L1, as well as high conversion levels of TMB, are positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, while the expression levels of ADAM12 and TOP2A are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy.

[0142] In some embodiments, step (a) further comprises the step of: iv) median centering the measured expression levels of at least two, at least three, or all of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12, before or after normalizing the expression levels of the measured RNA transcripts.

[0143] In some embodiments, step (a) further comprises v) log2 transforming the measured expression levels, median centered expression levels, normalized expression levels or median centered normalized expression levels of at least two, at least three, or all of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12, wherein the expression levels utilized to calculate the IRS in step (c) are the transformed expression levels, transformed median centered expression levels, transformed normalized expression levels or transformed median centered normalized expression levels. In some embodiments, the transformed expression levels, transformed median centered expression levels, transformed normalized expression levels or transformed median centered normalized expression levels of both PD-1 and PD-L1, and high transformed levels of TMB positively correlate with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the transformed, transformed median-centered, transformed normalized, or transformed median-centered normalized expression levels of both ADAM12 and TOP2A negatively correlate with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the transformed, transformed median-centered, transformed normalized, or transformed median-centered normalized expression levels of PD-L and PD-L1, and high transformed levels of TMB positively correlate with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, while the transformed, transformed median-centered, transformed normalized, or transformed median-centered normalized expression levels of ADAM12 and TOP2A negatively correlate with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy.

[0144] In some embodiments, the determination that the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy is based on the IRS exceeding a pre-established threshold. The IRS itself is derived from a Cox proportional hazards model, where the IRS is a function of the hazard function H(t), where H(t)=H0(t)×exp(b1x2+b2x2+···b n x n ) where H(t) is the baseline hazard and (x1, x2, . . . , x n ) are the covariates that determine the hazard at time (t), and (b1, b2, . . . , b n ) are coefficients indicating the influence of different covariates in determining the hazard at time (t). In some embodiments, the IRS may be derived from the Cox proportional hazards model and represent the relative risk H(t) / H0(t), where the IRS is H(t) / H0(t)=exp(b1x2+b2x2+···b n x n In some further embodiments, the IRS may be derived from the Cox proportional hazards model and represent the natural logarithm of the relative risk H(t) / H0(t), such that the IRS is ln[H(t) / H0(t)]=(b1x2+b2x2+···b n x n ) As such, one of skill in the art will recognize that the threshold utilized to determine that the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy will vary based on how the IRS is derived from the Cox proportional hazards model.

[0145] In some embodiments, the IRS is calculated as follows: IRS=approximately 0.27×[transformed TMB measurement]+approximately 0.11×[transformed PD-1 level]+approximately 0.06×[transformed PD-L1 level]−approximately 0.06[transformed ADAM12 level]−approximately 0.077×[transformed TOP2A level], where the transformed PD-1 level, transformed PD-L1 level, transformed ADAM12 level, transformed TOP2A level is each one of the transformed expression level, transformed median centered expression level, transformed normalized expression level, or transformed median centered normalized expression level.

[0146] In some embodiments, the IRS is calculated as follows: IRS=0.27×[transformed TMB measurement]+0.11×[transformed PD-1 level]+0.06×[transformed PD-L1 level]−0.06[transformed ADAM12 level]−0.077×[transformed TOP2A level], where the transformed PD-1 level, transformed PD-L1 level, transformed ADAM12 level, transformed TOP2A level are each one of the transformed expression level, transformed median centered expression level, transformed normalized expression level, or transformed median centered normalized expression level.

[0147] In some embodiments, the IRS is calculated as follows: IRS=0.273758×[transformed TMB measurement]+0.112641×[transformed PD-1 level]+0.061904×[transformed PD-L1 level]−0.057991[transformed ADAM12 level]−0.077011×[transformed TOP2A level], where the transformed PD-1 level, transformed PD-L1 level, transformed ADAM12 level, transformed TOP2A level is one of the transformed expression level, transformed median centered expression level, transformed normalized expression level, or transformed median centered normalized expression level.

[0148] In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.70, 0.75, 0.80, 0.85, 0.90, 0.95, 1.00, 1.05, 1.10, 1.15, 1.20, 1.25, 1.30 or more. In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.80, 0.81, 0.82, 0.83, 0.84, 0.85, 0.86, 0.87, 0.88, 0.89, 0.90, 0.91 or more. In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.87 or more. In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.873569 or more.

[0149] Some aspects of the present invention are directed to methods of treatment comprising: a. measuring expression levels of RNA transcripts for PD-1 and PD-L2 obtained from a tumor specimen from a subject; b. measuring the tumor mutational burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; c. calculating an Immunotherapy Response Score (IRS) from the expression levels or normalized levels of the RNA transcripts of PD-1 and PD-L2 and the transformed TMB measurement, obtaining an IRS that positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and d. administering checkpoint inhibitor therapy to the subject.

[0150] In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0151] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM12. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]).

[0152] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, each of the transformed TMB expression, PD-1 expression, PD-L1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized PD-L1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0153] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0154] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured. In some embodiments, each of the converted TMB expression, PD-1 expression, PD-L1, and PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the converted TMB expression, normalized PD-1 expression, normalized PD-L1, and normalized PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0155] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4 and ADAM12 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0156] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, each of the transformed TMB expression, PD-1 expression, VTCN1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized VTCN1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0157] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12 and VTCN1 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the normalized CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0158] In some embodiments of each of the methods disclosed herein, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is 6 or more, 7 or more, 8 or more, 9 or more, 9.5 or more, 10 or more, 10.5 or more, 11 or more, 12 or more, 13 or more, 14 or more, 15 or more, or 20 or more. In some embodiments of the methods disclosed herein, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is 8 or more. In some embodiments of the methods disclosed herein, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is 10 or more. In some embodiments of the methods disclosed herein, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is 12 or more. In some embodiments of the methods disclosed herein, the IRS is calculated as 8, 10, or 12 times the inverse of the hazard ratio of a patient compared to the median hazard ratio using a Cox model. In some embodiments of the methods disclosed herein, the IRS is calculated as 10 times the inverse of the hazard ratio of a patient compared to the median hazard ratio using a Cox model. In some embodiments of the methods disclosed herein, the IRS is calculated as 8 to 12 times the inverse of the hazard ratio of a patient compared to the median hazard ratio using a Cox model.

[0159] In some embodiments, the tumor specimen is a formalin-fixed paraffin-embedded (FFPE) tumor specimen. In some embodiments, the tumor specimen contains at least 20% tumor content. In some embodiments, the tumor specimen is breast cancer, central or peripheral nervous system cancer, cancer of unknown primary site, colorectal cancer, endometrial cancer, gastrointestinal stromal tumor, glioma, hepatobiliary cancer, neuroendocrine cancer, ovarian cancer, pancreatic cancer, prostate cancer, salivary gland cancer, sarcoma, or thyroid cancer. In some embodiments, the tumor specimen is adrenal cancer, biliary tract cancer, bladder cancer, brain cancer, breast cancer, cervical cancer, colon cancer, rectal cancer, endometrial cancer, esophageal cancer, head and neck cancer, kidney cancer, liver cancer, non-small cell lung cancer, lung cancer, lymphoma, melanoma, meninges cancer, non-melanoma skin cancer, ovarian cancer, pancreatic cancer, prostate cancer, sarcoma, small intestine cancer, or gastric cancer.

[0160] In some embodiments, the tumor specimen is evaluated as having low frequency microsatellite instability or microsatellite stability. In some embodiments, the tumor specimen is evaluated as having low tumor mutation burden, and low tumor mutation burden is classified as having less than 10 mutations per megabase (mut / Mb). In some embodiments, the expression level of RNA transcripts is measured using PCR and next generation sequencing.

[0161] In some embodiments, the checkpoint inhibitor is an anti-PD-1 antibody, an anti-CTLA-4 antibody, an anti-PD-L1 antibody, or an anti-PD-L2. In some embodiments, the checkpoint inhibitor is nivolumab, pembrolizumab, atezolizumab, cemiplimab, durvalumab, pidilizumab, atezolimumab, PDR001, BMS-936559, avelumab, ipilimumab, or SHR-1210.

[0162] In some embodiments, the checkpoint inhibitor therapy is administered as a monotherapy. In some embodiments, the checkpoint inhibitor therapy is administered in combination with one or more other chemotherapeutic agents.

[0163] In some embodiments, the tumor specimen exhibits a TPS score of 1-49%. In some embodiments, the checkpoint inhibitor is administered as part of a first line treatment regimen. In some embodiments, the checkpoint inhibitor is administered as part of a second line or subsequent treatment regimen.

[0164] In some embodiments of the method disclosed herein, step a. further comprises measuring the expression level of the RNA transcript for at least one reference gene in the biological sample, step b. comprises normalizing the measured expression levels of the other measured RNA transcripts to the level of the RNA transcript of the at least one reference gene to provide a normalized level of the other RNA transcript, and step c. comprises calculating the IRS from the normalized level. In some embodiments, the one or more reference genes comprise three genes selected from LRP1, MRPL13, TBP, HMBS, ITGB7, MYC, CIAO1, CTCF, EIF2B1, GGNBP2, SLC4A1AP. In some embodiments, the one or more reference genes comprise three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC and SLC4A1AP. In some embodiments, the one or more reference genes comprise a combination of CIAO1, EIF2B1 and HMBS.

[0165] Some aspects of the disclosure include a method of identifying a subject who would benefit from checkpoint inhibitor therapy, comprising the steps of: (a) measuring expression levels of RNA transcripts for PD-1, TOP2A, PD-L1 and ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from the subject; (b) log2-transform, median-center, and normalize the measured expression levels of the RNA transcripts to the levels of the RNA transcripts of the at least one reference gene to provide normalized levels of the RNA transcripts; (c) measuring tumor mutational burden (TMB) in the biological sample; (d) calculating an Immunotherapy Response Score (IRS) from the transformed normalized levels of RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12, and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and (e) identifying subjects as benefiting from checkpoint inhibitor therapy.

[0166] In some embodiments, the tumor specimen is from a cancer that is not approved for the indicated use of checkpoint inhibitor therapy. In some embodiments, the tumor specimen is assessed as having low frequency microsatellite instability or microsatellite stability. In some embodiments, the tumor specimen is assessed as having low tumor mutational burden, where low tumor mutational burden is classified as less than 10 mutations per megabase (mut / Mb). In some embodiments, the tumor specimen exhibits a TPS score of 1-49%.

[0167] Some aspects of the disclosure include a method of identifying a subject that would benefit from checkpoint inhibitor therapy, comprising the steps of: (a) receiving, by a processor, measured expression levels of RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; (b) log2 transforming, median centering, and normalizing, by the processor, the measured expression levels of the RNA transcripts to the level of the RNA transcript of the at least one reference gene to provide transformed normalized levels of the RNA transcripts; (c) receiving, by the processor, measured expression levels of the RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from the subject; The present invention is directed to a method comprising the steps of: receiving a measured tumor mutation burden (TMB) in a body sample; (d) log2 transforming, by a processor, the TMB measurement to provide a transformed TMB measurement; (e) calculating, by the processor, an immunotherapy response score (IRS) from the transformed normalized levels of RNA transcripts of at least two of PD-1, TOP2A, PD-L1, and ADAM12, and the transformed TMB measurement, which positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy; and (f) providing a determination that the subject has a checkpoint inhibitor responsive cancer.

[0168] In some embodiments, the IRS is calculated as follows: IRS=0.27×[transformed TMB measurement]+0.11×[transformed normalized level of PD-1]+0.06×[transformed normalized level of PD-L1]−0.06[transformed normalized level of ADAM12]−0.077×[transformed normalized level of TOP2A].

[0169] In some embodiments, the IRS is calculated as follows: IRS=0.273758×[transformed TMB measurement]+0.112641×[transformed normalized level of PD-1]+0.061904×[transformed normalized level of PD-L1]−0.057991[transformed normalized level of ADAM12]−0.077011×[transformed normalized level of TOP2A].

[0170] In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.87 or greater.In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.873569 or greater.In some embodiments, the one or more reference genes comprise three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC and SLC4A1AP.In some embodiments, the one or more reference genes comprise a combination of CIAO1, EIF2B1 and HMBS.

[0171] Some aspects of the disclosure include a method of treating a subject in need of treatment with checkpoint inhibitor therapy, comprising administering checkpoint inhibitor therapy to the subject, wherein the subject in need of treatment comprises: (a) measuring expression levels of RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; (b) log2-transforming, median-centering, and normalizing the measured expression levels of the RNA transcripts to the level of the RNA transcript of the at least one reference gene to obtain the transformed normalized levels of the RNA transcripts. (c) measuring tumor mutation burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; (d) calculating an Immunotherapy Response Score (IRS) from the transformed normalized levels of RNA transcripts of at least two of PD-1, TOP2A, PD-L1, and ADAM12, and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and obtaining an IRS having a value indicative of a beneficial response to checkpoint inhibitor therapy.

[0172] In some embodiments, the IRS is calculated as follows: IRS=0.27×[transformed TMB measurement]+0.11×[transformed normalized level of PD-1]+0.06×[transformed normalized level of PD-L1]−0.06[transformed normalized level of ADAM12]−0.077×[transformed normalized level of TOP2A].

[0173] In some embodiments, the IRS is calculated as follows: IRS=0.273758×[transformed TMB measurement]+0.112641×[transformed normalized level of PD-1]+0.061904×[transformed normalized level of PD-L1]−0.057991[transformed normalized level of ADAM12]−0.077011×[transformed normalized level of TOP2A].

[0174] In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.87 or greater. In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.873569 or greater.

[0175] In some embodiments, the one or more reference genes comprise three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC and SLC4A1AP.In some embodiments, the one or more reference genes comprise a combination of HMBS, CIAO1 and EIF2B1.

[0176] In some embodiments, the tumor specimen is from a cancer that is not approved for the indicated use of checkpoint inhibitor therapy. In some embodiments, the tumor specimen is evaluated as having low frequency microsatellite instability or microsatellite stability. In some embodiments, the tumor specimen is evaluated as having low tumor mutation burden, where low tumor mutation burden is classified as less than 10 mutations per megabase (mut / Mb).

[0177] In some embodiments, the checkpoint inhibitor therapy is administered as a monotherapy. In some embodiments, the checkpoint inhibitor therapy is administered in combination with one or more other chemotherapeutic agents. In some embodiments, the tumor specimen exhibits a TPS score of 1-49%.

[0178] Some embodiments of the disclosure provide a method of treatment comprising the steps of: (a) measuring expression levels of RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from a subject; (b) log2-transforming, median-centering, and normalizing the measured expression levels of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12 to the levels of the RNA transcripts of the at least one reference gene to provide transformed normalized levels of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12; (c) measuring expression levels of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12 in the biological sample. (d) calculating an immunotherapy response score (IRS) from the transformed normalized levels of RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12, and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and (e) administering checkpoint inhibitor therapy to the subject.

[0179] In some embodiments, the IRS is calculated as follows: IRS=0.27×[transformed TMB measurement]+0.11×[transformed normalized level of PD-1]+0.06×[transformed normalized level of PD-L1]−0.06[transformed normalized level of ADAM12]−0.077×[transformed normalized level of TOP2A].

[0180] In some embodiments, the IRS is calculated as follows: IRS=0.273758×[transformed TMB measurement]+0.112641×[transformed normalized level of PD-1]+0.061904×[transformed normalized level of PD-L1]−0.057991[transformed normalized level of ADAM12]−0.077011×[transformed normalized level of TOP2A].

[0181] In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.87 or greater. In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is approximately 0.873569 or greater.

[0182] In some embodiments, the one or more reference genes comprise three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC and SLC4A1AP.In some embodiments, the one or more reference genes comprise a combination of HMBS, CIAO1 and EIF2B1.

[0183] In some embodiments, the tumor specimen is a formalin-fixed paraffin-embedded (FFPE) tumor specimen. In some embodiments, the tumor specimen is breast cancer, central or peripheral nervous system cancer, cancer of unknown primary, colorectal cancer, endometrial cancer, gastrointestinal stromal tumor, glioma, hepatobiliary cancer, neuroendocrine cancer, ovarian cancer, pancreatic cancer, prostate cancer, salivary gland cancer, sarcoma, or thyroid cancer.

[0184] In some embodiments, the expression levels of RNA transcripts are measured using PCR and next generation sequencing.

[0185] In some embodiments, the checkpoint inhibitor is an anti-PD-1 antibody, an anti-CTLA-4 antibody, an anti-PD-L1 antibody, or an anti-PD-L2. In some embodiments, the checkpoint inhibitor is nivolumab, pembrolizumab, atezolizumab, cemiplimab, durvalumab, pidilizumab, atezolimumab, PDR001, BMS-936559, avelumab, ipilimumab, or SHR-1210.

[0186] Some aspects of the disclosure include a method of identifying a subject who would benefit from checkpoint inhibitor therapy, comprising the steps of: (a) measuring expression levels of RNA transcripts for PD-1, TOP2A, PD-L1, and ADAM12, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; (b) log2-transforming, median-centering, and normalizing the measured expression levels of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12 to the levels of the RNA transcripts of the at least one reference gene to provide transformed normalized levels of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12. (c) measuring tumor mutation burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; (d) calculating an Immunotherapy Response Score (IRS) from the transformed normalized levels of RNA transcripts of PD-1, TOP2A, PD-L1 and ADAM12, and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and (e) identifying a subject as benefiting from checkpoint inhibitor therapy.

[0187] In some embodiments, the tumor specimen is from a cancer that is not approved for the indicated use of checkpoint inhibitor therapy. In some embodiments, the tumor specimen is assessed as having low frequency microsatellite instability or microsatellite stability. In some embodiments, the tumor specimen is assessed as having low tumor mutation burden, where low tumor mutation burden is classified as less than 10 mutations per megabase (mut / Mb). In some embodiments, the tumor specimen contains at least 20% tumor content.

[0188] In some embodiments, the checkpoint inhibitor therapy is administered as a monotherapy. In some embodiments, the checkpoint inhibitor therapy is administered in combination with one or more other chemotherapeutic agents. In some embodiments, the tumor specimen exhibits a TPS score of 1-49%.

[0189] In some embodiments of each of the methods disclosed herein, a subject is identified as benefiting from or is being treated with a checkpoint inhibitor therapy (e.g., pembrolizumab) if the tumor type is not approved for use with a checkpoint inhibitor (off-label) and the IRS value is 0.90 or greater. In some embodiments, a subject is identified as benefiting from or is being treated with a checkpoint inhibitor therapy (e.g., pembrolizumab) if the tumor type is not approved for use with a checkpoint inhibitor (off-label) and the IRS value is 0.88 or greater. In some embodiments, a subject is identified as benefiting from or treated with a checkpoint inhibitor therapy (e.g., pembrolizumab) if the tumor type is not approved for use with a checkpoint inhibitor (off-label), the TMB is 10 mutations per megabase (MPM) or greater, and the IRS value is 0.87 or greater. In some embodiments, a subject is identified as benefiting from or treated with a checkpoint inhibitor therapy (e.g., pembrolizumab) if the tumor type is not approved for use with a checkpoint inhibitor (off-label), the TMB is 10 MPM or greater, and the IRS value is 0.873 or greater. In some embodiments, a subject is identified as benefiting from or being treated with checkpoint inhibitor therapy (e.g., pembrolizumab) if the TMB is 10 MPM or greater and the IRS value is 0.8736 or greater.In some embodiments, a subject is identified as benefiting from or being treated with checkpoint inhibitor therapy (e.g., pembrolizumab) if the TMB is 10 MPM or greater and the IRS value is 0.873569 or greater.

[0190] Some aspects of the disclosure include a method of identifying a subject who would benefit from checkpoint inhibitor therapy, comprising the steps of: a. measuring expression levels of RNA transcripts for PD-1, PD-L2, and optionally one or more of CD4, ADAM12, PD-L1, and VTCN1, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; b. log2-transform, median-center, and normalize the measured expression levels of the RNA transcripts to the level of the RNA transcript of the at least one reference gene to provide a normalized level of the RNA transcript; c. determining the tumor mutation burden ( d. measuring normalized levels of RNA transcripts of one or more of PD-1, PD-L2, and optionally CD4, ADAM12, PD-L1, and VTCN1, and the transformed TMB measurement, obtaining an IRS that positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and e. identifying subjects as benefiting from checkpoint inhibitor therapy.

[0191] In some embodiments, the tumor specimen is from a cancer that is not approved for the indicated use of checkpoint inhibitor therapy. In some embodiments, the calculated IRS value indicates that the median time to next treatment (TNTT) is 24 months or longer.

[0192] In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0193] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM12. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]).

[0194] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, each of the transformed TMB expression, PD-1 expression, PD-L1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized PD-L1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0195] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0196] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured. In some embodiments, each of the converted TMB expression, PD-1 expression, PD-L1, and PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the converted TMB expression, normalized PD-1 expression, normalized PD-L1, and normalized PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0197] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4 and ADAM12 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0198] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, each of the transformed TMB expression, PD-1 expression, VTCN1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized VTCN1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0199] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12 and VTCN1 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the normalized CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0200] Some aspects of the disclosure include a method of identifying a subject who would benefit from checkpoint inhibitor therapy, comprising the steps of: a. receiving, by a processor, measured expression levels of RNA transcripts for PD-1, PD-L2, and optionally expression levels of RNA transcripts for one or more of CD4, ADAM12, PD-L1, and VTCN1, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; b. log2 transforming, median centering, and normalizing, by the processor, the measured expression levels of the RNA transcripts to the level of the RNA transcript of the at least one reference gene to provide normalized levels of the RNA transcripts; c. receiving, by a processor, a measured tumor mutation burden (TMB) in the biological sample; d. log2 transforming, by a processor, the TMB measurement to provide a transformed TMB measurement; e. calculating, by the processor, an immunotherapy response score (IRS) from normalized levels of RNA transcripts of PD-1, PD-L2, and optionally one or more of CD4, ADAM12, PD-L1, and VTCN1, and the transformed TMB measurement, which positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy; and f. providing a determination that the subject has a checkpoint inhibitor responsive cancer.

[0201] In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0202] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM12. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]).

[0203] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, each of the transformed TMB expression, PD-1 expression, PD-L1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized PD-L1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0204] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0205] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured. In some embodiments, each of the converted TMB expression, PD-1 expression, PD-L1, and PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the converted TMB expression, normalized PD-1 expression, normalized PD-L1, and normalized PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0206] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4 and ADAM12 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0207] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, each of the transformed TMB expression, PD-1 expression, VTCN1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized VTCN1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0208] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12 and VTCN1 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the normalized CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0209] Some aspects of the present disclosure include a method of treating a subject in need of treatment with checkpoint inhibitor therapy, comprising administering checkpoint inhibitor therapy to the subject, wherein the subject in need of treatment comprises: a. measuring expression levels of RNA transcripts for PD-1, PD-L2, and optionally expression levels of RNA transcripts for one or more of CD4, ADAM12, PD-L1, and VTCN1, and at least one reference gene, in a biological sample obtained from a tumor specimen from the subject; b. log2 transformed, median centered, and normalized the measured expression levels of the RNA transcripts to the level of the RNA transcript of the at least one reference gene to obtain a PD-1 expression level. c. measuring tumor mutation burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; d. calculating an Immunotherapy Response Score (IRS) from the normalized levels of RNA transcripts of one or more of PD-1, PD-L2, and optionally CD4, ADAM12, PD-L1, and VTCN1, and the transformed TMB measurement, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and obtains an IRS having a value indicative of a beneficial response to checkpoint inhibitor therapy.

[0210] In some embodiments, the IRS is calculated as 10 times the inverse of the hazard ratio for a patient compared to the median hazard ratio using a Cox model.

[0211] In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0212] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM12. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]).

[0213] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, each of the transformed TMB expression, PD-1 expression, PD-L1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized PD-L1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0214] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0215] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured. In some embodiments, each of the converted TMB expression, PD-1 expression, PD-L1, and PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the converted TMB expression, normalized PD-1 expression, normalized PD-L1, and normalized PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0216] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4 and ADAM12 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0217] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, each of the transformed TMB expression, PD-1 expression, VTCN1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized VTCN1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0218] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12 and VTCN1 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the normalized CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0219] In some embodiments, an IRS value of 10 or greater indicates a beneficial response to checkpoint inhibitor therapy.

[0220] Some embodiments of the present disclosure are directed to a method of treatment comprising the steps of: a. measuring expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from a subject; b. log2 transforming, median centering, and normalizing the measured expression levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12 to levels of the RNA transcripts of the at least one reference gene to provide normalized levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12; c. measuring expression levels of the RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 in a biological sample obtained from a tumor specimen from a subject; measuring mutational burden (TMB) and log2 transforming the TMB measurement to provide a transformed TMB measurement; d. calculating an Immunotherapy Response Score (IRS) from normalized levels of PD-1, PD-L2, CD4, and ADAM12 RNA transcripts and the transformed TMB measurement, obtaining an IRS that positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and e. administering checkpoint inhibitor therapy to the subject.

[0221] In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0222] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM12. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]).

[0223] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, each of the transformed TMB expression, PD-1 expression, PD-L1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized PD-L1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0224] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0225] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured. In some embodiments, each of the converted TMB expression, PD-1 expression, PD-L1, and PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the converted TMB expression, normalized PD-1 expression, normalized PD-L1, and normalized PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0226] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4 and ADAM12 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0227] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, each of the transformed TMB expression, PD-1 expression, VTCN1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized VTCN1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0228] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12 and VTCN1 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the normalized CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0229] In some embodiments, the IRS value indicating a beneficial response to checkpoint inhibitor therapy is 10 or greater. In some embodiments, the one or more reference genes include three genes selected from LRP1, MRPL13, TBP, HMBS, ITGB7, MYC, CIAO1, CTCF, EIF2B1, GGNBP2, SLC4A1AP. In some embodiments, the tumor specimen is a formalin-fixed paraffin-embedded (FFPE) tumor specimen. In some embodiments, the tumor specimen is adrenal cancer, biliary tract cancer, bladder cancer, brain cancer, breast cancer, cervical cancer, colon cancer, rectal cancer, endometrial cancer, esophageal cancer, head and neck cancer, kidney cancer, liver cancer, non-small cell lung cancer, lung cancer, lymphoma, melanoma, meningeal cancer, non-melanoma skin cancer, ovarian cancer, pancreatic cancer, prostate cancer, sarcoma, small intestine cancer, or gastric cancer. In some embodiments, the expression level of the RNA transcript is measured using PCR and next-generation sequencing. In some embodiments, the checkpoint inhibitor is an anti-PD-1 antibody, an anti-CTLA-4 antibody, an anti-PD-L1 antibody, or an anti-PD-L2. In some embodiments, the checkpoint inhibitor is nivolumab, pembrolizumab, atezolizumab, durvalumab, pidilizumab, PDR001, BMS-936559, avelumab, or SHR-1210.

[0230] Some aspects of the present disclosure include a method of identifying a subject who would benefit from checkpoint inhibitor therapy, comprising the steps of: a. measuring expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from the subject; b. log2-transform, median-center, and normalize the measured expression levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12 to the levels of the RNA transcripts of the at least one reference gene to provide normalized levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12; and c. measuring tumor mutation burden (TMB) in a biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; d. calculating an Immunotherapy Response Score (IRS) from normalized levels of PD-1, PD-L2, CD4, and ADAM12 RNA transcripts and the transformed TMB measurement, obtaining an IRS that positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy; and e. identifying a subject as benefiting from checkpoint inhibitor therapy.

[0231] In some embodiments, the tumor specimen is from a cancer that is not approved for the indicated use of checkpoint inhibitor therapy. In some embodiments, the calculated IRS value indicates that the median time to next treatment (TNTT) is 24 months or longer.

[0232] In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0233] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM12. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]).

[0234] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, each of the transformed TMB expression, PD-1 expression, PD-L1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized PD-L1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0235] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0236] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured. In some embodiments, each of the converted TMB expression, PD-1 expression, PD-L1, and PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the converted TMB expression, normalized PD-1 expression, normalized PD-L1, and normalized PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0237] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4 and ADAM12 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0238] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, each of the transformed TMB expression, PD-1 expression, VTCN1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized VTCN1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0239] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12 and VTCN1 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the normalized CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0240] Some aspects of the disclosure include a method of identifying a subject that would benefit from checkpoint inhibitor therapy, comprising the steps of: a. receiving, by a processor, measured expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from the subject; b. log2 transforming, median centering, and normalizing, by the processor, the measured expression levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12 to levels of the RNA transcripts of the at least one reference gene to obtain normalized levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12. providing, by a processor, a measured tumor mutation burden (TMB) in the biological sample; d. log2 transforming, by a processor, the TMB measurement to provide a transformed TMB measurement; e. calculating, by a processor, an immunotherapy response score (IRS) from the normalized levels of PD-1, PD-L2, CD4, and ADAM12 RNA transcripts and the transformed TMB measurement, which positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy; and f. providing a determination that the subject has a checkpoint inhibitor responsive cancer.

[0241] In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0242] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM12. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]).

[0243] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, each of the transformed TMB expression, PD-1 expression, PD-L1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized PD-L1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0244] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0245] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured. In some embodiments, each of the converted TMB expression, PD-1 expression, PD-L1, and PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the converted TMB expression, normalized PD-1 expression, normalized PD-L1, and normalized PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0246] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4 and ADAM12 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0247] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, each of the transformed TMB expression, PD-1 expression, VTCN1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized VTCN1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0248] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12 and VTCN1 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the normalized CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0249] Some aspects of the present disclosure include a method of treating a subject in need of treatment with checkpoint inhibitor therapy, comprising administering checkpoint inhibitor therapy to the subject, wherein the subject in need of treatment comprises the steps of: a. measuring expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and at least one reference gene in a biological sample obtained from a tumor specimen from the subject; b. log2-transform, median-center, and normalize the measured expression levels of the RNA transcripts of PD-1, PD-L2, CD4, and ADAM12 to the levels of the RNA transcripts of the at least one reference gene to determine the expression levels of PD-1, PD-L2, CD4, and ADAM12 in a biological sample obtained from a tumor specimen from the subject; providing a normalized level of an ADAM12 RNA transcript; c. measuring tumor mutational burden (TMB) in the biological sample and log2 transforming the TMB measurement to provide a transformed TMB measurement; and d. calculating an Immunotherapy Response Score (IRS) from the normalized levels of PD-1, PD-L2, CD4, and ADAM12 RNA transcripts, and the transformed TMB measurement, resulting in an IRS that positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and has a value indicative of a beneficial response to checkpoint inhibitor therapy.

[0250] In some embodiments, the IRS is calculated as 10 times the inverse of the hazard ratio of a patient compared to the median hazard ratio using a Cox model. In some embodiments, the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]). In some embodiments, an IRS value indicating a beneficial response to checkpoint inhibitor therapy is 10 or greater.

[0251] In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.97×exp(0.301×[transformed TMB measurement]+0.110×[normalized level of PD-1]+0.078×[normalized level of PD-L2]).

[0252] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for ADAM12. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the altered TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized ADAM12 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.307×[transformed TMB measurement]+0.115×[normalized level of PD-1]+0.106×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]).

[0253] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for PD-L1. In some embodiments, each of the transformed TMB expression, PD-1 expression, PD-L1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized PD-L1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.91×exp(0.296×[transformed TMB measurement]+0.097×[normalized level of PD-1]+0.056×[normalized level of PD-L2]+0.0043×[normalized level of PD-L1]).

[0254] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for CD4. In some embodiments, each of the transformed TMB expression, PD-1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression negatively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=4.10×exp(0.281×[transformed TMB measurement]+0.139×[normalized level of PD-1]+0.112×[normalized level of PD-L2]+−0.128×[normalized level of CD4]).

[0255] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured. In some embodiments, each of the converted TMB expression, PD-1 expression, PD-L1, and PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the converted TMB expression, normalized PD-1 expression, normalized PD-L1, and normalized PD-L2 expression is positively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and PD-L1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.280×[transformed TMB measurement]+0.134×[normalized level of PD-1]+0.122×[normalized level of PD-L2]+−0.070×[normalized level of ADAM12]+−0.154×[normalized level of CD4]+0.052×[normalized level of PD-L1]).

[0256] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4 and ADAM12 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and CD4 and ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and normalized CD4 expression and normalized ADAM12 expression are negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, and ADAM12 are measured, and the IRS is calculated as follows: IRS=4.03×exp(0.287×[transformed TMB measurement]+0.147×[normalized level of PD-1]+0.143×[normalized level of PD-L2]−0.138×[normalized level of CD4]−0.073×[normalized level of ADAM12]).

[0257] In some embodiments, step a. further comprises measuring the expression level of an RNA transcript for VTCN1. In some embodiments, each of the transformed TMB expression, PD-1 expression, VTCN1 expression, and PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, each of the transformed TMB expression, normalized PD-1 expression, normalized VTCN1 expression, and normalized PD-L2 expression positively correlates with the likelihood that the patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the IRS is calculated as follows: IRS=3.90×exp(0.309×[transformed TMB measurement]+0.104×[normalized level of PD-1]+0.087×[normalized level of PD-L2]+0.021×[normalized level of VTCN1]).

[0258] In some embodiments, the expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12 and VTCN1 are measured. In some embodiments, the converted TMB expression, PD-1 expression and PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, the converted TMB expression, normalized PD-1 expression and normalized PD-L2 expression are each positively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and the normalized CD4 expression is negatively correlated with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy. In some embodiments, expression levels of RNA transcripts for PD-1, PD-L2, CD4, ADAM12, and VTCN1 are measured and the IRS is calculated as follows: IRS=3.95×exp(0.295×[transformed TMB measurement]+0.142×[normalized level of PD-1]+0.150×[normalized level of PD-L2]+0.020×[normalized level of VTCN1]+−0.070×[normalized level of ADAM12]+−0.139×[normalized level of CD4]).

[0259] In some embodiments of each of the methods disclosed herein, a subject is identified as benefiting from or is being treated with a checkpoint inhibitor therapy (e.g., pembrolizumab) if the tumor type is not approved for use with a checkpoint inhibitor (off-label) and the IRS value is 10 or greater. In some embodiments, a subject is identified as benefiting from or is being treated with a checkpoint inhibitor therapy (e.g., pembrolizumab) if the tumor type is not approved for use with a checkpoint inhibitor (off-label) and the IRS value is 12 or greater. In some embodiments, a subject is identified as benefiting from or is being treated with a checkpoint inhibitor therapy (e.g., pembrolizumab) if the tumor type is not approved for use with a checkpoint inhibitor (off-label), the TMB is 10 mutations per megabase (MPM) or greater, and the IRS value is 10 or greater. In some embodiments, a subject is identified as benefiting from or is being treated with a checkpoint inhibitor therapy (e.g., pembrolizumab) if the tumor type is not approved for use with a checkpoint inhibitor (off-label), the TMB is 10 MPM or greater, and the IRS value is 12 or greater. In some embodiments, a subject is identified as benefiting from or being treated with checkpoint inhibitor therapy (e.g., pembrolizumab) if the TMB is 10 MPM or greater and the IRS value is 10 or greater.In some embodiments, a subject is identified as benefiting from or being treated with checkpoint inhibitor therapy (e.g., pembrolizumab) if the TMB is 10 MPM or greater and the IRS value is 12 or greater.

[0260] In some embodiments of each of the methods disclosed herein, determining whether a tumor is responsive or likely to be responsive to immune checkpoint therapy includes collecting or providing a tumor specimen from a subject. In some embodiments, the tumor specimen is a fresh tumor specimen or a formalin-fixed paraffin-embedded (FFPE) tumor specimen. However, the preparation of the specimen is not limited and can be any suitable preparation known in the art. In some embodiments, the method does not include collecting or providing a tumor. Instead, data (e.g., an IRS value) or a qualitative assessment (e.g., a determination that the tumor has a suitable IRS value) is provided. In some embodiments, the data or qualitative assessment is provided to a physician or other medical professional, who uses such data or assessment to decide whether or not to administer immune checkpoint therapy. The data or qualitative assessment provided can be calculated or determined by any of the methods disclosed herein.

[0261] In some embodiments of each of the methods disclosed herein, the tumor may be from any cancer, without limitation. As used herein, the term "cancer" refers to a malignant neoplasm (Stedman's Medical Dictionary, 25th ed.; Hensyl ed.; Williams & Wilkins: Philadelphia, 1990). Exemplary cancers include, but are not limited to, acoustic neuroma; adenocarcinoma; adrenal cancer; anal cancer; angiosarcoma (e.g., lymphangiosarcoma, lymphangioendothelial sarcoma, hemangiosarcoma); appendix cancer; benign monoclonal gammopathy; biliary tract cancer (e.g., cholangiocarcinoma); bladder cancer; breast cancer (e.g., adenocarcinoma of the breast, papillary carcinoma of the breast, mammary carcinoma, ovarian ... cancer), medullary carcinoma of the breast); brain cancer (e.g., meningioma, glioblastoma, glioma (e.g., astrocytoma, oligodendroglioma), medulloblastoma); bronchial carcinoma; carcinoid tumor; cervical cancer (e.g., cervical adenocarcinoma); choriocarcinoma; chordoma; craniopharyngioma; colorectal cancer (e.g., colon carcinoma, rectal carcinoma, colorectal adenocarcinoma); connective tissue carcinoma; epithelial carcinoma; ependymoma; endothelial sarcoma (e.g., Kaposi's sarcoma, multiple idiopathic hemorrhagic sarcoma; endometrial cancer (e.g., uterine cancer, uterine sarcoma); esophageal cancer (e.g., esophageal adenocarcinoma, Barrett's adenocarcinoma); Ewing's sarcoma; eye cancer (e.g., intraocular melanoma, retinoblastoma); familial hypereosinophilia; gallbladder cancer; gastric cancer (e.g., gastric adenocarcinoma); gastrointestinal stromal tumor (GIST); germ cell cancer; head and neck cancer (e.g., head and neck squamous cell carcinoma, oral cancer (e.g., oral squamous cell carcinoma), pharyngeal cancer (throat cancer) (e.g., laryngeal cancer, pharyngeal cancer, nasopharyngeal cancer, oropharyngeal cancer); hematopoietic cancer (e.g., leukemia, e.g., acute lymphocytic leukemia (ALL) (e.g., B-cell ALL, T-cell ALL), acute myeloid leukemia (AML) (e.g., B-cell AML, T-cell AML), chronic myeloid leukemia (CML) (e.g., B-cell CML, T-cell CML), and chronic lymphocytic leukemia (CLL) (e.g., B-cell CLL, T-cell CLL));Lymphomas, such as Hodgkin's lymphoma (HL) (e.g., B-cell HL, T-cell HL) and non-Hodgkin's lymphoma (NHL) (e.g., B-cell NHL, such as diffuse large cell lymphoma (DLCL) (e.g., diffuse large B-cell lymphoma), follicular lymphoma, chronic lymphocytic leukemia / small lymphocytic lymphoma (CLL / SLL), mantle cell lymphoma (MCL), marginal zone B-cell lymphoma (e.g., mucosa-associated lymphoid tissue (MALT) lymphoma, nodal marginal zone B-cell lymphoma, splenic marginal zone B-cell lymphoma), primary mediastinal B-cell lymphoma, Burkitt's lymphoma, lymphoplasmacytic lymphoma (i.e., Waldenstrom's hypergammaglobulinemia), hairy cell leukemia (HCL), immunoblastic large cell lymphoma, precursor B-lymphoblastic lymphoma and primary central nervous system (CNS) lymphoma; and T-cell NHL, e.g., precursor T-lymphoblastic lymphoma / leukemia, peripheral T-cell lymphoma (PTCL) (e.g., cutaneous T-cell lymphoma (CTCL) (e.g., mycosis fungoides, Sézary syndrome), angioimmunoblastic T-cell lymphoma, extranodal natural killer T-cell lymphoma, enteropathy-type T-cell lymphoma, subcutaneous panniculitis-like T-cell lymphoma, ... follicular lymphoma, anaplastic large cell lymphoma, and anaplastic large cell lymphoma; mixed leukemia / lymphoma of one or more of the above; and multiple myeloma (MM)), heavy chain disease (e.g., alpha chain disease, gamma chain disease, mu chain disease); hemangioblastoma; hypopharyngeal carcinoma; inflammatory myofibroblastic tumor; immune cell amyloidosis; kidney cancer (e.g., nephroblastoma, also known as Wilms' tumor, renal cell carcinoma); liver cancer (e.g., hepatocellular carcinoma (HCC), malignant hepatoma); lung cancer (e.g., bronchogenic carcinoma, small cell lung cancer (SCLC), non-small cell lung cancer (NSCLC), adenocarcinoma of the lung); leiomyosarcoma (LMS); mastocytosis (e.g., systemic mastocytosis); muscle cancer; myelodysplastic syndromes (MDS); mesothelioma; myeloproliferative disorders (MPDs) (e.g., polycythemia vera (PV), essential thrombocytosis (ET), primary myelofibrosis (AMM), also known as myelofibrosis (MF), chronic idiopathic myelofibrosis, chronic myeloid leukemia (CML), chronic neutrophilic leukemia (CNL), hypereosinophilic syndrome (HES)); neuroblastoma; neurofibromas (e.g., neurofibromatosis (NF) type 1 or 2, schwannoma); neuroendocrine cancers (e.g., gastroenteropancreatic neuroendocrine tumors (GEP-NETs), carcinoid tumors);Osteosarcoma (e.g., bone cancer); ovarian cancer (e.g., cystadenocarcinoma, ovarian embryonal carcinoma, ovarian adenocarcinoma); papillary adenocarcinoma; pancreatic cancer (e.g., pancreatic andenocarcinoma, intraductal papillary mucinous neoplasm (IPMN), pancreatic islet tumor); penile cancer (e.g., Paget's disease of the penis and scrotum); pinealoma; primitive neuroectodermal tumor (PNT); plasma cell neoplasm; paraneoplastic syndromes; intraepithelial neoplasm; prostate cancer (e.g., prostatic adenocarcinoma); rectal cancer; rhabdomyosarcoma; salivary gland cancer; skin cancer (e.g., squamous cell carcinoma (SCC), keratoacanthoma (KA), melanoma, basal cell carcinoma (BCC)); small intestinal cancer (e.g., appendix cancer); soft tissue sarcoma (e.g., malignant fibrous histiocytoma (MFH), liposarcoma, malignant peripheral nerve sheath tumor (MPNST), chondrosarcoma, fibrosarcoma, myxosarcoma); sebaceous gland carcinoma; small intestine cancer; sweat gland carcinoma; synovium; testicular cancer (e.g., seminoma, testicular embryonal carcinoma); thyroid cancer (e.g., papillary thyroid carcinoma, papillary thyroid carcinoma (PTC), medullary thyroid carcinoma); urethral cancer; vaginal cancer; and vulvar cancer (e.g., Paget's disease of the vulva). In some embodiments, the cancer is a solid cancer.;

[0262] In some embodiments of each of the methods disclosed herein, the cancer is not a blood-bone marrow cancer or a hematopoietic cancer. In some embodiments, the cancer is not an MSI-H cancer. In some embodiments, the cancer is not one, two, three, four, five, six or all seven of melanoma, lung cancer, kidney cancer, bladder cancer, head and neck cancer, and Hodgkin's lymphoma. In some embodiments, the cancer is adrenal cancer, biliary tract cancer, bladder cancer, brain cancer, breast cancer, cervical cancer, colon cancer, rectal cancer, endometrial cancer, esophageal cancer, head and neck cancer, kidney cancer, liver cancer, non-small cell lung cancer, lung cancer, lymphoma, melanoma, meningeal cancer, non-melanoma skin cancer, ovarian cancer, pancreatic cancer, prostate cancer, sarcoma, small intestine cancer, or gastric cancer. In some embodiments, the cancer is not a TMB-H cancer. In some embodiments, the cancer is not all one, two, three, four, five, six, seven, eight, nine, or ten of melanoma, lung cancer, kidney cancer, bladder cancer, head and neck cancer, cervical cancer, esophagogastric cancer, hepatobiliary cancer, non-melanoma skin cancer, and Hodgkin's lymphoma. In some embodiments, the cancer is adrenal cancer, biliary tract cancer, bladder cancer, brain cancer, breast cancer, cervical cancer, colon cancer, rectal cancer, endometrial cancer, esophageal cancer, head and neck cancer, kidney cancer, liver cancer, non-small cell lung cancer, lung cancer, lymphoma, melanoma, meningeal cancer, non-melanoma skin cancer, ovarian cancer, pancreatic cancer, prostate cancer, sarcoma, small intestine cancer, or gastric cancer.

[0263] In some embodiments of each of the methods disclosed herein, determining or calculating whether the tumor is responsive or likely to be responsive to immune checkpoint therapy comprises calculating, collecting or determining immune response-related data (e.g., IRS value) derived from the tumor. In some embodiments, the methods described herein comprise obtaining immune response-related data (quantitative or qualitative) derived from the tumor from a third party and determining whether the tumor is responsive or likely to be responsive to immune checkpoint therapy. In some embodiments, the immune response-related data is collected or determined via NGS and / or multiplex PCR. In some embodiments, the immune response-related data is obtained from NGS and / or multiplex PCR performed by a third party.

[0264] In some embodiments of each of the methods disclosed herein, PD-1 expression is determined or calculated via NGS of gene expression transcripts using multiplex PCR (amplicon). In some embodiments, PD-1 expression is obtained from NGS of gene expression transcripts using multiplex PCR (amplicon) data. In some embodiments, PD-1 expression is verified, confirmed or combined using multiplex PCR and a second amplicon. In some embodiments, PD-1 verification or confirmation requires that the percentile value of the second amplicon is 70%, 75%, 80%, 85% or higher of the calculated PD-1 percentile value. In some embodiments, PD-1 verification or confirmation requires that the percentile value of the second amplicon is 80% or higher of the calculated PD-1 percentile value.

[0265] In some embodiments of each of the methods disclosed herein, PD-L2 expression is determined or calculated via NGS of gene expression transcripts using multiplex PCR (amplicon). In some embodiments, PD-L2 expression is obtained from NGS of gene expression transcripts using multiplex PCR (amplicon) data. In some embodiments, PD-L2 expression is verified, confirmed or combined using multiplex PCR and a second amplicon. In some embodiments, PD-L2 verification or confirmation requires that the percentile value of the second amplicon is 70%, 75%, 80%, 85% or higher of the calculated PD-L2 percentile value. In some embodiments, PD-L2 verification or confirmation requires that the percentile value of the second amplicon is 80% or higher of the calculated PD-L2 percentile value.

[0266] In some embodiments of each of the methods disclosed herein, CD4 expression is determined or calculated via NGS of gene expression transcripts using multiplex PCR (amplicon). In some embodiments, CD4 expression is obtained from NGS of gene expression transcripts using multiplex PCR (amplicon) data. In some embodiments, CD4 expression is verified, confirmed, or combined using multiplex PCR (amplicon) to measure GZMA, GZMB, GZMK, PRF1, IFNG, or CD8B expression. In some embodiments, CD4 expression is verified, confirmed, or combined using multiplex PCR (amplicon) to measure GZMA expression. Both CD4 and GZMA are part of the interferon-gamma gene signature. In some embodiments, CD4 verification, confirmation, or combination requires that the percentile value of the second amplicon measurement is 80% or higher than the calculated percentile value of CD4.

[0267] In some embodiments of each of the methods disclosed herein, ADAM12 expression is determined or calculated through NGS of gene expression transcripts using multiplex PCR (amplicon). In some embodiments, ADAM12 expression is obtained from NGS of gene expression transcripts using multiplex PCR (amplicon) data. In some embodiments, ADAM12 expression is verified, confirmed or combined using multiplex PCR and a second amplicon. In some embodiments, validation or confirmation of ADAM12 requires that the percentile value of the second amplicon is 70%, 75%, 80%, 85% or higher of the calculated percentile value of ADAM12. In some embodiments, validation or confirmation of ADAM12 requires that the percentile value of the second amplicon is 80% or higher of the calculated percentile value of ADAM12.

[0268] In some embodiments of each of the methods disclosed herein, PD-L1 expression is determined or calculated via NGS of gene expression transcripts using multiplex PCR (amplicon). In some embodiments, PD-L1 expression is obtained from NGS of gene expression transcripts using multiplex PCR (amplicon) data. In some embodiments, PD-L1 expression is verified, confirmed or combined using multiplex PCR and a second amplicon. In some embodiments, PD-L1 verification or confirmation requires that the percentile value of the second amplicon is 70%, 75%, 80%, 85% or higher of the calculated PD-L1 percentile value. In some embodiments, PD-L1 verification or confirmation requires that the percentile value of the second amplicon is 80% or higher of the calculated PD-L1 percentile value.

[0269] In some embodiments of each of the methods disclosed herein, VTCN1 expression is determined or calculated via NGS of gene expression transcripts using multiplex PCR (amplicon). In some embodiments, VTCN1 expression is obtained from NGS of gene expression transcripts using multiplex PCR (amplicon) data. In some embodiments, VTCN1 expression is verified, confirmed or combined using multiplex PCR and a second amplicon. In some embodiments, VTCN1 verification or confirmation requires that the percentile value of the second amplicon is 70%, 75%, 80%, 85% or higher of the calculated percentile value of VTCN1. In some embodiments, VTCN1 verification or confirmation requires that the percentile value of the second amplicon is 80% or higher of the calculated percentile value of VTCN1.

[0270] In some embodiments of each of the methods disclosed herein, TOP2A expression is determined or calculated via NGS of gene expression transcripts using multiplex PCR (amplicon). In some embodiments, TOP2A expression is obtained from NGS of gene expression transcripts using multiplex PCR (amplicon) data. In some embodiments, TOP2A expression is verified, confirmed or combined using multiplex PCR and a second amplicon. In some embodiments, TOP2A validation or confirmation requires that the second amplicon percentile value is 70%, 75%, 80%, 85% or higher of the calculated PD-L1 percentile value. In some embodiments, TOP2A validation or confirmation requires that the second amplicon percentile value is 80% or higher of the calculated TOP2A percentile value.

[0271] In some embodiments of each of the methods disclosed herein, the TMB is determined or calculated by NGS of tumor DNA. In some embodiments, the TMB is obtained from a third party. The method of detecting the mutation (e.g., TMB) is not limited. In some embodiments, the mutation is detected, calculated or obtained via NGS. In some embodiments, the TMB includes non-coding (at highly characterized genomic loci) and coding, synonymous and non-synonymous, single and multi-nucleotide (2-base) variants present at a variant allele frequency (VAF) of >10%. In some embodiments, the mutations per megabase (Mb) are estimated and the associated 90% confidence interval is calculated via the total number of positions with sufficient depth of coverage (maximum possibility of 1.7 Mb) required for final evaluation.

[0272] In some embodiments of each of the methods disclosed herein, the checkpoint inhibitor administered is an antibody against at least one checkpoint protein, such as PD-1, CTLA-4, PD-L1 or PD-L2. In some embodiments, the checkpoint inhibitor administered is an antibody that is effective against two or more of the checkpoint proteins selected from the group of PD-1, CTLA-4, PD-L1 and PD-L2. In some embodiments, the checkpoint inhibitor administered is a small molecule, non-proteinaceous compound that inhibits at least one checkpoint protein. In one embodiment, the checkpoint inhibitor is a small molecule, non-proteinaceous compound that inhibits a checkpoint protein selected from the group consisting of PD-1, CTLA-4, PD-L1 and PD-L2. In some embodiments, the checkpoint inhibitor administered is nivolumab (Opdivo®, BMS-936558, MDX1106, commercially available from BristolMyers Squibb, Princeton NJ), pembrolizumab (Keytruda® MK-3475, lambrolizumab, commercially available from Merck and Company, Kenilworth NJ), atezolizumab (Tecentriq®, Genentech / Roche, South San Francisco CA), durvalumab (MEDI4736, Mediimmune / AstraZeneca), pidilizumab (CT-011, CureTech), PDR001 (Novartis), BMS-936559 (MDX1105, BristolMyers Squibb), avelumab (MSB0010718C, Merck Serono / Pfizer), or SHR-1210 (Incyte).Additional antibody PD1 pathway inhibitors for use in the methods described herein include those described in U.S. Pat. No. 8,217,149, issued July 10, 2012 (Genentech, Inc.), U.S. Pat. No. 8,168,757, issued May 1, 2012 (Merck Sharp and Dohme Corp.), U.S. Pat. No. 8,008,449, issued August 30, 2011 (Medarex), and U.S. Pat. No. 7,943,743, issued May 17, 2011 (Medarex, Inc.).

[0273] In some embodiments of each of the methods disclosed herein, the disclosed methods include performing one or more normalization processes, for example, to make the sequencing output (e.g., associated with any suitable biomarker described herein, etc.) comparable to a threshold and / or across different sequencing runs. In an example, determining the IRS value may include background subtracting the sequence read counts and normalizing the background-subtracted sequence read counts to normalized reads per million (nRPM). In a specific example, a fold change ratio can be determined for a given gene (and / or suitable biomarker) according to a ratio=background-subtracted read count / reads per million (RPM) profile. In a specific example, the RPM profile can be determined based on the average RPM (and / or other suitable aggregate RPM metric) of multiple replicates of a biological sample across different validation sequencing runs. In a specific example, the median of the determined ratios can be used for normalization ratios for a given biological sample, where nRPM can be calculated according to nRPM=background subtracted read counts / normalization ratio.Housekeeping genes useful for the normalization process (e.g., as described herein) can include any one or more of LRP1, MRPL13, TBP, HMBS, ITGB7, MYC, CIAO1, CTCF, EIF2B1, GGNBP2, SLC4A1AP, and / or other suitable housekeeping genes (and / or any suitable genes).In some embodiments, 2, 3, 4, 5, 6, 7, or 8 of LRP1, MRPL13, TBP, HMBS, ITGB7, MYC, CIAO1, CTCF, EIF2B1, GGNBP2, SLC4A1AP are used for the normalization process. In some embodiments, three of LRP1, MRPL13, TBP, HMBS, ITGB7, MYC, CIAO1, CTCF, EIF2B1, GGNBP2, and SLC4A1AP are used for the normalization process. In some embodiments, EIF2B1, HMBS, and CIAO1 are used for the normalization process.In another specific example, the median of the determined ratios can be used for normalization ratios for a given biological sample, where nRPM can be calculated according to nRPM=background subtracted read counts / normalization ratio. Housekeeping genes useful for the normalization process (e.g., as described herein) can include any one or more of CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC, and SLC4A1AP. In some embodiments, three or more of CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC, and SLC4A1AP are used for the normalization process. In some embodiments, the one or more reference genes include a combination of CIAO1, EIF2B1 and HMBS, CTCF, GGNBP2, ITGB7, MYC, SLC4A1AP, and / or other suitable housekeeping genes (and / or any suitable genes). In some embodiments, 2, 3, 4, 5, 6, 7, or all 8 of CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC, and SLC4A1AP are used for normalization process.In some embodiments, 3 of CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC, and SLC4A1AP are used for normalization process.In some embodiments, EIF2B1, HMBS, and CIAO1 are used for normalization process.Additionally or alternatively, any suitable backgrounding and / or normalization process can be performed (e.g., for comparison with threshold value; for comparison of value across sequencing runs; etc.).

[0274] In some embodiments of each of the methods disclosed herein, the measurement of housekeeping genes is omitted, and internal standards are used for normalization.For example, when using real-time quantitative PCR, it measures the accumulation of PCR products by dual-labeled fluorogenic probes (i.e., TaqMan® probes).Real-time PCR is compatible with both quantitative competitive PCR, in which an internal competitor for each target sequence is used for normalization, and quantitative comparative PCR, which uses normalization genes contained in samples or housekeeping genes for RT-PCR.For further details, see, for example, Held et al., Genome Research 6:986-994 (1996).Alternatively, in some embodiments of the methods disclosed herein, normalization is based on the average or median signal (CT) of all assayed genes or a large subset thereof (global normalization approach).

[0275] In particular embodiments, the methods of the claimed invention include collecting a set of biological samples (e.g., FFPE tumor specimens) from a set of patients (e.g., cancer patients; etc.); generating one or more sequencing libraries (e.g., suitable for generating sequencing output indicative of biomarkers associated with patients responsive to one or more therapies; etc.) based on processing of the biological samples; determining a set of sequencing reads for the set of patients based on the one or more sequencing libraries (e.g., for cDNA sequences derived from cDNA converted from mRNA indicative of expression levels for the biomarkers provided herein, and optionally at least one reference gene); determining immune response-related data (e.g., PD-L2 gene expression levels; PD-1 gene expression levels; CD4, ADAM12, TOP2A, PD-L1, and VTCN1 gene expression levels); processing the sequencing reads to determine one or more of the current levels; cDNA sequence data, e.g., cDNA sequence data from cDNA converted from mRNA; DNA sequence data; TMB-associated data; MSI-associated data; etc.; determining a treatment response signature (e.g., IRS) for the set of patients based on the immune response-associated data (e.g., based on independent and / or combined analysis of different types of immune response-associated data; etc.); and facilitating the provision of treatment to one or more patients of the set of patients based on the treatment response signature (e.g., identifying a subset of patients exhibiting a positive response to therapy for a clinical trial, e.g., for enrollment in a clinical trial; providing the treatment response signature to one or more care providers, e.g., to guide care decisions by the one or more care providers; etc.).

[0276] In specific examples, the IRS values ​​can be used for clinical trials (e.g., clinical trial enrollment and patient selection; stratifying patient populations, e.g., based on different combinations of biomarkers; therapy signatures; outcome analysis; and / or other suitable purposes related to clinical trials; etc.), care provision (e.g., providing treatment response signatures to care providers to guide care decisions regarding the patient; therapy decisions for the patient; etc.), and / or other suitable applications. Additionally or alternatively, embodiments of the methods and systems disclosed herein may function to preserve valuable biological samples, e.g., lung cancer tissue biopsies, tumor specimens, and / or suitable types of biological samples. In specific examples, collection of immune response related data can be based on RNA sequencing, e.g., PD-L1 gene expression levels; PD-1 gene expression levels; one or more of CD4, ADAM12, TOP2A, PD-L2, and VTCN1 gene expression levels; cDNA sequence data, e.g., cDNA sequence data from cDNA converted from mRNA; DNA sequence data; TMB-related data; MSI-related data; etc.), and / or other suitable processing approaches as alternatives to sample processing approaches that may require relatively large amounts of biological sample usage (e.g., immunohistochemistry; etc.). However, embodiments of the methods and systems disclosed herein can include any suitable functionality.

[0277] The method embodiments disclosed herein preferably apply, include, and / or are otherwise associated with next-generation sequencing (NGS) (e.g., processing biological samples to create sequence libraries for sequencing using a next-generation sequencing system; etc.). The method embodiments disclosed herein may include, apply, and / or are otherwise associated with semiconductor-based sequencing technology. Additionally or alternatively, the method embodiments disclosed herein may include, apply, and / or are otherwise associated with any suitable sequencing technology (e.g., sequencing library preparation technology; sequencing system; sequencing output analysis technology; etc.). The sequencing technology preferably includes next-generation sequencing technology. Next generation sequencing techniques may include any one or more of high throughput sequencing (e.g., facilitated by high throughput sequencing techniques; massively parallel signature sequencing, Polony sequencing, 454 pyrosequencing, Illumina sequencing, SOLiD sequencing, Ion Torrent semiconductor sequencing and / or other suitable semiconductor based sequencing technologies, DNA nanoball sequencing, Heliscope single molecule sequencing, single molecule real time (SMRT) sequencing, nanopore DNA sequencing, etc.), any generation number of sequencing technologies (e.g., second generation sequencing technologies, third generation sequencing technologies, fourth generation sequencing technologies, etc.), sequencing by synthesis, tunneling current sequencing, sequencing by hybridization, mass spectrometry sequencing, microscopy based techniques, and / or any suitable next generation sequencing technology.In specific examples, method embodiments disclosed herein may include applying next generation sequencing techniques to the prepared sequence library to facilitate the generation of sequence reads associated with multiple biomarkers for responsiveness to one or more immune checkpoint therapies (e.g., PD-1 / PD-L1 inhibitors; etc.).

[0278] Additionally or alternatively, the sequencing technique may include any one or more of capillary sequencing, Sanger sequencing (e.g., microfluidic Sanger sequencing, etc.), pyrosequencing, nanopore sequencing (e.g., Oxford nanopore sequencing, etc.), and / or any other suitable type of sequencing facilitated by any suitable sequencing technique.

[0279] Embodiments of the methods disclosed herein may include, apply, perform, and / or otherwise be associated with any one or more of a sequencing operation, an alignment operation (e.g., sequencing read alignment; etc.), a melting operation, a cleavage operation, a tagging operation (e.g., using barcodes; etc.), a ligation operation, a fragmentation operation, an amplification operation (e.g., helicase-dependent amplification (HDA), loop-mediated isothermal amplification (LAMP), self-sustained sequence replication (3SR), nucleic acid sequence-based amplification (NASBA), strand displacement amplification (SDA), rolling circle amplification (RCA), ligase chain reaction (LCR), etc.), a purification operation, a cleaning operation, an operation suitable for sequencing library preparation, an operation suitable to facilitate sequencing and / or downstream analysis, a suitable sample processing operation, and / or any suitable sample- and / or sequence-related operation. In specific examples, sample processing operations can be performed to process a biological sample to generate a sequencing library to facilitate characterization of multiple biomarkers associated with responsiveness to one or more immune checkpoint therapies.

[0280] Additionally or alternatively, the data described herein (e.g., immune response-related data, thresholds, models, parameters, normalized data, IRS values, treatment decisions, sample data, sequencing data, etc.) can be associated with any suitable time indicator (e.g., seconds, minutes, hours, days, weeks, time periods, timestamps, etc.), including one or more of: a time indicator indicating when the data was collected, determined, transmitted, received, and / or otherwise processed; a time indicator providing a context for what is described by the data; changes in the time indicator (e.g., data over time; changes in the data; patterns in the data; trends in the data; data extrapolation and / or other predictions; etc.); and / or any other suitable indicator related to time. In specific examples, characterization of treatment response may be performed over time for one or more patients to facilitate patient monitoring, evaluation of therapeutic efficacy, facilitating the provision of additional treatment, and / or other suitable purposes.

[0281] Additionally or alternatively, the parameters, metrics, inputs, outputs, and / or other suitable data can be associated with a type of value including any one or more of a binary value (e.g., determining a binary status of presence or absence of one or more biomarkers associated with positive responsiveness to an immune checkpoint therapy and / or other suitable therapy, etc.), a score (e.g., aggregating a score indicative of the probability and / or degree of responsiveness to a therapy described herein; etc.), a value indicative of the presence, absence, degree of responsiveness to one or more therapies described herein, a classification (e.g., patient classification for sensitivity to a therapy described herein; patient classification based on the absence or presence of different biomarkers of a set of biomarkers associated with responsiveness to a therapy described herein; etc.), an identifier (e.g., a sample identifier; a sample label indicative of association with a different cancer condition; a patient identifier; a biomarker identifier; etc.), a value along a spectrum, and / or any other suitable type of value. Any suitable type of data described herein can be used as input (e.g., for different models; for comparison against a threshold value), can be produced as output (e.g., for different models; for use in characterizing treatment response; etc.), and / or can be manipulated in any suitable manner for any suitable components associated with the method embodiments disclosed herein.

[0282] One or more examples and / or portions of the method embodiments disclosed herein can be performed asynchronously (e.g., serially), simultaneously (e.g., in parallel; simultaneously on different threads for parallel computing to improve system throughput for immune response related data processing and / or generation of treatment response characteristics; multiplexed sample processing; multiplexed sequencing, e.g., combined with a minimum number of sequencing runs, e.g., multiplexed sequencing for multiple biomarkers; etc.), in temporal relationship to a triggering event (e.g., operation of a portion of a method disclosed herein), and / or by and / or using one or more examples of the embodiments of the invention described herein, at any suitable time and frequency, and in any other suitable order.

[0283] Embodiments of systems for performing the methods described herein may include one or more of a sample handling system (e.g., for processing samples; for sequencing library creation; etc.); a sequencing system (e.g., for sequencing one or more sequencing libraries; etc.); a computing system (e.g., for sequencing output analysis; for collecting and / or processing immune response related data; for creating treatment response signatures; for any suitable computer process; etc.); a treatment system (e.g., for providing treatment recommendations; for facilitating patient selection for clinical trials; for providing therapy; etc.); and / or any other suitable components.

[0284] The system embodiments and / or portions of the embodiments described herein may, in whole or in part, execute, host, communicate with, and / or otherwise include one or more of a remote computing system (e.g., a server, at least one networked computing system, stateless, stateful; etc.), a local computing system, a user device (e.g., a mobile phone device, other mobile device, personal computing device, tablet, wearable, head-mounted wearable computing device, wrist-mounted wearable computing device, etc.), a database (e.g., including sample data and / or analysis, sequencing data, user data, data described herein, etc.), an application program interface (API) (e.g., for accessing data described herein, etc.), and / or any suitable components. Communication by and / or between any components of the system and / or other suitable components may include wireless communication (e.g., WiFi, Bluetooth, radio frequency, Zigbee, Z-wave, etc.), wired communication, and / or any other suitable type of communication.

[0285] The components of the method and system embodiments described herein may be physically and / or logically integrated in any manner (e.g., with any suitable distribution of functionality across the components). Portions of the method and system embodiments described herein are preferably performed by a first party, but may additionally or alternatively be performed by one or more third parties, users, and / or any suitable entity. However, the methods and systems described herein may be arranged in any suitable manner.

[0286] Embodiments of the methods disclosed herein may include a step of collecting immune response-related data from one or more biological samples, which may function to collect (e.g., generate, determine, receive, etc.) data associated with immune response functionality to enable characterization of one or more patients in relation to responsiveness to immune checkpoint therapy (e.g., calculating an IRS value by one or more processors).

[0287] Immune response-related data preferably includes data indicative of biological phenomena associated with (e.g., affecting components of, being affected by, relating to, being part of, including, etc.) the immune response and / or immune system; however, immune response-related data may include any suitable data related to the immune response and / or immune system (e.g., derived by sample processing techniques, bioinformatics techniques, statistical techniques, sensors, etc.).

[0288] Any of the variations (e.g., embodiments, variations, examples, specific examples, figures, etc.) described herein and / or any portions of the variations described herein may be additionally or alternatively combined, assembled, eliminated, used, performed sequentially, performed in parallel, and / or otherwise applied.

[0289] Portions of the method and system embodiments may be embodied and / or executed, at least in part, as a machine (e.g., a processor) configured to receive a computer-readable medium having computer-readable instructions stored thereon. The instructions may be executed by a computer-executable component that may be integrated with the system and method embodiments described herein. The computer-readable medium may be stored in any suitable computer-readable medium, such as a RAM, a ROM, a flash memory, an EEPROM, an optical device (CD or DVD), a hard drive, a floppy drive, or any suitable device. The computer-executable component may be a general or application-specific processor, although any suitable dedicated hardware or combination hardware / firmware device may alternatively or additionally execute the instructions.

[0290] As one skilled in the art will recognize from the foregoing detailed description, and from the figures and claims, modifications and variations can be made to the embodiments and / or variations of the methods and systems disclosed herein without departing from the scope defined in the claims. The variations described herein are not meant to be limiting. Certain features included in the drawings may be exaggerated in size, and other features may be omitted for clarity, and are not meant to be limiting. The figures are not necessarily to scale. Section headings herein are used for convenience of organization and are not meant to be limiting. The description of any variation does not necessarily limit any section of this specification.

[0291] As used herein, the terms "comprising" or "comprises" are used in reference to compositions, methods, and individual components thereof that are essential to the method or composition, and further permit the inclusion of unspecified elements, whether essential or not.

[0292] The term "consisting of" refers to compositions, methods, and individual components thereof described herein, which excludes any element not recited in that description of an embodiment.

[0293] As used herein, the term "consisting essentially of" refers to elements required for a given embodiment. The term permits the presence of elements that do not materially affect the basic and novel or functional characteristics of the embodiment.

[0294] The term "statistically significant" or "significantly" refers to statistical significance, and generally means a "p" value greater than 0.05 (as calculated by the relevant statistical test). Those skilled in the art will readily recognize that the relevant statistical test for any particular experiment depends on the type of data being analyzed. Additional definitions are provided in the text of the individual sections below.

[0295] Definitions of common terms in cell and molecular biology can be found in "The Merck Manual of Diagnosis and Therapy", 19th Edition, published by Merck Research Laboratories, 2006 (ISBN 0-911910-19-0); Robert S. Porter et al. (eds.), The Encyclopedia of Molecular Biology, published by Blackwell Science Ltd., 1994 (ISBN 0-632-02182-9); The ELISA guidebook (Methods in molecular biology 149) by Crowther JR (2000); Immunology by Werner Luttmann, published by Elsevier, 2006. Definitions of common terms in molecular biology can also be found in Benjamin Lewin, Genes X, published by Jones & Bartlett Publishing, 2009 (ISBN-10: 0763766321); Kendrew et al. (eds.), Molecular Biology and Biotechnology: a Comprehensive Desk Reference, published by VCH Publishers, Inc., 1995 (ISBN 1-56081-569-8) and Current Protocols in Protein Sciences 2009, Wiley Intersciences, Coligan et al., eds.

[0296] Unless otherwise stated, the present invention was carried out using standard procedures as described, for example, in Sambrook et al., Molecular Cloning: A Laboratory Manual (3 ed.), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, USA (2001) and Davis et al., Basic Methods in Molecular Biology, Elsevier Science Publishing, Inc., New York, USA (1995), both of which are incorporated herein by reference in their entireties.

[0297] The description of the embodiments of the present disclosure is not intended to be exhaustive or to limit the present disclosure to the precise form disclosed. Specific embodiments and examples of the present disclosure are described herein for illustrative purposes, but various equivalent modifications are possible within the scope of the present disclosure, as those skilled in the art will recognize. For example, while steps or functions of a method are presented in a given order, alternative embodiments may perform the functions in a different order, or may perform the functions substantially simultaneously. The teachings of the present disclosure provided herein may be applied to other procedures or methods, where appropriate. The various embodiments described herein may be combined to provide further embodiments. Aspects of the present disclosure may be modified, if necessary, to further provide further embodiments of the present disclosure using the compositions, functions and concepts of the above references and applications. These and other changes may be made to the present disclosure in light of the detailed description.

[0298] Specific elements of any of the foregoing embodiments can be combined with or substituted for elements in other embodiments. Furthermore, although advantages associated with certain embodiments of the present disclosure have been described in the context of those embodiments, other embodiments may also exhibit such advantages, and all of the embodiments need not necessarily exhibit such advantages to be within the scope of the present disclosure.

[0299] All patents and other publications identified are expressly incorporated herein by reference for the purpose of describing and disclosing, for example, the methodologies described in such publications that may be used in connection with the present invention. These publications are provided solely for their disclosure prior to the filing date of this application. Nothing in this regard should be construed as an admission that the inventors are not entitled to antedate such disclosure by reason of prior invention or prior art, or for any other reason. All references to dates or representations as to the contents of these documents are based on the information available to the applicants and do not constitute any admission as to the accuracy of the dates or contents of these documents.

[0300] Those skilled in the art will readily recognize that the present invention is well adapted to carry out the objects and obtain the aims and advantages mentioned, as well as those inherent therein. The details of the descriptions and examples herein are representative of certain embodiments, are illustrative, and are not intended as limitations on the scope of the invention. Modifications therein and other uses will occur to those skilled in the art. These modifications are encompassed within the spirit of the invention. Those skilled in the art will readily recognize that various substitutions and modifications can be made to the invention disclosed herein without departing from the scope and spirit of the invention.

[0301] As used herein, the articles "a" and "an" should be understood in the present specification and claims to include plural referents unless clearly indicated to the contrary. A claim or description including "or" between one or more members of a group is considered to be satisfied when one, more than one, or all of the members of the group are present in, used in, or otherwise relevant to a given product or process, unless indicated to the contrary or otherwise clear from the context. The invention includes embodiments in which exactly one member of a group is present in, used in, or otherwise relevant to a given product or process. The invention also includes embodiments in which more than one, or all of the members of a group are present in, used in, or otherwise relevant to a given product or process. Furthermore, it should be understood that the present invention provides all variations, combinations, and permutations of one or more limitations, elements, clauses, descriptive terms, etc., from one or more listed claims that are introduced into another claim that is dependent on the same base claim (or any other claim, if relevant), unless otherwise indicated or unless a contradiction or inconsistency would arise, is apparent to one skilled in the art. It is contemplated that all embodiments described herein are applicable to all different aspects of the present invention, where appropriate. It is also contemplated that any of the embodiments or aspects may be freely combined with one or more other such embodiments or aspects, where appropriate. When elements are presented as lists, e.g., in a Markush group or similar format, it should be understood that each subgroup of elements is also disclosed, and any element may be removed from the group. In general, when the invention or aspects of the invention are described as including certain elements, features, etc., it should be understood that a certain embodiment of the invention or aspect of the invention consists of or consists essentially of such elements, features, etc. For purposes of brevity, those embodiments are not specifically set forth in so many words in every instance herein.It should also be understood that any embodiment or aspect of the invention may be explicitly excluded from the claims, regardless of whether a specific exclusion is recited herein. For example, any one or more active agents, additives, ingredients, optional agents, organisms, disorders, subjects, or combinations thereof may be excluded.

[0302] Where a claim or description is directed to a composition of matter, it should be understood that methods of making or using the composition of matter by any of the methods disclosed herein, and methods of using the composition of matter for any of the purposes disclosed herein, are aspects of the invention, unless otherwise indicated or it would be apparent to one of ordinary skill in the art that a contradiction or inconsistency would arise. Where a claim or description is directed to a method, it should be understood that, for example, methods of making compositions useful for carrying out the method, and products produced according to the method, are aspects of the invention, unless otherwise indicated or it would be apparent to one of ordinary skill in the art that a contradiction or inconsistency would arise.

[0303] When ranges are given herein, the invention includes embodiments in which the endpoints are included, in which both endpoints are excluded, and in which one endpoint is included and the other is excluded. It is to be assumed that both endpoints are included unless otherwise indicated. Furthermore, unless otherwise indicated or otherwise clear from the context and the understanding of one of ordinary skill in the art, values ​​expressed as ranges should be understood to assume any specific value or subrange within the stated range in different embodiments of the invention, to the tenth of the unit of the lower limit of the range, unless the context clearly indicates otherwise. It is also understood that when a series of numerical values ​​is stated herein, the invention includes embodiments that similarly relate to any intermediate value or range defined by any two values ​​in the series, the lowest value may be the minimum and the highest value may be the maximum. Numeric values, as used herein, include values ​​expressed as percentages. For any embodiment of the invention in which a numerical value is prefaced by "about" or "approximately", the invention includes embodiments in which that exact value is recited. For any embodiment of the invention where a numerical value is not prefaced by "about" or "approximately," the invention includes embodiments where the value is prefaced by "about" or "approximately."

[0304] "Approximately" or "about" generally includes numbers in either direction (greater or less than that number) within 1%, or in some embodiments, within 5% of the number, or in some embodiments, within 10% of the number, unless otherwise stated or clear from the context (unless such number unacceptably exceeds 100% of possible values). Unless expressly indicated to the contrary, in any method claimed herein that includes two or more acts, it should be understood that the order of the acts of the method is not necessarily limited to the order in which the acts of the method are recited, but the invention includes embodiments in which the order is so limited. It should also be understood that any product or composition described herein may be considered to be "isolated" unless otherwise indicated or clear from the context. EXAMPLES

[0305] Working Example

[0306] method

[0307] Biomarker testing

[0308] PCR-based comprehensive genomic profiling, including assessment of tumor mutation burden, was performed in formalin-fixed, paraffin-embedded solid tumor tissues using StrataNGS (Strata Oncology, Ann Arbor, MI) as previously described (Tomlins et al, Journal of Precision Oncology, 2020). In parallel, immune gene expression levels were quantified by analytically validated research assays (Strata Oncology, Ann Arbor, MI). Briefly, exon-spanning PCR amplicons were selected for each target gene and three housekeeping genes. After 20 or 30 cycles of PCR amplification, Ion Torrent-based next-generation sequencing was performed, targeting approximately 1,000,000 reads per sample. Target gene expression was normalized to housekeeping genes and reads per million (nRPM) compared to normal control samples.

[0309] patient

[0310] Subjects with advanced solid tumors who were treated with pembrolizumab and had molecular data were identified from the Strata Clinical Molecular Database. Seven hundred and eight subjects with solid tumors were identified who had tumor mutation burden and immune gene expression data that passed QC, tumor content ≥ 20%, and pembrolizumab initiated after sample collection. Patients treated with pembrolizumab monotherapy or the combination of pembrolizumab plus chemotherapy were included.

[0311] Determining the endpoint

[0312] Real-world time to next treatment (TTNT) was defined as the time (in months) from the start of therapy to the date of initiation of new therapy after stopping the initial therapy or the date of death. To validate the endpoint as a surrogate for overall survival, TTNT was compared with the time to death for patients with both events using Pearson correlation.

[0313] Algorithm Development

[0314] Standard Cox proportional hazards regression was performed to evaluate single and combined biomarkers (Statistical Models and Methods for Lifetime Data, by JF Lawless. 1982, John Wiley & Sons, New York.) using the software at statpages.info / prophaz.html. For multivariate model construction, backward stepwise regression was used, first including all variables in the model, then selectively removing the least significant variables as long as the overall model significance was improved. TMB measurements were log2 transformed, and gene expression measurements were log2 transformed and median centered before analysis. The statistical significance of the models was tested using chi-square test.

[0315] A patient's immunotherapy response score (IRS) was derived using a Cox model as 10 times the inverse of the patient's hazard ratio compared to the median hazard ratio.

[0316] The Kaplan-Meier (KM) method was used to visualize TTNT across patient groups and treatments. Differences in TTNT curves were tested using the log-rank test.

[0317] IRS groups were established by dividing the data set into eight equal IRS bins and then combining the bins based on the overlaid TTNT curves.

[0318] result

[0319] Biomarker analysis

[0320] 708 pembrolizumab-treated patients (from 24 cancer types, 481 pembro+chemo; 227 pembro mono; 170 2nd-line pembro mono and prior chemo) with StrataIO score were analyzed. With the exception of a few outliers, pembrolizumab TTNT correlated with OS (Spearman: n=43 r=0.74, or with outliers, n=45, r=0.61).

[0321] [Table 1]

[0322] Real-world time to next treatment (TTNT) was estimated for each subject as the time from initiation of pembrolizumab to the time pembrolizumab was stopped and new therapy was started or death occurred. To establish the suitability of TTNT to study the outcome of pembrolizumab treatment, TTNT was compared with overall survival and found to be more highly correlated (n=43 r=0.74, or, with outliers, n=45, r=0.61, FIG. 6).

[0323] First, we evaluated the association of TMB and 10 gene expression biomarkers with the outcome of pembrolizumab treatment (Table 2 ).

[0324] [Table 2]

[0325] Backward stepwise regression was then performed to fit a multivariate Cox proportional hazards model. The final model, which included 5 of the 10 input variables, was more significantly associated with the outcome of pembrolizumab treatment than any individual variable or other models (p=1.2e-7). Notably, TMB, PD-L1, pd-1, and PD-L2 were all independent predictors of the outcome of pembrolizumab treatment (Table 2).

[0326] Analyzing the results

[0327] To evaluate the applicability of the Cox model for predicting the outcome of pembrolizumab treatment in patients, the patient's Immunotherapy Response Score (IRS) was derived as follows:

[0328] IRS = 4.03 × exp(0.287 × [transformed TMB measurement] + 0.147 × [normalized level of PD-1] + 0.143 × [normalized level of PD-L2] - 0.138 × [normalized level of CD4] - 0.073 × [normalized level of ADAM12]).

[0329] Here, an IRS of 10 is equal to the median hazard ratio observed in the dataset, with values ​​greater than 10 representing a decreased hazard (i.e., more benefit from pembrolizumab) and values ​​less than 10 representing an increased hazard (i.e., less benefit from pembrolizumab). Patients were assigned to one of three IRS groups to compare patient outcomes (Table 3).

[0330] [Table 3]

[0331] Kaplan-Meier analysis showed that outcomes of pembrolizumab treatment varied widely across groups, with median TTNT ranging from greater than 24 months in the high group to just 7 months in the low group (Figure 8).

[0332] [Table 4]

[0333] [Table 5]

[0334] The continuous StrataIO score, including pembro mono in NSCLC (p=9.37e-05), pembro+chemo in NSCLC (p=3.02e-03), pembro mono in non-NSCLC (p=6.57e-06) and pembro+chemo in non-NSCLC (p=2.88e-02), predicted TTNT of pembrolizumab across the cohort (p=3.67e-12) and within all examined subsets, but not TTNT of chemotherapy (p=4.10e-01). * StrataIO's "high" vs. "low" was based on low / medium / high groupings.

[0335] Consideration

[0336] Herein, we developed a highly significant multivariate model combining TMB and immune gene expression to predict real-world pembrolizumab treatment outcomes in 700+ patients with diverse solid tumor types. Model inputs were generated simultaneously from a clinically validated next-generation sequencing platform, requiring a single small formalin-fixed paraffin-embedded biopsy specimen for testing. An immunotherapy response score (IRS) was derived to predict individual patients' likelihood of benefit from pembrolizumab, demonstrating that patients in the high IRS group had much better treatment outcomes compared to chemotherapy, while patients in the low IRS group did not. The association between IRS group and treatment outcomes was stable after separating NSCLC from other tumors, monotherapy from combination therapy, and TMB-high tumors from TMB-low tumors, suggesting that the model captures universal biological features of pembrolizumab benefit.

[0337] Notably, when applied to over 25,000 advanced solid tumors, high IRS groups were more prevalent in tumor types known to benefit from immunotherapy, but also occurred in a subset of nearly all tumor types. The proportion of tumors within the high IRS group predicted the objective response rates observed in independent studies, indicating that the IRS group may be identified or enriched for patients who are likely to benefit from ICIs.

[0338] The model has several potentially interesting biological insights. First, TMB, PD-L1 and PD-L2 were all independent predictors of benefit, and showed a multiplicative predictive effect across the three biomarkers, with increased antigenicity (TMB) and increased immune checkpoint activity (PD-L1 and / or PD-L2) benefiting from immune checkpoint blockade. While many past studies have established TMB and PD-L1 as predictive biomarkers, and recent studies have established that PD-L2 is also predictive, this is the first to combine and optimize these three variables in a single model, potentially providing a more comprehensive predictor of immunotherapy benefit.

[0339] IRS has potential applications both for refining the use of pembrolizumab in tumor types for which immunotherapy is indicated, and for selecting patients for immunotherapy (e.g., off-label) in tumor types for which it is not indicated. In the low IRS group, among tumor types for which pembrolizumab is approved, such as NSCLC, pembrolizumab shows little to no benefit compared to chemotherapy. Considering that immunotherapy is expensive and can cause severe toxicity, its use should be considered more carefully in the low IRS group. Among tumor types for which immunotherapy is not indicated, patients in the high IRS group should be considered for treatments that offer the potential for significant benefit compared to chemotherapy.

[0340] Pembrolizumab was recently approved for TMB-high tumors (>10 mutations per megabase), regardless of tumor type, inducing an objective response rate of 25%. The disclosed data suggest that an integrated model combining TMB with immune gene expression provides better predictions for stratifying treatment outcomes within TMB-high and TMB-low patients.

[0341] In summary, we disclose a biologically rational predictive model of immunotherapy response that integrates expression biomarkers based on DNA mutations and RNA expression. We utilized a single clinically validated NGS platform that can simultaneously read mutations and quantify gene expression, providing a clear diagnostic pathway for clinical application. Upon further validation in an independent tumor cohort, this study has the potential to extend the benefits of ICIs to additional patients, as well as reduce unnecessary toxicity and economic burden in patients unlikely to benefit from ICI treatment.

[0342] Additional Examples

[0343] cohort

[0344] The Strata Trial (NCT03061305) is an observational clinical trial evaluating the impact of molecular profiling on patients with advanced solid tumors. It was reviewed and approved by the Advarra Institutional Review Board (IRB; IRB Pro00019183) prior to study initiation. All adult patients with locally advanced (stage III), unresectable or metastatic (stage IV) solid tumors, and with available FFPE tumor tissue, were eligible when enrolling in the healthcare system, and the protocol also permitted enrollment of patients with rare early-stage tumors.

[0345] The Strata Clinical Molecular Database (SCMD) contains de-identified subject, molecular profiling, treatment, and survival data for all enrolled NCT03061305 participants. Previous antineoplastic therapy was collected for study participants at the time of study entry, including start and stop dates. Antineoplastic therapy data and survival status were collected prospectively for up to 3 years from the time of enrollment and / or informed consent. First, a series analysis of cases was performed here, focusing on the development of a PD-(L)1 benefit predictor based on integrated CGP+qTP for the exploratory purpose of the study. No post-hoc power analysis was performed to determine the sample size of this discovery cohort. A power analysis was then performed to determine the required cohort size for the independent validation cohort described below. Patients in SCMD tested by a version of StrataNGS assessing TMB (see biomarker data below) with parallel gene expression testing data completed between January 25, 2017 and July 12, 2022 were eligible for analysis using a data cutoff of July 12, 2022; for the discovery cohort, only patients tested through May 4, 2021 were eligible, and the data cutoff date was the same as for the entire cohort. A general validity analysis of SCMD is described in Supplementary Methods.

[0346] For both discovery and validation cohorts, common inclusion criteria were: valid TMB measurements from StrataNGS testing (including meeting the requirement of 20% tumor content overall), valid immune gene expression quantification from an exploratory multiplex PCR-based transcriptome profiling test, and documented treatment with at least one antineoplastic agent. For the discovery cohort, additional inclusion and exclusion criteria included: 1) treatment with pembrolizumab including a line of systemic therapy, 2) tissue specimens tested were collected before the systemic pembrolizumab line initiation date, and 3) patients had no prior anti-PD-(L)1 or CTLA4 blockade therapy before the pembrolizumab line initiation date. For the validation cohort, additional inclusion and exclusion criteria included: 1) treatment with systemic non-pembrolizumab anti-PD-(L)1 monotherapy, 2) tissue specimens examined were collected before the start date of PD-(L)1 therapy, 3) no prior anti-PD-(L)1 or CTLA4 blockade therapy before the non-pembrolizumab PD-(L)1 line start date, and 4) patients were not from the discovery cohort. Additional inclusion / exclusion criteria for other analyses are described below and in the overall study diagram (Figure 40). Patients with samples collected after the start date of the line of therapy analyzed were excluded from all analyses, except for analyses specifically evaluating the performance of IRS in samples collected after PD-(L)1 therapy.

[0347] Source data validation in the Strata Trial was performed on high-risk data fields, e.g., demographics and treatment history, according to an approved trial monitoring plan. Data completeness, consistency, and quality assurance checks were performed across the Strata electronic data capture (EDC) system according to an approved data management plan, and 100% source data validation was performed on the discovery cohort. Additional details on the Strata Trial experience and Strata molecular profiling are described. 54~56 .

[0348] Real-world clinical treatment data

[0349] Patient treatment histories from electronic health records (EHRs) or manual updates were standardized to allow derivation of real-world clinical progression-free survival (rwPFS) by time to next therapy (TTNT), and overall survival (OS). All drug therapies were classified as antineoplastic or non-antineoplastic treatments, and all antineoplastic treatments were further subdivided (e.g., chemotherapy, immuno-oncology [IO; PD-(L)1 or CTLA4], oncogene TKI, hormones, etc.), with non-antineoplastic treatments excluded from further consideration. Allocation of lines of therapy was performed in two stages: first, single-dose treatments with consecutive doses administered within 90 days were combined into courses of treatment with a single start and end date, and then non-overlapping lines of treatment were estimated by considering the courses of each drug therapy in order of their start date. Subsequent courses of treatment that started more than 30 days after the start of a given line of treatment or had less than 50% duration of overlap with the line were considered to establish a new line of treatment. Any line of treatment with two or more antineoplastic therapies administered between lines was considered combination therapy. First-line chemotherapy and / or hormonal therapy administered 180 days or earlier before the start of a subsequent therapy was considered adjuvant.

[0350] To determine rwPFS, the last effective date was defined for each course of treatment as either a) the date of last record if treatment was ongoing (censored), b) the date of death (event), c) the date of initiation of a subsequent line of therapy (event), or d) the latest last available date (censored if there was no subsequent line of therapy or death). rwPFS was calculated as the difference (in months) between the date of initiation of a line of treatment and the last effective date. OS was calculated as the difference (in months) between the date of initiation of a line of treatment and the date of death (or censored).

[0351] Biomarker Data

[0352] Multiplex PCR-based comprehensive genomic profiling (PCR-CGP) including TMB assessment was performed on FFPE solid tumor tissues using StrataNGS (Strata Oncology, Ann Arbor, MI). The current version of StrataNGS is a 437-gene non-regulated test (LDT) for FFPE tumor tissue samples performed on simultaneously isolated DNA and RNA, which has been validated in over 1,900 FFPE tumor samples and is available for use in patients with covered Medicare beneficiaries. 55 Earlier StrataNGS versions were also used during the study period and all had similar performance for TMB assessment (and MSI) as used herein. 56 In parallel, immune gene expression was assayed as described. 54 Determined by analytically and clinically validated multiplex PCR-based qTP via survey / supplementary tests performed on the same co-isolated RNA, different versions of this quantitative transcriptome profiling test were performed in parallel with StrataNGS (assessing 26, 46, and currently 103 expression targets) with panel-specific scaling validated by concordance analysis performed where necessary. PCR amplicons spanning one or more exons were selected for each target gene, and multiple housekeeping genes (see Supplementary Methods) were included along with the three pan-cancer stable housekeeping genes used for clinical testing. qTP was performed using Ampliseq after reverse transcription followed by Ion Torrent-based next-generation sequencing. Expressed target transcripts were measured in normalized reads per million (nRPM), whereby the raw expressed target read counts were normalized by a factor that yielded the median housekeeping gene expression value matched to the standard reads per million standard for the same genes in a reference FFPE normal cell line sample (GM24149) run in parallel with all clinically tested samples. 54 The relevant components of the analytical and clinical validation of the current version of the integrated CGP+qTP LDT, including the IRS model, are described in Supplementary Methods.

[0353] statistical analysis

[0354] Unadjusted rwPFS and overall survival (OS) across groups and treatments were visualized using the Kaplan-Meier method. Adjusted rwPFS and OS analyses were performed to compare group outcomes (by adjusted hazard ratios and two-sided p-values) using Cox proportional hazards models unless otherwise specified. Common covariate adjustments across all models included age and sex. Repeated measures took into account situations in which participants had multiple records (e.g., prior treatment followed by pembrolizumab monotherapy). When appropriate, analysis-dependent covariates included IRS group, tumor type (most relevant for cohort vs. all other types), number of systemic therapy lines, TMB status (high vs. low), type of therapy (monotherapy or combination), type of PD-(L)1 therapy (PD-1 or PD-L1 therapy), CDKN2A status (wild type or deep deletion) and tumor content (continuous). The analysis shown compares the MSKCC definition of TMB-sensitive tumor types (MSI-H as TMB sensitive, POLE as TMB sensitive) instead of the most relevant tumor type versus all others. mutant , NSCLC, head and neck cancer, or melanoma; all other samples as TMB non-sensitive) 57 were used. Performance status (or surrogates) were not available from data collected as part of the Strata Trial. The proportional hazards assumption was checked for each model and cohort of interest using Schoenfeld residuals. Results of non-stratified analyses are presented throughout (discovery cohorts) as proportional hazards-preserving stratified analyses that yielded similar covariate effect sizes where the assumption was not met; analyses of the discovery cohort for all monotherapy and analyses of the validation cohorts met the proportional hazards assumption. Where specified, two-sided log-rank tests were used to test for differences in rwPFS and OS curves (adjusted with Benjamini-Hochberg where appropriate).

[0355] For prospective analyses using an internal comparator cohort considering rwPFS in immediately preceding systemic therapy versus subsequent pembrolizumab monotherapy, an adjusted Cox proportional hazards model was utilized to examine the interaction between rwPFS of pembrolizumab versus prior chemotherapy within the same patient and IRS status (IRS-high versus low). Likelihood ratio tests for interaction compared a reduced model that excluded IRS by treatment interactions to a competing full model that included IRS by treatment interactions.

[0356] To determine the performance of IRS in situations where both PD-(L)1 monotherapy and combination therapy are used in the same line, we restricted the discovery cohort to the subset of patients with NSCLC treated with first-line pembrolizumab monotherapy or pembrolizumab plus chemotherapy combination therapy. Because PD-L1 IHC status and performance status were not available and the confounding factors driving this treatment decision were not known, we used nearest-neighbor propensity score matching (applied a 0.25 standard deviation caliper) using age, sex, TMB, IRS, and the normalized PD-L1 expression component of the IRS biomarker. 58 (See Supplementary Results for validation of this biomarker vs. IHC in a separate cohort.) The standard deviation of the propensity scores for monotherapy patients was 0.25 * All patients in the combination therapy cohort who were not matched within were dropped. Confirmation was made that the final matched monotherapy and combination therapy were not significantly different (two-tailed t-test for continuous variables and two-tailed Fisher's exact test for categorical variables; both significant at p<0.05). Kaplan-Meier analysis was used to visualize rwPFS of monotherapy versus combination therapy in the separate IRS-H and IRS-L populations, using two-tailed log-rank tests to compare outcomes of therapy groups.

[0357] Correlation between rwPFS and overall survival (OS) was calculated using Spearman's p within patients with both a documented death event and at least two lines of therapy. Throughout this study, TMB-H was defined as ≥10 Mut / Mb by StrataNGS, given previous validation of TMB by StrataNGS and high concordance with TMB estimates from FoundationOne tissue testing (see Supplementary Methods). 55 All statistical analyses were performed in R (v.4) and SAS (v.9.4). For all cohort analyses, a two-sided p-value <0.05 was considered statistically significant.

[0358] Model development and validation of the Immunotherapy Response Score (IRS)

[0359] The association of TMB and 23 candidate immune and proliferation gene expression biomarkers with rwPFS of pembrolizumab was determined using Cox proportional hazards regression in the 648 patient pembrolizumab (both monotherapy and combination) discovery cohort. TMB measurements were log2 transformed and gene expression measurements were log2 transformed and median centered according to laboratory workflow prior to analysis. Feature selection was performed via Lasso penalized Cox proportional hazards regression in this 648 patient discovery cohort, with the Lasso penalty condition selected as the value that maximized the concordance index via 5-fold cross-validation. Model coefficients for the five features with non-zero coefficients in the Lasso model were finalized via standard Cox regression. Individual patient IRS were derived from the Cox model as follows:

[0360] IRS=0.273758×TMB+0.112641×PD-1+0.061904×PD-L1-0.077011×TOP2A-0.057991×ADAM12

[0361] We assigned patients to one of two IRS groups to compare patient outcomes based on a balance between minimizing the hazard ratio for IRS-H versus IRS-L and maximizing the IRS-H monotherapy population (i.e., low (L) < 0.873569 and high (H) ≥ 0.873569; more likely to be beneficial).

[0362] After locking the IRS model (and the -H vs. -L threshold), a power analysis was performed to determine the appropriate size of the independent validation cohort. In the entire discovery cohort, 46% of patients were IRS-H, and we observed an adjusted hazard ratio of 0.49 (47% event rate) for rwPFS for IRS-H vs. IRS-L, therefore, assuming a 1:1 ratio of IRS-H vs. IRS-L and a 50% event rate, a validation cohort of 180 patients had 90% power to detect a similar (0.5) hazard ratio. We then identified all (n=248) patients in SCMD who met the validation cohort inclusion / exclusion criteria described above (same as the discovery cohort, except only that it included any non-pembrolizumab PD-(L)1 monotherapy treatment), and then applied the locked IRS model (and the -H vs. -L threshold) to these subjects.

[0363] result

[0364] Clinical and Molecular Data

[0365] The Strata Trial (NCT03061305) is an observational clinical trial evaluating the impact of tumor molecular profiling on patients with advanced solid tumors. De-identified demographic, clinical, and molecular data from patients in the Strata Trial have been maintained in the Strata Clinical Molecular Database (SCMD). With a data cutoff of July 12, 2022, SCMD will include patients who have undergone routine FFPE tumor tissue molecular profiling with the Strata NGS CGP test. 55、56, containing clinical and molecular data from a total of 57,648 unique patients with advanced solid tumors (from 47 tumor types) from 59 U.S. healthcare systems and 9,899 Strata Trial patients (from 43 tumor types) from 30 U.S. healthcare systems with treatment data from at least one systemic anti-neoplastic agent ( Figure 40 , and Tables S1 and S2 ).

[0366] For all Strata Trial patients with treatment data in SCMD, antineoplastic treatment start and stop dates (all prior therapies and up to 3 years after Strata trial enrollment) were obtained from automated electronic health record queries or manual entries, with data updated periodically by the submitting site, and dates of death were obtained similarly. Because of variability in real-world treatment patterns, time to next therapy (TTNT) as a measure of real-world progression-free survival (rwPFS) was determined directly from treatment start and stop dates for each line of therapy, taking into account adjuvant / systemic therapy, monotherapy / combination therapy, potential overlap in treatment start / stop dates, and repeated lines of therapy (whether monotherapy or combination) (Figure 41).

[0367] Among 9,899 patients, the median follow-up from initiation of first systemic treatment was 15.4 months [interquartile range (IQR) 6.9-29.4 months]. The median number of total lines of systemic therapy per patient was 1 (IQR 1-2), with 49.2% of systemic lines being monotherapy, and the median number of systemic therapies per line was 2 (IQR 1-2). The median number of total lines of systemic therapy per patient after enrollment in the Strata trial was 1 (IQR 0-1), with 47.2% of systemic lines being monotherapy, and the median number of systemic therapies per line was 2 (IQR 1-2). As expected, median rwPFS was shorter in subsequent lines of therapy (median rwPFS in first, second, and third lines, 9.4 [95% CI, 9.1–9.7], 8.7 [95% CI, 8.4–9.1], and 7.1 [95% CI, 6.7–7.4] months, respectively, unadjusted log-rank p<0.0001; Figure 42). 56 The frequency of molecular alterations in the first approximately 30,000 consecutive patients enrolled in the Memorial Sloan Kettering single institution pan-cancer profiling effort, MSK-IMPACT, was 59 We previously demonstrated that the results were similar to those observed in . Additional clinical and molecular analyses supporting the generalizability of SCMD and validity of SCMD are described in Supplementary Results and Figures 43a-c.

[0368] Biomarkers for analysis of benefit of anti-PD-1 / PD-L1 blockade

[0369] To develop a tumor-agnostic PD-(L)1 blockade predictive biomarker based on integrative CGP+qTP, we first limited the results to 648 of 9,899 (6.5%) patients in SCMD who met all of the inclusion / exclusion criteria (see Methods), including a valid TMB measurement from the StrataNGS test (including meeting the requirement of 20% tumor content overall), a valid immune gene expression quantification from an exploratory multiplex PCR-based qTP test, and a line of systemic therapy containing pembrolizumab (Figure 40). As shown in Figure 45a, this discovery cohort was composed of patients with 26 tumor types, of which NSCLC accounted for 265 (40.9%); tumor types and demographics are provided in Tables S1 and S2. rwPFS was estimated for each patient from the time they started a pembrolizumab-containing line of therapy to the time they stopped that line and started a new line of therapy or died, and both rwPFS and OS were used to study the outcome of treatment based on a comparison of these endpoints (Supplementary Results and Figure 44). Importantly, to confirm the validity of ≥10 Mut / Mb from StrataNGS testing to define TMB-H, we found that TMB-H patients (n=130) had significantly longer pembrolizumab monotherapy rwPFS vs. TMB-L patients (n=291; median rwPFS not reached [95% CI 16.6–NA] vs. 7.2 [95% CI 6.0–10.7] months; adjusted hazard ratio 0.37 [95% CI 0.25–0.54]; p<0.0001, when adjusted for age, sex, most common tumor type [NSCLC] vs. other, and line of systemic therapy; Fig. 45a) and significantly longer OS (median OS The results demonstrated that patients with CR had a median survival time of 16.7 [95% CI, 13.2-22.9] months (adjusted hazard ratio 0.44 [95% CI, 0.29-0.67], p=0.0001; Figure 45b) versus not reached [95% CI, NA-NA] vs. 16.7 [95% CI, 13.2-22.9] months (adjusted hazard ratio 0.44 [95% CI, 0.29-0.67], p=0.0001; Figure 45b).

[0370] To identify potential expression-based biomarkers of PD-(L) therapy benefit, we first considered 23 candidate immune and proliferation gene expression biomarkers (from 21 genes; two amplicons targeting separate inter-exon connections of PDCD1 [PD-1] and CD274 [PD-L1] were included) that were evaluated across clinical RNA test runs in parallel with the StrataNGS CGP test (creating the TMB). Data on housekeeping gene selection, correlation of independent PD-1 and PD-L1 amplicons, correlation of tumor type expression profiles for candidate gene expression biomarkers between SCMD and tumors profiled in The Cancer Genome Atlas (TCGA), as well as analysis and clinical validation of the qTP component of the CGP+qTP test (including qRT-PCR and clinical IHC data from >1,000 total FFPE tumors) are described in Supplementary Methods, Supplementary Results, Table S3, and Figures 46a–f and 47a–d. Therefore, we evaluated the association of 23 candidate immune / proliferative gene expression biomarkers with rwPFS of pembrolizumab by StrataNGS-derived TMB in a discovery cohort of 648 patients. As shown in table S4, significant (p<0.01) univariate predictors included TMB (HR=0.79; p<0.0001), PD-1 expression (HR=0.91; p=0.001) and PD-L1 expression by both amplicons (both HR=0.92; both p=0.005).

[0371] Integrated Immunotherapy Response Score (IRS) for predicting benefit of PD-(L)1 blockade

[0372] To develop an integrative model predicting the benefit of PD-(L)1 therapy, we performed Lasso-penalized Cox proportional hazards regression with 5-fold cross-validation in this discovery cohort of 648 patients, with the highest concordance index obtained using a 5-condition model including TMB, PD-1, PD-L1, ADAM12, and TOP2A (Figure 48). Increases in TMB, PD-1, and PD-L1 were associated with longer pembrolizumab rwPFS, whereas increases in ADAM12 and TOP2A were associated with shorter pembrolizumab rwPFS. The same feature set was also obtained via an exhaustive combinatorial search of all 5-condition models with standard Cox proportional hazards regression, so the 5-condition Cox proportional hazards model was used to create the final integrative model (multivariate analysis for the final 5-condition set is shown in Table S4). As shown in table S5, across 24,463 Strata Trial samples in SCMD with informative TMB and gene expression (regardless of treatment data availability), TMB was minimally correlated with all final model gene expression biomarkers (Spearman ρ = 0.032 [ADAM12] to 0.211 [TOP2A]), while correlations for individual gene expression biomarkers ranged from ρ = 0.033 (PD-1 vs. TOP2A) to ρ = 0.571 (PD-1 vs. PD-L1).

[0373] To evaluate the potential of multivariate models to predict outcome of PD-1 / PD-L1 blockade treatment, we derived individual immunotherapy response scores (IRS) from the final five-variable model and assigned 648 patients to either the IRS-high group [-H; n=298 (46.0%); associated with a higher benefit of PD-1 / PD-L1 blockade] or the IRS-low group (threshold set by balancing maximizing IRS-H group size vs. minimizing the IRS hazard ratio for unadjusted rwPFS) and compared outcomes between groups by Kaplan-Meier analysis and Cox proportional hazards modeling after adjusting for age, sex, most frequent tumor type (NSCLC) vs. other, type of line (monotherapy / combination) and line of systemic therapy. As shown in Figure 35b and c, IRS-H patients had significantly longer pembrolizumab rwPFS (IRS-H vs. IRS-L median rwPFS 16.8 [95% CI, 14.9-22.9] vs. 7.2 [95% CI, 6.2-8.4] months; adjusted hazard ratio 0.49 [95% CI, 0.39-0.63]; p<0.0001) and OS (IRS-H vs. IRS-L median OS not reached [95% CI, 29.9-NA] vs. 17.1 [95% CI, 13.4-22.8] months; adjusted hazard ratio 0.53 [95% CI, 0.40-0.70]; p<0.0001; Figure 49a and b). IRS-H also demonstrated a statistically significant improvement in survival when using restricted mean survival analyses, in the unadjusted analysis (IRS-H vs. IRS-L mean event-free rwPFS 15.70 [95% CI, 14.53 to 16.88] vs. 10.63 [95% CI, 9.61 to 11.65]; OS 25.50 [95% CI, 23.61 to 27.39] vs. 19.24 [95% CI, 17.48 to 21.00]) and when adjusting for the same CPH model covariates listed above (rwPFS IRS-H vs. IRS-L 4.80 [95% CI, 3.20 to 6.41]; p<0.0001; OS IRS-H vs. IRS-L 6.00 [95% CI 3.37 to 8.63], p<0.0001; Table S6) both demonstrated significant rwPFS and OS benefits (at prespecified times of 24 and 36 months, respectively).

[0374] PD-(L)1 combination therapy regimens vary by tumor type, with little evidence of additional benefit from currently approved PD-(L)1 combination regimens. 33 , TMB has been shown to be broadly predictive of benefit from monotherapy PD-(L)1. 25~32 , We also restricted the results of the discovery cohort to only patients treated with pembrolizumab monotherapy (n = 421; 46.1% IRS-H). As shown in Figure 36a and b, IRS-H patients had significantly longer pembrolizumab rwPFS (IRS-H vs. IRS-L median rwPFS 21.9 [95% CI, 16.1-NA] vs. 6.2 [95% CI, 5.2-8.2] months; adjusted [for entire cohort, excluding line type] hazard ratio 0.45 [95% CI, 0.33-0.61]; p<0.0001) and OS (IRS-H vs. IRS-L median OS not reached [95% CI, 29.9-NA] vs. 15.5 [95% CI, 11.8-23.2] months; adjusted hazard ratio 0.52 [95% CI, 0.37-0.74]; p=0.0002).

[0375] Validation of an integrated IRS model to predict benefit of PD-1 / PD-L1 blockade

[0376] We next sought to validate the ability of the IRS to predict the outcome of PD-(L)1 monotherapy treatment by both rwPFS and OS in an independent cohort. Based on a power analysis (see Methods), we identified a sufficient cohort of 248 of 9,899 (2.5%) eligible patients in SCMD who met the same inclusion / exclusion criteria as the discovery cohort (no patients in the discovery cohort) except that they were treated with systemic non-pembrolizumab anti-PD-(L)1 monotherapy (effective TMB, and gene expression with documented antineoplastic agent treatment). As shown in Figure 35a, the validation cohort of PD-(L)1 monotherapy (n=248; PD-1 n=194 [78%] and n=54 [22%] PD-L1) consisted of patients with 24 tumor types (25% melanoma [the most frequent tumor type]), with tumor types and demographics provided in Tables S1 and S2. All patients in the validation cohort were assigned to the IRS-H or IRS-L group using a locked IRS model (48.4% IRS-H), and group outcomes were compared after adjusting for the discovery monotherapy analysis (except for adding type of therapy [PD-1 vs. PD-L1] as a covariate). As shown in Figure 36c and d, by Kaplan-Meier analysis, IRS-H patients had significantly longer PD-(L)1 monotherapy rwPFS (IRS-H vs. IRS-L median rwPFS 23.1 [95% CI, 17.1-32.9] vs. 10.2 [95% CI, 8.7-14.8] months, adjusted hazard ratio=0.52 [95% CI, 0.34-0.80], p=0.003) and OS (IRS-H vs. IRS-L median OS 40.4 [95% CI, 32.9-NA] vs. 21.4 [95% CI, 17.0-46.8] months, adjusted hazard ratio=0.49 [95% CI, 0.30-0.80], p=0.005) compared with IRS-L patients. As described in the Supplementary Results and shown in Figures 50a-d, results were similar when patients were stratified by PD-1 vs. PD-L1 therapy. Taken together, these results demonstrate the development and validation of an integrated DNA- and RNA-based predictor of benefit of PD-(L)1 blockade, with IRS-H patients showing significantly longer rwPFS and OS in an independent validation cohort.

[0377] IRS versus TMB and the emergence of single-gene biomarkers to predict benefit of PD-1 / PD-L1 blockade

[0378] As described above, TMB has been shown to predict benefit from both monotherapy PD-1 (pembrolizumab and nivolumab) and PD-L1 (atezolizumab) through both retrospective and prospective studies, although ORR at the same TMB cutoff varies by agent and TMB cutoff. Hence, quantitative TMB is a component of the IRS model, but both TMB and IRS are reported as binary predictors (given the near requirement of categorical biomarkers for clinical implementation) and therefore have clinical utility, and the IRS model should identify a population of patients at least comparable to the TMB-H population with similar PD-(L)1 benefit. As shown in Figure 36e, in the monotherapy of 421 patients treated with a subset of the discovery pembrolizumab cohort, 194 (46.1%) and 130 (30.9%) patients were identified as IRS-H and TMB-H, respectively, whereas in the validation cohort of 248 patients, 120 (48.4%) and 78 (31.5%) patients were identified as IRS-H and TMB-H, respectively. In the pembrolizumab cohort, Cox proportional hazards analysis demonstrated that both classifications TMB (TMB-H vs. TMB-L) and IRS (IRS-H vs. IRS-L) were associated with a significant reduction in rwPFS (TMB-H vs. TMB-L adjusted hazard ratio 0.37 [95% CI, 0.25 to 0.54], p<0.0001; IRS-H vs. IRS-L adjusted hazard ratio 0.45 [95% CI, 0.33 to 0.61], p<0.0001) and OS (TMB-H vs. TMB-L adjusted hazard ratio 0.44 [95% CI, 0.29 to 0.67]; IRS-H vs. IRS-L adjusted hazard ratio 0.46 [95% CI, 0.27 to 0.66], p<0.0001) of pembrolizumab monotherapy in models adjusted separately for IRS and TMB. It was a significant predictor of glaucoma with adjusted hazard ratio 0.52 [95% CI 0.37 to 0.74]; Figure 36e).However, in the validation cohort, IRS, but not TMB, significantly increased the risk of PD-(L)1 rwPFS (TMB-H vs. TMB-L adjusted hazard ratio 0.87 [95% CI, 0.55 to 1.37], p = 0.54; IRS-H vs. IRS-L adjusted hazard ratio 0.52 [95% CI, 0.34 to 0.80], p = 0.003) and OS (TMB-H vs. TMB-L adjusted hazard ratio 0.86 [95% CI, 0.51 to 1.44], p = 0.56; IRS-H vs. IRS-L adjusted hazard ratio 0.87 [95% CI, 0.55 to 1.37], p = 0.02) in models adjusted separately for IRS and TMB. (Kaplan-Meier plots of rwPFS and OS stratified by TMB status are shown in Figures 45c and d.) As shown in Figure 36f, across 24,463 Strata Trial samples in SCMD with informative TMB and gene expression (regardless of treatment data availability), overall the IRS-H population was nearly twice as large as the TMB-H population (20.9% vs. 10.8%).

[0379] As shown in Figure 51a and b, Kaplan-Meier analysis of the pembrolizumab monotherapy cohort stratified by IRS and TMB status revealed that the median rwPFS was significantly longer in IRS-H / TMB-H vs. IRS-H / TMB-L (median rwPFS not reached [95% CI 16.6-NA] vs. 16.7 [95% CI 8.8-22.9] months, pairwise log-rank p=0.02 with Benjamini-Hochberg adjustment), but the median OS was not significantly different between IRS-H / TMB-H vs. IRS-H / TMB-L patients (median rwPFS not reached [95% CI NA-NA] vs. 22.9 [95% CI 15.3-NA] months, pairwise log-rank p=0.12 with Benjamini-Hochberg adjustment). As shown in Figure 51c and d, in the validation cohort, neither median PFS nor median OS was significantly different between IRS-H / TMB-H vs. IRS-H / TMB-L patients (IRS-H / TMB-H vs. IRS-H / TMB-L median rwPFS 21.0 [95% CI, 13.6 to NA] vs. 28.2 [95% CI, 17.1 to NA] months, pairwise log-rank adjusted with Benjamini-Hochberg p=0.31; median OS 40.4 [95% CI, 30.4 to NA] vs. not reached [95% CI, 32.9 to NA] months, pairwise log-rank adjusted with Benjamini-Hochberg p=0.53). Of interest, only a small minority of patients in both the discovery and validation cohorts (and in the overall SCMD population described below) were IRS-L / TMB-H (2.1% and 4.0% in the discovery [monotherapy] and validation cohorts, respectively), and therefore their benefit from PD-(L)1 monotherapy was unclear.

[0380] Despite the fundamental limitations of all genomic biomarkers (including TMB) for predicting response to PD-(L)1 therapy, several single genes that may add to the ability of TMB and / or PD-L1 IHC to predict PD-(L)1 benefit have also been identified from translational research studies. 60~68Two recent reports suggested that CDKN2A deep deletion (homozygosity loss) status could be improved in TMB alone to predict benefit of monotherapy PD-(L)1. 60、68 Therefore, we assessed whether inclusion of CDKN2A deep deletion status was an independent predictor of monotherapy PD-(L)1 benefit by adjusted Cox proportional hazards modeling in the subset of patients with valid CDKN2A deep deletion status (StrataNGS detection limit for deep deletions is 40% tumor content) from the discovery (n=310 [47.8%] of 648) and validation (n=199 [79.9%] of 249) cohorts, after also excluding those treated with combination therapy. As shown in Figure 52a-d, IRS status, but not CDKN2A deep deletion status, confirmed the findings when CDKN2A deep deletion status was added to the appropriately adjusted Cox proportional hazards model, confirming the limitations of genomic markers alone for predicting PD-(L)1 therapy response (rwPFS IRS-H vs. -L adjusted hazard ratio 0.48 [95% CI 0.33-0.69], p<0.0001; rwPFS CDKN2A deep deletion vs. CDKN2A wt adjusted hazard ratio 1.07 [95% CI 0.67-1.70], p=0.78; OS IRS-H vs. -L adjusted hazard ratio 0.48 [95% CI 0.32-0.74], p=0.0009; OS CDKN2A deep deletion vs. CDKN2A wt adjusted hazard ratio 1.02 [95% CI, 0.61 to 1.72], p=0.94) and the validation cohort (rwPFS IRS-H vs -L adjusted hazard ratio 0.47 [95% CI, 0.30 to 0.75], p=0.001; rwPFS CDKN2A deep deletion vs CDKN2A wt adjusted hazard ratio 1.58 [95% CI, 0.97 to 2.57], p=0.07; OS IRS-H vs -L adjusted hazard ratio 0.49 [95% CI, 0.29 to 0.83], p=0.008; OS CDKN2A deep deletion vs CDKN2A wt It was a significant predictor of rwPFS and OS in both cohorts (adjusted hazard ratio 0.99 [95% CI 0.56 to 1.75], p = 0.96).

[0381] Taken together, these results demonstrate that in both discovery and independent validation cohorts, IRS identifies a higher proportion of patients than TMB alone with similar benefit from PD-(L1) therapy, establishing the clinical utility of IRS biomarkers and demonstrating the value of integrating quantitative gene expression with TMB to predict benefit of PD-1 / PD-L1 monotherapy treatment. Additional analyses supporting the robustness of the IRS model to temporal sample collection (pre-CPI treatment) and variable tumor content are described in Supplementary Results and Figures 53a-b and 54a-e.

[0382] Confirmation of the predictive nature of the IRS

[0383] To establish the IRS model as predictive and non-prognostic, we first evaluated the internal comparator cohort for the pembrolizumab monotherapy cohort, consisting of 146 of 648 (22.5%) patients who received a previous line of systemic therapy before pembrolizumab monotherapy (demographics and type of therapy are shown in Table S7). For each patient, rwPFS was determined for the line of systemic therapy immediately preceding pembrolizumab and for the pembrolizumab monotherapy line, and rwPFS stratified by IRS status was assessed by Kaplan-Meier analysis (Figure 37a). Pembrolizumab monotherapy compared with the rwPFS of the previous line of therapy was not significantly different in IRS-L patients (IRS-L pembrolizumab vs. previous therapy median rwPFS 5.2 [95% CI 4.0-7.2] vs. 5.7 [95% CI 4.6-6.4] months, log-rank p=0.15; Fig. 37b), but rwPFS of pembrolizumab was significantly longer than the previous line of therapy in IRS-H patients (IRS-H pembrolizumab vs. previous therapy median rwPFS 34.8 [95% CI 11.9-NA] vs. 4.8 [95% CI 4.0-6.8] months, log-rank p<0.0001; Fig. 37c). Testing for an interaction between pembrolizumab versus previous line of treatment and IRS status (IRS-H vs. IRS-L) (model shown in Table S8) was significant (likelihood ratio test for interaction, p=0.001). Notably, when this analysis was restricted to 46 patients with non-MSI-H (StrataNGS clinical test) tumors among the approved tumor types for non-PD(L)1 monotherapy, only IRS-H patients still had a significantly longer rwPFS of pembrolizumab monotherapy than the previous line of therapy (IRS-H pembrolizumab versus previous line of therapy median rwPFS 11.9 [95% 7.8-NA] vs. 3.2 [95% CI 2.3-9.6] months, log-rank, p=0.005; Figure 55a and b). Taken together, these results confirm the predictive nature of the IRS biomarker across tumor types.

[0384] As only 76 subjects in the validation cohort received at least second-line PD-(L)1 blockade, we instead leveraged a compendium of treatment data across SCMD patients not included in the discovery or validation cohorts to further confirm the predictive nature of the IRS biomarker. Across all 3,184 patients in SCMD with non-PD-(L)1 or first-line CTLA4 systemic therapy who otherwise met the criteria for the discovery and validation cohorts (n=592 IRS-H), IRS status was not a significant predictor of non-PD-(L)1 (or rwPFS of CTLA4 systemic therapy by Cox proportional hazards modeling (IRS-H vs. -L) when adjusted for age, sex, most common tumor type (colorectal cancer) vs. other, and monotherapy vs. combination therapy. Median rwPFS 7.0 [95%CI 6.1-7.9] vs. 8.5 [95%CI 8.0-9.0] months, adjusted hazard ratio 1.05 [95%CI 0.92-1.19], p=0.45), confirming the predictive nature of the IRS model (Figure 55c). Second, given the mechanistic differences between CTLA4 and PD-(L)1 blockade and the lack of additive or synergistic treatment effects between these agents in melanoma, 33 , we evaluated the ability of IRS to stratify benefit of combination ipilimumab + nivolumab (CTLA4+PD-1) in a cohort of 70 patients (n=30 IRS-H) who had received combination ipilimumab + nivolumab treatment but were otherwise eligible for the validation cohort (8 tumor types; 47% melanoma). As shown in Figure 55d, after adjusting for age, sex, most common tumor type (melanoma) vs. other, and line of therapy, rwPFS of combination ipilimumab + nivolumab (median IRS-H vs. IRS-L rwPFS 11.4 [95% CI 8.4-NA] vs. 10.8 [95% CI 5.9-NA] months, adjusted hazard ratio 0.78 [95% CI 0.34-1.76], p=0.55) was not significantly different with IRS status.

[0385] Preliminary analysis of IRS in first-line NSCLC patients

[0386] As described above, despite little, if any, evidence for additive or synergistic benefit of PD-(L)1 and other agents in approved combination regimens, PD(L)-1 combination regimens are rapidly developing and moving to earlier lines of therapy, highlighting the need for improved biomarkers that can predict the benefit of PD-(L)1 monotherapy. For example, in first-line advanced NSCLC, both monotherapy pembrolizumab and pembrolizumab + chemotherapy are approved for patients with PD-L1 IHC(TPS) ≥ 1% and ≥ 50%, however, prospective data are not available to guide monotherapy vs. combination therapy decision-making. Therefore, in the pembrolizumab cohort, we identified 242 patients with NSCLC who were treated with first-line systemic pembrolizumab monotherapy (n = 109) or pembrolizumab + chemotherapy (n = 133; Figure 40). Although this cohort is limited by the lack of TPS data for PD-L1, the IRS includes PD-L1 qTP expression, and we validated the accuracy of this individual transcript vs. TPS in NSCLC FFPE tumor samples (Fig. 47a-d). Consistent with both TPS and performance status primarily driving monotherapy vs. combination treatment decisions, we confirmed that patients treated with monotherapy were significantly older and had higher PD-L1 qTP expression compared to patients treated with combination therapy (Table S9). Hence, we performed propensity score matching (see Methods) between monotherapy and combination therapy groups using patient age, PD-L1 qTP expression, TMB, sex, and IRS, followed by the exclusion of 88 unmatchable patients, resulting in a final cohort of 154 patients (77 patients in each group) with no significant differences in any of these variables (Table S9). In IRS-L patients, rwPFS was significantly shorter in those treated with monotherapy versus combination therapy, as shown by Kaplan-Meier analysis of matched cohorts (median rwPFS 6.1 [95% CI 4.6–12.1] vs. 9.8 [95% CI 8.4–NA] months, log-rank p = 0.006; Figure 38a).In contrast, in IRS-H patients, rwPFS was not significantly different in those treated with combination therapy versus monotherapy (median rwPFS 16.1 [95%CI 12.9-NA] vs. 16.8 [95%CI 12.1-NA] months, log-rank p=0.93; Fig. 38b). Taken together, these results support pembrolizumab monotherapy as a potentially reasonable treatment option for the 34% of patients with TPS scores 1-49% who are IRS-H (Fig. 38c and Fig. 47a-d), consistent with a recent report evaluating TMB across PD-L1 IHC strata in patients with first-line NSCLC treated with PD-(L)1 monotherapy. 27 , suggesting more broadly its potential utility in identifying patients who may benefit from monotherapy PD-(L)1 versus combination therapy in current indications.

[0387] Distribution of pan-solid tumors in the IRS group

[0388] In both discovery and separate validation cohorts, we demonstrated that IRS identifies a larger population of patients with similar PD-(L)1 monotherapy benefit than TMB, however this analysis is limited by the requirement that patients received PD-1 / PD-L1 treatment. Therefore, we sought to leverage IRS distribution across tumor types (and pan-cancer biomarkers) across SCMD to understand the potential impact of IRS both within and outside of currently approved PD-(L)1 monotherapy indications. We therefore determined IRS for 24,463 patients in SCMD (NCT03061305) with informative TMB and gene expression data, with 20.9% and 79.1% of all patients classified as IRS-H and -L, respectively (Figure 39a). Approved Tumor Types for PD-(L)1 Monotherapy 69IRS-H (without considering PD-L1 IHC status) had a significantly higher proportion of IRS-H patients (37.6%) than tumor types approved for non-PD-(L)1 monotherapy (11.7%) (Figure 39b). Tumor types with a higher proportion of IRS-H patients include several known to be highly responsive to PD-(L)1 therapy, including lymphoma, non-melanoma skin cancer, melanoma, NSCLC, and renal cell carcinoma (which almost always has low TMB) (Figure 39c).

[0389] We finally examined the distribution of pan-solid tumors in the IRS group by TMB status, given that the pan-tumor approval and prospective trials of pembrolizumab in TMB-H tumors demonstrate efficacy in other PD-(L)1 monotherapy patients with TMB-H. In both approved and unapproved tumor types for PD-(L)1 monotherapy, the majority of TMB-H patients were also IRS-H (although only 1.8% of patients overall were IRS-L / TMB-H [3.1% and 1.0% in approved and unapproved tumor types, respectively]), however, overall the IRS-H population was nearly twice as large as the TMB-H population (20.9% IRS-H vs. 10.8% TMB-H, overall (Fig. 36f); 37.7% IRS-H vs. 22.6% TMB-H in approved tumor types, and 11.7% IRS-H vs. 5.1% TMB-H in unapproved tumor types, respectively; Fig. 39d) with similar PD-(L)1 monotherapy benefits as established herein. Importantly, this analysis demonstrates that 7.6% of patients in unapproved tumor types are IRS-H / TMB-L, representing a sizable population predicted to benefit from PD-(L)1 monotherapy.

[0390] Consideration

[0391] Leveraging a robust clinical and molecular database from the Strata Trial (NCT03061305), we herein present a method for predicting rwPFS (by time to next therapy) of pembrolizumab (anti-PD-1) in 648 patients from 26 solid tumor types using simultaneous clinically validated multiplex PCR-based NGS of DNA and RNA (StrataNGS CGP, and a separate RNA panel for quantitative transcriptome profiling). 54~56 From this, we developed an Integrated Immunotherapy Response Score (IRS) algorithm that combines TMB and quantitative gene expres...

Claims

1. 1. A method for identifying a subject who would benefit from checkpoint inhibitor therapy, comprising: a. measuring expression levels of RNA transcripts for PD-1, TOP2A, PD-L1, and ADAM12, as well as one or more reference genes, in a biological sample obtained from a tumor specimen from the subject, wherein the one or more reference genes comprise three genes selected from CIAO1, EIF2B1, HMBS, CTCF, GGNBP2, ITGB7, MYC, and SLC4A1AP; b. log2-transform, median-center, and normalize the measured expression level of the RNA transcript to the level of the RNA transcript of at least one reference gene to provide a normalized level of the RNA transcript; c. Measuring tumor mutational burden (TMB) in said biological sample and log2 transforming said TMB measurement to provide a transformed TMB measurement; d. calculating an immunotherapy response score (IRS) from the converted normalized levels of the RNA transcripts of PD-1, TOP2A, PD-L1, and ADAM12 and the converted TMB measurements, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and obtaining an IRS having a value indicative of a beneficial response to checkpoint inhibitor therapy; and e. identifying said subject as benefiting from said checkpoint inhibitor therapy.

2. 10. The method of claim 1, wherein the tumor specimen is from a cancer for which the checkpoint inhibitor therapy is not approved for its indicated use.

3. 3. The method of claim 1 or 2, wherein the tumor specimen is assessed as having low microsatellite instability or microsatellite stability.

4. 1. A method for identifying a subject who would benefit from checkpoint inhibitor therapy, comprising: a. receiving, by a processor, measured expression levels of RNA transcripts for at least two of PD-1, TOP2A, PD-L1, and ADAM12, and at least one reference gene, in a biological sample obtained from a tumor specimen from said subject; b. log2 transforming, median-centering, and normalizing, by a processor, the measured expression levels of the RNA transcripts relative to the levels of the RNA transcripts of the at least one reference gene to provide transformed, normalized levels of the RNA transcripts; c. receiving, by a processor, a measured tumor mutation burden (TMB) in the biological sample; d. by a processor, log2 transforming the TMB measurement to provide a transformed TMB measurement; e. calculating, by a processor, an immunotherapy response score (IRS) from the converted normalized levels of the RNA transcripts of the at least two of PD-1, TOP2A, PD-L1, and ADAM12 and the converted TMB measurement, which positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy; and f. Providing a determination whether the subject has a checkpoint inhibitor responsive cancer.

5. 1. A method for identifying a subject who would benefit from checkpoint inhibitor therapy, comprising: a. measuring the expression levels of RNA transcripts for PD-1, PD-L2, and optionally one or more of CD4, ADAM12, PD-L1, and VTCN1, and at least one reference gene, in a biological sample obtained from a tumor specimen from said subject; b. log2-transform, median-center, and normalize the measured expression level of the RNA transcript relative to the level of the RNA transcript of the at least one reference gene to provide a normalized level of the RNA transcript; c. Measuring tumor mutational burden (TMB) in said biological sample and log2 transforming said TMB measurement to provide a transformed TMB measurement; d. calculating an immunotherapy response score (IRS) from the normalized levels of the RNA transcripts of one or more of PD-1, PD-L2, and optionally CD4, ADAM12, PD-L1, and VTCN1, and the converted TMB measurements, wherein the IRS positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy, and obtains an IRS having a value indicative of a beneficial response to checkpoint inhibitor therapy; and e. identifying said subject as benefiting from said checkpoint inhibitor therapy.

6. 6. The method of claim 5, wherein the tumor specimen is from a cancer for which the checkpoint inhibitor therapy is not approved for its indicated use.

7. 1. A method for identifying a subject who would benefit from checkpoint inhibitor therapy, comprising: a. receiving, by a processor, measured expression levels of RNA transcripts for PD-1, PD-L2, and optionally one or more of CD4, ADAM12, PD-L1, and VTCN1, and at least one reference gene, in a biological sample obtained from a tumor specimen from said subject; b. log2 transforming, median-centering, and normalizing, by a processor, the measured expression levels of the RNA transcripts relative to the levels of the RNA transcripts of the at least one reference gene to provide normalized levels of the RNA transcripts; c. receiving, by a processor, a measured tumor mutation burden (TMB) in the biological sample; d. by a processor, log2 transforming the TMB measurement to provide a transformed TMB measurement; e. calculating, by a processor, an immunotherapy response score (IRS) from the normalized levels of the RNA transcripts of one or more of PD-1, PD-L2, and optionally CD4, ADAM12, PD-L1, and VTCN1, and the converted TMB measurement, which positively correlates with the likelihood that a patient will have a beneficial response to checkpoint inhibitor therapy; and f. Providing a determination whether the subject has a checkpoint inhibitor responsive cancer.

Images