WO2021202917A1

WO2021202917A1 - A noninvasive multiparameter approach for early identification of therapeutic benefit from immune checkpoint inhibition for lung cancer

Info

Publication number: WO2021202917A1
Application number: PCT/US2021/025442
Authority: WO
Inventors: Maximilian Diehn; Barzin Y. NABET; Mohammad SHAHROKH ESFAHANI; Arash ASH ALIZADEH
Original assignee: The Board Of Trustees Of The Leland Stanford Junior University
Priority date: 2020-04-01
Filing date: 2021-04-01
Publication date: 2021-10-07

Abstract

An integrated analytic method is provided, which uses a Bayesian framework to derive a single biomarker from circulating tumor DNA profiling and analysis of circulating immune cells to generate a prognostic for patient responsiveness to immune checkpoint inhibition. The methods robustly identify which patients will achieve durable clinical benefit from immune checkpoint inhibition, in some embodiments using only noninvasive blood draws.

Description

A NONINVASIVE MULTIPARAMETER APPROACH FOR EARLY IDENTIFICATION OF THERAPEUTIC BENEFIT FROM IMMUNE CHECKPOINT INHIBITION FOR LUNG CANCER

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims the benefit of and priority to U.S. Provisional Patent Application No. 63/003,737, filed April 1 , 2020, and U.S. Provisional Patent Application No. 63/016,120, filed April 27, 2020 the entire disclosure of which is hereby incorporated by reference herein in its entireties for all purposes.

GOVERNMENT SUPPORT

[0002] This invention was made with Government support under contract CA188298 awarded by the National Institutes of Health. The Government has certain rights in the invention.

BACKGROUND OF THE INVENTION

[0003] Antibody-based blockade of programmed death 1 and programmed death-ligand 1 (PD-[L]1) signaling has shown remarkable promise for treatment of advanced non-small lung cancer (NSCLC) (Reck et al., 2016; Socinski et al., 2018). Clinical trials combining PD-(L)1 blockade with cytotoxic therapy or with other immune checkpoint inhibition (ICI) strategies have shown higher response rates at the risk of higher toxicity (Gandhi et al., 2018; Hellmann et al., 2018). However, only a minority of patients achieve durable benefit from ICI and reliable biomarkers that can accurately identify these patients before or early during treatment have thus far remained elusive (Camidge et al., 2019). The most heavily studied biomarkers for predicting response to ICI prior to therapy include assessment of tumor PD-L1 expression and tumor mutational burden (TMB) (Cristescu et al., 2018; Reck et al., 2016). Moreover, PD-L1 has several major shortcomings as a predictive biomarker of durable benefit while TMB is continuing to be evaluated clinically (Camidge et al., 2019; Cristescu et al., 2018; Rizvi et al., 2018).

[0004] Current practice in the US for front-line therapy of advanced, driver mutation negative NSCLC is to treat patients with PD-L1 > 50% with pembrolizumab alone, and patients with PD-L1 < 49% with concurrent chemotherapy plus pembrolizumab (Hanna et al., 2020). However, a significant subset of patients whose tumors have < 49% PD-L1 staining can respond to immunotherapy alone. For example, the KEYNOTE-042 trial which compared pembrolizumab versus chemotherapy in advanced non- squamous NSCLC demonstrated that pembrolizumab alone had an objective response rate of 16% in patients with PD-L1 1-49%, comprising of 32% of all objective responses observed (Mok et al., 2019). Thus, PD-L1 staining alone is not sufficiently accurate to identify all potential responders to ICI in NSCLC. [0005] In the search for better biomarkers to predict which patients will respond to immunotherapy, several groups have recently demonstrated that multivariable models based on molecular analyses of tumor biopsy tissue collected prior to treatment can predict response to ICI (Anagnostou et al., 2020; Auslanderetal., 2018; Cristescu et al., 2018; Jiang etal. , 2018). Each of these methods achieves modest improvement for predicting benefit (area under the curve, [AUC]= 0.7-0.8) over PD-L1 expression (AUC = 0.6-0.7) (Mok et al., 2019) or TMB alone (AUC = 0.6-0.7) (Rizvi et al., 2018). Moreover, few studies have applied this type of approach to NSCLC. Additionally, the requirement for tumor tissue can be problematic in advanced NSCLC, where obtaining sufficient tissue for molecular analysis is not possible in a significant subset of patients (Green et al., 2014; Lim et al., 2015). Reliance on single biopsy specimens also risks confounding of biomarker results due to intra-tumoral heterogeneity or low tumor content, leading to lower than desired inter-observer concordance (Camidge et al., 2019). A separate approach for predicting ultimate clinical benefit of ICI would be early response assessment. Clinically, response is currently evaluated by conventional imaging, usually performed six

[0006] Although treatment of cancer, including without limitation melanoma and non-small cell lung cancer (NSCLC), with ICI can produce remarkably durable responses, many patients still show disease progression after treatment. Initial response assessment by conventional imaging is often unable to identify which patients will achieve durable clinical benefit (DCB). The present disclosure provides improved methods for selection of a treatment regimen.

SUMMARY OF THE INVENTION

[0007] Compositions and methods are provided for determining whether an individual with cancer will have a durable clinical benefit from treatment with an immune checkpoint inhibitor. Provided is an integrated analytic method, where a Bayesian framework is used to derive a single biomarker from circulating tumor DNA profiling and analysis of circulating immune cells, to generate a prognostic for patient responsiveness to immune checkpoint inhibition (ICI). In some embodiments that use only noninvasive blood draws, the methods robustly identify which patients will achieve durable clinical benefit from immune checkpoint inhibition. In an embodiment, the methods further comprise selecting a treatment regimen for the individual based on the analysis. In some embodiments, the prediction is based on analysis of a pretreatment sample. In some embodiments, the prediction is based on samples shortly after a first ICI treatment.

[0008] An individual for assessment by the method of the invention has cancer. In some embodiments the individual has been previously diagnosed with the cancer. In some embodiments the cancer is a carcinoma, including without limitation non-small cell lung carcinoma, small cell lung carcinoma, adenocarcinoma, squamous cell carcinoma, hepatocarcinoma, basal cell carcinoma, etc., which may be breast cancer, colorectal cancer, bladder cancer, head and neck cancer, renal cell cancer, liver cancer, skin cancer, pancreatic cancer, etc. In some embodiments the cancer is a lymphoma, e.g. Hodgkin lymphoma, non- hodgkin lymphoma, etc. In some embodiments the cancer is a melanoma. In certain embodiments the individual has non-small cell lung cancer (NSCLC), which may be early stage, or advanced stage.

[0009] In an individual with cancer, it is shown herein that circulating tumor DNA (ctDNA), tumor mutational burden (TMB), and immune profiling, e.g. peripheral CD8⁺ T cell levels, are independently associated with durable clinical benefit (DCB) in response to ICI. Features that are associated with DCB are lower concentrations of ctDNA; higher TMB, which is determined from sequencing ctDNA; and lower levels of circulating CD8⁺ T cells. TMB can be normalized as a ratio of ctDNA, which may be referred to as normalized bTMB, and the resulting value integrated with one or both of: circulating CD8+ T cell levels, and tumor expression of immune checkpoint protein, e.g. PD-L1 , in a Bayesian framework to provide a single predictive biomarker.

[0010] By integrating a measurement of ctDNA and TMB, i.e. normalized bTMB, and immune profiling, e.g. circulating CD8+ T cell levels, methods are provided for an entirely noninvasive multi-analyte assay (DIREct-On, Durable Immunotherapy Response Estimation by immune profiling and ctDNA- On-treatment) that robustly predicts which patients will achieve DCB in response to ICI with high accuracy. DIREct-On predicts an individual’s response to ICI by analysis of a non-invasive blood sample taken shortly after a first ICI treatment, e.g. within about 4 weeks, within about 3 weeks, within about 2 weeks, within about 1 week following therapy, e.g. a first dose of ICI. These results demonstrate that integrated ctDNA and circulating immune cell profiling provides accurate, noninvasive, and early forecasting of outcomes for cancer patients receiving ICI. DIREct-On can also incorporate pre-treatment with early on- treatment ctDNA response assessment to most accurately classify the likelihood of durable benefit after one cycle of immunotherapy. The integration of pre-treatment parameters that can be measured noninvasively allows for optimal classification of response, with earlier and significantly better performance than ctDNA dynamics alone.

[0011] Methods are also provided for a pre-treatment composite model (DIREct-Pre) that integrates tumor immune checkpoint expression, e.g. PD-L1 , with pre-treatment ctDNA, e.g. normalized bTMB, and circulating immune cell profiling, e.g. circulating CD8+ T cell levels, which accurately predicts outcomes.

[0012] DIREct-Pre and DIREct-On can be used to facilitate personalized selection of treatment, including ICI if appropriate, for patients with a number of different cancers. Types of cancers that may be suitable for analysis using DIREct-Pre and DIREct-On include, but are not limited to, carcinomas, sarcomas, gliomas, lymphomas, melanomas, etc., although hematologic cancers, such as leukemias, are not excluded. In one embodiment, the cancer is a non-small cell lung cancer. An individual with a high score that is predicted to benefit from ICI, can be selected, and treated, with an ICI, usually in combination with additional therapeutic agents. An individual with a low score that is not predicted to benefit from ICI can be selected, and treated, with non-ICI therapy, e.g. chemotherapy, non-ICI immunotherapy, radiation therapy, and the like. ICI of interest include, without limitation, inhibitors of PD-1 and inhibitors of PD-L1.

[0013] In one embodiment, DIREct-Pre and DIREct-On facilitate personalized selection of therapy, which may include ICI, for patients with advanced NSCLC, to improve outcomes while minimizing toxicities. For example, patients with late stage disease can be treated with single-agent PD-1 blockade for one cycle irrespective of PD-L1 expression and then have DIREct-On measured. Patients with a high DIREct-On score (expected durable benefit) remain on single agent PD-1 blockade whereas patients with low DIREct-On scores (expected lack of benefit) would receive treatment escalation through the addition of chemotherapy.

[0014] DIREct-Pre and DIREct-On both utilize measurements of ctDNA as a metric. ctDNA may be measured and quantified using a number of techniques known in the art. Examples of suitable ctDNA measurement techniques include, but are not limited to, Droplet Digital PCR (ddPCR), Beads, Emulsification, Amplification, and Magnetics (BEAMing), CAncer Personalized Profiling by deep Sequencing (CAPP-Seq), Tagged AMplification deep Sequencing (TAM-seq), Safe-sequencing (Safe-seq) Duplex sequencing, Integrated Digital Error Suppression (iDES)-enhanced CAPP-seq, etc.

[0015] DIREct-Pre and DIREct-On both utilize immune profiling as a metric. Immune profiling may be performed using a number of techniques known in the art. Examples of suitable immune profiling techniques include, but are not limited to, flow cytometry, immunohistochemistry, single cell RNA-seq, CIBERSORT, CIBERSORTx, etc.

[0016] DIREct-Pre and DIREct-On both utilize relative blood tumor mutational burden (bTMB) as a metric for determining if an individual will receive DCB from ICI treatment. bTMB may be measured using a number of techniques known in the art. Examples of suitable bTMB measurement techniques include, but are not limited to, sequencing of ctDNA, tumor whole- exome sequencing (WES), Oncomine Tumor Mutation Load Assay, etc.

[0017] In some embodiments, DIREct-Pre or DIREct-On may omit immune profile, ctDNA levels or bTMB data from their analysis. For instance, DIREct-Pre or DIREct-On scores may be calculated in the absence of bTMB measurements or immune profiling data (ie. abundance of CD8 T cells).

[0018] In other embodiments of the invention a device or kit is provided for the analysis of patient samples. Such devices or kits will include reagents that specifically identify one or more cells and signaling proteins indicative of the status of the patient, including without limitation affinity reagents. The reagents can be provided in isolated form, or pre-mixed as a cocktail suitable for the methods of the invention. A kit can include instructions for using the plurality of reagents to determine data from the sample; and instuctions for statistically analyzing the data. The kits may be provided in combination with a system for analysis, e.g. a system implemented on a computer. Such a system may include a software component configured for analysis of data obtained by the methods of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The invention is best understood from the following detailed description when read in conjunction with the accompanying drawings. It is emphasized that, according to common practice, the various features of the drawings are not to-scale. On the contrary, the dimensions of the various features are arbitrarily expanded or reduced for clarity. Included in the drawings are the following figures.

[0020] Figure 1 : Assocation of radiologic response and durable clinical benefit in NSCLC.

A) Rate of DCB or NDB in advanced NSCLC patients receiving ICI achieving partial response (PR), stable disease (SD), or progressive disease (PD) at the first scan by RECIST v1 .1 criteria.

B) Study schematic. Blood from NSCLC patients receiving ICI was collected and fractionated for CAPP- Seq ctDNA analysis and RNA-seq for CIBERSORTx immune profiling. Two main models were created and validated with the following parameters: 1) A pre-treatment response predictor that utilizes tumor PD-L1 expression, ctDNA metrics, and circulating immune cell profiling ; 2) A noninvasive early classifier of response that utilizes pre-treatment ctDNA metrics, circulating immune cell profiling, and early (<4 weeks) ctDNA dynamics.

[0021] Figure 2: Clinical and molecular features of advanced NSCLC patients receiving ICI. Patient characteristics. Each column represents an individual patient. Tumor histology, smoking status, best overall response, tumor PD-L1 expression, and ICI therapy type (PD-(L)1 blockade alone; PD-(L)1 blockade with either CTLA-4 or chemotherapy) are indicated. PFS is shown in months, where asterisks signify ongoing responses. TMB is presented as the number of nonsynonymous mutations per megabase of the coding exome captured, measured in the blood (See Figure 9A-B). Mutations in recurrently mutated genes identified by CAPP-Seq ctDNA analysis are shown at the bottom.

[0022] Figure 3: Pre-treatment ctDNA-normalized tumor mutation burden predicts response to ICI. A) Outcomes of immunotherapy treated patients from Gandara et al. (Gfandara et al., 2018) (normalized bTMB Discovery Cohort) stratified by high bTMB/MB (³14) and low bTMB/MB (<14). P-value calculated by two-sided Fisher’s Exact Test (DCB n = 139; NDB n = 290). B) Pre-treatment ctDNA concentration (haploid genome equivalents per mL of plasma, hGE/mL) and C) ctDNA-normalized bTMB in immunotherapy patients. P-values were calculating using a Wilcoxon test. D) Area under the curve (AUC) for individual parameters in immunotherapy patients generated by leave- one-out cross-validation (LOOCV) ROC analysis. E) Probability of PFS for high normalized bTMB (median = 4.14 mo.) and low normalized bTMB (median = 2.16 mo.) immunotherapy patients stratified by the LOOCV-identified optimal cut-point in the normalized bTMB Discovery Cohort (n = 429). F) Outcomes of chemotherapy treated patients from Gandara et al. (Gandara et al., 2018) stratified by high bTMB/MB (>14) and low bTMB/MB (<14). P-value calculated by two-sided Fisher’s Exact Test (DCB n=118; NDB n = 306). G) Pre-treatment ctDNA concentration and H) normalized bTMB (normalized bTMB Validation Cohort) in chemotherapy patients. P-values were calculated using a Wilcoxon test. I) Probability of PFS for high normalized bTMB (median = 3.25 mo.) and low normalized bTMB (median = 4.00 mo.) chemotherapy patients stratified by the LOOCV-identified optimal cut-point in the Chemotherapy Cohort (n = 424). J) Probability of PFS for high normalized bTMB (median = 8.07 mo.) and low normalized bTMB (median = 2.29 mo.) patients in the normalized bTMB Validation Cohort (n = 94) stratified by the cut-point of bTMB to ctDNA concentration identified in the bTMB Discovery Cohort.

[0023] Figure 4: Pre-treatment circulating immune profiling predicts response to ICI. A)

Comparison of pre-treatment circulating immune cell populations by CIBERSORTx in patients achieving DCB (n = 20) vs. NBD (n = 17). Cell populations are restricted to those with >1% median frequency in these patients. Shown are AUCs and 95% confidence intervals generated by bootstrapping for classifying DCB versus NDB by each cell population. Shown on the top is the median relative abundance of each cell type in these patients. B) Pre-treatment relative CD8 T cell fraction in circulation of ICI DCB (n = 20) and NDB (n = 17) with available CIBERSORTx immune profiling. P-value was calculating using a Wilcoxon test. C) LOOCV- generated PPV of pre-treatment response classification using the indicated pre-treatment tumor, ctDNA, and/or immune parameters in cases with all data types available (n = 27).

[0024] Figure 5: Early on-treatment ctDNA analysis classifies response to ICI. A) Early on- treatment ctDNA concentration within 4 weeks normalized to pre-treatment ctDNA concentration (median = 2.4 weeks, n = 46, DCB = 27, NDB = 19). Colors indicate the ultimate clinical outcome. The dashed line indicates 50% of pre-treatment ctDNA concentration (ctDNA molecular response). B) Early on-treatment ctDNA concentration at the first timepoint after the first cycle of ICI normalized to pre-treatment ctDNA concentration, stratified by ultimate clinical outcome. Colors indicate if the sample was collected <2 or 2-3.3 weeks after treatment initiation. The dashed line indicates 50% of pre-treatment ctDNA concentration. C) Probability of PFS in ActDNA <50% (median = 22.4 mo.) and ActDNA >50% (median = 2.30 mo.) patients within 4 weeks of treatment initiation.

[0025] Figure 6: Multiparameter Bayesian framework enables fully noninvasive response classification. A) Hazard ratio for low scores (below LOOCV-generated cut-point) of each model with the indicated parameters only considering patients with all parameters available (n = 26). Error bars are the 95% confidence intervals. NS = not significant, ^** = P< 0.01. B) Proportion of patients expected to achieve DCB (blue) or NDB (orange) by the DIREct-Pre model stratified by clinical outcome determined by RECIST in the DIREct Discovery Cohort (n = 34). C) Probability of PFS for high DIREct-Pre score (median = 8.2 mo.) and low DIREct-Pre score (median = 2.0 mo.) patients in the DIREct Discovery Cohort, using the optimal cut- point identified by LOOCV analysis (n = 34). D) Proportion of patients expected to achieve DCB (blue) or NDB (orange) by DIREct-On stratified by clinical outcome determined by RECIST in the DIREct Discovery Cohort (n = 34). E) Probability of PFS for high DIREct-On score (median = 16.5 mo.) and low DIREct-On score (median = 1 .9 mo.) patients in the DIREct Discovery Cohort, using the optimal cut-point identified by LOOCV analysis (n = 34).

[0026] Figure 7: DIREct-On validates and performs significantly better than each individual parameter. A) Probability of PFS for high DIREct-Pre score (median = 8.4 mo.) and low DIREct-Pre score (median = 2.6 mo.) patients in the DIREct Validation Cohort using the cut-point defined in the DIREct Discovery Cohort (n = 38). B) Probability of PFS for high DIREct- On score (median = 8.5 mo.) and low DIREct-On score (median = 2.1 mo.) patients in the DIREct Validation Cohort using the cut-point defined in the DIREct Discovery Cohort (n = 38). C) Accuracy (left) and hazard ratio (right) for each individual parameter and DIREct-On in the combined DIREct Discovery and Validation Cohorts for those cases with all data types available, using the cut- points identified the DIREct Discovery Cohort (n = 58). Error bars are 95% confidence intervals generated by bootstrapping. ^** = P < 0.01 , ^*** = P < 0.001 . D) Net reclassification improvement of DIREct-On compared to Bayesian models with each feature that comprises DIREct-On removed (top) and DIREct-On compared to each individual feature (bottom). Errors bars are 95% confidence intervals generated by bootstrapping. ^* = P< 0.01 , ^** = P < 0.01. E) DIREct-On score in the combined DIREct Discovery and Validation Cohorts (indicated by shape and color) stratified by response measured by RECIST and DCB versus NDB. The horizontal line indicates the threshold identified in the discovery cohort to best classify DCB versus NDB. F) Probability of PFS for high DIREct-On score (median = 16.49 mo.), low DIREct-On score (median = 3.50 mo.), in those patients with RECIST stable disease at the first available scan in the combined DIREct Discovery and Validation Cohorts (n = 18). G) Probability of PFS for high DIREct-On score (median = 11.69 mo.) patients, actual DCB patients measured by RECIST (median = 11.69 mo.), low DIREct-On score (median = 1.94 mo.) patients, and actual NDB patients measured by RECIST (median = 1.94 mo.) in the combined DIREct Discovery and Validation cohorts (n = 72). H) Potential application of DIREct-On to personalize immunotherapy in front-line treatment of advanced NSCLC. Patients would begin by receiving single agent PD-1 blockade for one cycle and would then either remain on PD- 1 blockade if DIREct-On forecasts durable response, or undergo treatment escalation by the addition of chemotherapy or additional ICI if DIREct-On forecasts lack of durable benefit. [0027] Figure 8: Data related to Figure 1. A) Study structure and CONSORT diagram for patient inclusion and allocation in this study.

[0028] Figure 9: Data related to Figure 2. A) Tumor mutation burden per megabase (TMB/MB) from tumor whole-exome sequencing (WES) as compared with cfDNA mutational burden per MB (n = 24). R = Pearson’s correlation coefficient. B) Validation of TMB/MB from tumor WES versus estimated TMB/MB from cfDNA using the linear regression relationship identified in B (n = 6).

[0029] Figure 10: Data related to Figure 3. A) Hazard ratios and 95% confidence intervals for each of the indicated variables in the Normalized bTMB Discovery Cohort (n=429) treating each as a standardized continuous variable. Normalized bTMB has a significantly greater hazard ratio compared to ctDNA or bTMB alone. B) Percent of patients stratified by high bTMB/MB (>14) and low bTMB/MB (<14) in ICI patients that achieve DCB (n = 39) versus NDB (n = 38) in our independent cohort (normalized bTMB Validation Cohort). P-value calculated by one-sided Fisher’s Exact Test. C) Pre-treatment ctDNA concentration and D) the ctDNA- normalized bTMB (normalized bTMB Validation Cohort). P-values were calculated using a Wilcoxon test. E) AUC for individual parameters in the normalized bTMB Validation Cohort generated by LOOCV ROC analysis.

[0030] Figure 11 : Data related to Figure 4. A) Comparison of cell proportions from flow cytometry of peripheral blood mononuclear cells (PBMCs) versus RNA-seq followed by CIBERSORTx in PBMCs (left) or plasma-depleted whole blood (PDWB, right) in matched samples from healthy donors (n = 3). R = Pearson’s correlation coefficient. B) Comparison of cell proportions from flow cytometry of PBMCs versus RNA-seq followed by CIBERSORTx in PBMCs (left) or PDWB (right) in matched samples from NSCLC patients (n = 10). R = Pearson’s correlation coefficient. C) Pre-treatment immune composition in circulation of ICI DCB (n = 20) and NDB (n = 17) with available CIBERSORTx immune profiling. Shown are major circulating immune cell populations, excluding CD8 T cells (shown in Fig. 3B). P-values were calculating using a Wilcoxon test.

[0031] Figure 12: Data related to Figure 5 ROC curve for classification of DCB and NDB by ctDNA dynamics within 4 weeks. Shown is sensitivity and specificity at the threshold of 50% pre-treatment ctDNA concentration (ctDNA molecular response).

[0032] Figure 13: Data related to Figure G A) Response prediction score by incorporating pre-treatment bTMB, ctDNA burden, CD8 T cell fraction, and the ctDNA change within 4 weeks with (x-axis) or without (y-axis) including tumor PD-L1 expression cases with all data types available (n = 26). B) ROC curve for classification of DCB and NDB by DIREct-Pre and DIREct- On.

[0033] Figure 14: Data related to Figure 7 A) Hazard ratios and 95% confidence intervals for each of models with the indicated feature removed in the combined DIREct Discovery and Validation Cohorts. DIREct-On high scores results in a significantly lower hazard ratio compared to each model with one feature removed. Errors bars are 95% confidence intervals generated by bootstrapping. B) Predicted probability of progression within 6 months versus observed probability of progression within 6 months in the DIREct Validation Cohort for DIREct-Pre (n = 38). C) Predicted probability of progression within 6 months versus observed probability of progression within 6 months in the DIREct Validation Cohort for DIREct-On (n = 38).

[0034] Table 1 : Summary of patient demographics, pathologic features, and treatment details

DETAILED DESCRIPTION

[0035] These and other features of the present teachings will become more apparent from the description herein. While the present teachings are described in conjunction with various embodiments, it is not intended that the present teachings be limited to such embodiments. On the contrary, the present teachings encompass various alternatives, modifications, and equivalents, as will be appreciated by those of skill in the art.

[0036] Most of the words used in this specification have the meaning that would be attributed to those words by one skilled in the art. Words specifically defined in the specification have the meaning provided in the context of the present teachings as a whole, and as are typically understood by those skilled in the art. In the event that a conflict arises between an art- understood definition of a word or phrase and a definition of the word or phrase as specifically taught in this specification, the specification shall control.

[0037] It must be noted that, as used in the specification and the appended claims, the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise.

[0038] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference.

[0039] Compositions and methods are provided for prognostic classification of patients according to their propensity for tumor benefit from immune checkpoint inhibitors. Once a classification or prognosis has been made, it can be provided to a patient or caregiver. The classification can provide prognostic information to guide clinical decision making, both in terms of institution of and escalation of treatment, and in some cases may further include selection of a therapeutic agent or regimen.

[0040] The term “immune checkpoint inhibitor” refers to a molecule, compound, or composition that binds to an immune checkpoint protein and blocks its activity and/or inhibits the function of the immune regulatory cell expressing the immune checkpoint protein that it binds (e.g., Treg cells, tumor-associated macrophages, etc.). Immune checkpoint proteins may include, but are not limited to, CTLA4 (Cytotoxic T-Lymphocyte-Associated protein 4, CD152), PD1 (also known as PD-1 ; Programmed Death 1 receptor), PD-L1 , PD-L2, LAG-3 (Lymphocyte Activation Gene- 3), 0X40, A2AR (Adenosine A2A receptor), B7-H3 (CD276), B7-H4 (VTCN1), BTLA (B and T Lymphocyte Attenuator, CD272), IDO (Indoleamine 2,3-dioxygenase), KIR (Killer-cell Immunoglobulin-like Receptor), TIM 3 (T-cell Immunoglobulin domain and Mucin domain 3), VISTA (V-domain Ig suppressor of T cell activation), and IL-2R (interleukin-2 receptor).

[0041 ] Immune checkpoint inhibitors are well known in the art and are commercially or clinically available. These include but are not limited to antibodies that inhibit immune checkpoint proteins. Illustrative examples of checkpoint inhibitors, referenced by their target immune checkpoint protein, are provided as follows. Immune checkpoint inhibitors comprising a CTLA- 4 inhibitor include, but are not limited to, tremelimumab, and ipilimumab (marketed as Yervoy).

[0042] Immune checkpoint inhibitors comprising a PD-1 inhibitor include, but are not limited to, nivolumab (Opdivo), pidilizumab (CureTech), AMP-514 (Medlmmune), pembrolizumab (Keytruda), AUNP 12 (peptide, Aurigene and Pierre), Cemiplimab (Libtayo). Immune checkpoint inhibitors comprising a PD-L1 inhibitor include, but are not limited to, BMS-936559/MDX-1105 (Bristol-Myers Squibb), MPDL3280A (Genentech), MED14736 (Medlmmune), MSB0010718C (EMD Sereno), Atezolizumab (Tecentriq), Avelumab (Bavencio), Durvalumab (Imfinzi).

[0043] Immune checkpoint inhibitors comprising a B7-H3 inhibitor include, but are not limited to, MGA271 (Macrogenics). Immune checkpoint inhibitors comprising an LAG3 inhibitor include, but are not limited to, IMP321 (Immuntep), BMS-986016 (Bristol-Myers Squibb). Immune checkpoint inhibitors comprising a KIR inhibitor include, but are not limited to, I PI-12101 (lirilumab, Bristol-Myers Squibb). Immune checkpoint inhibitors comprising an 0X40 inhibitor include, but are not limited to MEDI-6469 (Medlmmune). An immune checkpoint inhibitor targeting IL-2R, for preferentially depleting Treg cells (e.g., FoxP-3+ CD4+ cells), comprises IL- 2-toxin fusion proteins, which include, but are not limited to, denileukin diftitox (Ontak; Eisai).

[0044] The types of cancer that can be treated using the subject methods of the present invention include but are not limited to adrenal cortical cancer, anal cancer, aplastic anemia, bile duct cancer, bladder cancer, bone cancer, bone metastasis, brain cancers, central nervous system (CNS) cancers, peripheral nervous system (PNS) cancers, breast cancer, cervical cancer, childhood Non-Hodgkin's lymphoma, colon and rectum cancer, endometrial cancer, esophagus cancer, Ewing's family of tumors (e.g. Ewing's sarcoma), eye cancer, gallbladder cancer, gastrointestinal carcinoid tumors, gastrointestinal stromal tumors, gestational trophoblastic disease, hairy cell leukemia, Hodgkin's lymphoma, Kaposi's sarcoma, kidney cancer, laryngeal and hypopharyngeal cancer, acute lymphocytic leukemia, acute myeloid leukemia, children's leukemia, chronic lymphocytic leukemia, chronic myeloid leukemia, liver cancer, lung cancer, lung carcinoid tumors, Non-Hodgkin's lymphoma, male breast cancer, malignant mesothelioma, multiple myeloma, myelodysplastic syndrome, myeloproliferative disorders, nasal cavity and paranasal cancer, nasopharyngeal cancer, neuroblastoma, oral cavity and oropharyngeal cancer, osteosarcoma, ovarian cancer, pancreatic cancer, penile cancer, pituitary tumor, prostate cancer, retinoblastoma, rhabdomyosarcoma, salivary gland cancer, sarcomas, melanoma skin cancer, non-melanoma skin cancers, stomach cancer, testicular cancer, thymus cancer, thyroid cancer, uterine cancer (e.g. uterine sarcoma), transitional cell carcinoma, vaginal cancer, vulvar cancer, mesothelioma, squamous cell or epidermoid carcinoma, bronchial adenoma, choriocarinoma, head and neck cancers, teratocarcinoma, or Waldenstrom's macroglobulinemia.

[0045] Dosage and frequency may vary depending on the half-life of the agent in the patient. It will be understood by one of skill in the art that such guidelines will be adjusted for the molecular weight of the active agent, the clearance from the blood, the mode of administration, and other pharmacokinetic parameters. The dosage may also be varied for localized administration, e.g. intranasal, inhalation, etc., or for systemic administration, e.g. i.m., i.p., i.v., oral, and the like.

[0046] The terms "subject," "individual," and "patient" are used interchangeably herein to refer to a vertebrate, preferably a mammal, more preferably a human. Mammalian species that provide samples for analysis include canines; felines; equines; bovines; ovines; etc. and primates, particularly humans. Animal models, particularly small mammals, e.g. murine, lagomorpha, etc. can be used for experimental investigations. The methods of the invention can be applied for veterinary purposes.

[0047] As used herein, the term "theranosis" refers to the use of results obtained from a diagnostic method to direct the selection of, maintenance of, or changes to a therapeutic regimen, including but not limited to the choice of one or more therapeutic agents, changes in dose level, changes in dose schedule, changes in mode of administration, and changes in formulation. Diagnostic methods used to inform a theranosis can include any that provides information on the state of a disease, condition, or symptom.

[0048] The terms "therapeutic agent", "therapeutic capable agent" or "treatment agent" are used interchangeably and refer to a molecule or compound that confers some beneficial effect upon administration to a subject. The beneficial effect includes enablement of diagnostic determinations; amelioration of a disease, symptom, disorder, or pathological condition; reducing or preventing the onset of a disease, symptom, disorder or condition; and generally counteracting a disease, symptom, disorder or pathological condition.

[0049] Non-ICI cancer therapy may include Abitrexate (Methotrexate Injection), Abraxane (Paclitaxel Injection), Adcetris (Brentuximab Vedotin Injection), Adriamycin (Doxorubicin), Adrucil Injection (5-FU (fluorouracil)), Afinitor (Everolimus) , Afinitor Disperz (Everolimus) , Alimta (PEMET EXED), Alkeran Injection (Melphalan Injection), Alkeran Tablets (Melphalan), Aredia (Pamidronate), Arimidex (Anastrozole), Aromasin (Exemestane), Arranon (Nelarabine), Arzerra (Ofatumumab Injection), Avastin (Bevacizumab), Bexxar (Tositumomab), BiCNU (Carmustine), Blenoxane (Bleomycin), Bosulif (Bosutinib), Busulfex Injection (Busulfan Injection), Campath (Alemtuzumab), Camptosar (Irinotecan), Caprelsa (Vandetanib), Casodex (Bicalutamide), CeeNU (Lomustine), CeeNU Dose Pack (Lomustine), Cerubidine (Daunorubicin), Clolar (Clofarabine Injection), Cometriq (Cabozantinib), Cosmegen (Dactinomycin), CytosarU (Cytarabine), Cytoxan (Cytoxan), Cytoxan Injection (Cyclophosphamide Injection), Dacogen (Decitabine), DaunoXome (Daunorubicin Lipid Complex Injection), Decadron (Dexamethasone), DepoCyt (Cytarabine Lipid Complex Injection), Dexamethasone Intensol (Dexamethasone), Dexpak Taperpak (Dexamethasone), Docefrez (Docetaxel), Doxil (Doxorubicin Lipid Complex Injection), Droxia (Hydroxyurea), DTIC (Decarbazine), Eligard (Leuprolide), Ellence (Ellence (epirubicin)), Eloxatin (Eloxatin (oxaliplatin)), Elspar (Asparaginase), Emcyt (Estramustine), Erbitux (Cetuximab), Erivedge (Vismodegib), Erwinaze (Asparaginase Erwinia chrysanthemi), Ethyol (Amifostine), Etopophos (Etoposide Injection), Eulexin (Flutamide), Fareston (Toremifene), Faslodex (Fulvestrant), Femara (Letrozole), Firmagon (Degarelix Injection), Fludara (Fludarabine), Folex (Methotrexate Injection), Folotyn (Pralatrexate Injection), FUDR (FUDR (floxuridine)), Gemzar (Gemcitabine), Gilotrif (Afatinib), Gleevec (Imatinib Mesylate), Gliadel Wafer (Carmustine wafer), Halaven (Eribulin Injection), Herceptin (Trastuzumab), Hexalen (Altretamine), Hycamtin (Topotecan), Hycamtin (Topotecan), Hydrea (Hydroxyurea), lclusig (Ponatinib), Idamycin PFS (Idarubicin), Ifex (Ifosfamide), Inlyta (Axitinib), Intron A alfab (Interferon alfa-2a), Iressa (Gefitinib), Istodax (Romidepsin Injection), Ixempra (Ixabepilone Injection), Jakafi (Ruxolitinib), Jevtana (Cabazitaxel Injection), Kadcyla (Ado-trastuzumab Emtansine), Kyprolis (Carfilzomib), Leukeran (Chlorambucil), Leukine (Sargramostim), Leustatin (Cladribine), Lupron (Leuprolide), Lupron Depot (Leuprolide), Lupron DepotPED (Leuprolide), Lysodren (Mitotane), Marqibo Kit (Vincristine Lipid Complex Injection), Matulane (Procarbazine), Megace (Megestrol), Mekinist (Trametinib), Mesnex (Mesna), Mesnex (Mesna Injection), Metastron (Strontium-89 Chloride), Mexate (Methotrexate Injection), Mustargen (Mechlorethamine), Mutamycin (Mitomycin), Myleran (Busulfan), Mylotarg (Gemtuzumab Ozogamicin), Navelbine (Vinorelbine), Neosar Injection (Cyclophosphamide Injection), Neulasta (filgrastim), Neulasta (pegfilgrastim), Neupogen (filgrastim), Nexavar (Sorafenib), Nilandron (Nilandron (nilutamide)), Nipent (Pentostatin), Nolvadex (Tamoxifen), Novantrone (Mitoxantrone), Oncaspar (Pegaspargase), Oncovin (Vincristine), Ontak (Denileukin Diftitox), Onxol (Paclitaxel Injection), Panretin (Alitretinoin), Paraplatin (Carboplatin), Perjeta (Pertuzumab Injection), Platinol (Cisplatin), Platinol (Cisplatin Injection), PlatinolAQ (Cisplatin), PlatinolAQ (Cisplatin Injection), Pomalyst (Pomalidomide), Prednisone Intensol (Prednisone), Proleukin (Aldesleukin), Purinethol (Mercaptopurine), Reclast (Zoledronic acid), Revlimid (Lenalidomide), Rheumatrex (Methotrexate), Rituxan (Rituximab), RoferonA alfaa (Interferon alfa-2a), Rubex (Doxorubicin), Sandostatin (Octreotide), Sandostatin LAR Depot (Octreotide), Soltamox (Tamoxifen), Sprycel (Dasatinib), Sterapred (Prednisone), Sterapred DS (Prednisone), Stivarga (Regorafenib), Supprelin LA (Histrelin Implant), Sutent (Sunitinib), Sylatron (Peginterferon Alfa-2b Injection (Sylatron)), Synribo (Omacetaxine Injection), Tabloid (Thioguanine), Taflinar (Dabrafenib), Tarceva (Erlotinib), Targretin Capsules (Bexarotene), Tasigna (Decarbazine), Taxol (Paclitaxel Injection), Taxotere (Docetaxel), Temodar (Temozolomide), Temodar (Temozolomide Injection), Tepadina (Thiotepa), Thalomid (Thalidomide), TheraCys BCG (BCG), Thioplex (Thiotepa), TICE BCG (BCG), Toposar (Etoposide Injection), Torisel (Temsirolimus), Treanda (Bendamustine hydrochloride), Trelstar (Triptorelin Injection), Trexall (Methotrexate), Trisenox (Arsenic trioxide), Tykerb (lapatinib), Valstar (Valrubicin Intravesical), Vantas (Histrelin Implant), Vectibix (Panitumumab), Velban (Vinblastine), Velcade (Bortezomib), Vepesid (Etoposide), Vepesid (Etoposide Injection), Vesanoid (Tretinoin), Vidaza (Azacitidine), Vincasar PFS (Vincristine), Vincrex (Vincristine), Votrient (Pazopanib), Vumon (Teniposide), Wellcovorin IV (Leucovorin Injection), Xalkori (Crizotinib), Xeloda (Capecitabine), Xtandi (Enzalutamide), Yervoy (Ipilimumab Injection), Zaltrap (Ziv-aflibercept Injection), Zanosar (Streptozocin), Zelboraf (Vemurafenib), Zevalin (Ibritumomab Tiuxetan), Zoladex (Goserelin), Zolinza (Vorinostat), Zometa (Zoledronic acid), Zortress (Everolimus), Zytiga (Abiraterone).

[0050] Radiotherapy means the use of radiation, usually X-rays, to treat illness. X-rays were discovered in 1895 and since then radiation has been used in medicine for diagnosis and investigation (X-rays) and treatment (radiotherapy). Radiotherapy may be from outside the body as external radiotherapy, using X-rays, cobalt irradiation, electrons, and more rarely other particles such as protons. It may also be from within the body as internal radiotherapy, which uses radioactive metals or liquids (isotopes) to treat cancer.

[0051] As used herein, "treatment" or "treating," or "palliating" or "ameliorating" are used interchangeably. These terms refer to an approach for obtaining beneficial or desired results including but not limited to a therapeutic benefit and/or a prophylactic benefit. By therapeutic benefit is meant any therapeutically relevant improvement in or effect on one or more diseases, conditions, or symptoms under treatment. For prophylactic benefit, the compositions may be administered to a subject at risk of developing a particular disease, condition, or symptom, or to a subject reporting one or more of the physiological symptoms of a disease, even though the disease, condition, or symptom may not have yet been manifested. [0052] The term "effective amount" or "therapeutically effective amount" refers to the amount of an agent that is sufficient to effect beneficial or desired results. The therapeutically effective amount will vary depending upon the subject and disease condition being treated, the weight and age of the subject, the severity of the disease condition, the manner of administration and the like, which can readily be determined by one of ordinary skill in the art. The term also applies to a dose that will provide an image for detection by any one of the imaging methods described herein. The specific dose will vary depending on the particular agent chosen, the dosing regimen to be followed, whether it is administered in combination with other compounds, timing of administration, the tissue to be imaged, and the physical delivery system in which it is carried.

[0053] "Suitable conditions" shall have a meaning dependent on the context in which this term is used. That is, when used in connection with an antibody, the term shall mean conditions that permit an antibody to bind to its corresponding antigen. When used in connection with contacting an agent to a cell, this term shall mean conditions that permit an agent capable of doing so to enter a cell and perform its intended function. In one embodiment, the term "suitable conditions" as used herein means physiological conditions.

[0054] The term "inflammatory" response is the development of a humoral (antibody mediated) and/or a cellular response, which cellular response may be mediated by antigen-specific T cells or their secretion products), and innate immune cells. An "immunogen" is capable of inducing an immunological response against itself on administration to a mammal or due to autoimmune disease.

[0055] The terms “biomarker,” “biomarkers,” “marker” or “markers” for the purposes of the invention refer to, without limitation, proteins together with their related metabolites, mutations, variants, polymorphisms, modifications, fragments, subunits, degradation products, elements, and other analytes or sample-derived measures. Markers can include expression levels of an intracellular protein or extracellular protein. Markers can also include combinations of any one or more of the foregoing measurements, including temporal trends and differences. Broadly used, a marker can also refer to an immune cell subset.

[0056] To “analyze” includes determining a set of values associated with a sample by measurement of a marker (such as, e.g., presence or absence of a marker or constituent expression levels) in the sample and comparing the measurement against measurement in a sample or set of samples from the same subject or other control subject(s). The markers of the present teachings can be analyzed by any of various conventional methods known in the art. To “analyze” can include performing a statistical analysis, e.g. normalization of data, determination of statistical significance, determination of statistical correlations, clustering algorithms, and the like.

[0057] A “sample” in the context of the present teachings refers to any biological sample that is isolated from a subject, generally a sample comprising circulating immune cells. A sample can include, without limitation, an aliquot of body fluid, whole blood, PBMC (white blood cells or leucocytes), tissue biopsies, synovial fluid, lymphatic fluid, ascites fluid, and interstitial or extracellular fluid. "Blood sample" can refer to whole blood or a fraction thereof, including blood cells, white blood cells or leucocytes. Samples can be obtained from a subject by means including but not limited to venipuncture, biopsy, needle aspirate, lavage, scraping, surgical incision, or intervention or other means known in the art.

[0058] Samples for obtaining circulating cell-free DNA may include any suitable sample, often blood or blood-derived products, such as plasma, serum, etc. Alternative samples may include, for example, urine, ascites, synovial fluid, cerebrospinal fluid, saliva, and the like.

[0059] A “dataset” is a set of numerical values resulting from evaluation of a sample (or population of samples) under a desired condition. The values of the dataset can be obtained, for example, by experimentally obtaining measures from a sample and constructing a dataset from these measurements; or alternatively, by obtaining a dataset from a service provider such as a laboratory, or from a database or a server on which the dataset has been stored. Similarly, the term “obtaining a dataset associated with a sample” encompasses obtaining a set of data determined from at least one sample. Obtaining a dataset encompasses obtaining a sample, and processing the sample to experimentally determine the data, e.g., via measuring antibody binding, or other methods of quantitating a signaling response. The phrase also encompasses receiving a set of data, e.g., from a third party that has processed the sample to experimentally determine the dataset.

[0060] “Measuring” or “measurement” in the context of the present teachings refers to determining the presence, absence, quantity, amount, or effective amount of a substance in a clinical or subject-derived sample, including the presence, absence, or concentration levels of such substances, and/or evaluating the values or categorization of a subject's clinical parameters based on a control, e.g. baseline levels of the marker.

[0061] Classification can be made according to predictive modeling methods that set a threshold for determining the probability that a sample belongs to a given class. The probability preferably is at least 50%, or at least 60% or at least 70% or at least 80% or higher. Classifications also can be made by determining whether a comparison between an obtained dataset and a reference dataset yields a statistically significant difference. If so, then the sample from which the dataset was obtained is classified as not belonging to the reference dataset class. Conversely, if such a comparison is not statistically significantly different from the reference dataset, then the sample from which the dataset was obtained is classified as belonging to the reference dataset class.

[0062] The predictive ability of a model can be evaluated according to its ability to provide a quality metric, e.g. AUC or accuracy, of a particular value, or range of values. In some embodiments, a desired quality threshold is a predictive model that will classify a sample with an accuracy of at least about 0.7, at least about 0.75, at least about 0.8, at least about 0.85, at least about 0.9, at least about 0.95, or higher. As an alternative measure, a desired quality threshold can refer to a predictive model that will classify a sample with an AUC (area under the curve) of at least about 0.7, at least about 0.75, at least about 0.8, at least about 0.85, at least about 0.9, or higher.

[0063] As is known in the art, the relative sensitivity and specificity of a predictive model can be “tuned” to favor either the selectivity metric or the sensitivity metric, where the two metrics have an inverse relationship. The limits in a model as described above can be adjusted to provide a selected sensitivity or specificity level, depending on the particular requirements of the test being performed. One or both of sensitivity and specificity can be at least about at least about 0.7, at least about 0.75, at least about 0.8, at least about 0.85, at least about 0.9, or higher.

[0064] Compositions and methods, including methods of bioinformatic analysis, may include sensitive analysis of circulating tumor DNA (ctDNA), e.g. DNA sequences present in the blood of an individual that are derived from tumor cells, and TMB. These methods may be referred to as CAncer Personalized Profiling by Deep Sequencing (CAPP-Seq). See, for example, WO 2014/151117 and WO 2016/040901 , each herein specifically incorporated by reference. CAPP- Seq may comprise (a) obtaining sequence information of a cell-free DNA (cfDNA) sample derived from a subject; and (b) using sequence information derived from (a) to detect cell-free tumor DNA (ctDNA) in the sample, wherein the method is capable of detecting a percentage of ctDNA that is less than 2% of total cfDNA. CAPP-Seq may accurately quantify cell-free tumor DNA from early and advanced stage tumors, and can further be used to sequence and determine TMB.

[0065] The cell-free nucleic acid may be cell-free DNA (cfDNA). The cell-free nucleic acid may be cell-free RNA (cfRNA). The cell-free nucleic acids may be a mixture of cell-free DNA (cfDNA) and cell-free RNA (cfRNA). The tumor nucleic acid may be a nucleic acid originating from a tumor cell. The tumor nucleic acid may be tumor-derived DNA (tDNA). The tumor nucleic acid may be a circulating tumor DNA (ctDNA). The tumor nucleic acid may be tumor-derived RNA (tRNA). The tumor nucleic acid may be a circulating tumor RNA (ctRNA). The tumor nucleic acids may be a mixture of tumor-derived DNA and tumor-derived RNA. The tumor nucleic acids may be a mixture of ctDNA and ctRNA. [0066] For immune cell profiling, methods of interest include, without limitation, CIBERsort (see WO2016118860A1and Newman et al. (2015) Nat Methods 12:453-457, herein specifically incorporated by reference), flow cytometry, mass cyometry, and the like. An “affinity reagent”, or “specific binding member” may be used to refer to an affinity reagent, such as an antibody, ligand, etc. that selectively binds to a protein or marker of the invention, e.g. binding to CD8 for T cell profiling. The term "affinity reagent" includes any molecule, e.g., peptide, nucleic acid, small organic molecule, e.g. an antibody. For some purposes, an affinity reagent selectively binds to a cell surface marker. For other purposes an affinity reagent selectively binds to a cellular signaling protein, particularly one which is capable of detecting an activation state of a signaling protein over another activation state of the signaling protein.

[0067] The term "antibody" includes full length antibodies and antibody fragments, and can refer to a natural antibody from any organism, an engineered antibody, or an antibody generated recombinantly for experimental, therapeutic, or other purposes as further defined below. Examples of antibody fragments, as are known in the art, such as Fab, Fab', F(ab')2, Fv, scFv, or other antigen-binding subsequences of antibodies, either produced by the modification of whole antibodies or those synthesized de novo using recombinant DNA technologies. The term "antibody" comprises monoclonal and polyclonal antibodies. Antibodies can be antagonists, agonists, neutralizing, inhibitory, or stimulatory. They can be humanized, glycosylated, bound to solid supports, and possess other variations.

[0068] The methods the invention may utilize affinity reagents comprising a label, labeling element, or tag. By label or labeling element is meant a molecule that can be directly (i.e., a primary label) or indirectly (i.e., a secondary label) detected; for example a label can be visualized and/or measured or otherwise identified so that its presence or absence can be known. Labels include optical labels such as fluorescent dyes or moieties. Fluorophores can be either "small molecule" fluors, or proteinaceous fluors (e.g. green fluorescent proteins and all variants thereof). In some embodiments, activation state-specific antibodies are labeled with quantum dots as disclosed by Chattopadhyay et al. (2006) Nat. Med. 12, 972-977. Quantum dot labeled antibodies can be used alone or they can be employed in conjunction with organic fluorochrome — conjugated antibodies to increase the total number of labels available. As the number of labeled antibodies increase so does the ability for subtyping known cell populations.

[0069] The detecting, sorting, or isolating step of the methods of the present invention can entail fluorescence-activated cell sorting (FACS) techniques or flow cytometry, mass cytometry, etc., where FACS is used to select cells from the population containing a particular surface marker, or the selection step can entail the use of magnetically responsive particles as retrievable supports for target cell capture and/or background removal. A variety of FACS systems are known in the art and can be used in the methods of the invention (see e.g., W099/54494, filed Apr. 16, 1999; U.S. Ser. No. 20010006787, filed Jul. 5, 2001 , each expressly incorporated herein by reference).

[0070] Mass cytometry, or CyTOF (DVS Sciences), is a variation of flow cytometry in which antibodies are labeled with heavy metal ion tags rather than fluorochromes. Readout is by time- of-flight mass spectrometry. This allows for the combination of many more antibody specificities in a single samples, without significant spillover between channels. For example, see Bodenmiller at a. (2012) Nature Biotechnology 30:858-867.

[0071 ] Affinity reagents such as antibodies also find use in, for example, immunohistochemistry to determine expression of an immune checkpoint protein, such as CD274 (PD-L1), B7-1 , B7- 2, 4-1 BB-L, GITRL, etc. Alternatively, expression can be determined by any convenient method known in the art, e.g. mRNA hybridization, flow cytometry, mass cytometry, etc. A sample for analysis may include, for example, a tumor biopsy sample, such as a needle biopsy sample.

[0072] The present invention incorporates information disclosed in other applications and texts. The following patent and other publications are hereby incorporated by reference in their entireties: Alberts et al., The Molecular Biology of the Cell, 4th Ed., Garland Science, 2002; Vogelstein and Kinzler, The Genetic Basis of Human Cancer, 2d Ed., McGraw Hill, 2002; Michael, Biochemical Pathways, John Wiley and Sons, 1999; Weinberg, The Biology of Cancer, 2007; Immunobiology, Janeway et al. 7th Ed., Garland, and Leroith and Bondy, Growth Factors and Cytokines in Health and Disease, A Multi Volume Treatise, Volumes 1A and IB, Growth Factors, 1996.

[0073] Unless otherwise apparent from the context, all elements, steps or features of the invention can be used in any combination with other elements, steps or features.

[0074] General methods in molecular and cellular biochemistry can be found in such standard textbooks as Molecular Cloning: A Laboratory Manual, 3rd Ed. (Sambrook et al., Harbor Laboratory Press 2001); Short Protocols in Molecular Biology, 4th Ed. (Ausubel et al. eds., John Wiley & Sons 1999); Protein Methods (Bollag et al., John Wiley & Sons 1996); Nonviral Vectors for Gene Therapy (Wagner et al. eds., Academic Press 1999); Viral Vectors (Kaplift & Loewy eds., Academic Press 1995); Immunology Methods Manual (I. Lefkovits ed., Academic Press 1997); and Cell and Tissue Culture: Laboratory Procedures in Biotechnology (Doyle & Griffiths, John Wiley & Sons 1998). Reagents, cloning vectors, and kits for genetic manipulation referred to in this disclosure are available from commercial vendors such as BioRad, Stratagene, Invitrogen, Sigma-Aldrich, and ClonTech.

[0075] The invention has been described in terms of particular embodiments found or proposed by the present inventor to comprise preferred modes for the practice of the invention. It will be appreciated by those of skill in the art that, in light of the present disclosure, numerous modifications and changes can be made in the particular embodiments exemplified without departing from the intended scope of the invention. Due to biological functional equivalency considerations, changes can be made in protein structure without affecting the biological action in kind or amount. All such modifications are intended to be included within the scope of the appended claims.

[0076] The subject methods are used for prognostic, diagnostic and therapeutic purposes. As used herein, the term "treating" is used to refer to both prevention of relapses, and treatment of pre-existing conditions. The treatment of ongoing cancer to achieve durable clinical benefit is of particular interest.

[0077] The term "tumor mutation burden" (TMB) as used herein refers to the number of somatic mutations in a tumor's genome and/or the number of somatic mutations per area of the tumor's genome. Germline (inherited) variants are excluded when determining TMB, because the immune system has a higher likelihood of recognizing these as self. Tumor mutation burden (TMB) can also be used interchangeably with "tumor mutation load," "tumor mutational burden," or "tumor mutational load."

[0078] TMB is a genetic analysis of a tumor's genome and, thus, can be measured by applying sequencing methods well known to those of skill in the art. The tumor DNA can be compared with DNA from patient-matched normal tissue to eliminate germline mutations or polymorphisms.

[0079] In some embodiments, TMB is determined by sequencing ctDNA using a high- throughput sequence technique, e.g., next-generation sequencing (NGS) or an NGS-based method. In some embodiments, the NGS-based method is selected from whole genome sequencing (WGS), whole tumor exome sequencing (WES) like CAPP-seq or comprehensive genomic profiling (CGP) of cancer gene panels such as FOUNDATIONONE CDX.TM. and MSK-IMPACT clinical tests. In some embodiments, TMB, as used herein, refers to the number of somatic mutations per megabase (Mb) of DNA sequenced. In one embodiment, TMB is measured using the total number of nonsynonymous mutations, e.g., missense mutation (i.e. changing a particular amino acid in the protein) and/or nonsense (causing premature termination and thus truncation of the protein sequence), identified by normalizing matched tumor with germline samples to exclude any inherited germline genetic alterations. In another embodiment, TMB is measured using the total number of missense mutations in a tumor. In order to measure TMB, a sufficient amount of sample is required. In one embodiment, tissue sample (for example, a minimum of 10 slides) is used for evaluation. In some embodiments, TMB is expressed as mutations per megabase. 1 megabase represents 1 million bases.

[0080] In some embodiments, tumor mutational burden may be measured using CAPP-seq. When CAPP-seq is used for determining TMB in an individual, TMB may be measured from cell-free DNA (cfDNA)(i.e. plasma cell-free DNA derived from blood). When TMB is measured from cfDNA, TMB may be referred to as blood TMB (bTMB). bTMB may be calculated using the equation:

[0081] The TMB status can be a numerical value or a relative value, e.g., high, medium, or low; within the highest fractile, or within the top tertile, of a reference set.

[0082] In other embodiments, a "high TMB" refers to a TMB within the highest fractile of the reference TMB value. For example, all subject's with evaluable TMB data are grouped according to fractile distribution of TMB, i.e., subjects are rank ordered from highest to lowest number of genetic alterations and divided into a defined number of groups. In one embodiment, all subjects with evaluable TMB data are ranked ordered and divided into thirds, and a "high TMB" is within the top tertile of the reference TMB value. It should be understood that, once rank ordered, subjects with evaluable TMB data can be divided into any number of groups, e.g., quartiles, quintiles, etc.

[0083] In some embodiments, a "high TMB" refers to a TMB of at least about 14 mutations per megabase, at least about 20 mutations per megabase, at least about 25 mutations per megabase at least about 30 mutations per megabase, at least about 35 mutations per megabase, at least about 40 mutations per megabase, at least about 45 mutations per megabase, at least about 50 mutations per megabase, at least about 55 mutations per megabase, at least about 60 mutations per megabase, at least about 65 mutations per megabase, at least about 70 mutations per megabase, at least about 75 mutations per megabase, at least about 80 mutations per megabase, at least about 85 mutations per megabase, at least about 90 mutations per megabase, at least about 95 mutations per megabase, or at least about 100 mutations per megabase. In some embodiments, a "high TMB" refers to a TMB of at least about 105 mutations per megabase, at least about 110 mutations per megabase, at least about 115 mutations per megabase, at least about 120 mutations per megabase, at least about 125 mutations per megabase, at least about 130 mutations per megabase, at least about 135 mutations per megabase, at least about 140 mutations per megabase, at least about 145 mutations per megabase, at least about 150 mutations per megabase, at least about 175 mutations per megabase, or at least about 200 mutations per megabase.

Methods of the Invention

[0084] Compositions and methods are provided for determining whether an individual with cancer will have a durable clinical benefit from treatment with an immune checkpoint inhibitor. Provided is an integrated analytic method, where a Bayesian framework is used to derive a single biomarker from circulating tumor DNA profiling, e.g. normalized bTMB; analysis of circulating immune cells, e.g. levels of CD8+ T cells, and in some embodiments expression of immune checkpoint protein, e.g. PD-L1 , to generate a prognostic for patient responsiveness to immune checkpoint inhibition (ICI). In some embodiments that use only noninvasive blood draws, the methods robustly identify which patients will achieve durable clinical benefit from immune checkpoint inhibition. In an embodiment, the methods further comprise selecting a treatment regimen for the individual based on the analysis. In some embodiments, the prediction is based on analysis of a pre-treatment sample. In some embodiments, the prediction is based on samples shortly after a first ICI treatment.

[0085] A sample for circulating tumor DNA profiling and circulating immune cells can be any suitable type that allows for the analysis of one or more cells, preferably a blood sample. Samples can be obtained once or multiple times from an individual. Multiple samples can be obtained from different locations in the individual (e.g., blood samples, bone marrow samples and/or lymph node samples), at different times from the individual, or any combination thereof. In some embodiments a sample is obtained prior to ICI treatment. In some embodiments a sample is obtain following a first ICI treatment, and within about 4 weeks, 3 weeks, 2 weeks, 1 week, of a first ICI treatment. In some embodiments a sample is obtained both prior to and following ICI treatment.

[0086] One or more cells or cell types, or samples containing one or more cells or cell types, and circulating DNA can be isolated from body samples. The cells can be separated from body samples by red cell lysis, centrifugation, elutriation, density gradient separation, apheresis, affinity selection, panning, FACS, centrifugation with Hypaque, solid supports (magnetic beads, beads in columns, or other surfaces) with attached antibodies, etc. The samples are analyzed as described above for the specific metric of interest.

[0087] Fully robotic or microfluidic systems include automated liquid-, particle-, cell- and organism-handling including high throughput pipetting to perform all steps of screening applications. This includes liquid, particle, cell, and organism manipulations such as aspiration, dispensing, mixing, diluting, washing, accurate volumetric transfers; retrieving, and discarding of pipet tips; and repetitive pipetting of identical volumes for multiple deliveries from a single sample aspiration. These manipulations are cross-contamination- free liquid, particle, cell, and organism transfers. This instrument performs automated replication of microplate samples to filters, membranes, and/or daughter plates, high-density transfers, full-plate serial dilutions, and high capacity operation.

[0088] In some embodiments, platforms for multi-well plates, multi-tubes, holders, cartridges, minitubes, deep-well plates, microfuge tubes, cryovials, square well plates, filters, chips, optic fibers, beads, and other solid-phase matrices or platform with various volumes are accommodated on an upgradable modular platform for additional capacity. This modular platform includes a variable speed orbital shaker, and multi-position work decks for source samples, sample and reagent dilution, assay plates, sample and reagent reservoirs, pipette tips, and an active wash station. In some embodiments, the methods of the invention include the use of a plate reader.

[0089] In some embodiments, interchangeable pipet heads (single or multi-channel) with single or multiple magnetic probes, affinity probes, or pipetters robotically manipulate the liquid, particles, cells, and organisms. Multi-well or multi-tube magnetic separators or platforms manipulate liquid, particles, cells, and organisms in single or multiple sample formats.

[0090] In some embodiments, the instrumentation will include a detector, which can be a wide variety of different detectors, depending on the labels and assay. In some embodiments, useful detectors include a microscope(s) with multiple channels of fluorescence; plate readers to provide fluorescent, ultraviolet and visible spectrophotometric detection with single and dual wavelength endpoint and kinetics capability, fluorescence resonance energy transfer (FRET), luminescence, quenching, two-photon excitation, and intensity redistribution; CCD cameras to capture and transform data and images into quantifiable formats; and a computer workstation.

[0091] In some embodiments, the robotic apparatus includes a central processing unit which communicates with a memory and a set of input/output devices (e.g., keyboard, mouse, monitor, printer, etc.) through a bus. Again, as outlined below, this can be in addition to or in place of the CPU for the multiplexing devices of the invention. The general interaction between a central processing unit, a memory, input/output devices, and a bus is known in the art. Thus, a variety of different procedures, depending on the experiments to be run, are stored in the CPU memory.

Models

[0092] DIREct-Pre and DIREct-On are multi-parametric Bayesian biomarker models to determine whether or not an individual would receive DCB from ICI treatment. In some embodiments, the DIREct-Pre model incorporates normalized bTMB, immune checkpoint protein, e.g. PD-L1 , expression and the abundance of circulating CD8⁺ T cells. In other embodiments, DIREct-Pre may exclude any single factor from the model. For instance, the DIREct-Pre model may incorporate normalized bTMB and the abundance of CD8 T cells but leave out PD-L1 expresssion. In some embodiments. In some embodiments, the DIREct-On model may incorporate normalized bTMB, ActDNA following treatment and the abundance of circulating CD8⁺ T cells. In other embodiments, DIREct-On may exclude any single factor from the model. For instance, the DIREct-On model may incorporate ActDNA following treatment and the abundance of CD8 T cells but leave out normalized bTMB.

[0093] To classify DCB versus NDB a Bayesian probit approach was utilized. Here, likelihood of DCB for each sample was modeled as probability_Dc_B = F{c^tb), where F(. ) denotes the cumulative distribution function of a standard normal variable, x denotes the feature vector (i.e., normalized bTMB, PD-L1 , and CD8 fraction for DIREct-Pre and normalized bTMB, CD8 fraction, and ActDNAfor DIREct-On), and b denotes the model parameters from prior knowledge derived from the literature or from our discovery cohort (as described below). The model parameters were assumed to be governed by a Gaussian distribution, i.e. b ~ N(b₀, A₀ ^_1), where b₀ and L₀ ¹ denote the mean and covariance matrix of the distribution, respectively. This approach allows for uncertainty in the underlying distributions (as might be expected for example between different ctDNA profiling techniques or patient populations) and therefore expected to be more generalizable for response prediction.

[0094] To identify the parameters of the Gaussian distribution governing the probit coefficients, the features from Gandara et al. (Gandara et al., 2018) were denoted by X_prwr and the corresponding labels (DCB/NDB) by Y _prior- Here, n=100 cases were subsampled for m = 2000 times. For each iteration and for each set of covariates (e.g., PDL1 and normalized bTMB), a probit regression model was solved to generate a matrix of associated coefficients (/?). This matrix was used as a random realization of this multi variable normal distribution, and then estimated the Bayes prior probability parameters (i.e., hyper parameters).

[0095] A LOOCV framework was used to fully exploit the available data, and for the features not estimated by X_wi0r (e.g., CD8 T cell fraction). Specifically, n - 1 samples were used to find the hyperparameters as described above. The final full prior probability distributions were derived by concatenating two components: (1) the prior knowledge-based priors, and (2) the priors derived from our Discovery Cohort, such that

and b_{0, ourcohort} are the mean vectors of the variables, estimated from the prior set and our cohort, respectively. Similarly Ao_,wi0r ^_1 and L_{0 ,0lirc0¾0r£} ^_1 denote the covariance matrices. The same n - 1 samples were used with all available covariates in a MCMC step to calculate the posterior probabilities (i.e., updated prior probabilities). In cases where tumor PD-L1 expression was unavailable (Discovery Cohort, n = 11 ; Validation Cohort, n = 4), this parameter was left out.

[0096] To generate a single score representing the probability of DCB, we used the Markov- Chain Monte Carlo (MCMC) posterior samples, and generated 10,000 pseudo-samples with 1000 initial burn-in. These realizations of the b vectors (i.e. {/?i,/?₂,...,/?ioooo} along with the sample feature vector, x) were then used to calculate 10,000 realizations of the DCB score, such that F(c^tbi) for i e {1 ,...,10000}. We then used the median of these realizations as the final prediction score.

[0097] To select the optimal thresholds in the discovery cohort for the DIREct-Pre and DIREct- On scores, the optimal ROC corner point approach (Youden’s J) was used for classification of DCB, where the optimality was defined as the point with maximal unweighted accuracy, i.e. specificity plus sensitivity (Youden, 1950). These cut-points were then applied to the validation set for calculation of precision and survival analyses. [0098] We evaluated DIREct predictions for quantitative accuracy via model calibration regression (Steyerberg et al., 2010). To test calibration of the DIREct-Pre and DIREct-On models a bootstrap resampling with n=2000 was preformed and reported the “true event rate”, estimated via the Kaplan-Meier maximum likelihood estimates of observed events versus expected events from our models’ predictions. A perfect calibration leads to a slope of 1 , i.e. the predicted probability of the model is in fact representative of the true event rate in the population.

[0099] The model is used to generate a single biomarker value that is used to predict the probability of an individual developing a durable clinical benefit from ICI and to facilitate personalized selection of treatment, including ICI if appropriate, for patients with a number of different cancers. An individual with a high score that is predicted to benefit from ICI, can be selected, and treated, with an ICI, usually in combination with one or more non-ICI therapeutic modalities. An individual with a low score that is not predicted to benefit from ICI can be selected, and treated, with non-ICI therapy, e.g. chemotherapy, non-ICI immunotherapy, radiation therapy, and the like. The non-ICI regimen may be intensified if appropriate.

[00100] In some embodiments, the invention provides kits for the classification, diagnosis, prognosis, theranosis, and/or prediction of an outcome. The kit may further comprise a software package for data analysis of the cellular state and its physiological status, which may include reference profiles for comparison with the test profile and comparisons to other analyses as referred to above. The kit may also include instructions for use for any of the above applications.

[00101] Kits provided by the invention may comprise one or more of the affinity reagents described herein, reagents for isolation and sequencing analysis of ctDNA, etc. A kit may also include other reagents that are useful in the invention, such as modulators, fixatives, containers, plates, buffers, therapeutic agents, instructions, and the like.

[00102] Kits provided by the invention can comprise one or more labeling elements. Non-limiting examples of labeling elements include small molecule fluorophores, proteinaceous fluorophores, radioisotopes, enzymes, antibodies, chemiluminescent molecules, biotin, streptavidin, digoxigenin, chromogenic dyes, luminescent dyes, phosphorous dyes, luciferase, magnetic particles, beta-galactosidase, amino groups, carboxy groups, maleimide groups, oxo groups and thiol groups, quantum dots , chelated or caged lanthanides, isotope tags, radiodense tags, electron- dense tags, radioactive isotopes, paramagnetic particles, agarose particles, mass tags, e-tags, nanoparticles, and vesicle tags.

[00103] In some embodiments, the kits of the invention enable the detection of signaling proteins by sensitive cellular assay methods, such as IHC and flow cytometry, which are suitable for the clinical detection, classification, diagnosis, prognosis, theranosis, and outcome prediction. [00104] Such kits may additionally comprise one or more therapeutic agents. The kit may further comprise a software package for data analysis of the physiological status, which may include reference profiles for comparison with the test profile.

[00105] Such kits may also include information, such as scientific literature references, package insert materials, clinical trial results, and/or summaries of these and the like, which indicate or establish the activities and/or advantages of the composition, and/or which describe dosing, administration, side effects, drug interactions, or other information useful to the health care provider. Such information may be based on the results of various studies, for example, studies using experimental animals involving in vivo models and studies based on human clinical trials. Kits described herein can be provided, marketed and/or promoted to health providers, including physicians, nurses, pharmacists, formulary officials, and the like. Kits may also, in some embodiments, be marketed directly to the consumer.

Reports

[00106] In some embodiments, providing an evaluation of a subject for a classification, diagnosis, prognosis, theranosis, and/or prediction of an outcome includes generating a written report that includes the artisan’s assessment of the subject’s state of health i.e. a “diagnosis assessment”, of the subject’s prognosis, i.e. a “prognosis assessment”, and/or of possible treatment regimens, i.e. a “treatment assessment”. Thus, a subject method may further include a step of generating or outputting a report providing the results of a diagnosis assessment, a prognosis assessment, or treatment assessment, which report can be provided in the form of an electronic medium (e.g., an electronic display on a computer monitor), or in the form of a tangible medium (e.g., a report printed on paper or other tangible medium).

[00107] A “report,” as described herein, is an electronic or tangible document which includes report elements that provide information of interest relating to a diagnosis assessment, a prognosis assessment, and/or a treatment assessment and its results. A subject report can be completely or partially electronically generated. A subject report includes at least a diagnosis assessment, i.e. a diagnosis as to whether a subject will have a particular clinical response, and/or a suggested course of treatment to be followed. A subject report can further include one or more of: 1) information regarding the testing facility; 2) service provider information; 3) subject data; 4) sample data; 5) an assessment report, which can include various information including: a) test data, where test data can include an analysis of cellular signaling responses to activation, b) reference values employed, if any.

[00108] The report may include information about the testing facility, which information is relevant to the hospital, clinic, or laboratory in which sample gathering and/or data generation was conducted. This information can include one or more details relating to, for example, the name and location of the testing facility, the identity of the lab technician who conducted the assay and/or who entered the input data, the date and time the assay was conducted and/or analyzed, the location where the sample and/or result data is stored, the lot number of the reagents (e.g., kit, etc.) used in the assay, and the like. Report fields with this information can generally be populated using information provided by the user.

[00109] The report may include information about the service provider, which may be located outside the healthcare facility at which the user is located, or within the healthcare facility. Examples of such information can include the name and location of the service provider, the name of the reviewer, and where necessary or desired the name of the individual who conducted sample gathering and/or data generation. Report fields with this information can generally be populated using data entered by the user, which can be selected from among pre-scripted selections (e.g., using a drop-down menu). Other service provider information in the report can include contact information for technical information about the result and/or about the interpretive report.

[00110] The report may include a subject data section, including subject medical history as well as administrative subject data (that is, data that are not essential to the diagnosis, prognosis, or treatment assessment) such as information to identify the subject (e.g., name, subject date of birth (DOB), gender, mailing and/or residence address, medical record number (MRN), room and/or bed number in a healthcare facility), insurance information, and the like), the name of the subject's physician or other health professional who ordered the susceptibility prediction and, if different from the ordering physician, the name of a staff physician who is responsible for the subject's care (e.g., primary care physician).

[00111] The report may include a sample data section, which may provide information about the biological sample analyzed, such as the source of biological sample obtained from the subject (e.g. blood, type of tissue, etc.), how the sample was handled (e.g. storage temperature, preparatory protocols) and the date and time collected. Report fields with this information can generally be populated using data entered by the user, some of which may be provided as prescripted selections (e.g., using a drop-down menu).

[00112] The report may include an assessment report section, which may include information generated after processing of the data as described herein. The interpretive report can include a prognosis of the likelihood that the patient will develop tumor benefit from immune checkpoint inhibitors. The interpretive report can include, for example, results of the analysis, methods used to calculate the analysis, and interpretation, i.e. prognosis. The assessment portion of the report can optionally also include a Recommendation(s). For example, where the results indicate the subject’s prognosis for propensity to develop tumor benefit from immune checkpoint inhibitors.

[00113] It will also be readily appreciated that the reports can include additional elements or modified elements. For example, where electronic, the report can contain hyperlinks which point to internal or external databases which provide more detailed information about selected elements of the report. For example, the patient data element of the report can include a hyperlink to an electronic patient record, or a site for accessing such a patient record, which patient record is maintained in a confidential database. This latter embodiment may be of interest in an in-hospital system or in-clinic setting. When in electronic format, the report is recorded on a suitable physical medium, such as a computer readable medium, e.g., in a computer memory, zip drive, CD, DVD, etc.

[00114] It will be readily appreciated that the report can include all or some of the elements above, with the proviso that the report generally includes at least the elements sufficient to provide the analysis requested by the user (e.g., a diagnosis, a prognosis, or a prediction of responsiveness to a therapy).

Computer aspects

[00115] A computational system (e.g., a computer) may be used in the methods of the present disclosure to integrate and to analyze data generated from immune profiling, ctDNA analysis, PD-L1 expression, and TMB analysis. A computational unit may include any suitable components to analyze the measured images. Thus, the computational unit may include one or more of the following: a processor; a non-transient, computer-readable memory, such as a computer-readable medium; an input device, such as a keyboard, mouse, touchscreen, etc.; an output device, such as a monitor, screen, speaker, etc.; a network interface, such as a wired or wireless network interface; and the like.

[00116] The raw data from measurements, such as PD-L1 expression, immune profiling, ctDNA abundance and the like, can be analyzed and stored on a computer-based system. As used herein, “a computer-based system” refers to the hardware means, software means, and data storage means used to analyze the information of the present invention. The minimum hardware of the computer-based systems of the present invention comprises a central processing unit (CPU), input means, output means, and data storage means. A skilled artisan can readily appreciate that any one of the currently available computer-based system are suitable for use in the present invention. The data storage means may comprise any manufacture comprising a recording of the present information as described above, or a memory access means that can access such a manufacture.

[00117] The analysis may be implemented in hardware or software, or a combination of both. In one embodiment of the invention, a machine-readable storage medium is provided, the medium comprising a data storage material encoded with machine readable data which, when using a machine programmed with instructions for using said data, is capable of displaying a any of the datasets and data comparisons of this invention. Such data may be used for a variety of purposes, such as diagnosis, disease treatment and the like. In some embodiments, the invention is implemented in computer programs executing on programmable computers, comprising a processor, a data storage system (including volatile and non-volatile memory and/or storage elements), at least one input device, and at least one output device. Program code is applied to input data to perform the functions described above and generate output information. The output information is applied to one or more output devices, in known fashion. The computer may be, for example, a personal computer, microcomputer, or workstation of conventional design.

[00118] Each program is preferably implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the programs can be implemented in assembly or machine language, if desired. In any case, the language can be a compiled or interpreted language. Each such computer program is preferably stored on a storage media or device (e.g., ROM or magnetic diskette) readable by a general or special purpose programmable computer, for configuring and operating the computer when the storage media or device is read by the computer to perform the procedures described herein. The system can also be considered to be implemented as a computer-readable storage medium, configured with a computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner to perform the functions described herein.

[00119] A variety of structural formats for the input and output means can be used to input and output the information in the computer-based systems of the present invention. One format for an output means test datasets possessing varying degrees of similarity to a trusted profile. Such presentation provides a skilled artisan with a ranking of similarities and identifies the degree of similarity contained in the test pattern.

[00120] The data and analysis thereof can be provided in a variety of media to facilitate their use. “Media” refers to a manufacture that contains the signature pattern information of the present invention. The databases of the present invention can be recorded on computer readable media, e.g. any medium that can be read and accessed directly by a computer. Such media include, but are not limited to: magnetic storage media, such as floppy discs, hard disc storage medium, and magnetic tape; optical storage media such as CD-ROM; electrical storage media such as RAM and ROM; and hybrids of these categories such as magnetic/optical storage media. One of skill in the art can readily appreciate how any of the presently known computer readable mediums can be used to create a manufacture comprising a recording of the present database information. "Recorded" refers to a process for storing information on computer readable medium, using any such methods as known in the art. Any convenient data storage structure can be chosen, based on the means used to access the stored information. A variety of data processor programs and formats can be used for storage, e.g. word processing text file, database format, etc. [00121] A variety of structural formats for the input and output means can be used to input and output the information in the computer-based systems. Such presentation provides a skilled artisan with a ranking of similarities and identifies the degree of similarity contained in the test data.

[00122] Further provided herein is a method of storing and/or transmitting, via computer, sequence, and other, data collected by the methods disclosed herein. Any computer or computer accessory including, but not limited to software and storage devices, can be utilized to practice the present invention. Sequence or other data (e.g., immune repertoire analysis results), can be input into a computer by a user either directly or indirectly. Additionally, any of the devices which can be used to sequence DNA or analyze DNA or analyze immune repertoire data can be linked to a computer, such that the data is transferred to a computer and/or computer-compatible storage device. Data can be stored on a computer or suitable storage device (e.g., CD). Data can also be sent from a computer to another computer or data collection point via methods well known in the art (e.g., the internet, ground mail, air mail). Thus, data collected by the methods described herein can be collected at any point or geographical location and sent to any other geographical location.

EXPERIMENTAL

[00123] The following examples are given for the purpose of illustrating various embodiments of the invention and are not meant to limit the present invention in any fashion. The present examples, along with the methods described herein are presently representative of preferred embodiments, are exemplary, and are not intended as limitations on the scope of the invention. Changes therein and other uses which are encompassed within the spirit of the invention as defined by the scope of the claims will occur to those skilled in the art.

Example 1 :

[00124] To assess how accurately the first on-treatment imaging study identifies which patients ultimately achieve long-lasting clinical benefit from ICI, we analyzed response data from 237 NSCLC patients receiving ICI. Strikingly, we found that among patients achieving a partial response (PR, n = 31) or stable disease (SD, n = 100) at the first scan, nearly half (41%; n=54) did not ultimately achieve durable clinical benefit (DCB), defined as progression-free survival (PFS) of at least 6 months (Fig. 1A). Further, 39% of patients (32/82) who achieved SD as best overall response ultimately experienced DCB, suggesting objective responses assessed by imaging are not fully capturing patients that are benefitting. Therefore, there is considerable unmet need for early response assessment methods that more accurately identify who will achieve DCB or NDB. [00125] One such approach is comparison of circulating tumor DNA (ctDNA) levels before and after the start of treatment to assess treatment response. Indeed, several recent publications have demonstrated that ctDNA changes 4-8 weeks after therapy initiation can classify response to ICI in NSCLC with modest accuracy (AUC = 0.65-0.75) (Anagnostou et al., 2019; Goldberg et al., 2018; Raja et al., 2018). Each of these studies used a different ctDNA change threshold and timepoint to assess molecular responses, leaving questions about the optimal approach. Furthermore, even earlier classification of ultimate clinical benefit is desirable in order to optimally facilitate personalized medicine approaches.

[00126] We therefore set out to develop a non-invasive approach for early identification of which advanced NSCLC patients will achieve DCB upon treatment with ICI. We hypothesized that both tumor and immune factors can contribute to ICI response prediction and can be measured noninvasively. Specifically, we integrated ctDNA profiling, which measures tumor properties, and analysis of circulating immune cells, which reflect the immune milieu, using pre- and early on-treatment blood samples (Fig. 1 B). To combine these factors into a single biomarker, we applied a Bayesian framework for integrating diverse risk predictors that we recently demonstrated has significant advantages compare to Cox models for time- dependent biomarker development (Kurtz et al., 2019). Using a discovery and validation approach we demonstrate that integration of tumor-intrinsic and -extrinsic features assessable from noninvasive blood draws can robustly identify which NSCLC patients will achieve durable clinical benefit from ICI.

Results

[00127] Characteristics of advanced NSCLC patients receiving ICI. T o begin, we assembled a cohort of 99 advanced NSCLC patients receiving ICI. We characterized these patients for tumor cell PD-L1 expression using available biopsy specimens and also profiled their blood specimens pre-therapy and on-therapy ctDNA by CAPP-Seq genotyping of plasma (Fig. 1 B, 8, Tables 1) (Newman et al., 2014, 2016). Tumors in our cohort displayed an expected distribution of NSCLC driver mutations (Fig.2). We separately profiled available blood specimens for leukocyte immune composition using CIBERSORT (Newman et al., 2015, 2019), flow cytometry, or both (Fig. 1B). We then employed a discovery and validation approach to identify which individual and combined features best determine likelihood of response to ICI (Fig. 1 B, 8).

[00128] Pre-treatment ctDNA features are associated with durable clinical benefit from ICI.

Responses induced by ICI and their durability are known to be influenced both by the burden of mutations as neoantigenic substrates for immune recognition as captured by TMB (Mandal et al., 2019; Rizvi et al., 2015; Snyder et al., 2014), and by the total anatomic disease burden (Ito et al., 2019; Kaira et al., 2017). Several groups including ours have demonstrated that tissue-based TMB can be noninvasively estimated from ctDNA using sequencing panels of various sizes, as might be useful for predicting response to ICI (Chaudhuri et al., 2017; Gandara et al., 2018; Wang et al., 2019). To explore this further, we analyzed data from a recently published study of 853 advanced NSCLC patients treated with either the PD-L1 inhibitor atezolizumab or chemotherapy, and examined whether ctDNA metrics and PD-L1 expression could accurately classify DCB versus NDB (Fig.8) (Gandara et al., 2018). While high bTMB (>14 mutations per megabase) was significantly associated with durable benefit from ICI, we found the strength of this association to be modest (Fig. 3A, 3D). Separately, we reasoned that pretreatment ctDNA concentration may be associated with treatment benefit since it is a surrogate for total body disease burden (Chaudhuri et al., 2017; Ito et al., 2019). Indeed, patients ultimately achieving DCB had significantly lower ctDNA levels prior to ICI therapy (Fig. 3B, 3D). Furthermore, since these two blood- based measurements had reciprocal associations with DCB, we used their ratio to integrate them as ctDNA-normalized bTMB (normalized bTMB). Patients achieving DCB from ICI had significantly higher normalized bTMB (Fig. 3C-D). Accordingly, ICI-treated patients with higher normalized bTMB had significantly better ultimate clinical outcomes (Fig. 3E, P 0.001 , PFS HR = 1.46). Moreover, when comparing hazard ratios of normalized bTMB to either bTMB or ctDNA alone, we find that normalized bTMB is superior to both individual metrics in ICI-treated patients (Fig. S3A). To assess if normalized bTMB was predictive of durable benefit specifically for ICI, we separately analyzed its relationship to outcomes in patients who received chemotherapy alone (Fig. 3F). Interestingly, in chemotherapy treated patients, lower bTMB and lower ctDNA burden were individually associated with higher likelihood of DCB, such that normalized bTMB was not associated with outcome in the absence of ICI (P = 0.88, Fig. 3F-I). As expected, tumor PD- L1 expression was also associated with DCB only in the setting of ICI (P < 0.01 for ICI; P = 0.41 for chemotherapy). Therefore, ctDNA- normalized bTMB captures tumor-intrinsic factors that are predictive only in the context of ICI.

[00129] Next, we attempted to validate the association of ctDNA concentration and normalized bTMB in an independent cohort (Fig. 8), comprising 99 NSCLC patients who received ICI with available pre- treatment plasma and of whom 94 had detectable ctDNA by CAPP-Seq (95%, Tables S3-4). For 30 of these patients, we analyzed paired tumor specimens by whole- exome sequencing and pre-treatment plasma by CAPP-Seq and confirmed in a cross- validation framework that TMB measured in cell-free DNA robustly estimates tissue TMB (Fig. 10B-C). Overall, the associations between clinical benefit and ctDNA concentration or normalized bTMB in the validation cohort mirrored those observed in the discovery cohort (Fig. S3B-E). Utilizing the normalized bTMB threshold defined in the discovery cohort to stratify patients in the validation cohort, patients with high normalized bTMB had significantly better outcomes (Fig 3J, P = 0.020, PFS HR = 1.74). Collectively, these data suggest that pre- treatment normalized bTMB can noninvasively identify patients most likely to experience durable benefit in response to ICI. However, nearly two thirds of patients were misclassified using normalized bTMB alone: -25% of patients predicted to have NDB nevertheless achieved PFS >6 months, and conversely, -40% of patients predicted to have DCB had PFS < 6 months. This suggests that further improvement in predictive accuracy would be desirable.

[00130] Circulating immune profiles predict response to ICI. We therefore explored other biomarkers as potential determinants of durable benefit from ICI. Specifically, we reasoned that the frequency and representation of circulating leukocyte subsets might vary between patients achieving or failing to achieve DCB from ICI. We therefore profiled leukocyte transcriptomes using RNA-seq and applied CIBERSORTx (Newman et al., 2015, 2019) to quantify the relative proportions of major leukocyte subpopulations (Fig. 4A-B, 11) n = 37 patients). We confirmed highly correlated composition estimates of leukocyte composition, whether using RNA-seq deconvolution or conventional flow cytometric immunophenotyping (Fig. 11A-B). Interestingly, we observed fewer circulating CD8 T cells prior to ICI therapy in patients ultimately achieving DCB — while no other cell type was significantly associated with clinical benefit (Fig. 4A-B). Indeed, CD8 T cell levels alone had comparable classification accuracy for predicting durable benefit (AUC = 0.61 , precision [PPV] = 69%), compared to other pre-treatment features including PD-L1 and normalized bTMB (Fig. 4C).

[00131 ] Early on-treatment ctDNA dynamics classify durable ICI response Recent studies have reported that ctDNA responses within 4-8 weeks after starting ICI correlate with best radiographic responses and modestly with DCB (Anagnostou et al., 2019; Goldberg et al., 2018; Raja et al., 2018). We therefore explored whether ctDNA kinetics as early as after a single ICI infusion (<4 weeks after starting first cycle) could help determine likelihood of durable benefit from ICI. We measured post-treatment ctDNA responses early after ICI therapy initiation in patients with detectable ctDNA at baseline (Tables S3-4). Based on a prior study of intra-day ctDNA variation in untreated advanced NSCLC, we defined a threshold of 50% decrease in ctDNA concentration from pre-treatment as a “ctDNA molecular response” (Wang et al., 2017). Even at these early time points, ctDNA levels showed significant changes compared to baseline in most patients (Fig. 5A). ctDNA burden dropped in 59% of patients achieving DCB while it did not meet the 2-fold threshold in 95% of NDB patients (P < 0.0001 ; Fig. 5A-5B). Strikingly, ctDNA responses after a single cycle of ICI therapy distinguished the majority of DCB from NDB patients (Fig. 5B). Furthermore, early ctDNA dynamics outperformed all individual pre-treatment factors (Fig. 12, P < 0.05, AUC = 0.89) and patients with ctDNA molecular responses had significantly better clinical outcomes (Fig. 5C, P= 0.013, PFS HR = 2.28). Thus, early ctDNA molecular response after just one cycle of ICI appears to be a promising approach for early response assessment. However, given that 25% of patients are incorrectly classified, further improvements in predicting ultimate clinical outcomes would be desirable.

[00132] Multi-parameter models robustly classify durable clinical benefit from ICI Having identified tumor-cell intrinsic and -extrinsic determinants of ICI therapeutic benefit, we next combined these into integrated multi-parametric models that include normalized blood tumor mutation burden as well as circulating immune phenotypes. We developed two related Bayesian biomarker models, one relying on pre-treatment factors alone, and another incorporating early on-treatment ctDNA dynamics. These models were first trained in a discovery cohort with available blood samples both before ICI and also early after initial therapy (<4 weeks on-treatment; Fig. 8, n = 34 patients). We then validated these models in an independent cohort (Fig. 8, n = 38 patients). A model based on pre-treatment tumor PD-L1 expression, normalized bTMB, and circulating CD8 T cells (DIREct-Pre: Durable Immunotherapy Response Estimation by immune profiling and ctDNA- Pre-treatment) was significantly associated with clinical benefit (Fig. 6A, PFS HR = 2.86; P < 0.05). Cross-validation analyses revealed the DIREct-Pre model to achieve 88% sensitivity (classifying DCB) in our discovery cohort, but only 56% specificity (classifying NDB) (Fig. 6B, 13B). Patients with higher DIREct-Pre scores achieved significantly longer PFS (Fig. 6C). Given these observations, we reasoned that incorporation of early ctDNA dynamics would further improve response classification performance. As expected, addition of early ctDNA response to DIREct- Pre significantly improved classification accuracy (Fig. 6A, PFS HR = 6.65, P = 0.007). Furthermore, addition of early on-treatment ctDNA dynamics rendered tumor PD-L1 expression dispensable for response classification (i.e., DIREct-On, Durable Immunotherapy Response Estimation by immune profiling and ctDNA- On-treatment. Fig.6A, Fig. 13A-B). Cross-validation analyses revealed the DIREct- On model to achieve 94% sensitivity (classifying DCB) and 89% specificity (classifying NDB) in our discovery cohort (Fig. 6D, 13B). As expected, patients with higher DIREct-On scores had substantially longer PFS than those with lower scores (Fig. 6E, P< 0.0001 , HR = 8.93). Overall, DIREct-On achieved excellent classification performance (AUC = 0.93) with high precision (PPV = 0.88). Thus, integration of factors allows robust response classification using noninvasively measured biomarkers within a multi- analyte assay.

[00133] DIREct-Pre and DIREct-On validate in an independent cohort Having trained these two models, we next tested each in an independent validation cohort. Based on the performance of DIREct-On in the discovery cohort, a validation cohort of >20 provides >99% power to detect a similar difference between the two groups (one-sided two arm binomial with alpha = 0.05). We therefore assessed each of the parameters in an independent validation cohort of 38 additional advanced NSCLC patients receiving ICI. Using the threshold established in the discovery cohort, patients with higher Dl REct-Pre scores had significantly longer PFS than those with low DIREct-Pre scores (Fig. 7A, P = 0.03, PFS HR = 2.18). Remarkably, DIREct-On outperformed all other considered features, achieving 95% precision for classifying DCB in the validation cohort. Patients with higher DIREct-On scores had strikingly longer PFS than those with low DIREct-On scores, with median PFS of 8.1 months versus 2.1 months, respectively (Fig. 7B, P< 0.0001 , PFS HR = 7.11).

[00134] DIREct-On outperforms ctDNA dynamics alone To assess performance of DIREct- On compared to each individual feature, we compared accuracy of DCB versus NDB classification and hazard ratios of DIREct-On with that of individual features in the combined discovery and validation cohorts. DIREct-On had significantly better classification accuracy than each individual metric or tumor PD-L1 expression (Fig. 7C, left). Moreover, comparison of hazard ratios also demonstrates that patients with high DIREct-On scores had a significantly lower risk of progression than patients with ctDNA molecular responses, low circulating CD8 T cells, high normalized bTMB, or high tumor PD-L1 expression (P < 0.0001). (Fig. 7C, right). Similarly, we also assessed the superiority of DIREct-On to each of the individual features that comprise the model using Net Reclassification Improvement (NRI) (Fig. 7D, top). DIREct-On was significantly superior to normalized bTMB (NRI = -1.48, P < 0.05), CD8 T cell fraction (NRI = -1 .64, P < 0.01), and ctDNA dynamics (NRI = -1.13, P< 0.01). Thus, the multiparameter DIREct-On model significantly outperforms each individual feature. To assess if each of the features are required for optimal DIREct-On response classification we constructed models using the entire cohort (n = 72) and compared the full DIREct-On model to versions of the same model with individual parameters removed. This enables the measurement of the impact of each feature to the overall performance of DIREct-On. Indeed, each feature that comprises DIREct-On is required for optimal performance (Fig. 7D, bottom). As expected, removal of ctDNA dynamics had the largest NRI (NRI = -1.60, P < 0.001), but removal of either normalized bTMB or CD8 T cell fractions also significantly decreased performance (NRI = -1.25, P < 0.001 ; NRI = -1.13, P < 0.001 , respectively). Comparison of HRs for the full DIREct-On model and the versions with each parameter removed showed similar results (Fig. 14A). Thus, each feature is required for optimal response classification.

[00135] DIREct-On classifies patients with ambiguous imaging responses Next, we explored the potential utility of DIREct-On to reconciling response status in patients with advanced lung cancer with potentially ambiguous imaging results. First, we examined patients with a best overall response of SD. In both the discovery and validation cohorts, DIREct-On accurately classified the durability of ICI response for 94% of patients (17/18) who achieved SD as their best radiographic response (Fig. 7E). Moreover, in patients whose first scan after treatment initiation was classified as stable, DIREct-On correctly identified 94% of patients (17/18) ultimately achieving DCB (Fig. 7F). Therefore, DIREct-On allows early identification of patients with initially stable disease who are most likely to achieve durable clinical benefit from ICI. [00136] DIREct-On forecasts ultimate clinical outcomes for ICI-treated NSCLC patients

Finally, we also tested calibration of DIREct-Pre and DIREct-On in the validation cohort in order to evaluate how accurately they forecast ultimate clinical outcomes (Steyerberg et al., 2010) and found that both models were well-calibrated (slope = 0.64, 0.98, respectively; Fig. 14B- C). Similarly, the DIREct-Pre and DIREct-On scores forecasted PFS as continuous variables in the validation cohort (Cox likelihood ratio test: PFS HR = 6.97, P= 0.007; PFS HR = 11 .53, P < 0.0001 , respectively). Furthermore, we found no significant difference in PFS between patients who achieved DCB or NDB versus those expected to do so by DIREct-On (Fig. 7G).

[00137] Here, we describe novel parameters associated with durable clinical benefit from ICI in patients with advanced NSCLC that integrate both tumor-intrinsic and -extrinsic determinants and that can be measured noninvasively using blood. We found lower baseline circulating CD8 T cell levels in patients with lung cancer ultimately achieving durable benefit from ICI, a result consistent with observations in melanoma responders to immunotherapy (Krieg et al., 2018). We speculate that this observation might reflect increased CD8 T cell homing to tumor deposits in patients with strongly immunogenic tumors. We also demonstrate that a pretreatment composite model (DIREct-Pre) combining tumor PD-L1 expression with pre-treatment ctDNA and circulating immune cell profiling accurately predicts outcomes. Moreover, we developed DIREct-On, which is a fully noninvasive response classifier that incorporates pretreatment ctDNA and immune profiling with early on-treatment ctDNA response assessment to most accurately classify the likelihood of durable benefit after one cycle of immunotherapy. DIREct-On allows for classification of ultimate clinical outcomes significantly earlier than previously reported ctDNA-based approaches (Anagnostou et al., 2019; Goldberg et al., 2018; Raja et al., 2018).

[00138] Moreover, we demonstrate that DIREct-On has significantly better performance than ctDNA dynamics alone, both in our cohort or compared to prior approaches. Therefore, the integration of pre-treatment parameters that can be measured noninvasively allows for optimal classification of response. DIREct-On also outperforms previously described integrative models relying on tumor RNA and/or DNA-seq (Anagnostou et al., 2020; Auslander et al., 2018; Cristescu et al., 2018; Jiang et al., 2018). The prior approaches were largely focused on melanoma rather than NSCLC and have substantially inferior performance compared to DIREct-On (prior studies: HR = 0.30-0.37, AUC = 0.70-0.83; DIREct-On: HR = 0.04-0.11 , AUC = 0.89-0.93).

[00139] We envision that DIREct-Pre and DIREct-On could facilitate personalized selection of ICI for patients with advanced NSCLC, with the goal of improving outcomes while minimizing toxicities. For example, one approach that could be tested in clinical trials is for patients with Stage IV NSCLC to be treated with single-agent PD-1 blockade for one cycle (irrespective of PD-L1 expression) and to then have DIREct-On measured. Patients with a high DIREct-On score (expected durable benefit) would remain on single agent PD-1 blockade whereas patients with low DIREct-On scores (expected lack of benefit) would receive treatment escalation through the addition of chemotherapy (Fig. 7H). If successful, such a trial could optimize the number of patients receiving immunotherapy alone and reserve combination immunotherapy plus chemotherapy for patients who are destined not to respond to immunotherapy alone.

[00140] Finally, these models have utility in predicting response to ICI in other tumor types. Similar tumor-intrinsic and -extrinsic features have been identified to associate with response to ICI in other cancers, and therefore DIREct-On has value beyond NSCLC. Given the difficultly to date of developing robust predictive biomarkers for immunotherapy akin to somatic mutations for targeted therapies, an approach based on very early response assessment could improve personalization of therapy for patients treated with ICIs.

Materials and Methods

[00141] Data and materials availability Subject details Immune checkpoint inhibitor treated patients All patients had stage IV non-small cell lung cancer (NSCLC) and were treated with ICI (PD-(L)1 blockade alone or in combination with CTLA-4 blockade or chemotherapy) at Stanford University Cancer Center or Memorial Sloan Kettering Cancer Center (Figure 2 and Table 1). All patients consented to Institutional Review Board-approved protocols permitting specimen collection and genetic sequencing for analysis of tumor biopsies, leukocytes, and cfDNA.

[00142] Clinical efficacy analysis Response was quantified by investigator-assessed RECIST v1.1 (Eisenhauer et al., 2009). Durable clinical benefit (DCB) was defined as confirmed absence of progressive disease for at least 6 months after ICI; whereas, no durable benefit (NDB) was defined as patients experiencing progression or death within 6 months (Rizvi et al., 2015). Progression-free survival was determined from the start of PD-(L)1 blockade, with outcomes determined or censored as of the 01/22/2019 database lock.

[00143] Tumor samples Thirty of 82 patients had tumor tissue available for whole-exome sequencing. All tumor tissue was obtained prior to treatment with immunotherapy. Tumor PD- L1 expression was assessed by clinical immunohistochemistry and the percentage of PD- L1 positive tumor cells was used as input into the relevant Bayesian models.

[00144] Plasma and leukocyte samples Plasma was processed from whole blood samples collected in EDTA tubes. Tubes were centrifuged at 2500 x g at room temperature for 10 minutes. The plasma supernatant was collected and the remaining plasma-depleted whole blood was collected for DNA and/or RNA purification. In indicated cases, density gradient centrifugation was used to collect PBMC for flow cytometry, DNA, and/or RNA isolation. [00145] Method Details cfDNA extraction Cell-free DNA was extracted from three to six mL of plasma utilizing the QiaAmp Circulating Nucleic Acid Kit per manufacturer’s instructions. DNA was quantified using the Qubit dsDNA High Sensitivity Kit and quality and size was assessed by the Agilent 5400 Fragment Analyzer. For samples collected in CPT tubes, cfDNA was treated with Heparinase II (Sigma) for 2 hours at 37°C and re-purified by 1 .8X Ampure XP bead selection prior to quantitation and size analysis.

[00146] Germline DNA extraction Germline DNA was extracted from 100mI_ of PDWB or -30,000 PBMCs with the QiaAmp DNA Micro Kit per manufacturer’s instructions. 100-1000ng of genomic DNA was then sheared using the Covaris S2 Focused-ultrasonicator using the following settings: 10% duty cycle, intensity level 5, 200 cycles per burst, and 2 minute duration. After sonication, sheared DNA was re-purified using the QiaQuick PCR Purification Kit per manufacturer’s instructions.

[00147] CAPP-Seq CAPP-Seq was performed as previously described (Newman et al., 2014, 2016). Briefly, a 20- 55ng of cfDNA or sheared genomic DNA was utilized for library preparation with the KAPA HyperPrep Kit with some modifications to the manufacturer’s instructions, as described (Chabon et al., 2016; Chaudhuri et al., 2017). After library preparation, custom-designed biotinylated DNA oligonucleotides were utilized for hybridization and subsequent enrichment with a capture panel covering 355 kb and targeting 270 genes could. Following hybridization capture, samples were sequenced on an lllumina HiSeq4000 as 2 x 150bp lanes of 8-12 samples multiplexed per lane yielding -40 million paired-end reads with a median deduped depth of 3037X per case. Data were then processed using a custom bioinformatics pipeline (Newman et al., 2014, 2016). To abrogate the need for invasive biopsies, the CAPP-Seq analyses performed in this study were completed in a tumor-naive manner. For tumor na^'fve calling we: 1) limited variants to coding positions, 2) removed any variants with greater than 1 reads in the matched germline samples, 3) removed variants with more than 2 reads in 5% of healthy control plasma samples (n = 54), 4) removed variants present in 709 >0.05% of samples in the Genome Aggregation Database (Lek et al., 2016). For monitoring of these variants in early on-treatment samples we used a previously described Monte Carlo-based ctDNA detection index with a significance cut-point of P £ 0.05 (Newman et al., 2014). ctDNA fold change was calculated by dividing the ctDNA concentration at an on- treatment timepoint by the pre-treatment ctDNA concentration. In on-treatment cases where ctDNA was undetectable, the limit of detection for that sample was used as the ctDNA concentration, as described previously (Chabon et al., 2016). In order to define the threshold for a ctDNA molecular response we analyzed previously published data from 23 treatment- na^'rve advanced NSCLC patients who had three plasma collections over the course of 6 hours. The maximum deviation from the average ctDNA concentration amount these patients was 1.87-fold (Wang et al., 2017). Therefore, we considered a ctDNA molecular response as a >2 fold decrease from the pre-treatment time point.

[00148] Tumor mutation burden estimation by CAPP-Seq Tumor mutation burden was estimated by CAPP-Seq by identifying the relationship between the number of nonsynonymous mutations identified in pre-treatment tumors by whole-exome sequencing per megabase of coding exome captured to the number of coding mutations (nonsynonymous and synonymous) identified in the pre-treatment cfDNA by CAPP-Seq per megabase of coding exome captured in a discovery cohort of 24 patients and validated in 6 independent patients (Figures 8B-C). We derived a linear regression model in a leave-one-out cross validation (LOOCV) framework in the discovery cohort 24 patients and found that TMB measured in cell-free DNA was significantly correlated with tissue TMB (Figure 8B, R = 0.86, P < 0.0001). This resulted in the following relationship:

We then validated this model in the independent group of patients and again observed strong correlation (Figure 8C, R = 0.99, P < 0.001 , n=6).

[00149] RN A extraction We did not have access to viably preserved peripheral blood mononuclear cells (PBMCs) amenable to flow cytometry for most patients but did have frozen plasma depleted whole blood (PDWB) containing leukocyte RNA for the majority of our cohort. Leukocyte RNA was extracted from either PDWB or PBMCs. For PDWB, RNA was extracted from 200pL of PDWB by mixing with 600pL TRIzol LS Reagent (ThermoFisher), 200pL of chloroform was used for phase separation. For PBMCs, pellets were lysed with 600pL of TRIzol Reagent and 150pL of chloroform was used for phase separation. For both starting materials, the aqueous phase from the initial phase separation was then mixed with equal volume of 100% ethanol, then loaded onto columns for the RNeasy Micro Kit (Qiagen) and the protocol was followed per manufacturer’s instruction, including on-column DNase treatment. The GLOBINclear Kit (Ambion) was utilized to deplete globin mRNA per the manufacturer’s instructions in the PDWB samples. Before proceeding to RNA-seq, RNA was quantified using the Quant-iT RiboGreen RNA Assay Kit (ThermoFisher) and quality was confirmed by DV200 and RIN after RNA Pico Bioanalyzer Analysis. All RNA samples sequenced had DV200 > 80%.

[00150] RNA-seqT e lllumina TruSeq RNA Exome kit was used for RNA-seq library preparation with 20ng of input PBMC RNA or globin-depleted PDWB RNA per manufacturer’s instructions. In brief, RNA was fragmented and stranded cDNA libraries were created per the manufacturer’s protocol. The RNA libraries were then enriched for the coding transcriptome by exon capture. Hybridization captures were then pooled and samples were sequenced on an lllumina HiSeq4000 as 2 x 150bp lanes of 16-20 samples per lane, yielding ~20 million paired-end reads per case. After demultiplexing the data were aligned and expression levels summarized using Salmon to Gencode version 27 transcript models (Patro et al., 2017). Expression levels per gene were summarized as transcripts per million (TPM), and then used as input for CIBERSORTx for deconvolution with the LM22 signature matrix with B-mode batch correction (Newman et al., 2019).

[00151] Tumor whole-exome sequencing Whole-exome libraries were prepared using the Agilent Sure-Select Human All Exon v2.0, v4.0, or the lllumina Rapid Capture Exome Kit. Captured exome libraries for tumor and germline DNA were sequenced to equivalent depths on HiSeq2000, HiSeq2500, or HiSeq4000 platforms as 2x75bp or 2x150bp reads and aligned to the hg19 human genome build using the Burrows-Wheeler Aligner (Li and Durbin, 2009). Further indel realignment, base-quality score recalibration, and duplicate-read removal were performed utilizing the Genome Analysis Toolkit (DePristo et al., 2011). Single nucleotide variants were identified using Mutect2 with default parameters (Cibulskis et al., 2013). Tumor mutational burden per megabase for bTMB estimation was calculated by dividing the total number of nonsynonymous mutations by the coding region of each respective capture kit (Agilent Sure-Select Human All Exon v2.0 = 29.9MB, Agilent Sure-Select Human All Exon v4.0 = 34.2 MB, lllumina Rapid Capture Exome Kit = 42.7MB).

[00152] Flow cytometry Flow cytometry was used to confirm the fidelity of CIBERSORTx estimates for the major lymphoid and myeloid lineages among the cellular subsets captured by the LM22 signature matrix, whether viable or nonviable leukocytes were used as input RNA for sequencing and subsequent transcriptome deconvolution. To do so, single cell suspensions were prepared from viably frozen PBMC aliquots. Live/dead cell discrimination was performed using 7-AAD Viability Staining Solution (Biolegend). Human Fc receptors were blocked using Human TruStain FcX (Biolegend). Cell surface staining was performed for 30 min at 4°C with the following antibodies: CD14 (Clone: M5E2, Color: BV421 , Becton Dickinson), CD3 (Clone: UCHT1 , Color: PE, Biolegend), CD4 (Clone: RPA-T4, Color: PerCP-Cy5.5, Becton Dickinson), CD8 (Clone: RPA-T8, Color: FITC, Becton Dickinson), CD19 (Clone: SJ25C1 , Color: BV711 , Becton Dickinson), and CD56 (Clone: HCD56, Color: PE-Cy7, Biolegend). All data acquisition was done using a FACS Aria Fusion (Becton Dickinson) or a CytoFLEX Flow Cytometer (Beckman Coulter) and analyses were performed using FlowJo v10. CD4 T cells were gated by CD14-,CD19-,CD3+,CD8-,CD4+; CD8 T cells were gated by CD14-,CD19-,CD3+,CD8+,CD4- ; monocytes were gated by CD56-,CD3- ,CD19-,CD14+; B cells were gated by CD3-,CD19+; and NK cells were gated by CD14-,CD3-,CD19- ,CD56+.

[00153] Statistical Analyses and Modeling Statistical analysis The Wilcoxon test was used to compare groups where. Fisher’s exact test was used to compare bTMB high versus bTMB low in the normalized bTMB Discovery and Validation cohort. Receiver operating characteristic curves were used to generate area under the curve (AUC), and 95% confidence intervals for AUCs were generated by bootstrapping (Robin et al., 2011). Significance analysis of AUCs, NRI, hazard ratios, and PPVs were also computed by bootstrapping, where the difference in the metrics curves for each bootstrap replicate is computed and compared across conditions to generate an empiric P-value (Kurtz et al., 2019). For progression-free survival analysis, the log-rank test was used to compare Kaplan-Meier survival curves. Cox proportional hazards regression models were used to generate hazard ratios as stratified relative to cut-points defined in each figure, as captured in corresponding legends. Stratification thresholds were determined using optimal cut-point selection by ROC analysis only within the discovery cohorts. For ctDNA dynamics, the cut-point was not optimized and 0.5 was used as the cut-point in both discovery and validation cohorts. For PFS analysis of the validation cohorts, the cut-points from the discovery cohorts were applied to the respective validation sets to stratify response. DIREct-Pre and DIREct-On were also tested as continuous variables in the validation cohort by Cox likelihood ratio test. When comparing performance of individual biomarkers to each other or to Bayesian probit models analyses were restricted to patients with all data types available, as described in the corresponding legends. All analyses were conducted using fi v.3.5.1 using the pROC, survminer, MCMCpack, and survival packages.

Construction of Bayes priors from public data

[00154] To identify the parameters of the Gaussian distribution governing the probit coefficients, we denoted the features from Gandara et al. (Gandara et al., 2018), by X_w ,_r and the corresponding labels (DCB/NDB) by Y_wi0r. Here, we subsampled of n=100 cases for m = 2000 times. For each iteration and for each set of covariates (e.g., PDL1 and normalized bTMB), we solved a probit regression model to generate a matrix of associated coefficients (/?). We used this matrix as a random realization of this multi variable normal distribution, and then estimated the Bayes prior probability parameters (i.e., hyper parameters).

Prior construction within leave-one-out cross-validations and Markov-Chain Monte Carlo (MCMC)

[00155] We used a LOOCV framework to fully exploit the available data, and for the features not estimated by X_wi0r (e.g., CD8 T cell fraction). Specifically, we used n - 1 samples to find the hyperparameters as described above. We then derived the final full prior probability distributions by concatenating two components: (1) the prior knowledge-based priors, and (2) the priors derived from our Discovery Cohort, such that

bo, prior and b_{0, ourcohort} are the mean vectors of the variables, estimated from the prior set and our cohort, respectively. Similarly A_0,P™_r ^_1 and L_{0 , ourcohort} ^_1 denote the covariance matrices. We then used the same n - 1 samples with all available covariates in a MCMC step to calculate the posterior probabilities (i.e., updated prior probabilities). In cases where tumor PD-L1 expression was unavailable (Discovery Cohort, n = 11 ; Validation Cohort, n = 4), this parameter was left out.

Final prediction score for both DIREct models

[00156] To generate a single score representing the probability of DCB, we used the MCMC posterior samples, and generated 10,000 pseudo-samples with 1000 initial burn-in. These realizations of the b vectors (i.e. {/?i,/?₂,...,/?ioooo} along with the sample feature vector, x) were then used to calculate 10,000 realizations of the DCB score, such that F(c^tbi) for i e {1 ,...,10000}. We then used the median of these realizations as the final prediction score.

Threshold selection for DIREct models

[00157] To select the optimal thresholds in the discovery cohort for the DIREct-Pre and DIREct- On scores, we used the optimal ROC corner point approach (Youden’s J) for classification of DCB, where the optimality was defined as the point with maximal unweighted accuracy, i.e. specificity plus sensitivity (Youden, 1950). These cut-points were then applied to the validation set for calculation of precision and survival analyses.

Calibration testing of DIREct models

[00158] We evaluated DIREct predictions for quantitative accuracy via model calibration regression (Steyerberg et al., 2010). To test calibration of the DIREct-Pre and DIREct-On models we performed a bootstrap resampling with n=2000 and reported the “true event rate”, estimated via the Kaplan-Meier maximum likelihood estimates of observed events versus expected events from our models’ predictions. A perfect calibration leads to a slope of 1 , i.e. the predicted probability of the model is in fact representative of the true event rate in the population.

Claims

WHAT IS CLAIMED IS:

1 . A method for assessing propensity of an individual with cancer to develop durable clinical benefit from treatment with an immune checkpoint inhibitor (ICI), comprising:

(i) obtaining a patient cellular biological sample for analysis comprising immune cells for profiling; and determining the level of circulating CD8⁺ T cells in the cellular biological sample;

(ii) obtaining a patient sample comprising circulating cell-free DNA; determining concentration of circulating cell-free tumor DNA (ctDNA) in the sample;

(iii) determining a tumor mutational burden (TMB) value from a tumor sample or from ctDNA; integrating results of steps (i), (ii) and (iii) to generate a single biomarker that predicts DCB from ICI treatment.

2. The method of claim 1 , wherein the individual with cancer is treated with an ICI if DCB is predicted; and treated with non-ICI therapy if DCB is not predicted.

3. The method of claim 1 or claim 2, wherein the cancer is non-small cell lung carcinoma, small cell lung carcinoma, adenocarcinoma, squamous cell carcinoma, hepatocarcinoma, basal cell carcinoma, lymphoma, or melanoma.

4. The method of any of claims 1-3, wherein the cancer is non-small cell lung cancer.

5. The method of any of claims 1-4, wherein the immune checkpoint inhibitor is a PD-1 or PD-L1 inhibitor.

6. The method of any of claims 1-5, wherein the TMB value is determined from ctDNA with CAPP-seq.

7. The method of any of claims 1-6, wherein the TMB value is determined from ctDNA and is normalized as a ratio of ctDNA concentration.

8. The method of any of claims 1-7, wherein the cellular biological sample and the circulating cell-free DNA sample are obtained prior to ICI treatment.

9. The method of claim 8, further comprising:

(iv) determining expression of an immune checkpoint protein on tumor cells from the individual with cancer; integrating results of steps (i), (ii), (iii) and (iv) to generate a single biomarker that predicts DCB from ICI treatment.

10. The method of claim 9, wherein the immune checkpoint protein is PD-L1 .

11. The method of claim 9 or claim 10, wherein expression of an immune checkpoint protein is determined from a biopsy sample.

12. The method of any of claims 9-11 , wherein the expression of an immune checkpoint protein is determined by immunohistochemistry.

13. The method of any of claims 1-7, wherein the cellular biological sample and the circulating cell-free DNA sample are obtained within 4 weeks of a first ICI treatment.

14. The method of claim 13, further comprising

(iv) obtaining a patient sample comprising circulating cell-free DNA prior to ICI treatment; determining concentration of ctDNA in the sample; determining tumor mutational burden (TMB) value from the ctDNA; normalizing the TMB as a ratio of ctDNA to generate a single value; and integrating results of steps (i), (ii), (iii), (iv) to generate a single biomarker that predicts DCB from ICI treatment.

15. The method of any of claims 1-14, wherein immune profiling is conducted using CIBERSORTx, flow cytometry or mass cytometry.

16. The method of any of claims 1-15, wherein a Bayesian framework is used to integrate values to generate a single biomarker that predicts DCB from ICI treatment.

17. The method of any of claims 1-16, wherein one or more steps are implemented on a computer comprising a software component configured for analysis of data obtained by the methods.

18. A software product tangibly embodied in a machine-readable medium, the software product comprising instructions operable to cause one or more data processing apparatus to perform the method of any of the preceding claims.