Search

Aim: Investigate the relationship between response to pembrolizumab and expression of the 18-gene T cell-inflamed gene expression profile (TcellinfGEP) or PD-L1 combined positive score (CPS) in esophageal cancer. Materials & methods: This analysis included heavily pretreated patients with advanced/metastatic esophageal/gastroesophageal junction adenocarcinoma or squamous cell carcinoma who received pembrolizumab in the single-arm, phase II study KEYNOTE-180. PD-L1 CPS was evaluated with PD-L1 IHC 22C3 pharmDx. Results: In patients with squamous cell carcinoma, trends toward enrichment for responders were observed for patients with PD-L1 CPS ≥10 tumors. In patients with adenocarcinoma, a trend was observed for TcellinfGEP but not for PD-L1. Conclusion: TcellinfGEP and PD-L1 CPS may enrich for responders to pembrolizumab in patients with esophageal cancer. Clinical Trial Registration: NCT02559687 ( ClinicalTrials.gov )

Supplementary Table S1 from Breast Cancer Molecular Subtypes Respond Differently to Preoperative Chemotherapy

Supplementary Figure from Transcriptomic Determinants of Response to Pembrolizumab Monotherapy across Solid Tumor Types

Purpose:To explore relationships between biological gene expression signatures and pembrolizumab response.Experimental Design:RNA-sequencing data on baseline tumor tissue from 1,188 patients across seven tumor types treated with pembrolizumab monotherapy in nine clinical trials were used. A total of 11 prespecified gene expression signatures [18-gene T-cell–inflamed gene expression profile (TcellinfGEP), angiogenesis, hypoxia, glycolysis, proliferation, MYC, RAS, granulocytic myeloid-derived suppressor cell (gMDSC), monocytic myeloid-derived suppressor cell (mMDSC), stroma/epithelial-to-mesenchymal transition (EMT)/TGFβ, and WNT] were evaluated for their relationship to objective response rate (per RECIST, version 1.1). Logistic regression analysis of response for consensus signatures was adjusted for tumor type, Eastern Cooperative Oncology Group performance status, and TcellinfGEP, an approach equivalent to evaluating the association between response and the residuals of consensus signatures after detrending them for their relationship with the TcellinfGEP (previously identified as a determinant of pembrolizumab response) and tumor type. Testing of the 10 prespecified non-TcellinfGEP consensus signatures for negative association [except proliferation (hypothesized positive association)] with response was adjusted for multiplicity.Results:Covariance patterns of the 11 signatures (including TcellinfGEP) identified in Merck–Moffitt and The Cancer Genome Atlas datasets showed highly concordant coexpression patterns in the RNA-sequencing data from pembrolizumab trials. TcellinfGEP was positively associated with response; signatures for angiogenesis, mMDSC, and stroma/EMT/TGFβ were negatively associated with response to pembrolizumab monotherapy.Conclusions:These findings suggest that features beyond IFNγ-related T-cell inflammation may be relevant to anti–programmed death 1 monotherapy response and may define other axes of tumor biology as candidates for pembrolizumab combinations.See related commentary by Cho et al., p. 1479

Supplementary Tables S1, S2; Supplementary Figures S1, S2, S3, S4

Supplementary Data from Transcriptomic Determinants of Response to Pembrolizumab Monotherapy across Solid Tumor Types

Purpose:Molecular profiling of large databases of human tumor gene expression profiles offers novel opportunities for informing decisions in clinical development programs.Experimental Design:Gene expression profile of programmed death ligand 1 (PD-L1) was explored in a dataset of 16,000 samples, including approximately 4,000 metastatic tumors, across >25 tumor types prevalent in the United States, looking for new indications for the programmed death 1 (PD-1) inhibitor pembrolizumab. PD-L1 expression was highly concordant with several genomic signatures indicative of immune-inflamed tumor microenvironment. Prevalence of activated immune-inflamed tumors across all tumor types was explored and used to rank tumor types for potential response to pembrolizumab monotherapy.Results:The analysis yielded 3 tiers of indications in which high levels of PD-L1 and immune-inflamed signatures were found in up to 40% to 60%, 20% to 40%, and 0% to 20% of tumors. Tier 1 contained novel indications known at the time of analysis to be responsive to PD-1 checkpoint blockade in the clinic (such as melanoma and non–small cell lung cancer), as well as indications not studied in the clinic previously, including microsatellite instability–high colorectal, head and neck, bladder, and triple-negative breast cancers. Complementary analysis of an Asian/Pacific cancer dataset (gastric cancer) revealed high prevalence of immune-inflamed tumors in gastric cancer. These data contributed to prioritization of these indications for clinical development of pembrolizumab as monotherapy.Conclusions:Data highlight the value of molecular profiling in identifying populations with high unmet needs with potentially favorable response characteristics and accelerating development of novel therapies for these patients.See related commentary by Mansfield and Jen, p. 1443

Biomarker; Pembrolizumab; Tissue tumor mutational burden Biomarcador; Pembrolizumab; Càrrega mutacional del tumor tissular Biomarcador; Pembrolizumab; Carga mutacional del tumor tisular Introduction We evaluated tissue tumor mutational burden (tTMB) and mutations in STK11, KEAP1, and KRAS as biomarkers for outcomes with pembrolizumab plus platinum-based chemotherapy (pembrolizumab-combination) for NSCLC among patients in the phase 3 KEYNOTE-189 (ClinicalTrials.gov, NCT02578680; nonsquamous) and KEYNOTE-407 (ClinicalTrials.gov, NCT02775435; squamous) trials. Methods This retrospective exploratory analysis evaluated prevalence of high tTMB and STK11, KEAP1, and KRAS mutations in patients enrolled in KEYNOTE-189 and KEYNOTE-407 and the relationship between these potential biomarkers and clinical outcomes. tTMB and STK11, KEAP1, and KRAS mutation status was assessed using whole-exome sequencing in patients with available tumor and matched normal DNA. The clinical utility of tTMB was assessed using a prespecified cutpoint of 175 mutations/exome. Results Among patients with evaluable data from whole-exome sequencing for evaluation of tTMB (KEYNOTE-189, n = 293; KEYNOTE-407, n = 312) and matched normal DNA, no association was found between continuous tTMB score and overall survival (OS) or progression-free survival for pembrolizumab-combination (Wald test, one-sided p > 0.05) or placebo-combination (Wald test, two-sided p > 0.05) in patients with squamous or nonsquamous histology. Pembrolizumab-combination improved outcomes for patients with tTMB greater than or equal to 175 compared with tTMB less than 175 mutations/exome in KEYNOTE-189 (OS, hazard ratio = 0.64 [95% confidence interval (CI): 0.38‒1.07] and 0.64 [95% CI: 0.42‒0.97], respectively) and KEYNOTE-407 (OS, hazard ratio = 0.74 [95% CI: 0.50‒1.08 and 0.86 [95% CI: 0.57‒1.28], respectively) versus placebo-combination. Treatment outcomes were similar regardless of KEAP1, STK11, or KRAS mutation status. Conclusions These findings support pembrolizumab-combination as first-line treatment in patients with metastatic NSCLC and do not suggest the utility of tTMB, STK11, KEAP1, or KRAS mutation status as a biomarker for this regimen. All authors’ institutions received research funding from Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, for the conduct of this study.

We performed an integrated biomarker evaluation in pembrolizumab-treated patients with R/M HNSCC enrolled in KEYNOTE-012 or KEYNOTE-055. The relationship between biomarkers and HPV status was explored.We evaluated PD-L1 (combined positive score [CPS]), TMB, T-cell-inflamed gene expression profile (TcellTwo hundred and fifty-seven patients (KEYNOTE-012, n = 106; KEYNOTE-055, n = 151) had TMB data available; of these, 254 had PD-L1 and 236 had TcellTMB and the inflammatory biomarkers PD-L1 and Tcell

Background: 1L pembro + platinum-based chemo has shown clinical activity in metastatic NSCLC regardless of tissue tumor mutational burden (tTMB) status. bTMB assessed using circulating tumor DNA in plasma may be a surrogate for tTMB. The single-arm, phase 2 KEYNOTE-782 study (NCT03664024) evaluated the correlation of bTMB with efficacy of 1L pembro + chemo in nonsquamous NSCLC. Methods: Eligible patients (pts) had histologically or cytologically confirmed stage IV nonsquamous NSCLC with measurable disease per RECIST v1.1 and were not eligible for EGFR-, BRAF-, ROS1-, or ALK-directed therapy. Pts were not previously treated for advanced or metastatic disease, had an ECOG PS of 0 or 1, and had an evaluable biopsy sample for biomarker analysis. All pts received intravenous pembro 200 mg Q3W + platinum chemo doublet Q3W (pemetrexed 500 mg/m2 + 4 cycles of carboplatin AUC 5 mg/mL/min or cisplatin 75 mg/m2). The primary objective was to evaluate the association of baseline bTMB with ORR per RECIST v1.1 by investigator assessment. Secondary objectives were to determine safety and the association of baseline bTMB with PFS per RECIST v1.1 by investigator assessment and OS. A study-specific next-generation sequencing-based assay using a 1.9 Mb/654-gene cancer panel, which includes specific lung cancer-associated gene targets, was used to measure bTMB (continuous scale) in cell-free DNA extracted from baseline plasma samples. Paired white blood cell DNA sequencing was also performed to eliminate potential clonal hematopoiesis-derived somatic mutations. Database cutoff: November 5, 2021. Results: 117 pts were enrolled; median age was 64.0 years (range, 37-85), and the majority were male (60.7%), had an ECOG PS of 1 (69.2%), and no brain metastases at baseline (92.3%). Median time from first dose to data cutoff was 19.3 months (range, 1.0-35.5). ORR was 40.2% (95% CI, 31.2-49.6; 6 CRs, 41 PRs), median PFS was 7.2 months (95% CI, 5.6-9.8), and median OS was 18.1 months (95% CI, 13.5-25.6). Treatment-related adverse events (TRAEs) occurred in 113 pts (96.6%) and grade 3-5 TRAEs occurred in 56 (47.9%). Eight pts (6.8%) died due to a TRAE (febrile neutropenia and pneumonitis [n = 2 each], and septic shock, pulmonary sepsis, general physical health deterioration, and neutropenia [n = 1 each]). bTMB data were available for 101 pts. The area under the receiver operating curve for bTMB as a continuous variable to discriminate response was 0.47 (95% CI, 0.36-0.59). The posterior probabilities for a positive association of bTMB with PFS and OS were 16.8% and 7.8%, respectively. Conclusions: Baseline bTMB showed no evidence of an association with ORR, PFS, or OS in pts with nonsquamous NSCLC treated with 1L pembro + chemo. These findings indicate no clinical utility of bTMB in this patient population and treatment setting. No new safety signals were observed. Citation Format: Jair Bar, Emilio Esteban, Delvys Rodríguez-Abreu, Santiago Ponce Aix, Zsuzsanna Szalai, Enriqueta Felip, Maya Gottfried, Mariano Provencio Pulla, Andrew Robinson, Andrea Fülöp, Suman B. Rao, D. Ross Camidge, Giovanna Speranza, Steven M. Townson, Julie Kobie, Mark Ayers, Elisha J. Dettman, Robert McDaniel, Byoungsok Jung, David Burkhardt, Ruth Mauntz, Tibor Csőszi. Response to first-line (1L) pembrolizumab (pembro) + chemotherapy (chemo) in non-small cell lung cancer (NSCLC) by blood tumor mutational burden (bTMB): the phase 2 KEYNOTE-782 trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 2 (Clinical Trials and Late-Breaking Research); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(8_Suppl):Abstract nr CT216.

BackgroundTo characterize genomic determinants of response to pembrolizumab in recurrent/metastatic (R/M) head and neck squamous cell carcinoma (HNSCC) in the KEYNOTE-012 study.MethodsAssociations between biomarkers (tumor mutational burden (TMB), neoantigen load (NL), 18-gene T-cell-inflamed gene expression profile (TcellinfGEP), and PD-L1 combined positive score (CPS)) and clinical outcomes with pembrolizumab were assessed in patients with R/M HNSCC (n=192). Tumor human papillomavirus (HPV) status was also evaluated with the use of p16 immunohistochemistry and whole exome sequencing (WES; HPV+, mapping >20 HPV reads) in pretreatment tumor samples (n=106).ResultsTMB, clonality-weighted TMB, and TcellinfGEP were significantly associated with objective response (p=0.0276, p=0.0201, and p=0.006, respectively), and a positive trend was observed between NL and PD-L1 CPS and clinical response (p=0.0550 and p=0.0682, respectively). No correlation was observed between TMB and TcellinfGEP (Spearman ρ=–0.026) or TMB and PD-L1 (Spearman ρ=0.009); a correlation was observed between TcellinfGEP and PD-L1 (Spearman ρ=0.511). HPV status by WES and p16 immunohistochemistry showed concordance (84% ҡ=0.573) among patients whose HPV results were available using both methods.ConclusionsTMB and inflammatory biomarkers (TcellinfGEP and PD-L1) may represent distinct and complementary biomarkers predicting response to anti-programmed death 1 therapies in HNSCC; further study of these relationships in randomized clinical trials is needed.Trial registration numberNCT01848834.

BackgroundPembrolizumab is a standard-of-care first-line treatment for advanced/metastatic NSCLC, either as monotherapy (for patients with PD-L1 tumor proportion score [TPS] ≥1%) or combined with platinum chemotherapy. An improved OS benefit has been demonstrated for both pembrolizumab monotherapy and pembrolizumab plus chemotherapy in patients with higher tumor PD-L1 expression, and for pembrolizumab monotherapy in patients with higher tissue tumor mutation burden (tTMB). Mutations in KRAS occur relatively frequently in patients with nonsquamous NSCLC but infrequently in those with squamous NSCLC; most mutations are in codon 12. Notably, the pembrolizumab OS treatment effect was not diminished in patients with KRAS G12C mutations in phase 3 studies evaluating pembrolizumab monotherapy and pembrolizumab in combination with chemotherapy.1 2 Herein we describe prevalence of KRAS mutations among patients with advanced nonsquamous NSCLC from two phase 3 clinical studies evaluating first-line pembrolizumab (KEYNOTE-042 and KEYNOTE-189) and the relationship of such mutations with select patient characteristics.MethodsKEYNOTE-042 (NCT02220894) evaluated pembrolizumab versus platinum-based chemotherapy for advanced PD-L1–positive NSCLC (any histology) without EGFR/ALK alterations. KEYNOTE-189 (NCT02578680) evaluated pembrolizumab plus platinum-based chemotherapy versus platinum-based chemotherapy alone for metastatic nonsquamous NSCLC without EGFR/ALK alterations irrespective of tumor PD-L1 expression. Whole-exome sequencing of tumor tissue and matched normal DNA (blood) was performed for patients with nonsquamous histology. PD-L1 TPS was evaluated using the PD-L1 IHC 22C3 pharmDx assay (Agilent Technologies, Carpinteria, CA, USA). Prevalence of KRAS mutations and their relationships with TMB, PD-L1 TPS, and smoking status were analyzed descriptively.Results590 patients with nonsquamous NSCLC were included in these analyses (KEYNOTE-042, n=301; KEYNOTE-189, n=289). Overall, 42.9% of patients had tTMB ≥175 mut/exome, 81.4% were current/former smokers and, 40.3%, 42.7%, and 16.9% had PD-L1 TPS ≥50%, 1–49% and KRAS G12C, G12D, and G12V mutations occurred in 11.0%, 4.1%, and 5.4% of patients, respectively. Prevalence of KRAS mutations by patient characteristics is summarized in the table (table 1). KRAS G12C mutations occurred almost exclusively in current/former smokers. KRAS G12C was enriched in tumors with tTMB ≥175 mut/exome and tumors with PD-L1 TPS ≥50%. Prevalence was highest in tumors with both tTMB ≥175 mut/exome and PD-L1 TPS ≥50%.Abstract 364 Table 1KRAS Mutation PrevalenceConclusionsKRAS G12C mutations occurred with moderate frequency in patients with nonsquamous NSCLC, with most occurring in current/former smokers. KRAS G12C mutations occurred at higher frequency in patient subgroups defined by higher tTMB and PD-L1 TPS.AcknowledgementsMedical writing assistance was provided by Christabel Wilson, MSc, of ICON plc (North Wales, PA, USA), funded by Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA.Trial RegistrationKEYNOTE-042, ClinicalTrials.gov, NCT02220894; KEYNOTE-189, ClinicalTrials.gov, NCT02578680ReferencesGadgeel S, Rodriguez-Abreu D, Felip E, et al. KRAS mutational status and efficacy in KEYNOTE-189: pembrolizumab (pembro) plus chemotherapy (chemo) vs placebo plus chemo as first-line therapy for metastatic non-squamous NSCLC. Ann Oncol 2019;30(suppl 11):xi64-xi5.Herbst RS, Lopes G, Kowalski DM, et al. Association of KRAS mutational status with response to pembrolizumab monotherapy given as first-line therapy for PD-L1-positive advanced non-squamous NSCLC in KEYNOTE-042. Ann Oncol 2019;30(suppl 11):xi63-xi4.Ethics ApprovalFor both trials, the protocol and all amendments were approved by the appropriate ethics committee at each center, the study was conducted in accordance with the standards of Good Clinical Practice. Patients provided written informed consent before enrollment.

Purpose: To explore relationships between biological gene expression signatures and pembrolizumab response. Experimental Design: RNA-sequencing data on baseline tumor tissue from 1,188 patients across seven tumor types treated with pembrolizumab monotherapy in nine clinical trials were used. A total of 11 prespecified gene expression signatures [18-gene T-cell–inflamed gene expression profile (TcellinfGEP), angiogenesis, hypoxia, glycolysis, proliferation, MYC, RAS, granulocytic myeloid-derived suppressor cell (gMDSC), monocytic myeloid-derived suppressor cell (mMDSC), stroma/epithelial-to-mesenchymal transition (EMT)/TGFβ, and WNT] were evaluated for their relationship to objective response rate (per RECIST, version 1.1). Logistic regression analysis of response for consensus signatures was adjusted for tumor type, Eastern Cooperative Oncology Group performance status, and TcellinfGEP, an approach equivalent to evaluating the association between response and the residuals of consensus signatures after detrending them for their relationship with the TcellinfGEP (previously identified as a determinant of pembrolizumab response) and tumor type. Testing of the 10 prespecified non-TcellinfGEP consensus signatures for negative association [except proliferation (hypothesized positive association)] with response was adjusted for multiplicity. Results: Covariance patterns of the 11 signatures (including TcellinfGEP) identified in Merck–Moffitt and The Cancer Genome Atlas datasets showed highly concordant coexpression patterns in the RNA-sequencing data from pembrolizumab trials. TcellinfGEP was positively associated with response; signatures for angiogenesis, mMDSC, and stroma/EMT/TGFβ were negatively associated with response to pembrolizumab monotherapy. Conclusions: These findings suggest that features beyond IFNγ-related T-cell inflammation may be relevant to anti–programmed death 1 monotherapy response and may define other axes of tumor biology as candidates for pembrolizumab combinations. See related commentary by Cho et al., p. 1479

Purpose:Molecular profiling of large databases of human tumor gene expression profiles offers novel opportunities for informing decisions in clinical development programs.Experimental Design:Gene expression profile of programmed death ligand 1 (PD-L1) was explored in a dataset of 16,000 samples, including approximately 4,000 metastatic tumors, across >25 tumor types prevalent in the United States, looking for new indications for the programmed death 1 (PD-1) inhibitor pembrolizumab. PD-L1 expression was highly concordant with several genomic signatures indicative of immune-inflamed tumor microenvironment. Prevalence of activated immune-inflamed tumors across all tumor types was explored and used to rank tumor types for potential response to pembrolizumab monotherapy.Results:The analysis yielded 3 tiers of indications in which high levels of PD-L1 and immune-inflamed signatures were found in up to 40% to 60%, 20% to 40%, and 0% to 20% of tumors. Tier 1 contained novel indications known at the time of analysis to be responsive to PD-1 checkpoint blockade in the clinic (such as melanoma and non–small cell lung cancer), as well as indications not studied in the clinic previously, including microsatellite instability–high colorectal, head and neck, bladder, and triple-negative breast cancers. Complementary analysis of an Asian/Pacific cancer dataset (gastric cancer) revealed high prevalence of immune-inflamed tumors in gastric cancer. These data contributed to prioritization of these indications for clinical development of pembrolizumab as monotherapy.Conclusions:Data highlight the value of molecular profiling in identifying populations with high unmet needs with potentially favorable response characteristics and accelerating development of novel therapies for these patients.See related commentary by Mansfield and Jen, p. 1443

PURPOSE Biomarkers that can predict response to anti–programmed cell death 1 (PD-1) therapy across multiple tumor types include a T-cell–inflamed gene-expression profile (GEP), programmed death ligand 1 (PD-L1) expression, and tumor mutational burden (TMB). Associations between these biomarkers and the clinical efficacy of pembrolizumab were evaluated in a clinical trial that encompassed 20 cohorts of patients with advanced solid tumors. METHODS KEYNOTE-028 ( ClinicalTrials.gov identifier: NCT02054806) is a nonrandomized, phase Ib trial that enrolled 475 patients with PD-L1–positive advanced solid tumors who were treated with pembrolizumab 10 mg/kg every 2 weeks for 2 years or until confirmed disease progression or unacceptable toxicity occurred. The primary end point was objective response rate (ORR; by RECIST v1.1, investigator review). Secondary end points included safety, progression-free survival (PFS), and overall survival (OS). Relationships between T-cell–inflamed GEP, PD-L1 expression, and TMB and antitumor activity were exploratory end points. RESULTS ORRs (with 95% CIs) ranged from 0% (0.0% to 14.2%) in pancreatic cancer to 33% (15.6% to 55.3%) in small-cell lung cancer. Across cohorts, median (95% CI) PFS ranged from 1.7 months (1.5 to 2.9 months) to 6.8 months (1.9 to 14.1 months) in pancreatic and thyroid cancers, respectively, and median OS from 3.9 months (2.8 to 5.5 months) to 21.1 months (9.1 to 22.4 months) in vulvar and carcinoid tumors, respectively. Higher response rates and longer PFS were demonstrated in tumors with higher T-cell–inflamed GEP, PD-L1 expression, and/or TMB. Correlations of TMB with GEP and PD-L1 were low. Response patterns indicate that patients with tumors that had high levels of both TMB and inflammatory markers (GEP or PD-L1) represent a population with the highest likelihood of response. Safety was similar and consistent with prior pembrolizumab reports. CONCLUSION A T-cell–-inflamed GEP, PD-L1 expression, and TMB predicted response to pembrolizumab in multiple tumor types. These biomarkers (alone/in combination) may help identify patients who have a higher likelihood of response to anti–PD-1 therapies across a broad spectrum of cancers.

Background Key biomarkers under investigation for the ability to predict response to monotherapy PD-1 inhibitors such as pembrolizumab include PD-L1, TMB, MSI, and T-cell–inflamed gene expression profile (GEP). The KEYNOTE-180 trial (NCT02559687) was a single-arm phase 2 study of pembrolizumab as third-line or greater therapy in advanced/metastatic esophageal/gastroesophageal junction adenocarcinoma or squamous cell carcinoma (SCC). ORR was 9.9% and median DOR was NR at the primary analysis. We investigated the relationship in KEYNOTE-180 between response to pembrolizumab and T-cell–inflamed GEP or PD-L1 expression by histology. Methods Patients received pembrolizumab 200 mg Q3W for ≤2 years until disease progression, toxicity, or withdrawal. The end points for this analysis were ORR, DOR, and PFS per RECIST v1.1 by central review and OS in the SCC and adenocarcinoma populations by GEP (non-low [≥–1.540] or low [ Results Of 121 total patients, 118 had an evaluable GEP score and 121 had an evaluable PD-L1 CPS. Fifty-one patients (42.1%) had GEPnon-lowtumors, 58 (48.0%) had CPS ≥10 tumors, and 31 (25.6%) had GEPnon-low/CPS ≥10 tumors; 63 patients (52.1%) had SCC and 58 (47.9%) had adenocarcinoma. ORR was 15.4% with GEPnon-low and 13.5% with GEPlow among patients with SCC and 12% and 0% among patients with adenocarcinoma, respectively (table 1). ORR was 20% with CPS ≥10 and 7.1% with CPS Conclusions In KEYNOTE-180, data in a small number of patients suggested that measures of inflammation, like PD-L1 and GEP, may enrich for responses to pembrolizumab. In SCC, some trends toward enrichment were observed for both biomarkers, although the trend was stronger for PD-L1 CPS ≥10. In adenocarcinoma, a trend was observed for GEP but not for PD-L1; the small number of responders is limiting, and further studies are needed to understand molecular correlates in adenocarcinoma. Acknowledgements Medical writing and/or editorial assistance was provided by Tim Peoples, MA, ELS, and Holly C. Cappelli, PhD, CMPP, of the ApotheCom pembrolizumab team (Yardley, PA, USA). This assistance was funded by Merck Sharp & Dohme Corp, a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA. Trial Registration ClinicalTrials. gov, NCT02559687 Ethics Approval The study and the protocol were approved by the institutional review board or ethics committee at each site. Consent All patients provided written informed consent to participate in the clinical trial.

Programmed death-1-directed (PD-1-directed) immune checkpoint blockade results in durable antitumor activity in many advanced malignancies. Recent studies suggest that IFN-γ is a critical driver of programmed death ligand-1 (PD-L1) expression in cancer and host cells, and baseline intratumoral T cell infiltration may improve response likelihood to anti-PD-1 therapies, including pembrolizumab. However, whether quantifying T cell-inflamed microenvironment is a useful pan-tumor determinant of PD-1-directed therapy response has not been rigorously evaluated. Here, we analyzed gene expression profiles (GEPs) using RNA from baseline tumor samples of pembrolizumab-treated patients. We identified immune-related signatures correlating with clinical benefit using a learn-and-confirm paradigm based on data from different clinical studies of pembrolizumab, starting with a small pilot of 19 melanoma patients and eventually defining a pan-tumor T cell-inflamed GEP in 220 patients with 9 cancers. Predictive value was independently confirmed and compared with that of PD-L1 immunohistochemistry in 96 patients with head and neck squamous cell carcinoma. The T cell-inflamed GEP contained IFN-γ-responsive genes related to antigen presentation, chemokine expression, cytotoxic activity, and adaptive immune resistance, and these features were necessary, but not always sufficient, for clinical benefit. The T cell-inflamed GEP has been developed into a clinical-grade assay that is currently being evaluated in ongoing pembrolizumab trials.

INTRODUCTION: Immunotherapy targeting the programmed cell death protein–1 (PD-1) axis elicits durable antitumor responses in multiple cancer types. However, clinical responses vary, and biomarkers predictive of response may help to identify patients who will derive the greatest therapeutic benefit. Clinically validated biomarkers predictive of response to the anti–PD-1 monoclonal antibody pembrolizumab include PD-1 ligand 1 (PD-L1) expression in specific cancers and high microsatellite instability (MSI-H) regardless of tumor type. Tumor mutational burden (TMB) and T cell–inflamed gene expression profile (GEP) are emerging predictive biomarkers for pembrolizumab. Both PD-L1 and GEP are inflammatory biomarkers indicative of a T cell–inflamed tumor microenvironment (TME), whereas TMB and MSI-H are indirect measures of tumor antigenicity generated by somatic tumor mutations. However, the relationship between these two categories of biomarkers is not well characterized. RATIONALE: This study assessed the potential for TMB and a T cell–inflamed GEP to jointly predict clinical response to pembrolizumab in >300 patient samples with advanced solid tumors and melanoma across 22 tumor types from four KEYNOTE clinical trials. To assess the individual and joint clinical utility of TMB and GEP, patients were stratified in four biomarker–defined clinical response groups [GEP low and TMB low (GEP(lo) TMB(lo)), GEP low and TMB high (GEP(lo) TMB(hi)), GEP(hi) TMB(lo), and GEP(hi) TMB(hi)] based on predefined cutoffs for TMB and GEP. These patient–defined biomarker groups were further used to guide transcriptome and exome analyses of tumors in a large molecular database [The Cancer Genome Atlas (TCGA)] (n = 6384 tumors) to identify targetable patterns of biology that may modulate response and resistance. RESULTS: TMB and GEP exhibited only modest correlation and were independently predictive of response across the KEYNOTE clinical datasets. We found that objective response rates were strongest in patients with GEP(hi) TMB(hi) (37 to 57%), moderate in those with GEP(hi) TMB(lo) (12 to 35%) and GEP(lo) TMB(hi) (11 to 42%), and reduced or absent in those with GEP(lo) TMB(lo) (0 to 9%) (see the figure). Additionally, longer progression–free survival times were seen in patients with higher levels of both TMB and GEP. Findings were comparable when TMB and PD-L1 expression were jointly assessed. Within TCGA database, GEP and TMB again had a low correlation, demonstrating the potential to jointly stratify transcriptomic and genomic features across cancer types. Specific gene expression patterns reflective of TME biology showed significant associations with TMB, GEP, or both. In particular, gene set enrichment analysis identified proliferative and stromal, myeloid, and vascular biology corresponding to specific TMB-defined subgroups within GEP(hi) tumors. In TMB(hi) tumors, indication-dependent somatic DNA alterations in key cancer driver genes showed a strong negative association with GEP. CONCLUSION: This analysis shows that TMB and inflammatory biomarkers (T cell–inflamed GEP and PD-L1 expression) can jointly stratify human cancers into groups with different clinical responses to pembrolizumab monotherapy and identify patterns of underlying, targetable biology related to these groups. TMB and inflammatory biomarkers independently predict response and may capture distinct features of neoantigenicity and T cell activation, respectively. This approach may provide a precision medicine framework for rationally constructing and evaluating anti–PD-1– and/or –PD-L1–based combination therapy regimens.

e20593 Background: The receptor-ligand interaction between programmed cell death protein-1 (PD-1) and its ligands (PD-L1 and PD-L2) plays a critical role in tumor immune evasion and has become a therapeutic target in a number of cancer types, including extensive stage small cell lung cancer (ES-SCLC). While there are some published data on PD-L1 expression in ES-SCLC, the results vary with different laboratory assays or thresholds of positivity employed. Data on PD-L2 expression is very limited in ES-SCLC. The prognostic values of these biomarkers are not well understood. The current study aims to address this data gap. Methods: A retrospective cohort study of patients with ES-SCLC receiving usual care in the clinical setting in Denmark was conducted. Formalin-fixed paraffin embedded (FFPE) tumor tissue samples and data on patient characteristics and clinical outcomes were obtained for the patient population. Protein expression of PD-L1 and PD-L2 was determined by immunohistochemistry (IHC), and a combined positive score (CPS, the percentage of tumor cells and mononuclear inflammatory cells expressing the biomarker at any intensity out of the total number of tumor cells) of ≥1 was used to define biomarker positivity. Kaplan-Meier plots and Cox proportional hazard models were employed to assess the relationship between PD-L1 and PD-L2 protein expression and overall survival (OS). Results: Among the 80 patients with ES-SCLC included in this study, 25 and 29 were positive for PD-L1 and PD-L2, respectively (Table). A total of 68 patients had concordant expression status of PD-L1 and PD-L2 (21 as double positive and 47 as double negative). There was a significant association between PD-L1 positivity and longer OS, which persisted after adjustment for age and stage at SCLC diagnosis, performance score, prior exposure to chemotherapy, and lactate dehydrogenase levels (Table). Similar results were observed for PD-L2 (Table), and the data did not suggest any additional prognostic value of PD-L2 expression independent of PD-L1 expression. Conclusions: PD-L1 and PD-L2 positivity was both observed, and highly correlated with each other, in approximately one third of the ES-SCLC tumor samples tested. Patients with PD-L1 or PD-L2 positive tumors have significantly longer OS than those with negative tumors in this study, independent of other known prognostic factors.[Table: see text]

Limited data are available on the efficacy of anti-IGF-1R agents in KRAS mutant colorectal cancer (CRC). We analysed outcome of 69 chemorefractory, KRAS exon 2 mutant CRC patients who were enrolled in a double-blind, randomised, phase II/III study of irinotecan and cetuximab plus dalotuzumab 10 mg/kg once weekly (arm A), dalotuzumab 7.5 mg/kg every second week (arm B) or placebo (arm C). Objective response rate (5.6% vs. 3.1% vs. 4.8%), median progression-free survival (2.7 vs. 2.6 vs. 1.4 months) and overall survival (7.8 vs. 10.3 vs. 7.8 months) were not statistically significantly different between treatment groups. Most common grade ≥3 treatment-related toxicities included neutropenia, diarrhoea, hyperglycaemia, fatigue and dermatitis acneiform. Expression of IGF-1R, IGF-1, IGF-2 and EREG by quantitative real-time polymerase chain reaction was assessed in 351 patients from the same study with available data on KRAS exon 2 mutational status. Median cycle threshold values for all biomarkers were significantly lower (i.e., higher expression, p

Background: STK11 (also known as LKB1) and KEAP1 mutations have been associated with chemoresistance and poor outcomes and shown to be more frequent in PD-L1-negative tumors with high tumor mutational burden (TMB). We performed an exploratory analysis of KEYNOTE-042 (NCT02220894) to assess the prevalence of STK11 and KEAP1 mutations and their association with efficacy. Methods: STK11 and KEAP1 status and TMB were assessed by whole-exome sequencing (WES) in tumor tissue and matched-normal DNA. PD-L1 was assessed with the PD-L1 IHC 22C3 pharmDx assay. Descriptive analyses were performed to assess the correlation of STK11 and KEAP1 status with TMB and PD-L1 expression distributions and the association between STK11 and KEAP1 status and efficacy. Results: 429/1274 pts (34%) had evaluable WES data from tumor and normal DNA. STK11 and KEAP1 mutations were seen in 33 (8%) and 64 (15%) pts, respectively; 12 pts (3%) had both mutations. Pts with STK11 mutation tended to have a lower PD-L1 TPS than pts without (median [IQR] 15% [3-50] vs 40% [10-80]); TPS tended to be similar in pts with vs without KEAP1 mutation (40% [10-81] vs 40% [10-80]). TMB score tended to be higher in pts with vs without STK11 (median [IQR] 191 [104-272] vs 146 [72-253]) or KEAP1 (183 [114-283] vs 142 [68-252]) mutation. ORR, PFS, and OS with pembrolizumab were similar in pts with vs without STK11 or KEAP1 mutation; chemotherapy efficacy was lower in pts with vs without STK11 mutation (Table). Pembrolizumab was associated with better outcomes than chemotherapy regardless of STK11 or KEAP1 status (Table). 95% CIs were wide given the modest frequency of STK11 and KEAP1 mutations. Conclusions: The findings of this exploratory analysis suggest pembrolizumab monotherapy should be considered a standard first-line treatment option for advanced PD-L1-positive NSCLC regardless of STK11 or KEAP1 status. STK11KEAP1With MutationWithout MutationWith MutationWith MutationPembroChemoPembroChemoPembroChemoPembroChemo(n = 16)(n = 17)(n = 214)(n = 182)(n = 31)(n = 33)(n = 199)(n = 166)ORR, % (95% CI)31 (11-59)6 ( Citation Format: Byoung Chul Cho, Gilberto Lopes, Dariusz M. Kowalski, Kazuo Kasahara, Yi-Long Wu, Gilberto Castro, Hande Z. Turna, Razvan Cristescu, Deepti Aurora-Garg, Andrey Loboda, Jared Lunceford, Julie Kobie, Mark Ayers, M. Catherine Pietanza, Bilal Piperdi, Tony S. Mok. Relationship between STK11 and KEAP1 mutational status and efficacy in KEYNOTE-042: pembrolizumab monotherapy versus platinum-based chemotherapy as first-line therapy for PD-L1-positive advanced NSCLC [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr CT084.

Background: Mutations in the tumor suppressor genes STK11 (also known as LKB1) and KEAP1 have been associated with poorer clinical outcomes in patients (pts) with NSCLC. In an exploratory analysis, we assessed the prevalence of STK11 and KEAP1 mutations and their association with efficacy in KEYNOTE-189 (NCT02578680). Methods: STK11 and KEAP1 status and tumor mutational burden (TMB) were assessed by whole-exome sequencing (WES) in pts who had available tumor and matched-normal tissue. PD-L1 was assessed by the PD-L1 IHC 22C3 pharmDx assay. The association of STK11 and KEAP1 status with efficacy and their correlation with TMB and PD-L1 expression distributions were evaluated descriptively. Results: WES data from both tumor and normal DNA were evaluable for 289 (47%) of 616 pts, of whom 54 (19%) had an STK11 mutation and 68 (24%) had a KEAP1 mutation; 29 (10%) had both STK11 and KEAP1 mutations. PD-L1 TPS tended to be lower in pts with vs without STK11 mutation (median [IQR] 0% [0-16] vs 15% [0-75]), whereas TMB score tended to be higher in pts with mutation (209 [132-265] vs 146 [89-264]). Similar patterns were seen for pts with vs without KEAP1 mutation (PD-L1 TPS: 1% [0-13] vs 20% [0-75]; TMB: 173 [124-267] vs 147 [89-263]). Although ORR of pembrolizumab plus chemotherapy was lower and PFS and OS shorter in pts with vs without STK11 and KEAP1 mutation, pembrolizumab plus chemotherapy was associated with numerically better outcomes than placebo plus chemotherapy regardless of mutation status (Table). 95% CIs were wide given the modest mutation frequency and the 2:1 randomization in favor of pembrolizumab plus chemotherapy. Conclusions: Data from this exploratory analysis support use of pembrolizumab plus pemetrexed and platinum as standard first-line therapy for pts with metastatic nonsquamous NSCLC regardless of STK11 or KEAP1 status. STK11KEAP1With MutationWithout MutationWith MutationWithout MutationPembro + ChemoPlacebo + ChemoPembro + ChemoPlacebo + ChemoPembro + ChemoPlacebo + ChemoPembro + ChemoPlacebo + Chemo(n = 36)(n = 18)(n = 168)(n = 67)(n = 45)(n = 23)(n = 159)(n = 62)ORR, % (95% CI)31 (16-48)17 (4-41)49 (41-57)16 (8-27)36 (22-51)17 (5-39)48 (40-56)16 (8-28)PFS, median, mo (95% CI)6 (4-9)5 (5-9)10 (8-14)5 (5-5)5 (4-11)5 (5-9)10 (8-14)5 (5-5)PFS, HR (95% CI)0.81 (0.44-1.47)0.38 (0.27-0.52)0.65 (0.38-1.12)0.38 (0.28-0.53)OS, median, mo (95% CI)17 (5-NR)8 (7-NR)23 (20-NR)12 (8-25)13 (7-NR)9 (7-NR)24 (20-NR)12 (8-NR)OS, HR (95% CI)0.75 (0.37-1.50)0.59 (0.41-0.85)0.81 (0.44-1.49)0.57 (0.39-0.84) Citation Format: Shirish M. Gadgeel, Delvys Rodriguez-Abreu, Enriqueta Felip, Emilio Esteban, Giovanna Speranza, Martin Reck, Rina Hui, Michael Boyer, Edward B. Garon, Hidehito Horinouchi, Razvan Cristescu, Deepti Aurora-Garg, Andrey Loboda, Jared Lunceford, Julie Kobie, Mark Ayers, Bilal Piperdi, M. Catherine Pietanza, Marina C. Garassino. Pembrolizumab plus pemetrexed and platinum vs placebo plus pemetrexed and platinum as first-line therapy for metastatic nonsquamous NSCLC: analysis of KEYNOTE-189 by STK11 and KEAP1 status [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr LB-397.

9521 Background: In a previous analysis of KEYNOTE-189 (NCT02578680), we showed that tissue TMB (tTMB) assessed by whole-exome sequencing was not significantly associated with efficacy in either arm and that pembro + chemo improved outcomes vs placebo + chemo in both the tTMB ≥175 and tTMB < 175 mut/exome subgroups. Here, we explored the association of bTMB with efficacy in KEYNOTE-189. Methods: 616 patients (pts) were randomized 2:1 to pembro + chemo or placebo + chemo. bTMB was assessed in cfDNA using the Guardant Health Omni assay. Association of bTMB (continuous square root transformed) with outcomes in each arm was assessed using Cox proportional hazards models (OS, PFS) and logistic regression (ORR) adjusted for ECOG PS; statistical significance was determined at the 0.05 level without multiplicity adjustment. The clinical utility of bTMB on outcomes was assessed using the cutoff that most closely approximated the 175 mut/exome tTMB cutoff as determined by AUROC analysis. Data cutoff was 21 Sep 2018. Results: 235 (38%) treated pts had evaluable tTMB and bTMB: 160 in the pembro + chemo arm and 75 in the placebo + chemo arm. bTMB as a continuous variable was not significantly associated with OS or ORR for pembro + chemo (one-sided P = .229 and .051) or placebo + chemo (two-sided P = .641 and .069); bTMB was significantly associated with PFS in the pembro + chemo arm (one-sided P = .015) but not the placebo + chemo arm (two-sided P = .058). bTMB and tTMB scores were moderately correlated (r = .61). The bTMB cutoff that most closely approximated tTMB 175 mut/exome was 15 mut/Mb (AUROC 0.81, 95% CI 0.75-0.86). 178 (76%) pts had concordant bTMB and tTMB results—101 low and 77 high by both—whereas 57 (24%) had discordant results—21 high bTMB but low tTMB, 36 low bTMB but high tTMB. Pembro + chemo improved OS, PFS, and ORR vs placebo + chemo for bTMB ≥15 and < 15 mut/exome (Table). Conclusions: Similar to previous findings based on tTMB, bTMB has limited clinical utility in the setting of pembro with pemetrexed and platinum given as first-line therapy for metastatic nonsquamous NSCLC. Clinical trial information: NCT02578680 . [Table: see text]

Mining immunotherapy clinical trials Clinical trial data can provide a wealth of information about how drugs work. Yet such information often belongs to pharmaceutical companies and is rarely accessible to the scientific community at large. Cristescu et al. provide exploratory analysis of a cancer genomics dataset, collected from four separate clinical trials of Merck's PD-1 immunotherapy drug, pembrolizumab. This informative public resource examines more than 300 patient samples representing 22 different tumor types. Two widely used signatures that currently predict immunotherapy response are tumor mutational burden and a “hot” T cell–inflamed microenvironment. The study analyzed these two proposed biomarkers in combination to see what predictive clinical utility they may hold. Science , this issue p. eaar3593

Purpose We built an integrated analytic pipeline to robustly and comprehensively profile molecular features of pembrolizumab-treated tumors using whole exome sequencing (WES) data from clinical trials across different indications. Experimental design We implemented a computational framework for analysis of WES data generated by different sequencing vendors for 1467 samples from 8 major tumor types. Reads were first processed by BWA-MEM, Picard, and GATK (v2) to generate analysis-ready BAM files. Quality controls (QC) before downstream analysis included Y/X reads ratio for matched normal sample as patient sex prediction and tumor/normal concordance and contamination estimation by Conpair. Key molecular features were evaluated, including tumor mutation burden (TMB) by MuTect (v1) and VEP; HLA-I typing by OptiType; neoantigen burden by NetMHC (v3.4); mutation signature by deconstructSigs; allele-specific copy number by VarScan2 and Sequenza; clonality by PyClone; presence of oncogenic viruses (eg, EBV, HBV, HPV); and LOH score indicating homologous recombination deficiency (HRD-LOH). Results Concordance rate between predicted and clinical sex was 1446/1464 (98.8%), and 1420/1467 (96.8%) samples passed tumor-normal alignment QC. Highest TMB (median [range]) was detected in melanoma (245 [2-6246]) and urothelial carcinoma (UC) (124 [4-1579]), with lowest TMB in metastatic castration-resistant prostate cancer (52 [1-6143]) and ovarian cancer (OvCa) (49.5 [8-272]). The correlation of median TMB across 7 cancer types in our data and TCGA was Spearman R = 0.957. Within indications, there was no difference in TMB distribution for sequencing data originating from different sequencing vendors and TCGA data, which demonstrated concordance across data sets and robust TMB calling by our integrated pipeline. Neoantigen burden strongly correlated with TMB (N = 1420; Spearman R = 0.890). HPV was detected in 20/129 (15.5%) head and neck squamous cell carcinomas (HNSCC) and 4/6 (66.7%) anal cancers; EBV was detected in 8/129 (6.2%) HNSCC and 18/318 (5.7%) gastric cancers (GC). The dominant mutation signatures by disease included APOBEC for UC (135/236) and HNSCC (22/122), alcohol for HCC (19/35), HRD for OvCa (12/64) and triple-negative breast cancer (51/175), UV exposure for melanoma (145/176), and dMMR for GC (84/287) and CRPC (16/155). Samples with deleterious BRCA mutations showed significantly higher HRD-LOH score (N = 1420; AUROC = 0.61 [95% CI, 0.53-0.69]) and HRD mutational signature (N = 1316; DOR = 6.4 [95% CI, 3.7-11.1]). Conclusion We assembled heterogeneous computational modules into an integrated pipeline to reliably profile diverse molecular features from WES data of nearly 1500 clinical samples across different tumor types. These data serve as a foundation for translational research efforts supporting pembrolizumab development. Citation Format: Xiaoqiao Liu, Xinwei Sher, Hongchao Lu, Jun Zhuang, Weilong Zhao, Andrew Albright, Cinthia Umemoto, Christen Wudarski, Maureen Lane, Mark Ayers, Andrea L. Webber, Sandra C. Souza, Ping Qiu, Diane Levitan, Jennifer Cho, Deepti Aurora-Garg, Matthew Marton, Alexandra Snyder, Michael Morrissey, Andrey Loboda, Ronghua Chen, Razvan Cristescu. An integrated bioinformatics pipeline for profiling cancer-immune interaction from whole exome sequencing of pembrolizumab clinical samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2483.

BackgroundInfants with critical congenital heart disease who require cardiothoracic surgical intervention may have significant post-operative mortality and morbidity. Infants who are small for gestational age MethodsA single-institution retrospective review was performed in 230 infants with congenital heart disease who had cardiothoracic surgical intervention ResultsThere were 230 infants, 57 (23.3%) small for gestational age and 173 (70.6%) appropriate for gestational age. No significant difference was noted in pre-operative markers – gestational age, age at surgery, corrected gestational age, Society for Thoracic Surgeons and European Association for Cardiothoracic Surgery mortality score; or post-operative factors – length of stay, ventilation days, arrhythmias, need for extracorporeal membrane oxygenation, vocal cord dysfunction, hearing loss; or end-organ dysfunction – gastro-intestinal, renal, central nervous system, or genetic. Small for gestational age infants were more likely to have failed vision tests (p = 0.006). Small for gestational age infants were more likely to have increased 30-day (p = 0.005) and discharge mortality (p = 0.035). Small for gestational age infants with normal birth weight (>2500 g) were also at increased risk of 30-day mortality compared with appropriate for gestational age infants (p = 0.045).ConclusionsSmall for gestational age infants with congenital heart disease who undergo cardiothoracic surgery

Limited data are available on the efficacy of anti-IGF-1R agents in KRAS mutant colorectal cancer (CRC). We analysed the outcome of 69 chemorefractory, KRAS exon 2 mutant CRC patients who were enrolled in a double-blind, randomised, phase II/III study of irinotecan and cetuximab plus dalotuzumab 10 mg/kg once weekly (arm A), dalotuzumab 7.5 mg/kg every second week (arm B) or placebo (arm C). Objective response rate (5.6% vs. 3.1% vs. 4.8%), median progression-free survival (2.7 vs. 2.6 vs. 1.4 months) and overall survival (7.8 vs. 10.3 vs. 7.8 months) were not statistically significantly different between treatment groups. Most common grade ≥3 treatment-related toxicities included neutropenia, diarrhoea, hyperglycaemia, fatigue and dermatitis acneiform. Expression of IGF-1R, IGF-1, IGF-2 and EREG by quantitative real-time polymerase chain reaction was assessed in 351 patients from the same study with available data on KRAS exon 2 mutational status. Median cycle threshold values for all biomarkers were significantly lower (i.e., higher expression, p 0.05) among patients with KRAS wild-type compared to those with KRAS exon 2 mutant tumours. No significant changes were found according to location of the primary tumour with only a trend towards lower expression of IGF-1 in colon compared to rectal cancers (p = 0.06). Albeit limited by the small sample size, this study does not appear to support a potential role for anti-IGF-1R agents in KRAS exon 2 mutant CRC. Data on IGF-1R, IGF-1 and IGF-2 expression here reported may be useful for patient stratification in future trials with inhibitors of the IGF pathway.

Background Robatumumab (19D12; MK-7454 otherwise known as SCH717454) is a fully human antibody that binds to and inhibits insulin-like growth factor receptor-1 (IGF-1R). This multiinstitutional study (P04720) determined the safety and clinical efficacy of robatumumab in three separate patient groups with resectable osteosarcoma metastases (Group 1), unresectable osteosarcoma metastases (Group 2), and Ewing sarcoma metastases (Group 3). Procedure Robatumumab infusions were administered every 2 weeks and were well tolerated with minimal toxicity. Centrally reviewed response data were available for 144 patients. Results Low disease burden was important for osteosarcoma response: three of 31 patients had complete response or partial response (PR) by Response Evaluation Criteria in Solid Tumors (RECIST) in resectable patients (Group 1) versus zero of 29 in unresectable patients (Group 2); median overall survival was 20 months in Group 1 versus 8.2 months in Group 2. In centrally reviewed patients with Ewing sarcoma with PET-CT data (N = 84/115), there were six PR, 23 stable disease, and 55 progression of disease by RECIST at 2 months. Patients with Ewing sarcoma had a median overall survival of 6.9 months. However, responding patients with Ewing sarcoma were allowed to continue on treatment after study closure. A minority of patients with metastatic Ewing sarcoma showed clinical responses and have remained healthy after receiving 25–115 doses of robatumumab with remissions of >4 years duration (N = 6). Conclusions These findings show that although the IGF-1R remains an attractive treatment target, additional research is needed to identify responders and/or means to achieve durable remissions in order to successfully exploit IGF-1R signal blockade in Ewing sarcoma (clinicaltrials.gov: NCT00617890).

25 Background: Somatic tumor mutational burden (TMB) and a T-cell inflamed gene expression profile (GEP) predict response to anti-PD-1/PD-L1 immunotherapies in multiple tumor types. We assessed the potential for GEP and TMB to jointly predict clinical response to pembrolizumab and to identify distinct, targetable patterns of biology that may modulate response/resistance. Methods: To assess the individual and joint clinical utility of TMB and GEP in a pan-tumor context, pembrolizumab-treated patients with advanced solid tumors and melanoma were stratified as 4 biomarker-defined clinical response groups (GEP low/TMB low, GEP low/TMB high, GEP high/TMB low, GEP high/TMB high; N > 300) based on cutoffs for TMB (ROC Youden Index associated) and GEP (selected via analysis of pan cancer data). TMB and GEP were used to guide transcriptome and exome analysis of tumors in 2 large databases (Moffitt, n = 2944; TCGA, n = 6978). Results: TMB and GEP had a low, but significant, correlation in these clinical datasets. ORR was highest in GEP high/TMB high (37-57%), modest in GEP high/TMB low (12-35%) and GEP low/TMB high (11-42%), and lowest in GEP low/TMB low (0-9%) groups. Within the Moffitt and TCGA databases, GEP and TMB again had a low correlation, demonstrating their potential joint utility for stratifying additional transcriptomic and genomic features of these datasets. Specific gene modules showed strong positive or negative and highly statistically significant associations with TMB, GEP or both in each dataset, and patterns were consistent between datasets. In particular, gene set enrichment analysis identified proliferative, stromal and vascular biology corresponding to specific TMB-defined subgroups within GEP high tumors. In TMB-high tumors, indication-dependent somatic DNA alterations in key cancer driver genes showed a strong negative association ( P< 1e-5) with GEP. Conclusions: This analysis shows that TMB and T-cell inflamed GEP score can stratify human cancers into groups with different response rates to pembrolizumab monotherapy, and identify patterns of underlying, targetable biology related to these groups. This approach may provide a precision medicine framework for evaluating anti-PD-1/L1-based combination therapy regimens. Clinical trial information: NCT01848834; NCT02054806; NCT01295827; NCT01866319.

Purpose: Hepatocellular carcinoma (HCC) is the fifth most common primary neoplasm; surgery is the only curative option but 5-year survival rates are only 25% to 50%. Vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF) are known to be involved in growth and neovascularization of HCC. Therefore, agents that target these pathways may be effective in the treatment of HCC. The aim of this study was to determine the antineoplastic activity of brivanib alaninate, a dual inhibitor of VEGF receptor (VEGFR) and FGF receptor (FGFR) signaling pathways. Experimental Design: Six different s.c. patient-derived HCC xenografts were implanted into mice. Tumor growth was evaluated in mice treated with brivanib compared with control. The effects of brivanib on apoptosis and cell proliferation were evaluated by immunohistochemistry. The SK-HEP1 and HepG2 cells were used to investigate the effects of brivanib on the VEGFR-2 and FGFR-1 signaling pathways in vitro. Western blotting was used to determine changes in proteins in these xenografts and cell lines. Results: Brivanib significantly suppressed tumor growth in five of six xenograft lines. Furthermore, brivanib–induced growth inhibition was associated with a decrease in phosphorylated VEGFR-2 at Tyr1054/1059, increased apoptosis, reduced microvessel density, inhibition of cell proliferation, and down-regulation of cell cycle regulators. The levels of FGFR-1 and FGFR-2 expression in these xenograft lines were positively correlated with its sensitivity to brivanib-induced growth inhibition. In VEGF-stimulated and basic FGF stimulated SK-HEP1 cells, brivanib significantly inhibited VEGFR-2, FGFR-1, extracellular signal-regulated kinase 1/2, and Akt phosphorylation. Conclusion: This study provides a strong rationale for clinical investigation of brivanib in patients with HCC.

Collagen IV is a major component of the vascular basement membrane and may be a marker of angiogenesis. Serum levels of this protein are elevated in some human cancers. Our objectives were to compare collagen IV levels in the serum of breast cancer patients and healthy women and to examine changes during preoperative chemotherapy. Sera from 51 patients with stage II-III breast cancer and 55 healthy controls were analysed. Collagen IV level was measured by a commercially available sandwich enzyme link immunoassay. Baseline serum levels were compared between cancer patients and healthy women and paired pre- and post-chemotherapy measurements were also performed in 39 patients who received preoperative chemotherapy and were correlated with response to therapy. The median serum collagen IV concentration was significantly higher in cancer patients (166 microg l(-1)) than in healthy women (115 microg l(-1)), P0.001. Chemotherapy induced a significant further increase in serum collagen IV (167 microg l(-1) prechemo vs 206 microg l(-1) postchemo, P=0.001). There were no correlations between baseline collagen IV levels and response to therapy, age, clinical stage or HER2 status. In conclusion, patients with breast cancer have elevated levels of collagen IV compared to healthy women and collagen IV levels increase further during chemotherapy.

The process of neovascularization from preexisting blood vessels, referred to as angiogenesis, plays a critical role in both tumor growth and dissemination in multiple cancer types. Currently, there exists a need to identify biomarkers that can both indicate biological activity and predict efficacy at the molecular level for antiangiogenesis drugs which are anticipated to result in tumor stasis rather than regression. To identify such biomarkers, athymic mice bearing L2987 human tumor xenografts were treated with the antiangiogenic agent brivanib alaninate, which is currently under clinical evaluation. This is an orally available and selective tyrosine kinase inhibitor that targets the key angiogenesis receptors vascular endothelial growth factor receptor 2 (VEGFR-2) and fibroblast growth factor receptor 1. In the described studies, tumor samples were collected from these xenografts and RNA was extracted for gene expression profiling on Affymetrix 430A mouse GeneChips. Statistical analysis was done using a defined set of genes identified to be coexpressed with VEGFR-2 from a clinical tumor gene expression profiling database and between tumor samples isolated from brivanib alaninate–treated and untreated mice. Tyrosine kinase receptor 1 (Tie-1), collagen type IV α1 (Col4a1), complement component 1, q subcomponent receptor 1 (C1qr1), angiotensin receptor–like 1 (Agtrl1), and vascular endothelial-cadherin (Cdh5) were all identified to be significantly modulated by treatment with brivanib alaninate. These genes, which may be potentially useful as markers of brivanib alaninate activity, were further studied at the protein level in two separate in vivo human colon tumor xenograft models, HCT116 and GEO, using immunohistochemistry-based approaches. [Cancer Res 2007;67(14):6899–906]

Meeting abstracts Pembrolizumab (MK-3475), a humanized monoclonal IgG4 antibody against programmed death receptor 1 (PD-1), is currently being studied in clinical trials across more than 30 types of cancers. Immunotherapy with anti–PD-1 monoclonal antibodies such as pembrolizumab shows robust

Gastric cancer, a leading cause of cancer-related deaths, is a heterogeneous disease. We aim to establish clinically relevant molecular subtypes that would encompass this heterogeneity and provide useful clinical information. We use gene expression data to describe four molecular subtypes linked to distinct patterns of molecular alterations, disease progression and prognosis. The mesenchymal-like type includes diffuse-subtype tumors with the worst prognosis, the tendency to occur at an earlier age and the highest recurrence frequency (63%) of the four subtypes. Microsatellite-unstable tumors are hyper-mutated intestinal-subtype tumors occurring in the antrum; these have the best overall prognosis and the lowest frequency of recurrence (22%) of the four subtypes. The tumor protein 53 (TP53)-active and TP53-inactive types include patients with intermediate prognosis and recurrence rates (with respect to the other two subtypes), with the TP53-active group showing better prognosis. We describe key molecular alterations in each of the four subtypes using targeted sequencing and genome-wide copy number microarrays. We validate these subtypes in independent cohorts in order to provide a consistent and unified framework for further clinical and preclinical translational research.

The pathogenesis of breast cancers that do not express estrogen receptors or Her-2/neu receptors (ER−/HER2− phenotype) is incompletely understood. We had observed markedly elevated gene expression of gamma-aminobutyric acid type A (GABAA) receptor subunit π (GABAπ, GABRP) in some breast cancers with ER−/HER2− phenotype. In this study, transcriptional profiles (TxPs) were obtained from 82 primary invasive breast cancers by oligonucleotide microarrays. Real-time reverse transcription–polymerase chain reaction (RT–PCR) was used to measure GABAπ gene expression in a separate cohort of 121 invasive breast cancers. GABAπ gene expression values from TxP and RT–PCR were standardized and compared with clinicopathologic characteristics in the 203 patients. GABAπ gene expression was increased in 16% of breast cancers (13/82 TxP, 20/ 121 RT–PCR), particularly in breast cancers with ER−/HER2− phenotype (60%), and breast cancers with basal-like genomic profile (60%). The profile of genes coexpressed with GABAπ in these tumors was consistent with an immature cell type. In multivariate linear regression analysis, the level of GABAπ gene expression was associated with ER−/HER2− phenotype (PP=0.0003), and shorter lifetime duration of breastfeeding (≤ 6 months) in all women (P=0.017) and specifically in parous women (P=0.013). GABAπ gene expression was also associated with combinations of high grade with ER−/HER2− phenotype (P=0.002), and with Hispanic ethnicity (P=0.036). GABAπ gene expression is increased in breast cancers of immature (undifferentiated) cell type and is significantly associated with shorter lifetime history of breastfeeding and with high-grade breast cancer in Hispanic women.

We examined how well differentially expressed genes and multigene outcome classifiers retain their class-discriminating values when tested on data generated by different transcriptional profiling platforms. RNA from 33 stage I-III breast cancers was hybridized to both Affymetrix GeneChip and Millennium Pharmaceuticals cDNA arrays. Only 30% of all corresponding gene expression measurements on the two platforms had Pearson correlation coefficient ror= 0.7 when UniGene was used to match probes. There was substantial variation in correlation between different Affymetrix probe sets matched to the same cDNA probe. When cDNA and Affymetrix probes were matched by basic local alignment tool (BLAST) sequence identity, the correlation increased substantially. We identified 182 genes in the Affymetrix and 45 in the cDNA data (including 17 common genes) that accurately separated 91% of cases in supervised hierarchical clustering in each data set. Cross-platform testing of these informative genes resulted in lower clustering accuracy of 45 and 79%, respectively. Several sets of accurate five-gene classifiers were developed on each platform using linear discriminant analysis. The best 100 classifiers showed average misclassification error rate of 2% on the original data that rose to 19.5% when tested on data from the other platform. Random five-gene classifiers showed misclassification error rate of 33%. We conclude that multigene predictors optimized for one platform lose accuracy when applied to data from another platform due to missing genes and sequence differences in probes that result in differing measurements for the same gene.

Breast cancers show variable sensitivity to paclitaxel. There is no diagnostic test to identify tumors that are sensitive to this drug. We used U133A chips to identify genes that are associated with pathologic complete response (pCR) to preoperative paclitaxel-containing chemotherapy in stage I-III breast cancer ( n = 82). Tau was the most differentially expressed gene. Tumors with pCR had significantly lower ( P < 0.3 × 10 -5 ) mRNA expression. Tissue arrays from 122 independent but similarly treated patients were used for validation by immunohistochemistry. Seventy-four percent of pCR cases were tau protein negative; the odds ratio for pCR was 3.7 (95% confidence interval, 1.6-8.6; P = 0.0013). In multivariate analysis, nuclear grade ( P < 0.01), age P = 0.03), and tau-negative status ( P = 0.04) were independent predictors of pCR. Small interfering RNA experiments were performed to examine whether down-regulation of tau increases sensitivity to chemotherapy in vitro . Down-regulation of tau increased sensitivity of breast cancer cells to paclitaxel but not to epirubicin. Tubulin polymerization assay was used to assess whether tau modulates binding of paclitaxel to tubulin. Preincubation of tubulin with tau resulted in decreased paclitaxel binding and reduced paclitaxel-induced microtubule polymerization. These data suggest that low tau expression renders microtubules more vulnerable to paclitaxel and makes breast cancer cells hypersensitive to this drug. Low tau expression may be used as a marker to select patients for paclitaxel therapy. Inhibition of tau function might be exploited as a therapeutic strategy to increase sensitivity to paclitaxel.

In this second part of the two-part review of breast cancer biomarkers and molecular medicine, the first section will consider additional breast cancer prognostic factors, including oncogenes, tumor suppressor genes, cell adhesion molecules, invasion-associated proteins and proteases, hormone receptor proteins, drug resistance proteins, apoptosis regulators, transcription factors, telomerase, DNA repair and methylation and transcriptional profiling using high-density genomic microarrays. The second section will consider the prediction of therapy response using the techniques of pharmacogenetics and pharmacogenomics.

BACKGROUND Gene expression profiling should be applicable to needle biopsy samples if microarray technology is to become practically useful for clinical research or management of breast carcinoma. This study compared gene expression profiles derived from fine-needle aspiration biopsy (FNAB) and from core needle biopsy (CBX). METHODS Total RNA was extracted from single FNAB and CBX samples. Corresponding pairs of FNAB and CBX were analyzed for similarity of gene expression profiles using cDNA microarrays that contain 30,721 human sequences. A subset of genes that distinguished CBX samples from FNAB samples was evaluated in a larger group of needle biopsy samples and in a published genomic database derived from 78 sporadic breast carcinomas with known clinical outcome. RESULTS Sixty-eight patients with newly diagnosed breast carcinoma were included in the current study. Sixty-five patients underwent FNAB (17 had both FNAB and CBX) and 3 underwent CBX only. Extracted RNA was of suitable quality for hybridization in 46 (71%) FNABs and 15 (75%) CBXs. Total RNA yield in those samples was similar for single-pass FNAB (mean = 3.6 μg and median = 2.2 μg; n = 46) and CBX (mean = 2.8 μg and median = 2.0 μg; n = 15), with 1 μg or more of total RNA in all cases. Transcriptional profiling was performed successfully in all cases when it was attempted, in a total of 50 samples (38 FNABs and 12 CBXs), including matched FNAB and CBX samples from 10 patients. There were differences in gene expression profiles in 10 matched FNAB and CBX sample pairs. Genes that were expressed differently in CBX samples, compared with FNAB samples, were recognized as being predominantly from the endothelium, fibroblasts, myofibroblasts or smooth muscle, and histiocytes. Corresponding microscopic cell counts from FNABs demonstrated means of 80% tumor cells, 15% lymphocytes, and 5% stromal cells, whereas CBXs contained 50% tumor cells, 20% lymphocytes, and 30% stromal cells. Considering that CBXs are approximately six-fold richer in nonlymphoid stromal cells than FNABs and that CBXs differentially express a set of recognized stromal genes, the authors used these biopsies to define a transcriptional profile of breast carcinoma stroma. A set of 120 genes differentially expressed in CBXs was assessed independently in a published breast carcinoma genomic database to classify breast carcinomas based on stromal gene expression. Subgroups of tumors with low or high stromal signal were identified, but there was no correlation with the development of systemic metastases within 5 years. CONCLUSIONS Both FNAB and CBX yield a similar quality and quantity of total RNA and are suitable for cDNA microarray analyses in approximately 70–75% of single-pass samples. Transcriptional profiles from FNAB and CBX of the same tumor generally are similar and are driven by the tumor cell population. The authors concluded that each technique has relative advantages. The FNABs provide transcriptional profiles that are a purer representation of the tumor cell population, whereas transcriptional profiles from CBXs include more repre sentation from nonlymphoid stromal elements. Selection of the preferred needle biopsy sampling technique for genomic studies of breast carcinomas should depend on whether variable stromal gene expression is desirable in the samples. Cancer 2003;97:2960–71. © 2003 American Cancer Society. DOI 10.1002/cncr.11435

6009 Background: Somatic mutational load (ML) is associated with response to anti CTLA-4 and PD-1/-L1 immunotherapies in select tumors, due to formation of neoepitopes not subject to central immune tolerance. Neoepitopes specific to HPV, EBV virus infection are also present in some HNSCC. An IFNγ gene expression profile (GEP) characteristic of tumor inflammation is also related to response to anti PD-1/-L1 therapy. This study evaluated relationships between ML and clinical outcome and independent predictive values of ML and GEP in patients with HNSCC treated with pembrolizumab. Methods: Whole exome sequencing (WES) and GEP were assessed in FFPE tumor specimens of patients with HNSCC (KEYNOTE 012; subsets of B1 [PD-L1+, n = 34] and B2 [PD-L1+/‒, n = 73] cohorts). ML, neoantigen load (NL), HPV/EBV status and clonality were assessed by standard WES analytical methods. GEP score is a weighted sum of normalized expression values of 18 genes. Statistical testing of ML and response, and ML and GEP relationship by HPV/EBV status was prespecified. Results: There were 73 patients identified as HPV‒ and EBV‒ (n = 25 in B1; n = 48 in B2). In HPV‒ and EBV‒ patients in B1 and B2 cohorts, respectively, associations between ML and objective response (OR) (P = 0.029 and 0.055; AUROC 0.89 and 0.63), and GEP and OR (P = 0.064 and 0.01; AUROC 0.82 and 0.74) were statistically significant. In combined cohorts of HPV‒ and EBV‒ patients, ML and GEP were significantly associated with OR (P = 0.009 and 0.002; AUROC 0.70 and 0.76, respectively). ML and GEP were only weakly correlated (r = 0.173). In a joint model, ML was significantly associated with response (p = 0.020) after adjusting for GEP (also significant, p = 0.006). NL and clonality weighted ML were also significantly associated with response (P = 0.026 and 0.006, respectively). In HPV+ or EBV+subjects, OR association was not significant for ML, possibly due to a dominance of viral vs somatic neoepitopes; GEP was significant, likely due to tumor inflammation. Conclusions: ML and GEP are independently predictive of response to pembrolizumab in HPV‒/EBV‒ patients with HNSCC; GEP was predictive regardless of viral status. ML and GEP may have utility in characterizing responses to anti PD-1 therapies and novel cancer regimens in HNSCC. Clinical trial information: NCT01848834.

1 Background: ML is associated with response to anti CTLA-4 and PD-1/L1 immunotherapies in some cancers, likely due to increased tumor antigenicity via generation of neoepitopes not subject to central immune tolerance. An IFNγ centric gene expression profile (GEP), characteristic of a T cell inflamed microenvironment, is also related to response to PD-1/PD-L1 therapies. The association between ML and outcome, and independent predictive values of ML and GEP in pembrolizumab treated patients across 21 tumor types were assessed. Methods: Whole exome and RNA sequencing were performed on FFPE specimens pooled from 2 multicohort advanced solid tumor trials (KEYNOTE 012 [n = 39] and 028 [n = 80]). The previously defined GEP score is a weighted sum of normalized expression values for 18 genes. ML and neoantigen analysis tools included GATK, MuTect and NetMHC. Statistical testing was adjusted for multiple testing. ML and GEP association was also explored in Moffitt and TCGA datasets. Results: ML and neoantigen load were significantly associated with objective response (OR) (AUROC = 0.76 and 0.78; p = 0.0036 and 0.0083, respectively). Median numbers of mutations were 180 in responders and 61 in nonresponders. The overall response rate (ORR) in all patients was 15%. At a ML cutoff of 102 (based on ROC Curve YoudenIndex), the ORR was 32.3% (prevalence 31.0%; NPV 92.8%). GEP was also significantly associated with OR (AUROC = 0.76; p = 0.0071). ML and GEP were modestly correlated (r = 0.28), consistent with associations between ML and GEP in Moffitt (r = 0.11) and TCGA databases (r = 0.29). When jointly modeled, ML was significantly associated with response (p = 0.0078) after adjusting for GEP (also significant; p = 0.0251). Data from additional patients will also be presented. Conclusions: These data show that ML and GEP are independently predictive biomarkers of response to pembrolizumab in multiple tumor types, suggesting that tumor antigenicity and T cell infiltration may provide complementary information for expected pembrolizumab activity. As measures of distinct, yet common features of a PD-1 checkpoint blockade responsive tumor, ML and GEP may have utility in characterizing responses to anti PD-1 and other cancer immunotherapies. Clinical trial information: NCT01848834; NCT02054806.

Background Insulin-like growth factor type 1 receptor (IGF-1R) mediates resistance to epidermal growth factor receptor (EGFR) inhibition and may represent a therapeutic target. We conducted a multicenter, randomized, double blind, phase II/III trial of dalotuzumab, an anti-IGF-1R monoclonal antibody, with standard therapy in chemo-refractory, KRAS wild-type metastatic colorectal cancer. Methods Eligible patients were randomly assigned to dalotuzumab 10mg/kg weekly (arm A), dalotuzumab 7.5mg/kg every alternate week (arm B), or placebo (arm C) in combination with cetuximab and irinotecan. Primary endpoints were progression-free survival (PFS) and overall survival (OS). Secondary endpoints included exploratory biomarker analyses. All statistical tests were two-sided. Results The trial was prematurely discontinued for futility after 344 eligible KRAS wild-type patients were included in the primary efficacy population (arm A = 116, arm B = 117, arm C = 111). Median PFS was 3.9 months in arm A (hazard ratio [HR] = 1.33, 95% confidence interval [CI] = 0.98 to 1.83, P = .07) and 5.4 months in arm B (HR = 1.13, 95% CI = 0.83 to 1.55, P = .44) compared with 5.6 months in arm C. Median OS was 10.8 months in arm A (HR = 1.41, 95% CI = 0.99 to 2.00, P = .06) and 11.6 months in arm B (HR = 1.26, 95% CI = 0.89 to 1.79, P = .18) compared with 14.0 months in arm C. Grade 3 or higher asthenia and hyperglycaemia occurred more frequently with dalotuzumab compared with placebo. In exploratory biomarker analyses, patients with high IGF-1 mRNA tumors in arm A had numerically better PFS (5.6 vs 3.6 months, HR = 0.59, 95% CI = 0.28 to 1.23, P = .16) and OS (17.9 vs 9.4 months, HR = 0.67, 95% CI = 0.31 to 1.45, P = .31) compared with those with high IGF-1 mRNA tumors in arm C. In contrast, in arm C high IGF-1 mRNA expression predicted lower response rate (17.6% vs 37.3%, P = .04), shorter PFS (3.6 vs 6.6 months, HR = 2.15, 95% CI = 1.15 to 4.02, P = .02), and shorter OS (9.4 vs 15.5 months, HR = 2.42, 95% CI = 1.21 to 4.82, P = .01). Conclusions Adding dalotuzumab to irinotecan and cetuximab was feasible but did not improve survival outcome. IGF-1R ligands are promising biomarkers for differential response to anti-EGFR and anti-IGF-1R therapies.

Dinaciclib is a potent CDK1, 2, 5 and 9 inhibitor being developed for the treatment of cancer. Additional understanding of antitumor mechanisms and identification of predictive biomarkers are important for its clinical development. Here we demonstrate that while dinaciclib can effectively block cell cycle progression, in vitro and in vivo studies, coupled with mouse and human pharmacokinetics, support a model whereby induction of apoptosis is a main mechanism of dinaciclib's antitumor effect and relevant to the clinical duration of exposure. This was further underscored by kinetics of dinaciclib-induced downregulation of the antiapoptotic BCL2 family member MCL1 and correlation of sensitivity with the MCL1-to-BCL-xL mRNA ratio or MCL1 amplification in solid tumor models in vitro and in vivo. This MCL1-dependent apoptotic mechanism was additionally supported by synergy with the BCL2, BCL-xL and BCL-w inhibitor navitoclax (ABT-263). These results provide the rationale for investigating MCL1 and BCL-xL as predictive biomarkers for dinaciclib antitumor response and testing combinations with BCL2 family member inhibitors.