Galon, Jérôme

614 Background: A consensus interpretation of the IDEA colon cancer study results suggested that risk categories based on T/N stage grouping be used to guide decision-making for duration (3 vs 6 months) of adjuvant FOLFOX or CapeOX chemotherapy. Given the prognostic potential of immune biomarkers, we examined the immunoscore and individual T and B lymphocyte markers in low and high risk T/N subsets of stage III colon carcinoma patients (N=600) treated with adjuvant FOLFOX. Methods: Immunoscore (CD3+, CD8+) and individual T-cell and CD20+ B-cell immunostain densities in central tumor (CT) and invasive margin (IM) of FFPE sections were quantified by image analysis. A predetermined immunoscore categorization was used [high (2-4) vs low (0-1)]. Individual markers were analyzed by backwards selection wherein CD3+ IM was most robust for prognosis and an optimized cutoff was then determined. Associations with disease-free survival (DFS) were analyzed by multivariable Cox regression adjusting for age, T/N stage, sidedness, KRAS/BRAF, and DNA mismatch repair. Results: In low and high risk T/N patient subsets, the immunoscore and CD3+ IM were each significantly discriminant for prognosis. Among low risk (T1-3N1) patients, a high vs low immunoscore was associated with a 91% vs 77% 3-year DFS [HR 0.57, 95% confidence interval (CI) 0.34-0.95, adjusted (adj) P= 0.026]. Among high risk (T4 or N2) patients, a high vs low immunoscore was associated with a 68% vs 54% 3-year DFS (HR 0.64, 95% CI 0.42-0.98, Padj= 0.034]. Similarly, a high vs low intratumoral CD3+ density at the invasive margin (IM) was significantly associated with prognosis in low risk [HR 0.37, 95% CI 0.21- 0.66), Padj< 0.0003] and in high risk [HR 0.47, 95% CI, 0.27- 0.80), Padj< 0.0028] patient subsets. Conclusions: Immunoscore and CD3+ IM were shown to prognostically stratify FOLFOX-treated patients within both low and high risk T/N subsets. These data underscore limitations of the T/N risk classification for adjuvant treatment decisions in stage III patients, and demonstrate the ability of T-cell markers to enhance prognostication to guide clinical decision-making.

3579 Background: Tumor infiltrating lymphocytes (TIL) indicate a host immune response that may influence survival. Immunoscore was developed using CD3+ and CD8+density and location in primary CC pts with pooled stages, varying treatment and follow-up. We determined if individual immune markers and/or Immunoscore are prognostic in resected stage III CC pts (N=600). Methods: CD3+ and CD8+ T-cell or CD20+ B lymphocyte density in central tumor (CT) and invasive margin (IM) was evaluated by immunostaining and quantified by image analysis. Immunoscore was calculated on a scale of I0 to I4 with high densities of CD3+ and CD8+in both CT and IM scored as I4; low densities scored as I0. Associations with disease-free survival (DFS) were evaluated by multivariable Cox regression adjusting for covariates. Results: Data for CD3+, CD8+ and CD20+ were generated (N=595). Higher density of CD3+ CT, CD3+ IM and CD8+ IM were associated with longer DFS adjusting for covariates (Table). CD3+ IM had the strongest association with DFS, and was stronger in left-sided (HRadj.=0.81, 95% CI, 0.70-0.94, padj.= 0.0049) vs right-sided (HRadj.=0.93, 95% CI, 0.85-1.0 padj.=0.52) tumors (pinteraction=0.039). Higher density of CD3+ IM was associated with older age (p=0.034), T1/2 (p<.0001), N1 (p=0.017), right-sided (p=0.013), high TILs (p=0.0008), and deficient MMR (p=0.0003). Using a prior Immunoscore risk stratification, higher scores were associated with better DFS (HR.=0.62, CI, 0.44-0.87, p=0.006) (Table). Conclusions: Densities of CD3+ and CD8+, especially at IM, are individually prognostic in FOLFOX-treated pts. Association of CD3+IM with prognosis differed by primary CC site. Immunoscore was strongly prognostic, and this result provides validation in a clinical trial cohort. [Table: see text]

3025 Background: Axi-cel is an autologous anti-CD19 CAR T cell therapy. ZUMA-1 is a multicenter, registrational trial of axi-cel in patients (pts) with refractory/aggressive B-cell non-Hodgkin lymphoma (NHL). In a pre-specified interim analysis, ZUMA-1 met its primary endpoint with 76% objective response rate and 47% complete response (Blood 2016;128:LBA-6). We describe, for the first time, a tumor microenvironment immune gene signature associated with CAR T cell treatment (tx) of NHL pts. Methods: Paired biopsies, pre- and within 3 weeks post-axi-cel tx, were analyzed by digital gene expression followed by a pre-specified bioinformatics algorithm applied to IGES15 and IGES21 genes involved in immune-mediated tumor regression (Immunosign; Galon Immunity 2013). Immunosign profiles expression of a pre-defined set of effector T cell, Th1, chemokine, and cytokine genes. Expression analysis and hierarchical clustering were used to define an axi-cel-related tumor immune gene signature. Wilcoxon signed rank test with multiple test correction by FDR (Benjamini-Yekutieli) was used. Results: Gene expression profile comparisons of pre- and post-axi-cel tx biopsies from 14 pts showed profound changes in gene expression within the tumor environment after infusion. The most upregulated genes post-axi-cel tx were CCL5 (RANTES), CTLA4, and GZMA (log2 fold change > 2, P< 0.05, FDR < 0.050). Immune checkpoints PD-L1 and LAG3 were also upregulated post-axi-cel (log2 fold change > 1.6, P< 0.05, FDR < 0.055). Other genes associated with T cell proliferation, homing, and effector function were also upregulated: IL-15, GZMK, CXC3CL1 (Fractalkine), CD8A, and STAT4 (log2 fold change > 1.6; P< 0.05, FDR < 0.074). Additional baseline tumor characteristics and associative analysis will be presented. Conclusions: We define a mechanistic tumor immune gene signature in NHL pts associated with axi-cel tx. This signature comprises upregulation of T cell activation, effector, chemokine, and immune checkpoint genes. These data will potentially lead to rational optimization of T cell interventions in cancer Clinical trial information: NCT02348216.

Abstract Introduction: Axicabtagene ciloleucel (axi-cel) is an autologous anti-CD19 CAR T cell therapy approved by the US Food and Drug Administration for patients with relapsed or refractory large B cell lymphoma with ≥2 prior systemic therapies. Results from the primary analysis of ZUMA-1 demonstrated an objective response rate of 82%, with 54% complete responses (Neelapu & Locke et al. N Engl J Med. 2017). Grade ≥3 cytokine release syndrome and neurologic events occurred at 13% and 28%, respectively, and were generally reversible. Building on previous reports of CAR treatment-related tumor microenvironment (TME) changes (Galon et al. ASCO 2017. #3025), here, we examine associations between the pretreatment TME and clinical outcomes. Methods: Patients received axi-cel (target dose, 2.0 × 106 cells/kg) in phases 1 and 2 of ZUMA-1 (NCT02348216). Baseline tumor biopsy samples were analyzed by digital gene expression (NanoStringTM). Prespecified bioinformatics algorithm and cutoffs were applied to immune-mediated tumor regression genes (Immunosign®; Galon et al. Immunity. 2013), which included T cell-specific (effector T cell, Th1) genes, interferon pathway-related genes, chemokines, cytokines, and immune checkpoints. Expression analysis and scoring examined potential associations between TME features and response. A broader, hypothesis-generating analysis was performed using all PanCancer Immune Profiling Panel genes (770 genes) by NanoStringTM. Fisher's exact test and Wilcoxon signed rank test with multiple test correction by false discovery rate (Benjamini-Hochberg) were used. Results: Tumor samples from all immediate pre-treatment biopsy samples that passed quality control (6 nonresponders; 21 responders) were analyzed. Pretreatment TME Immunosign21 scores were elevated in responders vs nonresponders. Responses were seen in 18/20 patients (90%) with high and 3/7 patients (43%) with low Immunosign21 scores. Immune-related genes from the Immunosign21 or Pan Cancer profile that showed higher expression at baseline in responders included CTLA4, CD3γ, CD3ϵ, CD27, and ICOSL. Genes with lower expression in responders at baseline included MHC class II genes and cancer-testis antigens. Additional correlative analyses between TME characteristics and CAR T cell activity will be presented. Conclusions: Increased baseline expression of TME genes related to T cell activity, along with decreased expression of MHC class II and cancer testis antigens (CTAs), may be associated with a response to axi-cel. These findings should be validated in ongoing and subsequent studies. Citation Format: John M. Rossi, Jerome Galon, Sarah Turcan, Corinne Danan, Frederick L. Locke, Sattva S. Neelapu, David B. Miklos, Caron A. Jacobson, Lazaros J. Lekakis, Yi Lin, Armin Ghobadi, William Y. Go, Adrian Bot. Characteristics of the pretreatment tumor microenvironment may influence clinical response in patients with refractory large B cell lymphoma treated with axicabtagene ciloleucel (axi-cel) in the pivotal ZUMA-1 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr LB-016.

Abstract The American Joint Committee on Cancer/Union Internationale Contre le Cancer ( AJCC / UICC ) TNM staging system provides the most reliable guidelines for the routine prognostication and treatment of colorectal carcinoma. This traditional tumour staging summarizes data on tumour burden (T), the presence of cancer cells in draining and regional lymph nodes (N) and evidence for distant metastases (M). However, it is now recognized that the clinical outcome can vary significantly among patients within the same stage. The current classification provides limited prognostic information and does not predict response to therapy. Multiple ways to classify cancer and to distinguish different subtypes of colorectal cancer have been proposed, including morphology, cell origin, molecular pathways, mutation status and gene expression‐based stratification. These parameters rely on tumour‐cell characteristics. Extensive literature has investigated the host immune response against cancer and demonstrated the prognostic impact of the in situ immune cell infiltrate in tumours. A methodology named ‘Immunoscore’ has been defined to quantify the in situ immune infiltrate. In colorectal cancer, the Immunoscore may add to the significance of the current AJCC / UICC TNM classification, since it has been demonstrated to be a prognostic factor superior to the AJCC / UICC TNM classification. An international consortium has been initiated to validate and promote the Immunoscore in routine clinical settings. The results of this international consortium may result in the implementation of the Immunoscore as a new component for the classification of cancer, designated TNM ‐I ( TNM ‐Immune). © 2013 The Authors. Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.