Density, Distribution, and Composition of Immune Infiltrates Correlate with Survival in Merkel Cell Carcinoma

Abstract

Purpose: Merkel cell carcinoma (MCC) is an aggressive cancer with frequent metastasis and death with few effective therapies. Because programmed death ligand-1 (PD-L1) is frequently expressed in MCC, immune checkpoint blockade has been leveraged as treatment for metastatic disease. There is therefore a critical need to understand the relationships between MCPyV status, immune profiles, and patient outcomes.

Experimental design: IHC for CD3, CD8, PD-1, PD-L1, and MCPyV T-antigen (to determine MCPyV status) was performed on 62 primary MCCs with annotated clinical outcomes. Automated image analysis quantified immune cell density (positive cells/mm2) at discrete geographic locations (tumor periphery, center, and hotspot). T-cell receptor sequencing (TCRseq) was performed in a subset of MCCs.

Results: No histopathologic variable associated with overall survival (OS) or disease-specific survival (DSS), whereas higher CD3+ (P = 0.004) and CD8+ (P = 0.037) T-cell density at the tumor periphery associated with improved OS. Higher CD8+ T-cell density at the tumor periphery associated with improved DSS (P = 0.049). Stratifying MCCs according to MCPyV status, higher CD3+ (P = 0.026) and CD8+ (P = 0.015) T-cell density at the tumor periphery associated with improved OS for MCPyV+ but not MCPyV- MCC. TCRseq revealed clonal overlap among MCPyV+ samples, suggesting an antigen-specific response against a unifying antigen.

Conclusions: These findings establish the tumor-associated immune infiltrate at the tumor periphery as a robust prognostic indicator in MCC and provide a mechanistic rationale to further examine whether the immune infiltrate at the tumor periphery is relevant as a biomarker for response in ongoing and future checkpoint inhibitor trials in MCC. Clin Cancer Res; 22(22); 5553-63. ©2016 AACR.

Conflict of interest statement

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

©2016 American Association for Cancer Research.

Figures

Figure 1

The relative composition, density, and distribution of immune infiltrates in primary Merkel cell carcinoma. A, scanning magnification shows tumor extending into subcutaneous tissue. Inset shows morphology of tumor cells (H&E, 20×; inset: 200×). B, Merkel cell polyoma virus T-antigen (20×). C–F, schematic of how immune cell density was assessed. Red box designates “hotspot” (single 1-mm2 box); green boxes designate tumor periphery (5 × 1-mm2 boxes); and yellow boxes designate central areas (5 × 1-mm2 boxes). C, CD3 (20×). D, CD8 (20×). E, PD1 (20×). F, PD-L1 (20×). In some instances, the red hot spot box overlaps with a green peripheral box as seen for PD-1.

Figure 2

Kaplan-Meier analysis of overall survival among all MCC patients according to the density, composition, and distribution of the inflammatory infiltrate. OS according to immune cell density/mm2 at the periphery (left), center (middle), and hotspot (right) of MCCs. Top row, OS according to CD3+ T cells/mm2. Bottom row, OS according to CD8+ T cells/mm2. Legend: Patients with cell densities above the median indicated by dashed line (-----) and patients with cell density below the median indicated by solid line (———).

Figure 3

Clonality and clonal overlap of MCPyV + or − primary Merkel cell carcinomas. A, clonality of the tumor-infiltrating T cells in MCPyV+ (n = 10) or MCPyV− (n = 9) primary MCC. The bottom and top of the box represent the 25th and 75th percentile respectively for all patients, with the bar indicating the median value. B, positive correlation of (CD8+/mm2) IHC and clonality was assessed in MCC in which both analyses had been performed (n = 19). Red = MCPyV+ MCCs, Black = MCPyV− MCCs. C, heatmap demonstrates clonal overlap in primary MCC grouped according to MCPyV status. The lowest clonal overlap is represented by white with red being the highest clonal overlap between patients. Black boxes are used to block out all self-comparisons. Legend: P1+, Patient 1 MCPyV+; P2+, Patient 2 MCPyV+, etc; P1−, Patient 1 MCPyV− MCC; P2−, Patient2 MCPyV− MCC, etc.