Incorporating the 40-Gene Expression Profile (40-GEP) Test within Each Clinicopathologic Staging System Improves Metastatic Risk-Stratification in Patients Diagnosed with Cutaneous Squamous Cell Carcinoma (cSCC) and One or More High Risk Factors

Abstract

Purpose: Three clinicopathologic-based risk assessment systems (Brigham and Women’s Hospital (BWH) staging, American Joint Committee on Cancer 8^th Edition (AJCC8) staging or National Comprehensive Cancer Network (NCCN)) are used to guide treatment pathway decisions, including nodal assessment and adjuvant radiation therapy. Each of these systems suffers from accuracy limitations. The 40-GEP test was developed to use tumor biology to predict regional/distant metastasis and has previously been validated as an independent risk stratification test for cSCC patients who are diagnosed with one or more clinicopathologic risk factors. The purpose of this study was to evaluate the performance of the 40-GEP test in the prognostication of metastasis when results are used in the context of each of the clinicopathologic risk assessment system.

Study: Under an IRB-approved protocol, centralized pathology review, and sample analysis were performed in a CAP-accredited, CLIA-certified laboratory using formalin-fixed paraffin-embedded archival primary SCC tumor specimens with one or more risk factors (n=897).  Verified clinicopathologic information and outcomes data were collected from 58 contributing centers. Three-year risk stratification for regional and distant metastasis was assessed using Kaplan-Meier analysis for metastasis-free survival (MFS) for Class 1 (low risk), Class 2A (moderate risk), and Class 2B (high risk) groups. Nested Cox regression models and accuracy metrics were used to compare risk prediction of clinicopathologic risk assessment vs. clinicopathologic risk assessment in combination with 40-GEP.

Results: The overall cohort had a metastasis rate of 13.2%, substantiating the high-risk status of this population. For the overall cohort, the 40-GEP test showed clinically and statistically significant differences in MFS rates between Class 1 (94.1%), Class 2A (81.1%), and Class 2B (56.8%) (p<0.001, log-rank). Combining the results of the 40-GEP test with staging systems or NCCN risk categories led to increases in both NPV and PPV, indicating that a 40-GEP result improves the prediction of metastatic risk over the use of staging alone Table 1. Inclusion of 40-GEP with different clinicopathologic risk assessment methods (BWH and AJCC8, or NCCN risk subgroups) yielded significant improvements in prediction of metastatic events compared to each risk assessment method that did not include 40-GEP (ANOVA for deviance, p<0.0001 for each comparison).

Conclusions: Incorporating the 40-gene expression profile (40-GEP) test within each clinicopathologic risk assessment system improves risk-stratification, enabling more accurate treatment pathway decisions.