Clinical performance of Immunoscore in stage III colorectal cancer patients in the SCOT and IDEAHORG cohorts

Abstract

Background: The ESMO clinical practice guidelines recommend consideration of Immunoscore (IS) for risk assessment of early colon cancer patients. IS clinical performance was assessed in the SCOT and IDEA-HORG trials evaluating 3 vs. 6 months (3m vs. 6m) of mFOLFOX6 adjuvant chemotherapy in stage III colorectal cancer (CRC). Methods: 1,002 formalin-fixed paraffin-embedded (FFPE) tumor samples (762 from SCOT;240 from HORG) were collected, of which 851 were eligible for biomarker analysis. Eligible samples were classified into 2 groups using pre-defined cut-offs (IS-Low, IS- High) and the performance of IS to predict 3 year disease-free survival (3y-DFS) was evaluated. Results: IS was successfully assessed in 846 cases (99%). 615 (72.7%) samples were classified as IS-High (311 and 304 in 3m and 6m arm, respectively). No significant association between IS and patients’ gender, age, PS, BMI or primary tumour location was observed. However, a significant difference between IS-High (43.7%) and IS Low (57.1%) was observed in the proportion of high risk (T4 and/or N2) tumours (p=0.001). Patients with IS-High tumors had significantly longer 3y-DFS (79.4%, 95%CI: 75.9%-82.4%) compared to those with IS-Low tumors (65.0%, 95%CI: 58.3%-70.9%); adjusted hazard ratio (HR) 1.9 (95%CI: 1.46-2.46; p<0.0001). Similarly, IS-High was significantly correlated with longer 3y-DFS in both treatment arms: 78.5% (95% CI 73.4%-82.7%) for IS-High and 65.8% (95% CI 56.1%-73.9%) for IS-Low in 3m arm; 80.3% (95% CI 75.3%-84.5%) for IS-High and 64.4% (95% CI 54.8%-72.6%) for IS-Low in 6m arm. The estimated HRs according to treatment duration and IS classification were 1.80 (95% CI 1.25-2.60) in 3m arm, 2.00 (95% CI 1.38-2.92) in 6m arm and 1.89 (95% CI 1.46-2.47) in the total study population; interaction p = 0.687. Conclusions: The results of this study confirm the prognostic value of IS observed in the IDEA-France trial (Pagès F et al 2020). However, this analysis was not powered to determine the predictive value of IS for treatment duration. Similar analysis of patients treated with CAPOX is warranted.