Predictors of response to adjuvant chemotherapy for colorectal cancer.

Abstract

Background: It is well recognized that not all patients with stage C colorectal cancer (CRC) derive a survival benefit from adjuvant chemotherapy. It would therefore be advantageous to identify factors that define a target group for treatment. It has been suggested that those most likely to benefit are women with proximal tumours. Recent work has suggested microsatellite instability (MSI) may be a useful marker however the limited studies performed are conflicting. Aim: To determine if gender, site, tumour histology or microsatellite (MSI) status predict survival benefit from 5FU-based adjuvant chemotherapy in stage C CRC. Method: Data was collated on stage C colorectal cancer cases that underwent curative resection over a 20-year period (inclusive of years prior to standard chemotherapy). Pathology was re-evaluated, DNA extracted from the formalin fixed paraffin specimen and MSI status established. Primary endpoint was cancer-related death. Kaplan-Meier curves were constructed for univariate analysis and differences analysed by log rank test. Multivariate analysis was performed using Cox proportional hazard model adjusting for age, gender, site, distinct pathological variables and MSI. A compounding effect between these factors and chemotherapy benefit was measured by interaction testing Results: 811 unselected cases were included in the study. Thirty-seven percent received chemotherapy. Chemotherapy significant improved cancer-specific survival (HR of dying 0.66 (95% CI 0.52-0.83 p=0.0003). Female gender offered a survival advantage overall (HR 0.81 95% CI 0.68-0.97; p=0.02) however site did not influence outcome (HR 1.03). On interaction testing, gender, site and tumour histology did not significantly influence the survival effect of chemotherapy. 802 cases were included in the MSI analysis of which 77 exhibited MSI. MSI status did not influence prognosis (HR of cancer death 1.45, 95% CI 0.90-2.21; p= 0.13). However, in the non-chemotherapy cohort, MSI conferred a significantly less favourable outcome (HR 1.89, 95%CI 1.13-3.16; p= 0.02). Chemotherapy produced a survival benefit in both the MSI (HR 0.08 95% CI 0.02-0.27; p=<0.0001) and the microsatellite stable (MSS) cohort (HR 0.62, 95% CI 0.47-0.81; p=0.001). On interaction testing, neither compounded the benefit of chemotherapy, however of all the tested parameters, MSI came closest to significance (p=0.08). Conclusion: These results suggest that 5FU-based adjuvant chemotherapy for stage C colorectal cannot be targeted using gender, tumour site, histological characteristics or MSI.