Evaluating the correlation between clonal hematopoiesis and response to regorafenib and nivolumab in patients with refractory metastatic colorectal cancer.

Abstract

133 Background: Clonal hematopoiesis (CH) has emerged as a possible biomarker of immune modulation in cancer. Preliminary data have suggested an important role of inflammatory processes in CH with possible benefit of immunotherapy in mismatch repair proficient (MMRp) colorectal cancer. The combination of regorafenib (REGO) and nivolumab (REGONIVO) has shown limited activity in MMRp metastatic colorectal cancer (mCRC), but predictive biomarkers of benefit remain undefined. Methods: We conducted a retrospective cohort study of 55 patients with refractory MMRp mCRC treated with regorafenib alone or in combination with nivolumab at Moffitt Cancer Center. Patients were stratified by the presence or absence of CH based on tissue/blood based next-generation sequencing (NGS). Clinical and molecular characteristics including KRAS/BRAF status, tumor sidedness, prior line of treatment, liver metastasis were collected. Progression-free survival (PFS) was assessed using Kaplan-Meier analysis. Results: CH mutations were detected in 31 patients (56%) and 24 patients (44%) were classified as non-CH mutations. The median lines of treatment before REGO or REGONIVO was similar between both groups. Among CH-negative patients, median PFS was 2.9 months with REGO and 2.4 months with REGONIVO. Among CH-positive patients, median PFS was 3.1 months with REGONIVIO and 1.8 months with REGO with no statistically significant improvement on univariate analysis (p = 0.05) but statistically significant on multivariate analysis (HR = 0.18, 95% CI = 0.1 – 0.87). Presence of liver metastasis was found to be significant adverse risk factor in patients with CH mutations. Conclusions: Although CH status did not show a clear benefit on univariate analysis, multivariate analysis suggested that patients with CH status may derive greater benefit from REGONIVO therapy. These findings indicate that CH may influence response to combined regorafenib and nivolumab treatment and warrants further validation in larger prospective studies. Multivariate analysis. Age HR P value 18 – 40 1 41 – 60 0.1 (0.01 – 2.3) 0.1 61 – 75 0.05 (0. 00 – 1.6) 0.09 Sex  Male 1  Female 1.04 (0.22 -4.9) 0.05 KRAS  Wild 1  Mutated 0.4 (0. 9 – 1.7) 1.44 Prior lines of treatment  < 2 1  3 - 5 1.4 (0.3 – 7.4) 0.7  More than 5 0.7 (0.8 – 5.3) Liver metastasis  No 1  Yes 25 (1.2 – 545) 0.03 Treatment  Regorafenib 1  Regorafenib + nivolumab 0.18 (0.04 – 0.88) 0.03