Ipilimumab and nivolumab after disease progression on single agent immunotherapy in mismatch repair deficient metastatic colorectal cancer.

Abstract

e15523 Background: Pembrolizumab and ipilimumab with nivolumab (IPI/NIVO) are acceptable options for mismatch repair deficient (MMRd) metastatic colorectal cancer (mCRC) based on KEYNOTE-177 and CheckMate 8HW, respectively. Current guidelines recommend fluoropyrimidine-based chemotherapy as second line. Our study aims to explore the efficacy of IPI/NIVO in patients with disease refractory to single agent immunotherapy (IO). Methods: A database search was conducted including all patients who had received IPI/NIVO from 2014 to present at Moffitt Cancer Center (Tampa/FL). All patients without a primary diagnosis of mCRC excluded. Individual chart checking found 10 patients with MMRd who were treated with single agent IO prior to IPI/NIVO. Results: Out of 10 patients, four had germline mutations consistent with Lynch syndrome. The primary tumor of one Lynch syndrome patient was questionable (appendix vs cecum). Seven patients initially had liver metastasis. Four had been treated with single agent IO as the last line prior to starting IPI/NIVO however all had received single agent IO. One patient was on pembrolizumab and FOLFOX simultaneously. Two patients had partial response, two had stable disease, and five had progressive disease as the best response on single agent IO. Insurance denied treatment of one patient after two cycles. The two patients with partial response on single agent IO have been on IPI/NIVO for six and seven months without disease progression. Of the patients who had stable disease as their best response on single agent IO, one progressed three months after starting IPI/NIVO while the other is on IPI/NIVO four months later without disease progression. Disease progression was seen at four weeks, six weeks, and five months on IPI/NIVO in patients who had progressive disease as their best response on single agent IO. Two other patients had progressive disease as the best response on single agent IO; one pursued hospice after two cycles of IPI/NIVO, the other has been on IPI/NIVO for nine months without disease progression. The patients with the best response on IPI/NIVO (four, six, seven, and nine months without disease progression) all had low or intermediate risk characterization of disease progression while on single agent IO. Two had no liver metastasis and had Lynch syndrome. One other patient had similar features; they progressed on IPI/NIVO after three months. All patients with worse response to IPI/NIVO had no liver metastasis. Conclusions: The optimal treatment for MMRd mCRC after single agent IO is unknown. In this small retrospective cohort in patients who received IO rechallenge with IPI/NIVO after disease progression to single agent IO we identified three patients with stable disease and one patient with partial response. Prospective studies exploring this approach are needed given inferior efficacy with chemotherapy in this population.