Real-world outcomes of locally advanced mismatch repair deficient colorectal tumours treated with immune checkpoint inhibitors.

Abstract

75 Background: Immunohistochemical analysis of mismatch repair deficient (dMMR) status in colorectal (CRC) cancers is standard of care due to superiority of immune checkpoint blockade (ICB) over chemotherapy in the metastatic and potentially early-stage disease setting. Cercek et al reported complete clinical response in 100% rectal, 65% colon dMMR patients (pts) treated with ICB, most proceeding with nonoperative management. NICHE-2 showed 95% major pathological response (MPR), 68% pathological complete response (pCR) post dual ICB. NEOPRISM-CRC stratified pts to 1 or 3 cycles ICB with 59% pCR. However optimal ICB duration has not been defined and longer-term consequences of nonoperative management remain to be confirmed. Methods: A national, multicentre retrospective study involving data from 5 sites in Republic of Ireland. Standardised data collection tools were used, data was anonymised and centrally compiled by 3 independent assessors. Pts were eligible if had histologically confirmed locally advanced dMMR colon or rectal cancer and received minimum 1 cycle ICB. Results: Analysis data was available for 32 pts, with collection ongoing. Origin 25% (N=8) rectal, 75% (n=24) colon. 17 had proceeded to/ awaited surgery, 6 ongoing treatment, 4 active surveillance, 5 stopped treatment due to progression of disease (POD) or death. 40% (n=13) had germline testing with 18% (n=6) confirmed Lynch syndrome. 31% (n=10) BRAF V600 mutated. Four received chemotherapy pre ICB due to access. Median number of cycles was 4 (range 2-12). At radiological restaging after minimum 2 cycles ICB, 8 had complete radiological response (CRR), 9 partial radiological response (PRR). 3 of 4 pts on surveillance had CRR. At surgery 9 had pCR, 3 MPR. Ten grade 1-3 toxicities reported, 1 Grade 4 after 1 cycle ICB. Five pts died after 1, 2, 3, 4 cycles respectively; 3 due to POD, 1 after stopped treatment post grade 4 toxicity, 1 due to bowel obstruction. Conclusions: This real-world cohort dMMR CRC had 100% MPR, 75% pCR at surgery post ICB. Toxicity profile was comparable to literature. 5 pts had pCR with 4 or less cycles, indicating optimal treatment duration is still unclear. Concordance of radiological response with pathological response was not 100%, reflecting a challenge of selecting pts for nonoperative management. Duration of treatment by cycle (n=) 2-4 = 156-8 = 610-12 = 3Ongoing = 5NA = 3 MMR protein loss (n=) MLH1/PMS2 = 16MSH2/MSH6 = 6Single protein = 6NA = 1 Somatic mutation status (n=) BRAF = 10KRAS = 6 Radiological response (n=) Complete = 8Partial = 9Progression = 3NA/ Pending = 13 Pathological response (n=) Complete = 9Residual = 3Pending = 13NA = 7 Adverse events by grade (n=) G1 = 2 TFTd, 1 rashG2 = 3 TFTd, 2 adrenal insufficiencyG3 = 1 rash, 1 face swellingG4 = 1 myositis, myasthenia gravis, myocarditis Surgical complications (n=) Stricture = 3Perforation = 2Post-operative collection = 1 TFTd = thyroid dysfunction.