Metastatic site pattern as predictor of outcome of first-line alternating oxaliplatin-based chemotherapy and nivolumab for patients with microsatellite-stable (MSS) colorectal cancer (CRC).

Abstract

3533 Background: The randomized METIMMOX trial evaluated short-course oxaliplatin-based chemotherapy (FLOX) alternating with nivolumab for previously untreated, unresectable abdominal metastases (mets) from MSS CRC. A subgroup of patients assigned to this experimental (exp) treatment had remarkably extended progression-free survival (PFS) compared to the control group patients given standard FLOX chemotherapy with median PFS 9.3 months. We explored if the extent of involved organs might be decisive for responsiveness to the METIMMOX regimen. Methods: Patients with measurable infradiaphragmatic (liver, peritoneal, nodal) mets were randomly assigned to the control group of FLOX (oxaliplatin, 5-fluorouracil, folinic acid) Q2W or the exp group of alternating 2 cycles each of FLOX Q2W and nivolumab Q2W, with prespecified break periods. Radiologic response assessment was done every 8 weeks with PFS as the primary endpoint. For this post hoc analysis, at baseline, the principal metastatic site was defined by the 2 largest mets (main lesions) of the dominant infradiaphragmatic organ and the global metastatic pattern by the main and subsidiary lesions of all involved organs. Patients without adverse events leading to treatment discontinuation, thus with conclusive end of treatment (EoT) tumor data, were categorized into discrete outcome groups. Results: Of 36 exp group patients reaching the first radiologic reassessment, enabling formal evaluation, 31 proceeded to EoT tumor data. Of these, 6 patients (3 of 25 with liver main lesions, 3 of 4 with lymph node main lesions) had complete response (CR), including 3 of 3 BRAF -V600E cases. The remaining 3 CR cases had tumor mutational burden (TMB) 9.4-11.8. Of all 25 patients with liver mets, 13 (52%) had objective response and 5 (20%) stable disease. All 16 patients with objective response had improved PFS (median 15.5 months, 95% CI 12.4-18.5; p < 0.001, log-rank test). None of main or subsidiary lesions in peritoneum or lungs responded to the treatment. The 3 outcome groups comprised 7 patients with PFS 19.8-41.6 months (longer than twice the median), 8 with PFS 9.9-16.4 months (above median), and 16 with PFS 1.9-9.2 months (below median). At baseline, the best outcome group cases would have been predicted by the combination of right-sided primary, small main lesions (sum of diameters 42 mm or less), and all mets confined to the liver and/or lymph nodes; the mid group cases by left-sided or rectal primary along with peritoneal or lung subsidiary lesions; and the poor outcome cases by extended organ mets. Conclusions: Alternating short-course oxaliplatin-based chemotherapy and nivolumab was particularly efficient in treating unresectable liver or lymph node mets from right-sided MSS CRC with intermediate TMB or the BRAF driver mutation, but inefficient at peritoneal and lung mets. Clinical trial information: NCT03388190 .