Comment on Le DT et al. Phase II Open-Label Study of Pembrolizumab in Treatment-Refractory, Microsatellite Instability-High/Mismatch Repair-Deficient Metastatic Colorectal Cancer: KEYNOTE-164. J Clin Oncol. 2019 Nov 14:JCO1902107. doi: 10.1200/JCO.19.02107. PMID: 31725351 1 by Prof. Dr. R.-D. Hofheinz, Head of the Outpatient Treatment at Interdisciplinary Tumor Center, University Hospital Mannheim, University of Heidelberg, Germany.
Mismatch-repair deficiency in tumors arises through two routes, either from inherited germline defects in mismatch-repair genes or from sporadic inactivation of both alleles of a mismatch-repair gene by somatic mutations or by epigenetic silencing. In either case, these tumors harbor hundreds or thousands of mutations. Four and a half years ago, Le and coworkers were the first to test the hypothesis that these tumors may respond to checkpoint-inhibitors. 2 In their pivotal study 9 out of 17 evaluable, heavily pre-treated patients with microsatellite instability-high/mismatch repair-deﬁcient (MSI-H/dMMR) tumors responded to pembrolizumab. Meanwhile, several trials have confirmed this finding in MSI tumors. Not surprisingly, a certain focus in the investigation of checkpoint inhibitors in MSI tumors has been placed on metastatic colorectal cancer (mCRC).
Recently, the long-term results of the KEYNOTE-164 study have been reported. This two-cohort study recruited patients with pre-treated MSI-H/dMMR mCRC to receive three-weekly pembrolizumab at a dose of 200 mg. 1 A total of 14 patients had BRAF mutations. Response rates were 33% in patients pre-treated with ≥2 (cohort A; n=61 patients) or ≥1 (cohort B; n=63) lines of treatment, respectively. Responses were also noted in BRAF mutated tumors (6/14; 43%). After a median follow-up of 31 (cohort A) and 24 months (cohort B), long-term benefit was seen: two-year PFS rates were 31% and 37%, respectively. The lines of prior therapy administered for cohorts A and B largely overlap. However, adding response rates up according to lines of pre-treatment, a response rate of 30% (9/30) is seen in the group of patients with one prior treatment line, response rate was 42% (29/48) in those with two prior treatment lines, and 26% (12/46) in patients with three or more prior therapies. Thus, the author’s interpretation that treatment with programmed cell death 1 (PD-1) blockade in earlier lines could be more effective is not clearly substantiated by the presented response data. Although median survival was 31.4 months in cohort A, it had not been reached in cohort B. Treatment emergent adverse events (TEAEs) of grades 3-4 were reported in 18 out of 123 patients. No fatal TEAEs were observed. Grade 3-4 immune-mediated adverse events and infusion reactions were reported in six patients, including pancreatitis (n=2), pneumonitis (n=2), hepatitis (n=1), colitis (n=1), skin toxicity (n=1), and pneumonitis (n=1).
The authors correctly summarize that KEYNOTE-164 confirms the durable clinical beneﬁt of pembrolizumab in patients with previously treated MSI-H/dMMR mCRC. Comparably, in the CheckMate 142 study, patients with pre-treated MSI-H/dMMR mCRC receiving monotherapy with nivolumab after ≥1 line of pre-treatment exhibited a response rate of 31% and a one-year survival rate of about 70%. 3 Thus, although checkpoint inhibitors have not yet been approved by the European Medicines Agency, these drugs represent a valuable treatment option for patients with MSI-H/dMMR mCRC and other tumors.
However, despite excellent response rates and long-term disease control several open questions remain. First, it has to be noticed that the rate of primary progression ranged between 26% and 46% in both studies. 1,3 Preliminary data of patients with MSI-H/dMMR mCRC treated with first-line nivolumab in another cohort of the CheckMate 142 study indicate that 16% of patients exhibited primary progression. 4 These data underscore the need for additional biomarkers beyond MSI-H/dMMR to minimize the number of patients exposed to checkpoint inhibitors instead of potentially active chemotherapy regimens. This holds true especially for those patients scheduled to receive checkpoint inhibitors in earlier lines of treatment where other treatment options are available. Recently, Schrock and coworkers suggested that tumor mutational burden (TMB) might serve as an additional biomarker. 5 In a cohort of 22 patients with mCRC treated with checkpoint inhibitors, a cut-point of 37-41 mutations/Mb for TMB was reported to distinguish between responders and nonresponders. Although all 13 patients with TMB values above this threshold exhibited long-term benefit, 6 out of 9 patients with lower values showed primary progression. Clearly, given the fact that there is no consensus on how to determine TMB and what cut-offs should be used, further investigations are awaited. Second, it remains unclear if combinations of checkpoint inhibitors (e.g. nivolumab/ipilimumab) should be preferred. 6 Moreover, checkpoint inhibitors were also able to induce responses and long-term disease control in patients with BRAF V600E mutated MSI-H/dMMR mCRC. Given the new treatment options on the horizon for these patients (i.e. the combination of encorafenib, binimetinib, and cetuximab), 7 one may wonder which of these treatment regimens should be preferred.
Finally, the results of KEYNOTE-177 (NCT02563002) comparing first-line pembrolizumab with the standard of care chemotherapy are eagerly awaited.