Single-Agent Immunotherapy Made a Difference in the MSI-H Subtype of Stage 4 Colon and Rectal Cancer

Colorectal (colon and rectal) cancers account for 9% of all cancer-related deaths. Despite increased cancer screening methods, 20% of patients are metastatic / stage 4 at diagnosis. After surgery, 40% of patients with early-stage disease will experience cancer recurrence later on. Today, advances in targeted therapies and the introduction of immunotherapy agents have changed the treatment of colorectal cancers, as they have for other cancer types.

As cancer biology becomes better understood and biomarkers that help predict which patients will benefit from which treatments are identified, colorectal cancer treatment has come to involve individualized plans based not only on tumor location but also on genetic features.

With the contribution of targeted therapies, biomarker / molecular testing in metastatic colorectal cancer has evolved over time. Molecular biomarkers in colon and rectal cancer include the following (let us recall that colorectal cancers are now also classified according to these biomarkers):

• KRAS, NRAS mutations determine the benefit of anti-EGFR therapy,
• BRAF V600 mutations guide corresponding targeted therapies (such as encorafenib, binimetinib and cetuximab),
• HER2 amplification,
• NTRK1-3 mutations have been targeted with drugs that have provided substantial survival advantages.

One mechanism our cells use to repair DNA damage is the “DNA mistmach repair” (mismatch repair) system. Mutations in the genes encoding this DNA repair system lead to DNA mismatch repair deficiency and the accumulation of mismatched DNA sequences called microsatellites, resulting in genomic instability. Tumors with a high concentration of microsatellites in this way are considered DNA mistmach repair deficient (dMMR) or microsatellite instability-high (MSI-H) tumors. These tumors have a high tumor mutational burden, high neoantigen levels and increased immune infiltration.

MSI-H / dMMR Positivity varies with the stage of colorectal cancer. It is common in early-stage disease (20% in stages I and II and 12% in stage III), but occurs in 4-5% of metastatic cases. Most of these patients have sporadic disease (that is, it is not hereditary). However, in 20%, it is hereditary, with an underlying condition such as Lynch syndrome. Therefore, testing DNA repair genes should be requested for all patients with colorectal cancer, both to detect Lynch syndrome and to predict the response to immunotherapy.

MSI-H / dMMR status had been shown to predict immunotherapy response not only in patients with colorectal cancer but in all solid / organ tumors. In light of these studies, pembrolizumab entered clinical use in 2017 after receiving FDA approval for MSI-H or dMMR tumors regardless of the organ in which the cancer originated (see the related approval). In the study supporting this approval, pembrolizumab produced an objective response rate of 40% among 90 patients with metastatic colorectal cancer who had previously received standard chemotherapy.

Launched on the basis of these studies, KEYNOTE-177, the clinical trial comparing pembrolizumab with chemotherapy as first-line treatment in patients with dMMR / MSI-H metastatic colorectal cancer, made a major impact at the ASCO 2020 Virtual Congress.

A total of 307 previously untreated patients with MSI-H metastatic colorectal cancer were enrolled. The median follow-up was 32.4 months.

• Study group A: Pembrolizumab 200 mg every 3 weeks, for up to 35 cycles.
• Study group B: The FOLFOX (oxaliplatin, fluorouracil and leucovorin) or FOLFIRI (fluorouracil, leucovorin and irinotecan) chemotherapy regimen was combined with cetuximab or bevacizumab.

Results

• Looking at treatment responses, median progression-free survival (PFS) was an additional 8 months longer in the pembrolizumab arm (16.5 months with pembrolizumab versus 8.2 months with chemotherapy).
• Both the 12-month progression-free survival rate (55% vs 37%) and the 24-month PFS rate (48% vs 19%) were better in the pembrolizumab-only arm. • The benefit of pembrolizumab was present in all patient subgroups except those with KRAS mutations.
• Objective response rates were 11% better in the pembrolizumab arm (43.8% with pembrolizumab vs 33.1% with chemotherapy).
• Among patients receiving pembrolizumab, 48% continued to live without progression. Responses lasted over 2 years in 83% with pembrolizumab versus 35% with chemotherapy. Median response duration was not reached with pembrolizumab, compared with 10 months with chemotherapy.
• Regarding adverse events, grade 3 or higher events occurred in 22% of the pembrolizumab arm versus 66% of the chemotherapy arm.
• The most common immune-related adverse events were hypothyroidism (12%) and colitis (7%). There were no deaths due to adverse events, while the rate of discontinuing treatment because of adverse events was 10% in the pembrolizumab arm.
• Although progression was more frequent in the pembrolizumab arm during the first 6 months, the benefit of pembrolizumab increased after 6 months (29% versus 12%). In the subgroup analysis, patients with KRAS mutations benefited more from chemotherapy. Of patients who progressed on chemotherapy, 36% crossed over to the pembrolizumab arm.

Commentary

The study’s overall survival data are still awaited. In addition to the progression-free survival data, the overall response rate, median response duration, continued treatment benefit beyond 2 years in 83% of responders, and 44% less toxicity favor pembrolizumab.

An important point to emphasize is resistance to immunotherapy in the group that developed progression in the first 6 months—that is, one in every 3 patients. Further biomarker analysis of immunotherapy resistance will reveal which patients will not benefit from treatment. Our approach to symptomatic patients or those with poor performance status in this currently uncertain group remains a question. Will we truly see an OS benefit in this study? Long-term follow-up will show.

Following this study, new research in MSH-H colorectal cancer will investigate whether this approach is also effective in the adjuvant and neoadjuvant settings.

The current findings suggest that this study will influence our daily practice and standard treatment approach.