Frontline pembrolizumab outdoes CT for advanced colorectal cancer

In individuals with metastatic MSI-H/dMMR* colorectal cancer (CRC), first-line treatment with the PD-1 inhibitor pembrolizumab led to durable responses and a better safety profile than chemotherapy (CT), the phase III KEYNOTE-177 trial has shown.

CT remains the standard of care in this setting; however, some cases remain refractory to CT, for which PD-1 inhibition has shown some promise. [Science 2017;357:409-413; J Clin Oncol 2020;38:11-19; Lancet Oncol 2017;18:1182-1191] “[P]embrolizumab … led to durable responses in some patients with previously treated metastatic MSI-H/dMMR CRC, a finding that contributed to [its] FDA approval … for [cases] that have progressed after treatment with a fluoropyrimidine, oxaliplatin, and irinotecan,” noted the researchers.

A total of 307 participants (median age 63 years) who had not received treatment were randomized 1:1 to receive pembrolizumab 200 mg Q3W or doublet chemotherapy (CT)** Q2W. Crossover from CT to pembrolizumab was allowed in case of disease progression. Pembrolizumab treatment ran for a maximum of 35 treatments or until disease progression, unacceptable toxicities, or withdrawal. [N Engl J Med 2020;383:2207-2218]

After a median follow-up of 32.4 months, progression-free survival (PFS) was doubled with pembrolizumab vs CT (median, 16.5 vs 8.2 months; hazard ratio, 0.60; p=0.0002). More pembrolizumab vs CT recipients were alive and progression-free at 12 months (55 percent vs 37 percent) and 24 months (48 percent vs 19 percent).

“After an initial crossing of the PFS Kaplan-Meier curves, a pronounced separation of the curves for pembrolizumab and CT was observed, which indicated a meaningful long-term benefit with pembrolizumab,” the researchers explained.

Moreover, the estimated restricted mean survival time – a complementary PFS analysis conducted upon violation of the proportional-hazards assumption – at 24 months preferred pembrolizumab over CT (13.7 vs 10.8 months).

The researchers noted that when the proportional-hazards assumption is violated, treatment effect may change over time. “[Therefore,] evaluation of treatment effect must consider multiple factors*** … to reflect totality of the data. Differences in these factors were consistently favourable for pembrolizumab vs CT in [our] trial.”

More pembrolizumab vs CT recipients had an overall response (44 percent vs 33 percent), with complete responses (CR) in 11 percent and 4 percent, respectively. These results correlate with data reflecting higher CR rates with pembrolizumab or other ICIs^# vs CT. [J Clin Oncol 2020;38:1-10; J Natl Cancer Inst 2020;doi:10.1093/jnci/djaa052] There were also more pembrolizumab vs CT recipients who had ongoing responses by month 24 (83 percent vs 35 percent).

The median PFS and overall response seen with CT correlate with data reflecting its limited efficacy in this setting. [J Clin Oncol 2019;37:1217-1227; Int J Cancer 2020;147:285-296]

The incidence of grade ≥3 adverse events (AEs) was lower with pembrolizumab vs CT (56 percent vs 78 percent), as were the rates of grade ≥3 treatment-related AEs (22 percent vs 66 percent; with one death in the CT arm) and AEs attributed by the investigators to treatment (80 percent vs 99 percent).

Another step forward

Overall survival (OS) was not reported, as these data were still evolving, noted the researchers. “The trial [will be deemed] successful if pembrolizumab was superior to CT with respect to either primary endpoint.”

However, with 59 percent of CT recipients crossing over to pembrolizumab, “it is conceivable that this high rate of crossover will make it difficult to ever achieve a significant difference in … OS,” noted Dr Axel Grothey from the West Cancer Center and Research Institute, Germantown, Tennessee, US, in an editorial. [N Engl J Med 2020;383:2283-2285]

Of note as well was the greater fraction of pembrolizumab vs CT recipients having progressive disease as best response (29 percent vs 12 percent). “This is mirrored by an early poorer performance [in the pembrolizumab vs the CT arm] until about 6.5 months after onset of therapy,” Grothey noted. However, this improved thereafter, he stressed.

“[As such,] for MSI-H/dMMR CRC, durability of response, better safety profile, and improved quality of life associated with immunotherapy [vs] CT make pembrolizumab the preferred choice,” said Grothey.

“[P]embrolizumab should be considered an option for initial therapy [in this setting],” the researchers echoed. “These data represent another step forward for biomarker-driven studies targeting MSI-H/dMMR CRCs.”