KEYNOTE-716: First Data on DMFS Show Additional Pembrolizumab Benefit in Patients With Melanoma

Evidence-Based Oncology, July 2022, Volume 28, Issue 5
Pages: SP254-SP255

New date from the KEYNOTE-716 trial show distant metastasis-free survival benefit from pembrolizumab use in melanoma.

The double-blind KEYNOTE-716 study (NCT03553836) is investigating the use of adjuvant pembrolizumab vs placebo among individuals with cutaneous stage IIb and IIc newly diagnosed, surgically resected (negative sentinel lymph node biopsy) high-risk melanoma. A third interim analysis of data from the ongoing 2-part study bears out the survival benefits of the monoclonal antibody (mAB) in the form of longer-term recurrence-free survival (RFS) and the first analysis of distant metastasis-free survival (DMFS).

“We know from retrospective studies that patients with resected stage IIb and IIc melanoma have a high risk of melanoma recurrence after 24 months, and these patients have very poor outcomes,” stated lead investigator Georgina V. Long, MBBS, PhD, co-medical director of the Melanoma Institute Australia (MIA) and chair of melanoma medical oncology and translational research at MIA and Royal North Shore Hospital, The University of Sydney. “We also know from KEYNOTE-716 that pembrolizumab significantly improves RFS vs placebo in resected stage IIb and IIc melanoma.”

Long presented the late-breaker abstract of this third interim analysis1 at the 2022 annual meeting of the American Society of Clinical Oncology (ASCO).

These newest findings—with a data cutoff of January 4, 2022—echo results from the first and second interim analyses, which demonstrate that pembrolizumab reduces the risk of disease recurrence compared with placebo and has a favorable benefit-risk profile in this disease setting. End points of the KEYNOTE-716 trial are RFS per investigator assessment (primary); DMFS, overall survival (OS), and safety (secondary); and quality of life (exploratory).

In KEYNOTE-716, participants were randomly assigned 1:1 to 200 mg of intravenous pembrolizumab (n = 487) every 3 weeks (adult dose) or 2 mg/kg (pediatric dose) or placebo (n = 489) every 3 weeks for 17 cycles. This is part 1 of the trial. Patient characteristics were balanced between the groups: the median (SD) age in the treatment group was 60 years (range, 16-84) and in the placebo group, 61 years (range, 17-87). The most common T subcategories were T3b or T4b, at 41% and 35%, respectively, in both groups; most had stage IIb disease, at 63% and 65%, respectively.

Part 2 of the trial occurred subsequent to a part 1 follow-up period and is a crossover/rechallenge of patients in both groups with disease recurrence. Part 2 data were not presented, but Long did note that patients in part 2 were unblinded and crossed over from placebo or continued treatment with the mAB if more than 6 months had passed since their last dose; treatment continued to disease progression or 2 years.

For the third interim analysis, the median follow-up was 27.4 months (range, 14.0-39.4). At the 12-month and 24-month marks for DMFS, pembrolizumab had produced superior outcomes to placebo, and the median was not reached in either group:

  • 12 months: 94.7% vs 90.2%
  • 24 months: 88.1% vs 82.2%

In addition, there were 63 distant metastases in the pembrolizumab group and 95 in the placebo group, translating to 24-month DMFS rates of 87.1% and 80.6%, respectively. Overall, this represents a 36% reduction in the risk of distant metastasis (HR, 0.64; 95% CI, 0.47-0.88), favoring pembrolizumab, Long noted.

The trend of pembrolizumab’s superiority continued when comparing DMFS rate results among the T subcategories based on cancer stage, as shown by the following:

  • T3b: at 12 months, 95.3% vs 93.9%; at 24 months, 90.4% vs 85.1%
  • T4a: at 12 months, 96.2% vs 91.3%; at 24 months, 92.7% vs 85.4%
  • T4b: at 12 months, 94.0% vs 86.6%; at 24 months, 83.0% vs 77.9%

Again, the median DMFS was not reached in any of these T subcategories, and the risk of distant metastasis was lower in the pembrolizumab group than in the placebo group:

  • For patients classified into the T3b subgroup, there were 23 vs 31 distant metastases, respectively, for an overall 29% reduced risk of distant metastasis (HR, 0.71; 95% CI, 0.411.22) and 24-month DMFS rates of 88.5% and 84.5%, respectively
  • For those classified into the T4a subgroup, there were 8 vs 20 distant metastases, respectively, for an overall 58% reduced risk of distant metastasis (HR, 0.42; 95% CI, 0.19-0.96) and 24-month DMFS rates of 92.7% and 82.8%
  • For those classified into the T4b subgroup, there were 30 vs 41 distant metastases, respectively, for an overall 30% reduced risk of distant metastasis (HR, 0.70; 95% CI, 0.44-1.13) and 24-month DMFS rates of 82.5% and 75.7%

The most common site of first distant metastasis was the lung in both study groups (49% in the pembrolizumab group vs 73% in the placebo group). This was followed by the brain (16%), lymph nodes (14%), and other soft tissue (13%) in the pembrolizumab group, and other soft tissue (13%), lymph nodes (12%), and brain (9%) in the placebo group. Regarding the higher percentages in the pembrolizumab group, Long clarified the fact that some patients had multiple sites of distant metastasis.

“In terms of patterns of distant metastases, we can see that for the pembrolizumab arm, there is a marked reduction in the number of patients who had distant metastasis as first event,” she stated. “A similar number of patients had distant metastases after locoregional recurrence.”

In fact, she continued, the DMFS favored pembrolizumab for all key subgroups. In addition to the T subcategories discussed above, the following results were seen for reduced risk of distant metastasis in this study cohort:

  • Age: younger than 65 years, 21% (HR, 0.79; 95% CI, 0.51-1.23); 65 years and older, 49% (HR, 0.51; 95% CI, 0.32-0.82)
  • Gender: female, 24% (HR, 0.76; 95% CI, 0.441.32); male, 42% (HR, 0.58; 95% CI, 0.39-0.86)
  • Geographic region: United States, 45% (HR, 0.55; 95% CI, 0.21-1.44); outside United States, 33% (0.48-0.94)
  • ECOG status: Pembrolizumab participants, 31% (HR, 0.69; 95% CI, 0.49-0.95)

As for how RFS was affected by occurrences of distant metastasis, pembrolizumab was shown to have produced sustained benefits at 24 months in the pembrolizumab vs the placebo group, Long noted. There was a 36% overall reduced risk of disease recurrence seen among those receiving pembrolizumab (HR, 0.64; 95% CI, 0.50-0.84), and the overall RFS rates were 81.2% and 72.8% in the pembrolizumab and placebo groups, respectively. In addition, the median RFS was 37.2 months in the pembrolizumab group and not reached in the placebo group.

When looking at RFS and DMFS in the context of the 3 T subcategories, the results favored pembrolizumab in all 3. Among these groups overall, there was an 81.2% RFS rate in the pembrolizumab group compared with 72.8% in the placebo group, or an overall 36% reduced risk of disease recurrence (HR, 0.64; 95% CI, 0.50-0.84) compared with placebo.

This reduced risk of disease recurrence with pembrolizumab was greatest among those in the T4a subcategory at 52% (HR, 0.48; 95% CI, 0.26-0.90), followed by T3b at 43% (HR, 0.57; 95% CI, 0.36-0.89) and T4b at 24% (HR, 0.74; 95% CI, 0.51-1.13). These totals correspond with 24-month RFS rates of 89.1%, 86.8%, and 71.3%, respectively, for pembrolizumab compared with 75.6%, 76.2%, and 67.8% with placebo. A median RFS of 37.2 months with pembrolizumab was only reached overall and among those in the T4a subgroup; it was not reached among those in the T3b or T4b subcategories nor among those in any T subcategories in the placebo group.

Among those with disease recurrence, fewer patients in the pembrolizumab group had distant recurrence or local, regional, and locoregional recurrence as their pattern of first recurrence, at 9% each, compared with 16% and 11%, respectively, among the placebo group. One percent of participants from each group whose disease recurred died. Overall, there were 95 instances of disease recurrence in the pembrolizumab group vs 139 in the placebo group.

For those in the pembrolizumab group, the top 3 most common treatment-related adverse events (TRAEs) were pruritus (25%), fatigue (21%), and diarrhea (19%). TRAEs of grade 3 or higher most commonly reported were pruritus, rash, and diarrhea (1% each). For the placebo group, fatigue (19%), diarrhea (12%), and pruritus (11%) were the top 3 TRAEs reported overall, and few grade 3 or higher TRAEs were reported; rash, diarrhea, and fatigue were reported at less than 1% each.

“Adjuvant pembrolizumab vs placebo for resected stage IIb and stage IIIc melanoma significantly improved DMFS,” Long concluded. “There were fewer distant metastases as instances of RFS, sustained and continued reduction in recurrence risk, and a benefit-risk profile that is consistent with the first and second interim analyses.” The estimated study completion date of KEYNOTE-716 is October 12, 2033.2


1. Long GV, Luke JJ, Khattak M, et al. Distant metastasis-free survival with pembrolizumab versus placebo as adjuvant therapy in stage IIB or IIC melanoma: the phase 3 KEYNOTE-716 study. Abstract presented at: 2022 ASCO Annual Meeting; June 3-7, 2022; Chicago, IL. Accessed June 5, 2022.

2. Safety and efficacy of pembrolizumab compared to placebo in resected high-risk stage II melanoma (MK-3475-716/KEYNOTE-716). Updated August 24, 2021. Accessed June 5, 2022. https://