Adjuvant Pembrolizumab Maintains RFS Benefit in Resected High-Risk Stage II Melanoma

Jason J. Luke, MD

Adjuvant pembrolizumab (Keytruda) continued to showcase improved relapse-free survival (RFS) over placebo in patients with resected high-risk stage II melanoma, according to findings from the second interim analysis of the phase 3 KEYNOTE-716 trial (NCT03553836).¹

Data, which were presented during the 2021 Society for Melanoma Research Congress, showed that the median RFS had not yet been reached (NR) with either pembrolizumab (95% CI, NR–NR) or placebo (95% CI, 29.9–NR; hazard ratio [HR], 0.61; 95% CI, 0.45-0.82). The 18-month RFS rates in the investigative and control arms were 85.8% and 77.0%, respectively.

Moreover, the HR for RFS at the time of the second interim analysis was further improved compared with that reported at the first analysis (HR, 0.65; 95% CI, 0.46-0.92; P = .00658), and RFS continued to favor the immunotherapy across all key subgroups evaluated.

“With 6 months of additional follow-up, pembrolizumab continued to be associated with a reduction in the risk of disease recurrence or death compared with placebo and had a manageable safety profile,” lead study author Jason J. Luke, MD, of UPMC Hillman Cancer Center, said in a presentation on the data.

It is known that patients who have stage IIB and IIC melanoma are at increased risk of disease recurrence and experience outcomes that are comparable to what is seen in those with stage IIIA and IIIB disease. For those with stage II disease, current guidelines call for wide excision and observation or participation on a clinical trial examining adjuvant therapy.

KEYNOTE-716 enrolled patients with newly diagnosed, resected, high-risk stage II melanoma who were at least 12 years of age and who had an ECOG performance status of 0 or 1. In part 1 of the study, participants were randomized 1:1 to receive either intravenous (IV) pembrolizumab at 200 mg every 3 weeks or 2 mg/kg for pediatric patients (n = 487), or IV placebo every 3 weeks (n = 489). Treatment was administered for 17 cycles.

Upon recurrence, patients moved to the second portion of the trial. Those in the investigative arm were rechallenged with pembrolizumab, given at the same dose and schedule as part 1, and those in the control arm crossed over to receive the immunotherapy. Treatment was given until disease progression or recurrence or for up to 2 years.

Patients were stratified based on T category (3b vs 4a vs 4b) and pediatric status.

The primary end point of the trial was investigator-assessed RFS, and key secondary end points included distant metastasis-free survival, overall survival, and safety. Health-related quality of life (HRQoL) served as an exploratory end point.

Of the 487 patients allocated to receive pembrolizumab, 483 received treatment. At the time of data cutoff, 24 were still receiving treatment, 297 completed treatment, and 162 patients discontinued. The majority of those who discontinued did so because of adverse effects (AEs; n = 85), followed by patient withdrawal (n = 39), relapse or recurrence (n = 25), physician decision (n = 9), or a protocol violation (n = 4).

Of the 489 patients who were randomized to the control arm, 486 received treatment with 17 patients still on treatment and 353 having completed therapy. A total of 116 patients discontinued treatment on this arm, with most patients doing so because of relapse or recurrence (n = 60), followed by patient withdrawal (n = 26), toxicities (n = 23), physician decision (n = 4), and a protocol violation or another reason (n = 3).

The median time from randomization to date of death or database cutoff was 20.5 months (range, 4.6-32.7). The median treatment duration was 11.1 months with pembrolizumab (range, 0.0-16.4) and 11.1 months (range, 0.0-15.6) with placebo.

Across the arms, the median age was 60.5 years, and most patients were between the ages of 18 years and 64 years. Moreover, most patients were male, White, not from the United States, had an ECOG performance status of 0, a T category of 3b, and had stage IIB disease.

Results from the first interim analysis showed that the immunotherapy significantly reduced the risk of disease recurrence or death vs placebo in this patient population.² At a median follow-up of 14.4 months, the proportion of patients with distant recurrence was 4.7% with pembrolizumab vs 7.8% with placebo. The rates of grade 3 or 4 treatment-related AEs (TRAEs) in the investigative and control arms were 16.1% and 4.3%, respectively. No grade 5 toxicities were reported, and HRQoL was noted to be maintained with the immunotherapy.

Additional updated data from a sensitivity analysis including primary melanoma showed that the median RFS had not yet been reached for either the pembrolizumab arm (95% CI, NR–NR) or the placebo arm (95% CI, 29.9–NR; HR, 0.60; 95% CI, 0.45-0.79).

Among those with T3b disease, the median RFS had not yet been reached in the investigative arm (95% CI, NR–NR) and in the control arm (NR–NR; HR, 0.40; 95% CI, 0.23-0.69). The 18-month RFS rates in the pembrolizumab and placebo arms were 91.1% and 77.4%, respectively.

In the subgroup of patients with T4a disease, the median RFS had not yet been reached (95% CI, 24.9–NR) with the immunotherapy vs 29.9 months (95% CI, 28.2–NR) with placebo (HR, 0.49; 95% CI, 0.24-1.00). The 18-month RFS rates in the investigative and control arms were 90.4% and 81.8%, respectively.

In those with T4b disease, the median RFS had not yet been reached (95% CI, 24.2–NR) with pembrolizumab or placebo (95% CI, 25.2–NR; HR, 0.82; 95% CI, 0.54-1.26). The 18-month RFS rate in the investigative arm was 78.9% vs 74.1% in the control arm.

When looking at patterns of recurrence, 85.2% of patients who received pembrolizumab (n = 487) did not experience a RFS event vs 76.5% of those who were given placebo (n = 489). Conversely, 14.8% of those on the investigative arm and 23.5% of those on the control arm experienced an RFS event.

Among those who did have an event and received the immunotherapy, 7.8% had local or regional or locoregional recurrence and 6.4% experienced distant recurrence; these rates were 10.2% and 12.3%, respectively, in those who were given placebo. Three patients on the investigative arm died vs 5 on the control arm.

Regarding safety, 95.4% of those on the pembrolizumab arm (n = 483) experienced any-grade AEs vs 91.4% of those on the placebo arm (n = 486); 28.2% and 19.1% of patients, respectively, reported grade 3 to 5 toxicities. Moreover, 82.8% of those who received the immunotherapy experienced treatment-related AEs (TRAEs) vs 63.4% of those who received placebo; 17.0% and 4.3% of patients, respectively, experienced grade 3 to 5 TRAEs.

Moreover, 16.4% of patients on the immunotherapy experienced a toxicity that resulted in treatment discontinuation vs 2.5% of those on placebo. No toxicities resulted in death.

Additionally, 37.7% of those on the investigative arm and 9.1% of those on the control arm reported immune-mediated AEs. Grade 3 or 4 immune-mediated AEs occurred in 10.1% of those who received pembrolizumab and 1.2% of those given placebo.

AEs of special interest included hypothyroidism, hyperthyroidism, severe skin reactions, colitis, hypophysitis, adrenal insufficiency, hepatitis, pneumonitis, thyroiditis, myositis, nephritis, sarcoidosis, infusion reactions, myasthenic syndrome, pancreatitis, T1DM, uveitis, myelitis, and myocarditis.

Among the 107 patients who had an event of interest in the investigative arm, 20.3% received hormonal therapy. Among the 18 patients with an event of interest in the control arm, 1.2% received hormonal therapy.

References

1. Luke JJ, Long GV, Khattak MA, et al. Pembrolizumab versus placebo after complete resection of high-risk stage II melanoma: updated results from the KEYNOTE-716 trial. Presented at: 2021 Society for Melanoma Research Congress; October 28-31, 2021; virtual.
2. Luke JJ, Rutkowsi P, Queirolo P, et al. Pembrolizumab versus placebo after complete resection of high-risk stage II melanoma: efficacy and safety results from the KEYNOTE-716 double-blind phase III trial. Ann Oncol. 2021;32(suppl 5):S1283-S1346. doi:10.1016/annonc/annonc741