Neoadjuvant Nivolumab Shows Strong Clinical Activity in Merkel Cell Carcinoma

Suzanne L. Topalian, MD, of Johns Hopkins Medicine

Suzanne L. Topalian, MD

Nivolumab (Opdivo), when administered approximately 4 weeks prior to surgery in patients with Merkel cell carcinoma (MCC), was found to be tolerable and to induce pathological complete responses (pCRs) as well as radiographic tumor regressions in about half of treated patients, according to results from the phase 1/2 CheckMate-358 trial. ¹

"Neoadjuvant nivolumab therapy in high-risk resectable MCC mediated major pathologic and radiographic responses in approximately one half of patients, with significantly improved recurrence-free survival among responders," wrote the lead investigator, Suzanne L. Topalian, MD, of Johns Hopkins Medicine, and colleagues in the paper. "This treatment regimen seemed to be generally tolerable and should be further explored as a potentially beneficial adjunct to surgery in patients with MCC."

The study enrolled 39 patients with American Joint Committee on Cancer stage IIa-IV resectable MCC who received ≥1 nivolumab dose. Of these patients, 7.7% (n = 3) did not undergo surgery because of either tumor progression (n = 1) or adverse events (AEs; n = 2). Any-grade treatment-related AEs occurred in 46.2% of patients (n = 18), and grade 3 to 4 events were reported in 7.7% of patients (n = 3). No unexpected toxicities were observed.

Furthermore, among 36 patients who underwent surgery, 47.2% experienced a pCR (n = 17). Among 33 patients who underwent surgery and were radiographically evaluable, 54.5% (n = 18) experienced tumor reductions of ≥30%. At a median follow-up of 20.3 months, median RFS and overall survival (OS) had not yet been reached. Notably, investigators found that RFS significantly correlated with pCR and radiographic response at the time of surgery. Moreover, no patient with a pCR experienced tumor relapse during observation.

"In CheckMate-358, the efficacy of neoadjuvant anti—PD-1 did not correlate with baseline MCC viral status or PD-L1 expression," the authors wrote. "Also consistent with prior studies, Merkel cell polyomavirus (MCPyV)–positive tumors had low tumor mutational burden, suggesting that a limited number of strong viral antigens can serve as tumor rejection antigens in MCC and potentially in other virus-associated cancers. Multiplex markers may provide increased sensitivity and specificity for treatment outcomes."

In the multicenter, open-label, multicohort CheckMate-358 trial, investigators are evaluating the use of nivolumab monotherapy or nivolumab-based combination therapies in patients with virus-associated solid tumors in the recurrent/metastatic or neoadjuvant setting.

The primary end point of the study was the safety and tolerability of neoadjuvant nivolumab measured by treatment-related AEs (TRAEs) and surgical delays. Exploratory end points included pCR rate as assessed by investigators and an independent pathologist, radiographic response, RFS, OS, immunologic changes in blood and tumor, and association of tumor MCPyV status and PD-L1 expression with efficacy.

Between January 2016 and March 2019, 39 patients received nivolumab 240 mg intravenously on days 1 and 15, and then underwent surgery on day 29. Delayed administration of the second dose of the drug was considered to be acceptable up to day 22. In the neoadjuvant MCC cohort, eligible patients were ≥18 years of age with an ECOG performance score of 0 or 1. The median age of the participants was 68 years (range, 22-88 years).

Most patients had American Joint Committee on Cancer stage III disease at the time of study enrollment (66.7%). Among 35 patients with tumors evaluable for MCPyV status, 62.9% were found to be positive. Among the 27 patients who had quantifiable tumor cell PD-L1 expression, 25.9% (n = 7) had expression of ≥1%. A total of 36 out of 39 patients received both planned doses of nivolumab; 3 patients received only 1 dose.

Results showed that any-grade TRAEs were observed in 46.2% of all patients (n = 18/39). Only 7.7% of patients (n = 3) reported TRAEs that were grade 3 to 4 in severity. Additionally, no treatment-related deaths occurred. A total of 15.4% of patients (n = 6) experienced select TRAEs with potential immunologic cause; these included 2 grade 3 to 4 events (1 skin and 1 gastrointestinal). The most commonly observed any-grade select TRAE, were skin reactions (10.3%).

Of the 39 treated patients on the trial, only 3 did not undergo surgery. One patient discontinued following tumor progression, another withdrew consent because of grade 2 nausea that was not determined to be related to study treatment, and the last patient discontinued because of a treatment-related rash that was grade 3 in severity.

Among the 36 patients who underwent surgery, the median interval between the first dose of nivolumab and the procedure was 4.3 weeks (range, 2.9-15.0). One patient (2.6%) had surgery delayed >4 weeks because of treatment-related autoimmune colitis, with the procedure performed on day 105. Three patients (7.7%) had surgery delayed >7 days but ≤4 weeks due to administrative reasons; their surgeries were performed on days 37, 39, and 44, respectively. Six deaths occurred on study; 4 of them were because of disease progression and the other 2 were due to AEs unrelated to the study treatment, according to the investigators.

All 36 patients who underwent surgery were assessed for pCR via the study investigators; of these patients, 47.2% (n = 17) achieved a pCR. Definitive resection specimens obtained from 26 patients were examined as well. Results showed that 46.2% of patients (n = 12) experienced a pCR and 15.4% (n = 4) achieved a major pathologic response (MPR) to treatment; this yielded a pCR plus MPR rate of 61.5%.

Of those who underwent surgery, 33 were determined to be evaluable for radiographic response. Results showed that 87.9% of patients (n = 33) experienced a radiographic tumor reduction and 54.5% (n = 18) experienced a tumor reduction of ≥30%; specifically, the median change in tumor burden from baseline was —32.8%. Radiographic reduction of ≥30% was noted in patients whose tumors were MCPyV-positive or -negative, and PD-L1–positive or –negative. No trends were observed.

In the patients who underwent surgery, the median RFS was not reached. At 12 and 24 months postoperatively, RFS rates were 77.5% (95% CI, 58.4%-88.7%) and 68.5% (95% CI, 47.5%-82.6%), respectively. When comparing patients who achieved pCR versus those who did not via site review (n = 36), the RFS at 12 months was 100.0% compared with 59.6%, respectively, and 88.9% versus 52.2%, respectively, at 24 months (HR, 0.12; 95% CI, 0.01-0.93). Similarly, when comparing patients with pCR/MPR versus without pCR/MPR, the RFS at 12 months was 100.0% versus 50.0%, respectively; at 24 months, the rates were 88.9% versus 50.0%, respectively (HR, 0.09; 95% CI, 0.01-0.80).

Moreover, among the 33 radiographically evaluable patients, when comparing those with radiographic tumor reduction ≥30% versus those with reduction <30% or progression, the RFS at 12 months was 100.0% versus 56.5%, respectively, and 90.9% versus 48.5%, respectively, at 24 months (HR, 0.11; 95% CI, 0.01-0.87). The investigators noted that no substantial difference in RFS was identified between subgroups on the basis of tumor PD-L1 expression or MCPyV status.

Among the 39 treated patients, the median OS was not reached. At 12 and 24 months following the first dose of nivolumab, OS rates were 93.2% (95% CI, 75.5%-98.3%) and 79.4% (95% CI, 56.9%-91.0%), respectively. Among patients who underwent surgery and were considered evaluable for pathologic response (n = 36) or radiographic response (n = 33), 100.0% and 88.9% of patients with pCR via site review and all patients with radiographic tumor reduction of ≥30% were alive at 12 and 24 months, respectively.

The fact that no tumor relapses occurred after pCR/MPR suggested that standard adjuvant radiotherapy may not be required in responders, the investigators noted. However, relapses occurred among those without pCR/MPR, which suggests that a period of postoperative anti—PD-1 may be appropriate for some, they added.

"The possibility of tailoring the extent of surgery and/or administration of postsurgical therapy according to radiographic and pathologic response status after neoadjuvant anti—PD-1 should be examined in future MCC trials, which have the potential to be practice changing," the authors wrote.

Topalian SL, Bhatia S, Amin A, et al. Neoadjuvant nivolumab for patients with resectable Merkel cell carcinoma in the CheckMate 358 trial [published ahead of print on April 23, 2020]. J Clin Oncol. 2020. doi:10.1200/JCO.20.00201