First-line ipilimumab plus nivolumab in advanced merkel cell carcinoma: a meta-analysis of prospective trials and real-world validation cohort.

Introduction
Merkel cell carcinoma (MCC) is an aggressive neuroendocrine skin cancer with high metastatic potential. Immune checkpoint blockade (ICB) targeting the programmed death-1 (PD-1) receptor or its ligand (PD-L1) is the current standard first-line therapy options for advanced MCC. Based on non-randomized phase II trials, monotherapy with anti-PD(L)1 therapeutics including pembrolizumab, avelumab, nivolumab, and retifanlimab results in objective response rates (ORRs) between ~ 40 and 60% in frontline MCC [1–4].
Combination checkpoint blockade with anti-PD-1 plus anti-cytotoxic T lymphocyte-associated protein-4 (CTLA-4) is a commonly utilized systemic treatment approach in ICB-responsive tumor types, typically with higher response rates as well as higher toxicity compared to monotherapy approaches [5, 6]. In MCC, the combination of ipilimumab plus nivolumab (IPI + NIVO) has meaningful disease activity in the anti-PD(L)1 refractory setting and is an increasingly commonly utilized salvage option in this disease [7]. However, the role of IPI + NIVO as a first-line treatment option for advanced MCC remains unclear and controversial [8]. Two independent, non-randomized studies that evaluated IPI + NIVO as a first-line MCC treatment reported divergent results. In one trial, all 22 patients treated with first-line IPI + NIVO achieved objected response, suggesting very promising efficacy of this regimen [7]. However, a second trial that included 33 treatment naïve patients observed a more modest ~ 60% ORR to IPI + NIVO in a similar patient population, more in line with expectations of anti-PD-1 monotherapy [3].
As there have been no randomized trials comparing ICB combination to monotherapy in MCC, clinicians must rely on best available non-randomized trial data when selecting a first-line regimen in this disease. We therefore conducted a systematic review and meta-analysis to summarize the current evidence of first-line ICB therapy in MCC and to compare outcomes between combination of IPI + NIVO and ICB monotherapy. Further, to add to the published experience and as real-world validation, we performed a retrospective chart review of MCC patients treated with first-line IPI + NIVO at our institution.

Material and methods

Meta-analysis
A systemic review was conducted using MEDLINE and by searching for relevant abstracts presented at the European Society of Medical Oncology, American Society of Clinical Oncology, and Society for Immunotherapy of Cancer published or presented prior to 04/01/2025. Search terms included ("merkel cell carcinoma"[Title]) AND ("immunotherapy"[Title] OR "pembrolizumab"[Title] OR "avelumab"[Title] OR "retifanlimab"[Title] OR "nivolumab"[Title] OR "ipilimumab"[Title]). We included for analysis studies that 1) prospectively evaluated anti-PD(L)1 monotherapy or anti-PD-1 + anti-CTLA-4 as first-line systemic therapy in advanced MCC and 2) reported ORR of the primary study population. Reviews, case reports, and retrospective studies were not eligible. When the same trial source was included in different publications or abstracts, the most recent and relevant was used. Reported response data were extracted and checked by three authors (T.R., C.P., and A.S.B.) with disagreements resolved by consensus. The primary outcome was the efficacy of ICB treatment by objective response rate (ORR). To assess the ORRs, we calculated the pooled proportions and 95% confidence intervals (CIs) using the Freeman–Tukey double arcsine transformation to stabilize variance and applied a fixed-effect model due to the limited number of studies. We performed two subgroup analyses: ORR by IPI + NIVO versus single-agent anti-PD(L)1 and ORR by IPI + NIVO versus anti-PD-1 (i.e., excluding anti-PD-L1 in this second analysis). The difference in pooled proportions between groups was evaluated with a chi-square test for heterogeneity. Results are reported with pooled proportion estimates, 95% confidence intervals (CI), and two-sided p values for comparison. Forest plots were generated to visualize the ORRs from each study and the pooled estimates. Statistical analyses and visualizations were performed using R. The review was not registered.

Real-world validation cohort
We conducted a retrospective review to identify all patients with advanced MCC who received IPI + NIVO as first-line IO systemic therapy at our institution outside of a prospective interventional trial, with a data cut of 05/16/2025. The clinical review was approved by the local institutional review board (Advarra IRB, protocol MCC19191) and conducted in accordance with national and local ethics guidelines. Objective tumor response was determined from radiographic and clinical assessments as documented in the medical record, following Response Evaluation Criteria in Solid Tumors (RECIST) V.1.1 guidelines. Progression-free survival (PFS) was calculated as the time from initiation of IPI + NIVO therapy to documented disease progression or death, with patients alive without progression censored at date of last follow-up. The Kaplan–Meier method was utilized to estimate PFS distribution, and a Kaplan–Meier curve is presented for visualization. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC).

Results

Meta-analysis
We identified five reports of first-line trials of anti-PD(L)1 or ipilimumab plus nivolumab in metastatic MCC that were included in our analysis [2–4, 7, 9] (Supplementary Fig. 1). In total, there were two reports of IPI + NIVO, three reports of anti-PD-1 monotherapy, and one report of anti-PD-L1 monotherapy. One study included independent arms for both anti-PD-1 monotherapy and IPI + NIVO combination therapy [3]. Both IPI + NIVO studies utilized a dosing schema of 1 mg/kg IV every 6 weeks for ipilimumab and 240 mg IV every 2 weeks for nivolumab. For reports that included both treatment-naïve and later-line treated patients, we restricted analysis to only the subgroup of patients treated in first line.
The pooled ORRs of first-line treatments were 81.0% (95% CI 69.3–90.6%) for IPI + NIVO, 49.6% (95% CI 43.7–55.6%) for all single-agent anti-PD(L)1 agents, and 57.0% (95% CI 49.6–64.5%) for single-agent therapy when restricted to anti-PD-1 only (Fig. 1). The pooled ORR of IPI + NIVO was significantly higher than that of either anti-PD(L)1 monotherapy when considering all anti-PD-1 and anti-PD-L1 agents (p = 0.0001) as well as monotherapy when restricted to anti-PD-1 agents (p = 0.0043). There was low heterogeneity between anti-PD-1 monotherapy studies (I2 = 2.3%), high heterogeneity between anti-PD-1 plus anti-PD-L1 monotherapy studies (I2 = 69.6%), and high heterogeneity between the two IPI + NIVO studies (I2 = 91.2%).

Real-world validation cohort
We identified eight patients who received first-line IPI + NIVO for advanced MCC outside of a clinical protocol at our institution (Table 1). The median patient age was 77 years (59–84). All patients were male. All patients had either unresectable stage IIIB (n = 2; multi-focal in-transit disease in both cases) or stage IV (n = 6) disease at treatment initiation. Three patients had previously been treated with surgery and adjuvant radiation for earlier stage disease, and the remaining five presented with advanced disease at diagnosis. All stage IV patients except one had symptomatic disease and/or multi-system visceral disease involvement. One patient was immunosuppressed due to a history of CAR T cell therapy for diffuse large B cell lymphoma. We included in the analysis one patient who received a single dose of carboplatin and etoposide but had residual bulky stage IV disease at the time of IPI + NIVO initiation. The remaining patients were systemic treatment-naïve. All patients were treated with the 1 mg/kg IV q6 week dosing regimen of ipilimumab, concordant with the two reported trials in MCC.
Following treatment with IPI + NIVO, seven of eight patients (87.5%) achieved an objective response, including two (25%) with radiographic complete response and five (62.5%) with partial response. The remaining patient achieved a mixed response with radiographic improvement in multiple nodal disease sites but progression of a soft tissue mass that required salvage therapy with radiation approximately 4 months after treatment initiation. Of the seven responding patients, one received concurrent radiation therapy to an isolated symptomatic bulky disease site (but also had objective response in other non-radiated disease sites), two received radiation therapy to “consolidate” treatment after achieving a sustained partial response, and the remaining four were treated with immunotherapy alone. With a median follow-up of 22.4 months (range 4.6–29.1 months), all seven responding patients remain progression-free at their last response assessment (Fig. 2). The median duration of IPI + NIVO therapy for responding patients was 5.5 months (3.8–15.6 months). One patient remains on treatment at the time of analysis, with the remaining discontinuing therapy due to toxicity (4) or elective discontinuation for sustained response (2).
Three of eight patients (37.5%) experienced grade 3 or higher immune-related adverse events to IPI + NIVO, including two cases of grade 3 colitis and one case of grade 3 pneumonitis. One additional patient experienced grade 2 colitis that required steroid management. All patients but one (87.5%) experienced at least low-grade toxicity that was felt to be immune-related (Table 2). Immune-related toxicities of any grade included colitis (3), rash (3), hypophysitis (3), pneumonitis (1), and arthritis (1). One patient who experienced grade 3 colitis died due to iatrogenic complications during hospitalization for management of autoimmune colitis. In brief, this 76-year-old patient with a history of congestive heart failure and a mechanical heart value was admitted for grade 3 colitis. He was managed with intravenous steroids and supportive care, including transition of full-dose anticoagulation from oral to subcutaneous. On hospital day 6, the patient had improved colitis symptoms but while anticipating discharge planning experienced rapid clinical decline with hypotension and acute abdominopelvic pain. Imaging revealed a massive acute intramuscular hematoma involving latissimus dorsi, quadratus lumborum, erector spinous, psoas, and iliacus muscles. Bloodwork revealed a 3 gm/dL drop in hemoglobin compared to laboratories performed 4 h prior. During initial evaluation and resuscitation attempts, the patient experienced cardiac arrest and expired.

Discussion
In this combined meta-analysis and retrospective institutional review, we evaluate the role of first-line therapy with ipilimumab plus nivolumab in advanced Merkel cell carcinoma. Combining data from reported prospective trials, first-line treatment with IPI + NIVO results in a pooled objective response rate of approximately 80%, statistically exceeding the pooled objective response rate of single-agent therapy. In a real-world validation of this approach, we observe concordant disease efficacy of first-line IPI + NIVO in a cohort of consecutive MCC patients treated with this approach at a single institution.
The observation that IPI + NIVO treatment results in a higher disease response rate compared to anti-PD-1 monotherapy in Merkel cell carcinoma is biologically plausible and perhaps unsurprising. First, this result would be concordant with the pan-cancer expectation of IPI + NIVO combination therapy versus monotherapy, particularly in other immunotherapy-responsive cancer types such as melanoma [5] and non-small cell lung cancer [6]. Further, in MCC specifically IPI + NIVO has established salvage activity in the anti-PD-1 refractory setting [7, 10], highlighting directly the increased activity of combination therapy over monotherapy in this disease.
Despite the encouraging efficacy signal reported herein, several important considerations remain when considering combination versus single-agent immunotherapies for advanced Merkel cell carcinoma. First, objective response rate is a surrogate endpoint that does not necessarily correlate with survival outcomes. Even in melanoma, where experts are guided by long-term randomized trial data, there remains some debate as to whether initial combination therapy improves disease-specific or overall survival over anti-PD-1 monotherapy, particularly if combination therapy is available for salvage if needed [5]. Importantly, combination therapy comes with the cost of higher toxicity that is not always “manageable,” as highlighted by a fatal case in our institutional series.
This study also has several important limitations. Most critically, prospective first-line data evaluating combination checkpoint blockade in Merkel cell carcinoma are limited, with only two published trials of IPI + NIVO available for meta-analysis [3, 7]. Despite similar patient demographics and baseline characteristics, these studies demonstrated markedly divergent efficacy results with high statistical heterogeneity (I2 = 91.2%), raising some uncertainty about the precise magnitude of benefit. We were unable to identify meaningful subgroups for further analysis given limited sample sizes. Additionally, our real-world cohort is limited by small sample size (n = 8), single-center treatment, an all-male population, and clinical heterogeneity including prior use of chemotherapy and the presence of immunosuppression in one patient each. These limitations highlight the need for additional prospective data, ideally from randomized trials with extended follow-up, to definitively establish the role of combination checkpoint blockade as first-line therapy in advanced Merkel cell carcinoma.
Nonetheless, it is both statistically and biologically probable that combination therapy yields higher response rates in the first-line advanced Merkel cell carcinoma setting as compared to monotherapy. Further, the magnitude of response rate benefit predicted by this pooled analysis is clinically meaningful, particularly for high-risk clinical scenarios such as symptomatic, rapidly progressive, and/or high burden disease. Additionally, we observed very encouraging response durability in IPI + NIVO responding patients in our institutional series, with all responding patients still in response at most current evaluation with a median follow-up of almost 2 years and despite a median of only 5.5 months of treatment. While additional experience and follow-up is needed given the small cohort size and patient heterogeneity, this observation is encouraging compared to expectations of durability of response with anti-PD-1 monotherapy, particularly after therapy discontinuation [11].
In summary, based on this meta-analysis of available clinical trial data, first-line treatment of advanced Merkel cell carcinoma with ipilimumab plus nivolumab results in a higher objective response rate compared to monotherapy. The current clinical decision to select combination therapy over monotherapy must weigh this response rate benefit with the unknown survival benefit and higher toxicity of this approach. To help improve generalizability of these findings, combination checkpoint inhibitor therapy for first-line therapy of advanced Merkel cell carcinoma should be further evaluated, ideally in a randomized controlled trial.

Ethics approval
The clinical review was approved by the local institutional review board (Advarra IRB, protocol MCC19191) and conducted in accordance with national and local ethics guidelines.

Supplementary Information
Below is the link to the electronic supplementary material.