Improved survival at the population level for patients with advanced Merkel cell carcinoma following availability of immunotherapy

Abstract

Background:

Merkel cell carcinoma (MCC) is an aggressive skin cancer with poor survival rates. Immune checkpoint inhibitors (ICIs) were FDA-approved for advanced MCC in 2017, but their real-world survival impact remains unclear.

Objective:

Evaluate whether ICI introduction in the US corresponded with improved survival.

Methods:

This cohort study analyzed SEER data for MCC patients diagnosed from 2010 to 2021, grouped by 3-year periods, to calculate 2-year overall and relative survival.

Results:

For 453 patients with metastatic MCC, 2-year relative survival improved from 23% (2010–2012) to 37% (2013–2015), 42% (2016–2018), and 54% (2019–2021) (p < 0.001). Median overall survival also increased from 9 to 16 months among these patients. In 4,786 MCC patients overall, 2-year relative survival rose from 73% (2010–2012) to 81% (2019–2021) (p = 0.004), while overall survival improved from 67% to 72% (p = 0.012).

Limitations:

SEER lacks case-level data to link ICI treatment directly to survival, although ICIs represent the major recent treatment advance for MCC.

Conclusions:

The introduction of ICIs aligns with a >2-fold increase in survival for advanced MCC patients at the population level, translating to ~220 fewer deaths per year in the U.S.

INTRODUCTION

Merkel cell carcinoma (MCC) is a high-grade neuroendocrine skin cancer with marked propensity for spread. Historically, survival outcomes were poor, especially for advanced stages. For example, a large 2016 analysis by Harms et al. reported a 5-year overall survival (OS) for metastatic MCC of just 14%.¹

Before immune checkpoint inhibitors (ICIs) targeting the PD-(L)1 pathway, treatment options for advanced MCC were limited to cytotoxic chemotherapy, adapted from small cell lung cancer regimens. While tumor regressions occurred, responses were typically short-lived.² ICIs were tested for advanced MCC in nationwide trials starting in 2014, alongside ‘off-label’ case reports and data from other skin cancers.³ In 2017, the FDA approved avelumab (anti-PD-L1), followed a year later by pembrolizumab (anti-PD-1) for advanced MCC.^4,5 These agents showed high response rates in treatment-naïve patients, with durable responses. Clinical practice and guidelines quickly shifted from chemotherapy to immunotherapy.⁶ Although ICIs haven’t been compared directly to chemotherapy in randomized trials, retrospective analyses show significantly improved survival for metastatic MCC patients treated with immunotherapy.⁷ By 2019, immunotherapy became the first-line treatment for about two-thirds of metastatic MCC patients.⁸

We hypothesized that overall and relative survival for advanced MCC patients in the U.S. would significantly improve following the widespread availability of ICIs. To evaluate this hypothesis, we analyzed survival trends using population-based data from the Surveillance, Epidemiology, and End Results (SEER) database, focusing on patients diagnosed before versus after the approval of ICIs.

METHODS

Human subjects:

Studies were performed in accordance with Declaration of Helsinki and institutional review board approved (Providence-St Joseph Health IRB2020000564).

Population and data:

De-identified cross-sectional U.S. registry data was extracted from the SEER⁹ on August 20, 2024. All cases of MCC (ICD-O-3 8247/3) as the first primary invasive malignancy with known age, sex and survival outcomes of the patient diagnosed during years 2010–2021 were included.

Statistical analyses:

Patients were grouped into 4 three-year-long periods (2010–2012, 2013–2015, 2016–2018, 2019–2021) due to the rarity of MCC. Survival duration was calculated from diagnosis, with the 24-month survival rate selected as the primary endpoint, as it was the furthest available date for the most recent cohort. Local-regional-distant staging classification was used since AJCC 8^th edition was available for <50% of patients. Patient demographics were compared using the Kruskal-Wallis test or Fisher’s exact test. OS and relative survival (RS) were determined, with RS serving as a surrogate for cancer-specific survival by comparing observed survival to that of age- and sex-matched persons in the general population. Survival calculations used the actuarial method. Standard errors and 95% confidence intervals (CI) were calculated using the complementary log-log transformation. Expected survival was based on socioeconomic status, age, geography, race, and sex.¹⁰ Cumulative expected survival was determined with the Ederer II method.¹¹ Survival was compared between time periods using z-tests after complementary log-log transformation. Analyses were performed with SEER*Stat software, version 8.4.3¹² and R, version 4.0 (R Foundation for Statistical Computing, Vienna, Austria). Graphs were made in GraphPad Prism version 10 (Dotmatics, San Diego, California).

RESULTS

Patients and demographics:

A total of 4,786 patients presented with MCC as their first or only malignancy between 2010 and 2021. Of these, 2,933 (61.3%) were male. Racial/ethnic and age distributions aligned with previous reports,¹³ with the majority (66.6%) being 70 years or older at diagnosis. No significant changes in sex (p = 0.56) or race/ethnicity (p = 0.20) were observed over the study period (Table 1). Small but statistically significant differences in age were noted (median: 74–76 years, p = 0.032). Among patients with documented disease extent (n=4,389), 59.3% had localized, 30.4% had regional, and 10.3% had distant metastatic disease.

Table 1.

Patient Characteristics by Time Period

Year	2010–2012	2013–2015	2016–2018	2019–2021	P-value*
Number of patients	1,026	1,216	1,258	1,286
Sex					0.56
Men	628 (61.2%)	760 (62.5%)	752 (59.8%)	793 (61.7%)
Women	398 (38.8%)	456 (37.5%)	506 (40.2%)	493 (38.3%)
Age at Diagnosis - Median	75	76	74	75	0.032
<55	64 (6.2%)	65 (5.3%)	67 (5.3%)	51 (4.0%)
55–64	176 (17.2%)	161 (13.2%)	207 (16.5%)	188 (14.6%)
65–74	248 (24.2%)	335 (27.5%)	383 (30.4%)	361 (28.1%)
75–84	323 (31.5%)	366 (30.1%)	336 (26.7%)	411 (32.0%)
85+	215 (21.0%)	289 (23.8%)	265 (21.1%)	275 (21.4%)
Extent of Disease					0.027
Localized	563 (61.5%)	688 (61.6%)	676 (58.6%)	675 (56.2%)
Regional	248 (27.1%)	321 (28.7%)	360 (31.2%)	405 (33.7%)
Distant	105 (11.5%)	108 (9.7%)	118 (10.2%)	122 (10.1%)
Unknown	110	99	104	84
Race/Ethnicity
Non-Hispanic White	911 (88.8%)	1,062 (87.3%)	1,082 (86.0%)	1,091 (84.8%)
Non-Hispanic Black	<16 (<1.6%)	23 (1.9%)	24 (1.9%)	26 (2.0%)
Non-Hispanic American Indian/Alaska Native	<16 (<1.6%)	<16 (<1.3%)	<16 (<1.3%)	<16 (<1.2%)
Non-Hispanic Asian/Pacific Islander	18 (1.8%)	40 (3.3%)	34 (2.7%)	48 (3.7%)
Hispanic (All races)	73 (7.1%)	79 (6.5%)	99 (7.9%)	99 (7.7%)
Unknown	<16 (<1.6%)	<16 (<1.3%)	<16 (<1.3%)	<16 (<1.2%)

Values are no. (%) unless otherwise specified; some cells are masked to conform to small group size policies of database;

^*Time periods compared using the Kruskal-Wallis test (age) or Fisher’s exact test (all other characteristics).

Survival outcomes:

Among all patients, 24-month RS from the time of diagnosis significantly increased from 73.1% (95% CI 69.7–76.1%) at the start of the study (2010–2012) to 80.5% (95% CI 76.5–83.8%) at the end (2019–2021) (p = 0.004, Figure 1, Supplemental Figure 1A). OS also significantly increased from 66.0% (95% CI 63.0–68.8%) in 2010–2012 to 71.6% (95% CI 68.3–74.7%) (p = 0.012, Supplemental Figure 2A).

Figure 1.

24-month relative survival rates for Merkel cell carcinoma patients from 2010–2012 to 2019–2021. Error bars correspond to 95% confidence intervals. *p < 0.05 ** p <0.01 *** p < 0.001

RS gains varied by disease extent. For localized MCC, 24-month RS was 86.4% (95% CI 82.0–89.8) in the early period and 89.6% (95% CI 83.9–93.4%) in the later period, without significant change (p = 0.32, Figure 1, Supplemental Figure 1B). 24-month OS was also stable (77.5 vs 79.0%; p = 0.58, Supplemental Figure 2B).

For regional disease, a trend towards improvement was seen with 24-month RS improving from 65.1% (95% CI 58.0–71.2%) to 77.1% (95% CI 69.7–82.9%) (p = 0.015, Figure 1, Supplemental Figure 1C) and similar trends in OS (60.1% to 70.2%; p = 0.023, Supplemental Figure 2). Median survival improved from 45.0 months in 2010–2012 to 54.7 months in 2016–2018, though the median was not reached for 2019–2021 due to limited follow-up.

For distant metastases, 24-month RS more than doubled, from 23.2% (95% CI 15.3–32.2%) to 53.5% (95% CI 40.9–64.5%) (p < 0.001, Figure 1). 24-month RS progressively improved over each time, from 23.2% to 36.8% to 41.6% to 53.5%, respectively, for years 2010–2012, 2013–2015, 2016–2018, and 2019–2021 (Figure 2, Supplemental Figure 1). OS gains were similarly significant, with 24-month OS increasing from 21.4% (95%CI 14.1–29.7%) to 49.7% (95%CI 38.2–60.2%) (p < 0.001, Figure 2, Supplemental Figure 2D), and median OS extended from 8.8 months (2010–2012) to 15.9 months (2019–2021) (Figure 2).

Figure 2.

Relative and overall survival of patients with distant metastatic Merkel cell carcinoma at diagnosis are displayed across four time periods: 2010–2012, 2013–2015, 2016–2018, and 2019–2021, with error bars representing 95% confidence intervals. The bar graph on the far right illustrates median overall survival for these periods. Dx, diagnosis

DISCUSSION

In a large, real-world cohort of 4,786 MCC patients diagnosed in the US from 2010–2021, survival outcomes for metastatic MCC significantly improved in the post-immunotherapy era after FDA approval of anti-PD-(L)1 agents. Two-year survival (both relative and overall) for patients with metastatic MCC more than doubled from 2010–2012 to 2019–2021. Specifically, the 2-year RS for MCC improved from 23% to 54%. These benefits were among all newly diagnosed MCC patients, with RS increasing from 73% to 81%, implying a 30% reduction in MCC-related deaths. This equates to about 220 fewer deaths per year in the U.S. based on the 2021 incidence of 2,850 cases.¹⁴ These findings parallel the survival improvements seen in melanoma, where Adhikari et al. reported median OS for metastatic melanoma increased from 8 months pre-ICI to 14 months in the combination ICI era.¹⁵

Our study utilized data from the SEER national registry, which has several limitations. Firstly, SEER does not provide case-level data on specific treatments, making it difficult to directly link ICI treatment to survival outcomes. However, adoption of immunotherapy was the only major practice-changing intervention for metastatic MCC across this period. Furthermore, other studies support that immunotherapy leads to better outcomes than chemotherapy,⁷ and immunotherapy uptake was brisk, reaching two-thirds of metastatic patients by 2019.⁸ Therefore, it is reasonable to attribute the improved survival primarily to immunotherapy. Secondly, SEER employs RS as a proxy for cancer-specific survival. However, the greater survival improvements observed in metastatic MCC, where ICIs were likely more frequently used than in localized disease, further suggest their role in enhancing disease-specific survival. Thirdly, SEER does not track recurrence, disease progression, or post-diagnosis metastasis. Therefore, the survival benefit of ICIs for patients who initially presented with local or regional MCC and later developed distant disease are analyzed within their initial diagnostic stage group, along with the majority of patients in those groups that did not recur. As such, SEER may not fully capture the impact of ICIs on survival outcomes for patients who develop metastatic disease after an initial diagnosis of local-only disease. Lastly, long-term survival data for the 2019–2021 cohort are not available, potentially underestimating the full impact of immunotherapy, as durable responses reported in clinical trials may continue to emerge.

In addition to the adoption of immunotherapy, advancements in MCC management and treatment of associated comorbidities may have contributed to improved outcomes across all stages. For instance, sentinel lymph node biopsy has been included in MCC management guidelines since 2000 and has been increasingly utilized in clinical practice ever since.^16,17 Additionally, multiple studies have demonstrated the benefit of adjuvant radiation therapy in reducing recurrences and potentially improving survival in locoregional MCC.^18,19 Research on optimizing radiation dosing and timing has further refined its use.^20,21 Regarding ICI indications, clinical trials have suggested potential benefits of neoadjuvant or adjuvant ICI in certain settings.^22,23 However, these improvements primarily affect early stage disease, and the disproportionate survival gains in the present study for metastatic disease point to immunotherapy as a key factor.

Survival outcomes for MCC in the U.S. have significantly improved in the post-immunotherapy era, reflecting the impact of immune checkpoint inhibitors. Because of the magnitude of these changes for patients with advanced disease, data from the post-ICI era should now preferentially be used as a reference when establishing new benchmarks for patient care, refining staging criteria, or developing future clinical trials.