Session Item

Merkel cell carcinoma (MCC) is a rare cutaneous neuroendocrine tumor that is sensitive to radiotherapy (RT). Limited data exists regarding the efficacy of curative hypofractionated RT (hypo-RT) regimens with shorter duration of treatment that are more convenient for patients (pts). Thus, we compared outcomes for hypo-RT (>2 Gy per fraction) and conventionally fractionated RT (conv-RT) for non-metastatic (M0) MCC.

The Dana-Farber/Harvard Cancer Center Institutional Review Board approved this retrospective cohort study of 241 pts diagnosed between January 2005 and June 2021, and received adjuvant (65.4%) or definitive RT (36.2%) at Dana-Farber/Brigham & Women’s Cancer Center. Follow-up time was defined from start of RT. Kaplan-Meier method and log-rank test were used to estimate and compare survival data. Cox proportional hazards regression was used for univariable and multivariable analyses (UVA/MVA). 

There were 243 courses of RT and 50 courses (20.6%) of hypo-RT. Median age was 72.6 years at diagnosis (range: 28.6-101.0) and median follow-up was 37.8 months (range: 1.2-194.5). Compared to the conv-RT group, the hypo-RT group had a similar proportion of immunosuppressed pts (12.0% vs. 13.0%, p=1.00), and use of definitive RT (48.0% vs. 33.2%, p=0.07), but was more likely to be ≥73 years at diagnosis (78.0% vs. 41.5%, p<0.01), receive an equivalent total dose in 2-Gy fractions (EQD2) <50 Gy (58.0% vs. 5.2%, p<0.01), and undergo RT after 2016 (72.0% vs. 38.9%, p<0.01). While 3-year overall survival (OS; 53.4% vs. 77.4%, p<0.01) was expectedly lower in the hypo-RT cohort, 3-year MCC-specific survival (84.5% vs. 82.6%, p=0.76) was not different, and only 2 (4.0%) and 2 (1.0%) pts recurred within the irradiated region in the hypo-RT and conv-RT groups, respectively. The most common sites of first recurrence were distant (14.0% vs. 15.0%) and in-transit metastases (8.0% vs. 7.8%) in hypo-RT and conv-RT groups, respectively. On UVA, male, immunosuppression, clinical stage III, and lymphovascular invasion (LVI) was associated with decreased time to any recurrence, but hypo-RT was not (hazard ratio [HR] 1.36, 95% confidence interval [CI]: 0.78-2.36, p=0.28), and definitive RT was not (HR: 1.23, 95% CI: 0.77-1.96, p=0.39). On MVA, immunosuppression (HR: 2.11, 95% CI: 1.21-3.69, p=0.01), clinical stage III disease (HR 1.85, 95% CI: 1.12-3.06, p=0.02), and LVI (HR 1.90, 95% CI: 1.17-3.07, p=0.01) remained significant when controlling for sex, age, and hypo-RT (Table).

While OS differences between the cohorts likely reflect selection bias, there was no difference in MCC-specific survival and very few in-field recurrences in the two cohorts, despite lower EQD2 dosing in the hypo-RT group. Most recurrences occurred out-of-field as in-transit or distant recurrences. Our data suggest that hypo-RT may be a reasonable and more convenient option for select pts, but future prospective studies are needed.