Citation
Miller DM, Emerick KS, Kaufman HL, McEvoy AM, Zaba L, Sharma A, Adelaja A, Silk AW, Khushalani NI, Brownell I, Coit DG, Nghiem P, Sondak VK. Local recurrence after excision of Merkel cell carcinoma: What do we actually know, and what should we do next? Journal of Cutaneous Oncology. Published July 16, 2026. doi:10.59449/joco.2026.07.16.
Overview
The natural history of local recurrence after complete excision of localized Merkel cell carcinoma remains incompletely defined. That uncertainty directly affects one of the most debated decisions in localized MCC care: whether postoperative radiation should be recommended after margin-negative excision.
This Perspectives on the Science article synthesizes the evidence surrounding local recurrence, reviews the Kavanagh and colleagues study, presents findings from a multidisciplinary Society of Cutaneous Oncology Journal Club, and considers how curated prospective data, causal inference, and probabilistic modeling could improve decision-making when definitive randomized evidence is unlikely to emerge.
The clinically relevant question is not simply whether radiation lowers local recurrence. It is how much absolute benefit treatment provides for a particular patient, what event is being prevented, how salvageable that event is, and what treatment burden is required to prevent it.
The Kavanagh study
Completely excised localized MCCs
All patients underwent complete surgical excision with pathologically negative margins.
No primary-site postoperative RT
The analysis focused on local recurrence among patients managed without primary-site radiation.
Estimated local recurrence rate
Seven of eight local recurrences occurred during the first year after treatment.
Why the recurrence event matters
Local recurrence was uncommon, but its clinical meaning depended on context. All six patients with isolated local recurrence achieved durable local control after additional local therapy. The two patients whose local recurrence occurred with synchronous regional and distant disease died of MCC, suggesting a more aggressive systemic phenotype rather than failure of local salvage alone.
These findings shift attention from recurrence prevention alone toward the effectiveness of rescue, the morbidity of treatment, and the distinction between an isolated salvageable local event and recurrence that reflects broader disease progression.
The decision framework
What is the true baseline risk?
Historical and contemporary estimates differ substantially because studies span different clinical eras, staging approaches, margin definitions, radiation practices, and recurrence classifications.
What does primary-site RT prevent?
Primary-site radiation is intended to reduce local recurrence, but many studies report broader recurrence or survival outcomes that do not isolate this specific event.
How salvageable is local recurrence?
The clinical value of prevention depends partly on whether an isolated local recurrence can be detected early and successfully treated without compromising long-term disease control.
What are the harms of treatment?
Treatment burden, toxicity, travel, fractionation, frailty, competing mortality, and patient preferences all influence the absolute value of postoperative radiation.
What the Journal Club revealed
- Clinicians varied substantially in their estimates of local recurrence risk after margin-negative excision.
- Practice patterns ranged from routine postoperative radiation to selective use or observation.
- Consensus was greatest at the extremes of risk and weakest in intermediate-risk scenarios.
- Tumor size, margins, lymphovascular invasion, immunosuppression, and sentinel-node status most strongly influenced recommendations.
- When frailty, travel, or treatment burden mattered, single-fraction 8 Gy emerged as an important abbreviated option.
- Variation extended beyond whether to treat and included radiation field, dose, fractionation, and assumptions about salvageability.
Building a better evidence base
Rare events, heterogeneous treatment, and strong selection effects limit what conventional retrospective studies can establish. The article argues for prospective, physician-curated, multi-institutional data that capture the date and site of first recurrence, radiation field and fractionation, salvage treatment, disease-specific outcomes, competing mortality, and the patient- and clinician-level factors that shape treatment decisions.
Causal-inference methods could help clarify comparative effectiveness when treatment is not randomized, while Bayesian and probabilistic models can express the range of plausible absolute benefit rather than implying false certainty. When untreated recurrence risk is very low, the scientific goal may shift from proving a large benefit to determining whether a clinically meaningful benefit can be confidently excluded for selected patients.
Move from debating a single recurrence rate toward improving decision quality: estimate individualized baseline risk, define the outcome radiation is intended to prevent, incorporate salvageability and treatment burden, and generate prospective evidence designed around clinically meaningful absolute benefit.
Authors
David M. Miller, Kevin S. Emerick, Howard L. Kaufman, Aubriana M. McEvoy, Lisa Zaba, Ajay Sharma, Adewunmi Adelaja, Ann W. Silk, Nikhil I. Khushalani, Isaac Brownell, Daniel G. Coit, Paul Nghiem, and Vernon K. Sondak.
Daniel G. Coit, Paul Nghiem, and Vernon K. Sondak contributed equally as senior authors.