Quantitative metastatic lymph node burden and survival in Merkel cell carcinoma

Background: Current lymph node (LN) staging for Merkel cell carcinoma (MCC) does not account for the number of metastatic LNs, which is a primary driver of survival in multiple cancers. Objective: To determine the impact of the number of metastatic LNs on survival in MCC. Methods: Patients with MCC undergoing surgery were identified from the National Cancer Database (NCDB). The association between metastatic LN number and survival was modeled with restricted cubic splines. A novel nodal classification system was derived by using recursive partitioning analysis. MCC patients undergoing surgery in the Surveillance, Epidemiology, and End Results (SEER) Program were used as validation cohort. Results: Among 3670 patients in the NCDB, increasing metastatic LN number was associated with decreased survival (P < .001). Mortality risk increased continuously with each additional positive LN when using multivariable, nonlinear modeling. According to a novel staging system derived via recursive partitioning analysis, the hazard ratio for death in multivariable regression compared with patients without LN involvement was 1.24 (P = .049), 2.08 (P < .001), 3.24 (P < .001), and 6.13 (P < .001) for the proposed N1a (1-3 metastatic LNs with microscopic detection), N1b (1-3 metastatic LNs with macroscopic detection), N2 (4-8 metastatic LNs), and N3 (>= 9 metastatic LNs), respectively. This system was validated in the SEER cohort and showed improved concordance compared with the American Joint Committee on Cancer, Eighth Edition. Limitations: Retrospective design. Conclusions: Number of metastatic LNs is the dominant nodal factor driving survival in patients with MCC.

Automated summary

The number of metastatic lymph nodes is the key factor affecting survival in patients with Merkel cell carcinoma (MCC). A novel nodal classification system was developed, which showed improved concordance compared with the American Joint Committee on Cancer, Eighth Edition.