30-Year Experience Managing Anorectal Melanoma With Sphincter-Sparing Local Excision and Hypofractionated Radiation Therapy

Abstract

Anorectal melanoma (ARM) is a rare malignancy with poor outcomes. We analyzed our 30-year experience using sphincter-sparing local excision (SSLE) and adjuvant radiation therapy (RT) to optimize local control (LC) across eras of changing nodal disease management and systemic therapy.We identified 111 ARM patients (pts) who underwent SSLE followed by adjuvant primary tumor RT at our institution between 1989-2020. Actuarial curves were calculated using the Kaplan-Meier method and Cox proportional hazards regression was performed for adjusted analyses.Median follow up was 31 months (mos) (IQR 15-56 mos). Most pts had locally-confined disease at diagnosis (n = 73, 66%) but 33% (n = 37) had nodal involvement; 20% (n = 22) had a positive SSLE margin. Nodal evaluation was performed for 75 pts (68%): 16 had lymph node dissection (LND) for gross nodal disease and 59 had sentinel lymph node biopsy (SLNB), of which 20 (34%) were positive and 4 proceeded to LND. All patients received adjuvant RT to the primary site (30 Gy in 5 fractions with goal 90% prescription dose coverage); 40 pts received inguinal nodal RT (36%) with 10 having known involved nodes and 30 receiving elective nodal RT. Neoadjuvant systemic therapy was administered to 17 pts (15%), including biochemotherapy (BCM; n = 8), chemotherapy (CTX; n = 2), and immunotherapy (IO; n = 7). Thirteen pts (76%) had evidence of any disease response at surgery. Adjuvant systemic therapy was administered to 43 pts (39%, 21 B.M. 8 CTX, 13 IO, 1 targeted therapy). Eighty pts (72%) relapsed. Distant metastasis (DM) was the most common site of first relapse (n = 55, 69%) followed by nodal relapse (n = 19, 24%). Three-year actuarial outcome rates were: 97% colostomy-free survival (CFS), 83% LC, 64% nodal control (NC), 36% DM-free survival (DMFS), 29% disease-free survival (DFS) and 49% melanoma specific survival (MSS). Multivariate analysis of disease and treatment factors showed that the only independent variable predicting for LC was melanoma in situ at any location in the resection specimen (HR 2.8, P = 0.026). Worse DFS was associated with > 4 mm tumor thickness (HR 1.8, P = 0.022), node+ disease at diagnosis (HR 1.7, P = 0.024), and positive margins at SSLE (HR 1.8, P = 0.035). Node+ disease at diagnosis was the only independent variable predicting for worse MSS (HR 2.2, P = 0.005). Systemic therapy was not significantly associated with improved outcomes, though this result is confounded by patient selection and small numbers. Nodal RT improved NC but had no effect on DFS or MSS. DFS was similar for pts treated in the IO era (2017-2020, n = 25, 3-yr DFS 36%) compared to those treated previously (1989-2016, n = 86, 3-yr DFS 29%; P = 0.38).SSLE and adjuvant RT provide effective LC with excellent CFS, an important quality of life outcome, but overall outcomes for ARM patients continue to be poor in the contemporary era.