P47: Predicting recurrence after a negative sentinel lymph node in melanoma: the model changes with time and sites of recurrence

Friday, March 24, 2023

3:26 PM – 3:31 PM

Location: SSO HUB: Melanoma & Sarcoma Zone

Poster Presenter(s)

SC

Shu-Ching Chang, PhD (she/her/hers)

Senior Biostatistician
Providence Health and Services
Portland, Oregon, United States

Disclosure information not submitted.

Author(s)

KL

Kristel Lourdault, PhD (she/her/hers)

senior scientist
Saint John's cancer institute
Santa Monica, California, United States

Disclosure information not submitted.
SC

Shih-Ting Chiu, PhD (she/her/hers)

Senior Biostatistician
Providence Health and Services, United States

Disclosure information not submitted.
GG

Gary Grunkemeier, PhD

Research Program Manager
Providence Oregon Regional Shared Services, United States

Disclosure information not submitted.
DH

Douglas Hanes, PhD (he/him/his)

Senior Biostatistician
Providence Health and Services, United States

Disclosure information not submitted.
SS

Stacey Stern, MS

Data Manager
Saint John's Cancer Institute, United States

Disclosure information not submitted.

Author(s)

RE

Richard Essner, MD (he/him/his)

Director of surgical oncology
Saint John's cancer institute
Santa Monica, California, United States

Disclosure information not submitted.

Introduction: Despite the favorable prognosis of AJCC stage I/II melanoma patients after a tumor-negative sentinel lymph node (SLN) biopsy, up to 20-30 % of patients will ultimately develop metastases. There have been several clinical models developed to predict recurrence, yet they lack specifics of time or sites which makes follow-up schedules difficult to standardize. The objective of this study is to investigate the relationship between clinical and pathologic risk factors and recurrence-free survival (RFS), and to develop a Risk Score (RS) predicting long-term RFS in early-stage melanoma patients.

Methods: This study is based on 2,668 AJCC stage I-II melanoma patients with tumor-negative SLN from the MSLT-I clinical trial and our own prospectively collected melanoma database. The median follow-up time was 9.2 years (IQR 6.0-11.6). Patient demographics and tumor characteristics were analyzed for their potential prognostic capability. The RS for individualized prediction of RFS was constructed using age, gender, primary site, Breslow thickness and ulceration. Discrimination was measured by time-dependent ROC analyses. Overall survival (OS) after the first recurrence was also evaluated by recurrence sites.

Results: Patients had a mean age of 54.7 years (SD,15.9) and median Breslow of 1.2 mm (IQR, 0.8-1.9). Most patients were male (59.7%), had primary tumors without ulceration (78.6%), located on the extremities (42.3%) or trunk (37.7%). 427 (16%) had recurrence during follow-up, with first recurrence of non-distant sites (53.7%), distant sites (42.8%) or unknown/multiple sites (4.5%). Time-dependent ROC analyses showed AUC values (95% CI) of 0.87 (0.83-0.90) and 0.82 (0.78-0.85) at follow-up years one and two, respectively, and then dropped to about 0.70. Population-level RS analysis showed significant difference in RFS, stratified by tertile of RS, with 10-year RFS of 95.5% (93.9-96.8), 87.0% (84.7-88.9) and 66.3% (62.9-69.8) for non-distant, distant and multiple sites of recurrence. OS following recurrence differed significantly by site of recurrence, with a worse 5-year OS (%) for distant vs. non-distant sites [15.0, 95% CI 10.2-22.1] vs. [50.0, 95% CI 43.5-57.5]; risk-adjusted HR: 2.60 (2.00-3.40).

Conclusions: Sentinel lymph node biopsy is an integral part of our modern management of melanoma, yet because most patients have a tumor-negative sentinel lymph node, they are left with an unclear prognosis. We have devised a risk score that is based on clinical and pathologic features, time from diagnosis and site of first recurrence to help clinicians better understand individual patients’ risk of recurrence and possibly lead to a more standardized approach to following these patients.

Learning Objectives:

Upon completion, the participant should understand the patterns of recurrence following a tumor negative sentinel lymph node in melanoma.
Upon completion, the participant should understand factors that contribute to recurrence and survival in melanoma.
Upon completion, the participant should understand the statistical model for predicting recurrence in melanoma and its relationship to time.