Is There a Relationship Between TILs and Regression in Melanoma?

Background: The relationship between tumor-infiltrating lymphocytes (TILs) and regression in melanoma is unknown. This report describes a large multicenter study assessing the association between TILs and regression. Methods: The Sentinel Lymph Node Working Group database was queried from 1993 to 2018 for cases with TILs and regression data. Clinicopathologic factors were correlated with regression and TIL status, sentinel lymph node (SLN) status, and overall survival (OS). Results: The study enrolled 2450 patients. In 1811 cases, TILs (73.9%) were present, with regression present in 328 of these 1811 (18.1%) cases and in 49 (7.7%) of 639 cases without TILs. The presence of TILs was significantly associated with regression (p < 0.0001) as well as a negative SLN (p < 0.05). However, when TILs were stratified by regression status, only absence or presence of both TILs and regression were significantly associated with SLN metastases (p = 0.038). Although the presence of TILs was associated with OS (p < 0.05), regression status by itself was not (p = 0.2058 and 0.252, respectively). Furthermore, when TILs were stratified by regression status, only the presence of TILs with or without regression was significantly associated with improved OS (p = 0.0081 and 0.0137, respectively) versus the absence of both TILs and regression, with regression status not significantly affecting OS for patients with or without TILs (p = 0.2314 and 0.65, respectively). Conclusions: Regression is highly correlated with TILs, but only TILs are significantly associated with SLN metastasis and OS in melanoma patients, whereas regression is not. The impact of regression on outcomes ultimately appears dependent upon the absence or presence of TILs.