Impact of assessment-to-treatment interval and metastatic biopsy site on the predictive value of PD-L1 expression at the 1 % cut-off level in melanoma

Background: Intratumoral PD-L1 expression at the 1 % cut-off predicts clinical outcomes and may guide first-line immune checkpoint inhibitor (ICI) selection for metastatic melanoma (MM). However, the impact of the interval between PD-L1 assessment and ICI initiation and the metastatic site used for PD-L1 evaluation, remains unclear. 

Methods: In this nationwide cohort study we used the Danish Metastatic Melanoma Database (DAMMED) and the Danish Pathology Registry to analyze patients with MM treated with anti-PD-1 or anti-PD-1 plus anti-CTLA-4 from January 2017 to February 2024. Progression-free survival (PFS) and overall survival (OS) were analyzed using Log-rank tests and Cox regression. 

Results: Data from 1137 patients were analyzed. Among patients with PD-L1 assessed within 90 days of treatment (n = 964; 55.2 % PD-L1 <1 %, 44.8 % PD-L1 ≥1 %), combination therapy improved outcomes in PD-L1 < 1 % (PFS adjusted (a)HR 0.62; 95 % CI 0.48–0.80; p < 0.001, OS aHR 0.64; 95 % CI 0.48–0.85; p = 0.002), while outcomes were comparable for PD-L1 ≥ 1 % patients (PFS aHR 0.90; 95 % CI 0.62–1.30; p = 0.57, OS aHR 0.97; 95 % CI 0.60–1.57; p = 0.89). For PD-L1 assessed > 90 days prior (n = 173), this pattern was less pronounced. Among 48 paired PD-L1 assessments from the same organ, discordance occurred in 25 %. Combination therapy improved PFS for patients with PD-L1 < 1 % skin/subcutaneous (aHR 0.51; 95 % CI 0.34–0.76; p < 0.001) and visceral metastases (aHR 0.65; 95 % CI 0.42–1.02; p = 0.060) while this association was not evident for lymph node metastases (aHR 0.79; 95 % CI 0.48–1.29; p = 0.35). 

Conclusions: PD-L1 seems a reliable predictive biomarker in MM, when assessed on tissue obtained within 90 days prior to ICI initiation. Non-nodal metastatic sites appear preferable.