Pretreatment metastatic growth rate determines clinical outcome of advanced melanoma patients treated with anti-PD-1 antibodies: a multicenter cohort study

BACKGROUND: Checkpoint inhibitors revolutionized the treatment of metastatic melanoma patients. Although tumor burden and lactate dehydrogenase (LDH) are associated with overall survival (OS), the impact of tumor growth kinetics remains elusive and in part contradictory. The aims of this study were to develop a novel simple and rapid method that estimates pretreatment metastatic growth rate (MGR) and to investigate its prognostic impact in melanoma patients treated with antiprogrammed death receptor-1 (PD-1) antibodies. METHODS: MGR was assessed in three independent cohorts of a total of 337 unselected consecutive metastasized stage IIIB–IV melanoma patients (discovery cohort: n=53, confirmation cohort: n=126, independent multicenter validation cohort: n=158). MGR was computed during the pretreatment period before initiation of therapy with anti-PD-1 antibodies nivolumab or pembrolizumab by measuring the increase of the longest diameter of the largest target lesion. Tumor doubling time served as quality control. Kaplan-Meier analysis and univariable as well as multivariable Cox regression were used to examine the prognostic impact of MGR. RESULTS: Pretreatment MGR >3.9 mm/month was associated with impaired OS in the discovery cohort (HR 6.19, 95% CI 2.92 to 13.10, p<0.0001), in the confirmation cohort (HR 3.62, 95% CI 2.19 to 5.98, p<0.0001) and in the independent validation cohort (HR 2.57, 95% CI 1.56 to 4.25, p=0.00023). Prior lines of systemic treatment did not influence the significance of MGR. Importantly, the prognostic impact of MGR was independent of total tumor burden, diameter of the largest metastasis, number of prior lines of systemic treatment, LDH, as well as liver and brain metastasis (discovery and confirmation cohorts: both p<0.0001). Superiority of MGR compared with these variables was confirmed in the independent multicenter validation cohort (HR 2.92, 95% CI 1.62 to 5.26, p=0.00036). CONCLUSIONS: High pretreatment MGR is an independent strong prognostic biomarker associated with unfavorable survival of melanoma patients receiving anti-PD-1 antibodies. Further investigations are warranted to assess the predictive impact of MGR in distinct systemic therapeutic regimens.