Drugs targeting programmed death receptor-1 and its ligand PD-L1 (aPD(L)1) have shown activity in multiple malignancies. Considering their novel mechanism of action, whether traditional prognostic scores also apply to patients receiving aPD(L)1 is unknown. We investigated whether a baseline 3-points (pts) CT-scan (PS3-CT) score and a 7-pts prognostic score (PS7) allowed identifying long-term survivors on aPD(L)1 as compared to molecularly targeted agents (MTA).
We reviewed all consecutive patients enrolled in phase I (PhI) trials evaluating aPD(L)1 between 12/2011 and 07/2015, and in PhI trials evaluating MTA between 03/2008 and 07/2012. PS3-CT was calculated using high tumor burden (TB > 9 cm), low-skeletal muscle index (SMI< 53 cm2.m-2) and non-pulmonary visceral metastases (NPVM) (1 pt each). PS7 was calculated by adding performance status >1, low albumin, high LDH and >2 metastases (1 pt. each). Effect on overall survival (OS) of each parameter was tested using Kaplan Meier and Multivariate Cox analyses.
461 patients (251 receiving aPD(L)1 and 210 receiving MTA) were included. Patients characteristics were comparable in both groups, excepted for melanoma patients (40% and 5% of the patients in the aPD(L)1 and MTA groups, respectively). PS3-CT was a significant independent predictor of OS in the aPD(L)1 (HR = 1.39 [95%CI = 1.07-1.81]; p = 0.01) and MTA (HR = 1.35 [95%CI = 1.08-1.68]; p = 0.01) groups, when compared to the Royal Marsden Hospital, Barbot and AJCC scores. High-TB and low-SMI were independent predictors of OS (HR = 2.00 [95%CI = 1.38-2.88], p
Objective and rapid risk scoring based on three CT-scan parameters allows identifying patients with prolonged OS on anti-PD(L)1 and MTA PhI trials, independently from conventional clinico-biological prognostic scores.
Clinical trial identification
Legal entity responsible for the study
All authors have declared no conflicts of interest.