The advent of immune check point inhibitors (ICIs) has improved prognosis of various cancers. To better select responding patients (pts) and for a more accurate management of economic resources, biological and biochemical factors have been investigated. To date, no predictive biomarkers have been validated. The aim of this study is to identify manageable and routinely detectable parameters to use in clinical practice to select pts with higher change of response to ICIs.
271 consecutive metastatic solid tumor pts treated in our Institute from 2013 to 2017 with ICIs were evaluated for baseline LDH serum level, ECOG score, age, type of ICI, number of metastatic sites, histology and sex. A training and validation set were used to build and test models, respectively. Variables’ effects were assessed through odds ratio estimates (OR) and area under the receive operating characteristic curves (AUC), from univariate and multivariate logistic regression models. The validated estimates were used to develop an Excel algorithm to calculate probabilities of response.
As best response, 55.4% of pts achieved disease control and 44.7% had progressive disease. On the training set, LDH, age and ECOG showed a significant OR (p:<.001, 0.009, 0.042, respectively) and were combined in a multivariate model with an AUC of 0.771 (95%CI: 0.701;0.842). These results were statistically validated on the validation set (AUC: 0.685, 95%CI: 0.569;0.801). By fitting the validated model on all pts, the 3 variables retained a significant OR and a satisfactory cross-validated AUC.
We confirm, as reported in literature, that baseline LDH serum levels are inversely associated with response probability. It’s reasonable to jointly consider age and ECOG, which give a significant contribution to model performance. The developed algorithm, once validated on an independent prospective series, might be a base to guide physicians in clinical practice to better plan ICI therapy tailored on pts characteristics.
Clinical trial identification
Legal entity responsible for the study
Fondazione IRCCS Istituto Nazionale Tumori.
Has not received any funding.
All authors have declared no conflicts of interest.