Targeted cancer therapies require the assessment of the effect of a given treatment on different populations of patients. One important issue is that the distribution of many biomarkers directly targeted by the treatment may also vary widely across patients, as well as before and after therapy. In this paper, we discuss a probabilistically coherent and unified framework for quantifying how a targeted therapy affects the distribution of a given biomarker and, a fortiori, the individual survival outcomes. In particular, we adopt a Bayesian nonparametric perspective and nest the assessment of the complete individual biomarker profiles in a general survival model. We exemplify our methods on a known clinical dataset, that reports the results of a therapeutic clinical trial targeting plateled-derived growth factor levels in prostate cancer subjects.
BAYESIAN NON PARAMETRIC IDENTIFICATION OF PRE-VERSUS-POST TREATMENT BIOMARKER EFFECTS ON PROGRESSION FREE SURVIVAL
GRAZIANI, REBECCA;GUINDANI, MICHELE;
2013
Abstract
Targeted cancer therapies require the assessment of the effect of a given treatment on different populations of patients. One important issue is that the distribution of many biomarkers directly targeted by the treatment may also vary widely across patients, as well as before and after therapy. In this paper, we discuss a probabilistically coherent and unified framework for quantifying how a targeted therapy affects the distribution of a given biomarker and, a fortiori, the individual survival outcomes. In particular, we adopt a Bayesian nonparametric perspective and nest the assessment of the complete individual biomarker profiles in a general survival model. We exemplify our methods on a known clinical dataset, that reports the results of a therapeutic clinical trial targeting plateled-derived growth factor levels in prostate cancer subjects.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
