Prognostic, predictive and potential surrogate markers in castration-resistant prostate cancer

Abstract

Several novel therapeutic agents have demonstrated improvement of overall survival and further patient relevant endpoints in the setting of castration-resistant prostate cancer. Nevertheless, two questions have become increasingly relevant: are there any prognostic or predictive markers that could ease clinical decision-making using risk stratification and risk-adapted treatment in order to provide a benefit for the patient? Furthermore, against the background of increasing possibilities of therapy sequencing in clinical practice and in the clinical trial landscape in castration-resistant prostate cancer, does an isolated evaluation of overall survival reliably mirror the benefit attributable to a single compound? To address both these questions, suitable parameters serving as surrogates for intermediate and long-term endpoints and reflecting individual benefit, respectively, need to be identified and proven.

Keywords:

• biomarker
• castration-resistant prostate cancer
• clinical decision-making
• clinical trials
• predictive marker
• prognostic marker
• surrogate endpoint
• surrogate marker

Financial & competing interests disclosure

S Brookman-May Janssen has affiliations with Pharma Research and Development, Beerse, Belgium. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Key issues

• Several validated prognostic markers are available for prostate cancer and castration-resistant prostate cancer. Subcategories of prognostic biomarkers are tumor-associated markers (PSA-levels, PSA-kinetics, Gleason score, TMPRSS2-ERG, circulating tumor cell count), bone turnover markers (alkaline phosphatase, serum TRAP-5b and others), patient related parameters (ECOG performance-status, pain), unspecific laboratory measures (albumin, lactate dehydrogenase, neutrophil-to-lymphocyte ratio) and type of progression.

• Besides prognostic markers, there is a clear need to also identify predictive biomarkers since reliable and validated predictive markers to ease clinical decision making are still lacking.

• Besides the need for predictive markers, surrogate markers also need to be developed, especially for efficacy assessment of compounds in clinical trials despite treatment sequencing.

• Before clinical utility of a biomarker can be confirmed, it needs to be shown that this marker provides additional information which is not already available.

• Major hurdles for appropriate biomarker development are the missing inclusion of biomarker assessment and validation in several clinical trials, the assessment of the right markers in the wrong setting or attributed to a discordant stage of disease, considerable tumor heterogeneity in prostate cancer and the lack of process from theory to practice in order implement a marker in clinical routine and make it clinically utile.

• Due to the tumor heterogeneity in prostate cancer, a single biomarker use is probably not the best way, and panels of biomarkers reflecting changes in independent pathways may lead to enhanced predictive ability.

• Future trials should integrate also proteomic and metabolomic biomarkers as well as circulating tumor cells.