Study supports need for diagnostic analysis of germline and tumor biomarker information

Germline biomarkers predicted drug toxicity: impact on clinical routine

A core tenet of precision medicine is that predictive biomarkers can enhance therapeutic decision-making. In a new pilot study, scientists at Molecular Health analyzed a randomly selected set of 250 patients with solid tumors and detected predictive biomarkers in more than 85% of tumors. The figures highlight the clinical benefit of tumor profiling: overall, around one out of four samples contained FDA-endorsed biomarkers, while around 70% contained clinically emergent or translational level biomarkers.

The study demonstrated the clinical relevance of germline (i.e., genetic information from healthy tissue in our bodies) biomarkers. In several cases, germline biomarkers predicted avoidable drug toxicities. "It is important to not characterize only somatic aberrations, as it is possible that a patient's germline mutations might interfere with underlying molecular mechanisms and effectiveness of drugs otherwise recommended solely on the basis of the tumor's profile," explains Theodoros Soldatos, Ph.D. and scientist at Molecular Health. "Somatic mutation profiling is essential but not sufficient for predicting cancer drug response, thereby supporting the need for diagnostic analysis of both germline and tumor biomarker information."

Mr Soldatos and his co-authors (Sandra Morandell, Ph.D., Francesca Diella, MSc., Sasha Badbanchi, MSc., Anja Doerks, Ph.D., Bernhard Sulzer, Ph.D., Ralf Stecher, BSc., Gabriel Bien-Willner, M.D./Ph.D., David Jackson, Ph.D., Martin Stein, Ph.D. - all Molecular Health) detected predictive biomarkers in a broad range of clinical validities and evidence levels.

The randomly selected set encompassed 20 different cancer indications. Biomarkers based on SNVs, indels and fusion proteins were identified using a 613-gene NGS panel and an analytical platform that screens identified aberrations against more than 5,500 peer-reviewed predictive biomarkers. Identified biomarkers were then classified according to lineage, clinical validity/evidence level, prevalence and potential functional interdependence. The pilot study was presented today as a poster at the WIN Symposium (Worldwide Innovative Networking in personalized cancer medicine - http://www.winsymposium.org) in Paris.