P14.127 From next generation sequencing and comparative genomic hybridization to personalised medicine for adult primary brain tumors: the profiler trial

Abstract

Abstract BACKGROUND Personalized anti-tumoral therapies may currently be proposed on the basis of immuno-histochemistry, but also next-generation sequencing and comparative genomic hybridization. ProfiLER trial explored the feasibility, efficacy and the impact of molecular profiling for patients with solid or hematological advanced cancers including brain tumors. MATERIAL AND METHODS Patients with primary brain tumors, pre-treated with at least one line of anti-cancer treatment, could be included in this multicentric prospective trial. A molecular profile (next-generation sequencing and comparative genomic hydridization) was established on fresh or archived sample. Weekly molecular tumor board analysed results to propose as far as possible a molecular targeted therapy. RESULTS between February 2013 and December 2015, 141 patients with primary brain tumor were enrolled. One hundred five samples were further analyzed as 30 samples were excluded, and 6 are on-going. The rate of screen failure was 16/33 for stereotactic biopsy (49%) versus 11/104 (11%) for removal. The main representative histologic type of tumors were glioblastoma (n=46, 43,8%), low grade glioma (n=26, 24,8%), high grade glioma (n=12, 11,4%) and atypical and anaplastic meningioma (n=8, 7,6%). Median delay between the diagnostic of the primitive tumor and the inclusion in ProfiLER study was 2.7 years (0.2 - 29 years). Median delay between the consent and the results of the multidisciplinary meeting was 2.8 months (1–7.1 months). Forty-three patients (41%) presented at least one “druggable molecular alteration”. The most frequently altered genes were CDKN2A (n=18, 29%), EGFR (n=12, 20%), PDGFRa (n=8, 13%), PTEN (n=8, 13%), CDK4 (n=7, 11%), KIT (n=6, 10%), PIK3CA (n=5, 8%), MDM2 (n=3, 5%). Sixteen patients could not have a proposition of specific treatment due to death before MBT (n=5, 31.3%), lack of available clinical trials (n=9, 56%), or ambiguous results (n=2, 12.5%). Among the 27 patients (26%) for whom a specific therapy has been proposed, only six patients ultimately received a medical targeted therapy (everolimus n=3, erlotinib n=1, ruloxitinib n=1, sorafenib n=1). Four patients discontinued the treatment for toxicity, the 2 others for clinical progression. CONCLUSION routine high-throughput sequencing is feasible for brain tumors but delays should be reduced to be able to propose targeted therapies to patients fit enough to benefit from experimental treatment. Macroscopic surgery is the best way to obtain workable samples. Specific panel genes for neurologic tumors should be developed, as well as change of practices concerning exclusion criteria in clinical trials.