Personalized Targeted Therapy for Refractory Pediatric Brain Tumors Program: Concept and Preliminary Clinical Experience

As former diagnoses are separated in smaller groups, many novel agents will never be studied for pediatric malignancies with traditional clinical research methods. In clinical care of refractory pediatric brain tumors, we have recently personalized the treatment by adding agents based on tumor markers. The algorithm starts with the treatment history and evidence-based drug rankings, and creating a first personalized treatment cycle. In addition, each tumor is analyzed for a panel of markers resulting in novel drugs to be added to the second cycle. Here we report a preliminary retrospective chart review of patients treated this way. In seven patients our charts contained outcome information. The histological diagnoses included: PNET (3), GBM (2), ATRT (1), and choroid plexus carcinoma (1). The markers resulting in the final treatment decision included ERK-P, Topo IIa, BCL2, VEGFa, p-STAT 3, ER, m-TORC1, and p-NF-kappaB. The drugs chosen included Sorafenib, Bevacizumab, Fluvestrant, Rapamycine, Bortezomib, and Curcumin. The early response was 0/2/2/1/and one classified as CR/PR/CCR/SD/PD/and undetermined, respectively. By January 2010 three patients lived in remission, one with tumor, one had passed away from tumor progression, and one had passed away for unknown reasons. When comparing response and EFS of these treatment attempts to the previous ones in the same patients, the response appeared superior, but data were insufficient to show a benefit for EFS. We conclude that both the algorithm and the analysis need to become more concrete as the treatment approach develops further to result in a new type of clinical studies.