Abstract Background: The knowledge of actionable somatic genomic alterations present in human tumors is enabling the new era of personalized cancer treatment. The great intellectual challenge lies in linking confirmed mutations to protein function. Personalized tumor graft models (Avatars) can aid in the process of genomic analyses interpretation to ultimately move from molecular profile to medication. Methods: Using massive parallel sequencing we performed whole exome sequencing analysis of tumor and matched normal blood samples of 23 patients (pts) with advanced solid tumors (7 lung cancer, 7 pancreatic cancer, 1 neuroendocrine tumor, 2 glioblastoma, 1 uveal melanoma, 2 melanomas and 3 colon cancer) to identify putatively actionable tumor-specific genomic alterations. Avatar models generated by direct engraftment of tumor samples from the pts into immunocompromised mice were used as an in vivo platform to test proposed treatment strategies. Results: Successful exome sequencing analyses has been obtained for 21 pts (1 patient died prematurely, 1 sample was insufficient). Tumor specific mutations (Muts) and copy number variations were identified ranging from 5 to 952 and 0 to 36 respectively. All samples profiled contained clinically meaningful genomic alterations. A successful Avatar model was generated for 10 out of 17 pts. Two engraftment failures corresponded to EGFR mutant lung tumors resected while pts were receiving erlotinib, which initially grew but then regressed. Some of the most relevant drugabble alterations were: KRAS, CHEK1, FGFR2, IGF1R, MET, BRCA1, XPC, NOTCH, CREB3LB, GNA11, SMAD4 and EGFR. In occasions druggable alterations such as muts in NF1, PTPRC, PI3KA and DDR2 failed to provide any benefit when a targeted drug was tested in the Avatar and accordingly treatment of the pts with these drugs was not effective. In one case, loss of STK11 lead to testing of mTOR and SRC inhibitors resulting in tumor regression both in the Avatar and in the clinic. At present time 10 pts have received a personalized treatment: 2 pts, as expected based on the Avatar model, did not response; 4 pts resulted in durable partial remissions and 4 pts are currently on treatment with disease stabilization. In one of the EGFR mutant lung pts the genomic analysis revealed traces of an acquired mutation and allowed decision making at an earlier time point, prior to relapse. Overall, there was a remarkable correlation between drug activity in the Avatar and clinical outcome in the pts, in terms of drug resistance and sensitivity. Conclusion: The detection of actionable tumor-specific genomic alterations in the clinical setting is feasible. However predicting treatment response to known oncogenes is complex and requires detailed information of how different genetic backgrounds function. Avatar models are a powerful investigational platform for therapeutic decision making and help to guide cancer treatment in the clinic. Citation Format: Elena Garralda, Keren Paz, Pedro P. López-Casas, Siân Jones, Amanda Katz, Lisa M. Kann, Fernando López-Rios, Francesca Sarno, Fátima Al-Shahrour, David Vasquez, Elizabeth Bruckheimer, Samuel V. Angiuoli, Luis A. Diaz, Alfonso Valencia, Victor E. Velculescu, David Sidransky, Manuel Hidalgo. Integrated next generation sequencing and patient-derived xenografts to personalized cancer treatment. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2205. doi:10.1158/1538-7445.AM2013-2205