Precision oncology in sarcoma drug development: Impact of genomic matching on response, clinical benefit, and survival in sarcoma patients on phase 1 trials.

Abstract

11018 Background: Genomic analyses have revealed many potentially actionable mutations across sarcoma subtypes. Whether sarcoma patients enrolled on genomically matched early phase trials have improved clinical outcomes over patients enrolled on non-genomically matched trials remains unclear. Methods: We analyzed clinical and next gen sequencing data from sarcoma patients on phase 1 trials at MD Anderson Cancer Center (MDACC) and performed logistic and Cox proportional hazards regression analyses to compare response rate (RR), median time to progression (mTTP), clinical benefit rate (CBR = CR, PR, or SD > 6 months), and median overall survival (mOS) between patients treated on genomically-matched and non-genomically matched trials. Results: Among the 406 patients with advanced sarcomas (321 soft tissue sarcoma [STS], 85 bone sarcomas) treated on phase 1 trials at MDACC from May 2006 to May 2018, median age was 53 (range 11-84), 48% were female, and patients had a median 3 prior lines of therapy (range 0-9). The most commonly treated STS subtypes were leiomyosarcoma (n = 66; 16%), liposarcoma (n = 52; 13%), GIST (n = 44; 11%), and synovial sarcoma (n = 11; 3%) and most commonly treated bone sarcomas were osteosarcoma (n = 34; 8%), chondrosarcoma (n = 28; 7%), and Ewing's sarcoma (n = 25; 6%). 23% (n = 93) of sarcoma patients treated on phase 1 trials were treated on genomically-matched trials. RR on non-genomically matched trials was 6% compared to 11% on genomically-matched trials, OR 1.97 (95% CI 0.88, 4.44), p = 0.10. Responses on genomically-matched trials were seen with novel agents targeting TRK, LRRC15, cMET, mTOR, VEGF, MDM2, KIT/PDGFRA, and FGFR. mTTP on non-genomically matched trials was 2.7 months compared to 3.7 months on genomically-matched trials, HR 0.72 (95% CI 0.57, 0.91), p = 0.0048. CBR on non-genomically matched trials was 19% compared to 41% on genomically-matched trials, OR 2.91 (95% CI 1.77, 4.80), p < 0.0001. mOS on non-genomically matched trials was 15.5 months compared to 22.1 months on genomically-matched trials, HR 0.70 (95% CI 0.50,0.98), p = 0.031. Conclusions: Enrollment on genomically-matched phase 1 trials is associated with an improvement in clinical benefit rate, time to progression, and overall survival in heavily pretreated, metastatic sarcoma patients. While RR remained low, we report the mutations associated with responses on genomically-matched trials. A prospective, biomarker-driven genomically-matched basket trial for these alterations is warranted in advanced sarcomas.