Circulating tumor DNA as a molecular biomarker in the phase II trial of imatinib plus binimetinib in patients with advanced gastrointestinal stromal tumors.

PURPOSE This study investigated factors that affected ctDNA detection and ctDNA as a molecular biomarker in a phase II trial of imatinib and binimetinib in newly diagnosed advanced GIST, including patients exposed to imatinib within 4 weeks of trial enrollment. EXPERIMENTAL DESIGN Plasma and tumor tissue samples were collected at baseline, on-treatment, and upon progression. DNAs extracted from plasma and tumor tissue were analyzed using genomic assays, MSK-ACCESS and MSK-IMPACT, respectively. Sequenced ctDNA detection of the primary oncogenic driver was determined and correlated with clinical characteristics. RESULTS Patients (n=31) included in this analysis had KIT mutant (n=29, 94%), metastatic disease (n=24, 77%), and achieved the best response of partial response (n=22, 71%), stable (n=8, 26%) or progressive disease (n=1, 3%). Sixteen patients (52%) were exposed to imatinib at baseline. The ctDNA detection rate of the primary oncogenic driver at baseline was 39% (n=12/31), and significantly more likely in patients that were treatment-naïve (n=15) or had ≤4.2 weeks of treatment (n=8) than otherwise (48% versus 13%, p-value = 0.004). Baseline ctDNA detection did not correlate with tumor burden or stage. The ctDNA serial analysis of the primary oncogenic driver paralleled and sometimes preceded radiographic response. CtDNA detected resistance mutations in KIT (n=4). Active treatment influenced detection of secondary KIT alterations in one patient. CONCLUSIONS Active therapy at the time of ctDNA collection negatively affected the ability to detect primary and secondary KIT alterations in sequenced ctDNA from patients with advanced GIST. CtDNA responses may precede radiographic responses, and merits further investigation.