e21560 Background: Prognosis in melanoma with brain metastasis is poor with a median survival of four months and a one-year survival rate of 10–20%. There is an unmet need for surveillance methods that can supplement imaging at regular intervals. Serial analysis of circulating tumor DNA (ctDNA) may aid surveillance and prognostication. A PCR-based, “specimen in/result out” testing device was employed to detect BRAF variants in plasma-derived ctDNA to evaluate the utility of rapid biomarker detection in the management of melanoma with brain metastasis. Methods: Serial blood samples from patients diagnosed with BRAF mutation-positive metastatic melanoma were collected at regular intervals. We employed a real-time PCR-based automated mutation detection system (Idylla; Biocartis, Belgium) to interrogate the plasma samples. The ctDNA mutation detection trend was analyzed relative to disease progression. Results: 39 patients with BRAF mutation positive melanoma were enrolled. 29 patients were treated in the metastatic setting, 10 in the adjuvant setting. 18 of the 29 patients with metastatic disease (62%) had brain metastases. Circulating BRAF mutation was detected in 17 of the 29 (59%) patients with metastatic disease, and was not detected in any patients treated adjuvantly. In the group with metastatic disease, this circulating biomarker changed from undetectable to detectable in eight (28%) and detectable to undetectable in three (10%). No change in circulating mutation status occurred in 18 (62%). In the eight patients who had an initial negative test that later became positive, seven (87%) had brain metastases. In three patients, ctDNA mutation detection occurred before the diagnosis of brain metastases on imaging, with a median lead time of five weeks (range, 3-12 weeks). In one patient with de novo metastatic disease admitted to the ICU, tissue was unavailable for BRAF testing but plasma was found to be positive for ctDNA BRAF detection. BRAF/MEK targeted therapy resulted in a sustained objective response. Five of six (83%) patients that had persistent ctDNA positivity had brain metastases. Among patients with brain metastases, median overall survival (mOS) of patients demonstrating >50% test positivity was numerically longer than those with <50% positivity (mOS 12.3 vs 53.5 months; p = 0.133). Conclusions: Plasma-based, rapid ctDNA testing may be useful as an aid in detecting progression and gauging prognosis in patients with melanoma treated in the metastatic setting. The dynamics of ctDNA test positivity may indicate a need for more urgent imaging, particularly of the brain. Blood-based, semi-automated ctDNA detection may serve as an attractive adjunct to scheduled imaging surveillance in melanoma.
Reliability of BRAF mutation detection using plasma sample
