Application of circulating cell-free tumor DNA profiles for therapeutic monitoring and outcome prediction of metastatic melanoma

Abstract

The implementation of precision oncology requires novel disease and therapy monitoring technologies that are non-invasive, sensitive and specific. Circulating cell-free tumor DNA (ctDNA) reflects the heterogeneous spectrum of specific mutations, especially in systemic disease. Specifically, this thesis aimed to establish and validate plasma-based assays that allow the dynamic quantitative detection of ctDNA as a prognostic biomarker for tumor load and prediction of therapy response in melanoma patients using droplet digital PCR (ddPCR) and next generation sequencing (NGS). ctDNA monitoring by ddPCR showed excellent sensitivity and high reproducibility. After technical validation of ddPCR with cell line-derived DNA (sensitivity 0.01%), plasma-derived ctDNA was analyzed from a large training cohort (N=96) of advanced stage melanoma patients with assays for the BRAFV600E and NRASQ61 driver mutations as well as TERTC250T and TERTC228T promoter mutations. An independent patient cohort (N=35) was used to validate the clinical utility of ctDNA monitoring under MAPK-targeted or immune checkpoint inhibition therapies. In contrast, the establishment of an amplicon based NGS protocol in ctDNA samples was hindered by technical issues related to low sensitivity, discrepancies in mutation detection, and poor agreement in mutational status of tumor tissues vs. plasma. In dilution series, using cell line-derived DNA, the lowest limit of detection was 1,000 mutant copies in the background of 10,000 wild-type copies resulting in only 10% analytical sensitivity. The ddPCR ctDNA results were evaluated with various statistical methods, including ROC and Kaplan Meier analyses, where ctDNA levels were correlated with radiologic treatment response and patient survival. Elevated plasma ctDNA at baseline (i.e. before treatment) was an independent prognostic factor of disease progression when compared with serum S-100 and LDH levels in multivariable analysis (HR 7.43, 95% CI 1.01-55.19, P=0.05). Changes in ctDNA levels during therapy correlated with treatment response. For example, in patients with the BRAFV600E mutation, which drives about half of the melanomas, increasing ctDNA levels were predictive for shorter progression free survival (PFS) (HR 7.28 95% CI 3.64-14.53, P<0.0001), and predicted earlier disease progression compared to routine radiological scans (P<0.05) with a mean lead-time of 3.5 months. In BRAFV600 patients treated with signaling targeted therapies, the occurrence of secondary NRASQ61 mutation is associated with treatment failure due to therapy resistance. Accordingly, NRAS mutant ctDNA was detected in a significant proportion of samples from patients with BRAF mutant tumors under therapy, but unexpectedly also already at baseline. In vitro sensitivity studies suggested that this represents higher than expected intra-tumoral heterogeneity with pre-existence of small subclones of NRAS-mutated cells. Furthermore, the detection of plasma NRASQ61 ctDNA in baseline samples of MAPKi-treated BRAFV600E patients significantly correlated with shorter PFS (HR 3.18 95% CI 1.31-7.68, P=0.03) and shorter overall survival (OS) (HR 4.08 95%CI 1.57-10.58, P=0.01). Overall, these results show the potential clinical utility of ddPCR based ctDNA assays as a sensitive monitoring tool, and that ctDNA assessment is a clinically applicable prediction tool for the early assessment of disease progression and therapeutic response in metastatic melanoma patients.