Introduction

Genomic tracking from liquid biopsy such as urine and blood is a novel technique with the potential to monitor presence of tumour material non-invasively. To identify truncal mutations within heterogeneous tumours, an individualised or large panel multi-target approach must be used to monitor response. The complete response rate to chemotherapy for metastatic bladder cancer (BC) is limited, but the ability to identify response could be of significance in the neo-adjuvant setting. The aim of this study is to test the feasibility of detecting and monitoring somatic mutations in cell-free DNA (cfDNA) found in plasma of patients with BC pre and post-treatment.

Methods

25 patients with muscle-invasive BC and 10 patients with metastatic BC were selected for this study. Patients with metastatic BC received at least 2 cycles of gemcitabine and cisplatin chemotherapy. 10 ml of blood was collected in EDTA tubes at each timepoint. Samples were taken at baseline for all patients, twelve weeks after surgery for the cystectomy cohort, and before each cycle of chemotherapy for the metastatic cohort. Plasma was separated by centrifugation, and cfDNA extracted with the QIAamp Circulating Nucleic Acid Kit. Targeted amplification was performed using duel indexed primers, followed by next generation sequencing (NGS) using a panel of known (TERT PIKC3, FGFR3, HRSA) and novel BC mutations. Changes in cfDNA mutation burden were compared with cross sectional imaging, and response was assessed using the RECIST criteria.

Results

cfDNA was isolated from all samples at baseline and subsequent treatment. Baseline DNA quantification was performed, and a mean of 77.5 and IQR 10.9-72.7 ng/ml (Qubit) collected. This did not correlate with baseline disease burden or subsequent response to chemotherapy. Mutation profiles varied between patients, but at least one mutation was identified the plasma cfDNA of all patients. Dynamic changes in the mutational burden were detected in several patients. In a few patients with nodal disease, mutational changes were still detectable post-cystectomy.

Conclusions

Plasma cfDNA is detectable in patients with BC but its quantity is not a reliable indicator of disease burden. Mutational analysis of plasma cfDNA is feasible for use as a non-invasive biomarker of therapeutic response to chemotherapy, and the presence of residual disease post-cystectomy. Dynamic changes in mutational burden for each patient are likely to be related to their individual tumour’s heterogeneity, and the expansion of sub-clonal populations in both primary and metastatic lesions.

Funding

none