Urothelial Bladder Cancer (UBC) is common and has a high recurrence rate. Recurrence is managed by regular monitoring by cystoscopy, however this is an invasive procedure. Detection of circulating tumour DNA (ctDNA) in liquid biopsies is non-invasive, and may identify a relapse earlier. Previously, we have shown that PI3KCA E542K and TP53 Y163C mutation assays can identify the presence of ctDNA in patient's plasma and urine by digital droplet PCR (ddPCR) (presented at EACR 3rd Cancer Genomics, June 2017). However, an overall lack of common mutations in UBC is a challenge, as patient coverage by the targeted assays would be low. Birkenkamp-Demtroder et al 2016 (European Urology, Volume 70, p75-82) suggested that patient-specific assays from whole exome sequencing (WES) could be an option for ctDNA-based monitoring of UBC. The aim of this preparatory pilot study was to investigate whether WES-based approach could be used for mutation identification to produce patient specific ddPCR assays, in our local hospital Queen Elizabeth University Hospital, Glasgow. The objectives were; 1) To evaluate the ability of WES to identify mutations in FFPE tumour tissues, 2) To assess how tumour heterogeneity influences the identification of assayable mutations in each patient, 3) To evaluate the efficiency in producing patient-specific assays, starting from assay design to optimization.
WES, carried out on FFPE pT2G3 UBC tumour tissues (n=3), identified 19674-39776 tumour-specific mutations. There were no common mutations among three tumours. Mutations from the three areas within one tumour contained 19618-21525 mutations, of which 142 were common. CTNAP4 G727* mutation was selected from the pool and the mutation was confirmed by Sanger sequencing. A designed ddPCR assay was able to detect the mutation in the tumour DNA.
This preliminary study showed that it was possible to correctly identify tumour mutations from patients' FFPE tumour tissues, and subsequently to produce ddPCR assays. Tumour heterogeneity will remain as a challenge, as WES is unlikely to provide any advantage in finding common assay targets. Target selection will need consideration of intra-tumour heterogeneity, in addition to the expected clonal pressures from treatments, which could cause assayable mutations to be lost in the course of follow-up.