Introduction. Despite advances in diagnostics and treatment, the pancreatic cancer prognosis remains poor. Circulating DNA analysis in liquid biopsy could enable better prediction of disease course and allow for personalized treatment. In our study, we aimed to evaluate the role of liquid biopsy in detecting tumour-specific mutations, monitoring treatment response, and predicting clinical outcome in patients with pancreatic cancer.
Hypotheses. 1. Analysis of cell-free DNA and extracellular vesicle-associated DNA achieves comparable success rate in identifying tumour-specific mutations as the analysis of DNA isolated from surgically resected tumour tissue. 2. A higher proportion of tumour-specific mutations in cell-free DNA or extracellular vesicle-associated DNA prior to tumour resection reflects greater disease burden or tumour unresectability and is associated with poorer overall survival. 3. After radical tumour resection, tumour-specific mutations cannot be detected in cell-free DNA or extracellular vesicle-associated DNA, while their presence enables early detection of disease recurrence or progression.
Methods. We included 83 patients with histologically confirmed pancreatic ductal adenocarcinoma who underwent surgery with curative intent. Blood samples were collected before surgery and at follow-up visits one, six, and twelve months postoperatively for cell-free DNA and extracellular vesicle-associated DNA isolation. Archived formalin-fixed paraffin-embedded tissue samples were used as reference material for identifying tumour-specific mutations. Molecular analyses were performed using competitive allele-specific PCR (castPCR) and digital PCR (dPCR) to determine the presence and frequency of selected common somatic mutations in four pancreatic cancer driver genes: KRAS, TP53, SMAD4, and CDKN2A.
Results. Using castPCR, the most frequently detected mutations in tumour tissue DNA were KRAS p.G12D and TP53 p.R273H. These mutations were selected for further analysis of DNA from tumour tissue and longitudinal samples of cell-free DNA using dPCR. Isolation and analysis of extracellular vesicle-associated DNA were performed in a subgroup of 10 patients; however, mutation detection was limited by low DNA yield. In contrast to primary tumour analysis, the detection of selected mutations in cell-free DNA was markedly lower. Due to the low frequency of mutant alleles in cell-free DNA, we did not observe a statistically significant correlation between the mutation frequency in tumour tissue and cell-free DNA at surgery. The frequency of mutations in cell-free DNA samples at surgery was partly associated with preoperative disease burden. No consistent association was observed between the mutation frequency in longitudinal cell-free DNA samples and the clinical course of the disease.
Conclusion. Liquid biopsy could not completely replace tissue DNA analysis, especially in cases of limited disease. A higher proportion of mutations in cell-free DNA at the time of surgery correlated with greater disease burden but was not associated with survival. The proportion of mutations in cell-free DNA was also not useful for monitoring disease progression. The results of our study, within the framework of the methodology used, do not support the use of liquid biopsy with analysis of circulating DNA forms in the treatment of patients with pancreatic cancer.
|
