Establishment of PCR based methods for detection of ctDNA in blood

Abstract

Cell-free DNA appears in the circulation when cells undergo apoptosis and necrosis. This is a natural process and small amounts of cell-free DNA can be found in the blood of healthy individuals. When tumors are present, a variable part of the cell-free DNA will derive from tumor. Monitoring ctDNA is attractive in solid cancers because a blood sample is less invasive than a biopsy and it might represent all lesions including metastasis. “Liquid biopsies” also give the opportunity for serial monitoring during therapy and follow up when tumor biopsies are difficult to achieve. Little is still known about the stability of ctDNA in vitro, so one of the subprojects in this thesis was to elucidate this. Cell-free DNA is extracted from plasma and ctDNA is often a minor fraction of the total plasma DNA. Our results showed that blood can be left for 24 hours in room temperature without significantly affecting the total plasma-DNA concentration. It was however not clear if ctDNA showed a slight decrease due to delayed processing or whether it was due to assay variation. More data needs to be collected. ERBB2 is amplified in 15-20 % of breast cancers and targeted therapy is available against its protein (HER2). Studies have shown that HER2-status might change during disease progression, so detection of the amplification in plasma would be desirable. Other methods like quantitative PCR (qPCR) and digital PCR (dPCR) have been used to detect the amplification, but these rely on the use of an unamplified reference gene. Breast cancers are heterogeneous and show a variety of genetic alterations. A common unamplified gene can therefore be hard to find. One of our subprojects detected ERBB2 amplification by targeted massively parallel sequencing (MPS). ERBB2 is a monoallelic event, and SNP ratios could be used to detect amplification. No reference gene was needed as the unamplified allele was the reference. The method achieved the same limit of detection as dPCR, 1.2 fold. Breast cancer also has few recurrent mutations. Massively parallel sequencing (MPS) has given the opportunity to detect the broad range of mutations that can be seen in breast tumors and to use these mutations as tumor markers for detection of ctDNA. The last subproject screened 17 genes frequently mutated in breast cancer with targeted MPS and detected somatic mutations in 94 % of the patients (49/52). 48 % of the patients had two mutations that could be tracked in serial blood samples taken during therapy. Both of the cases presented in this thesis showed evidence of subclonality. In one case, the KRAS mutation emerged during therapy. This might be due to acquisition of the mutation during treatment or that it belonged to minor subclone that was not detected at presentation. This emphasizes the importance of detecting all subclones at disease presentation and not only tracking the dominant clone in the primary tumor. Monitoring ctDNA is an important means to detect therapy response and early relapse and might be a step towards more personalized medicine.