posted on 2019-07-15, 09:56authored byLuke J. Martinson
Non-small cell lung cancer (NSCLC) and malignant pleural mesothelioma (MPM) represent two of the most aggressive solid tumours, where treatment is frequently ineffective, likely due to intratumour heterogeneity and associated resistance to therapy. This thesis primarily assessed circulating tumour DNA (ctDNA) in the characterisation of these diseases within a phylogenetic framework, specifically through the targeting of variants (SNVs) acquired early during tumour evolution (clonal variants). Another project evaluated whether targeted next-generation sequencing (NGS) could resolve diagnostic ambiguity by differentiating between metastases and independent second primary malignancies.
As part of Lung TRACERx, clonal tumour variants were detected within pre-surgical cfDNA of 13/41 (32%) early-stage NSCLC patients using a deep NGS approach targeting eleven genes (>150 hotspots). Of 14 stage I adenocarcinomas, none were detected with this technology. Longitudinal profiling provided evidence of minimal residual disease following surgery, with ctDNA detection antedating clinical relapse by up to five months. Despite being a generic approach (not patient-specific), this assay targeted ≥ 1 clonal tumour variants in 41/43 (95%) patients, indicating that the majority of “hotspot” SNVs are acquired early during NSCLC evolution and hence represent ideal targets for intervention.
In 11 patients with MPM, clonal variants selected from multiregional exome data, were analysed in pre-surgical cfDNA using patient-specific ddPCR assays. Four of 11 patients (36.3%) were detected, and interestingly these preliminary findings suggest ctDNA detection may have prognostic value in MPM and potentially could be used to preoperatively stratify patients with regards to invasive surgery.
Lastly, in an unambiguous diagnostic cohort, 19/20 (95%) cases were accurately diagnosed as metastatic or independent primary by inferring clonal origin through NGS. Subsequently, ambiguous case studies were analysed, where results suggest NGS could aid clinicians in providing precise diagnosis for these patients. Together, the results of this thesis highlight how NGS technologies are fast approaching the clinic, but magnify the challenges we face in bringing about the genomic revolution.