Github page for the project: Clinical utility of subclonal evolution inferred using liquidCNA in metastatic breast cancer

Thesis submitted for MPhil in Computational Biology, University of Cambridge

Abstract

Background: Liquid biopsy and the analysis of cell free DNA (cfDNA) in the blood offer a minimally invasive method to moni- tor cancer evolution. Recently, liquidCNA, a computational algo- rithm that infers subclonal architecture has been released. Un- like current methods which uses mutation profiles, liquidCNA utilises information on somatic copy number aberrations from low-pass whole genome sequencing (lpWGS) 0.1x. LiquidCNA offers an economical method to characterise subclonal events. The purpose of this study was to assess the clinical utility of liquidCNA’s output in monitoring response to therapy.

Methods: Data consisted cfDNA from 283 plasma samples from 80 advanced metastatic breast cancer patients. Using liquidCNA, tumour-fraction and the size of emergent subclones were esti- mated for each of the samples. These estimations were analysed with radiographic response as evaluated by Response Evaluation Criteria In Solid Tumours (RECIST).

Results: The size of emergent subclones were non-significantly associated with progression (p-value = 0.883) and metastasis (p-value = 0.302). Inferred subclonal events were also weak pre- dictors of recurrent metastasis (p-value = 0.334, HR = 1.243, CI = 0.799–1.935). Our study showed the non-significant results to arise from limitations of the algorithm regarding tumour-fraction estimation. Additionally, the study highlighted prominent limita- tions of liquidCNA and provided modifications to the algorithm that can be implemented by future users.