104P - The development of a classifier of somatic copy number alteration burden in liquid biopsy with potential clinical impact in advanced non-small cell lung cancer (NSCLC)

Background

Liquid biopsy has recently emerged as an important tool in precision medicine and in cancer management. Beyond the mutational profile, cell free DNA (cfDNA) analysis can provide information regarding copy number alterations, fragment composition and methylation marks.

Methods

We retrospectively evaluated results of plasma NGS analysis performed at our Institution by using a multigene panel of 77 genes that detects the 4 major classes of genetic alterations, according to clinical practice or spontaneous translational studies. We developed a support vector machines (SVM) classifier to automatically classify chromosomal profiles as stable (SCP) or unstable (UCP). In a subset of patients, validation of the classifier results was performed by shallow whole genome sequencing (sWGS), an established application for assessment of the tumor fraction (TF) in cfDNA.

Results

Amongst 177 samples analyzed, 28 (15.8%) were classified as unstable according to their chromosomal profile. High concordance was found between binary classification and TF evaluation by sWGS. Mean TF was 3.6% and 36.6% in SCP and UCP samples respectively. Among clinical features, patients with an UCP were more likely to have ≥3 metastatic sites (p=0.009) and liver metastases (p=0.010). Longitudinal analyses were performed in 33 advanced NSCLC patients treated with immune checkpoint inhibitors (ICIs), as exploratory analyses. Among these, baseline UCP was not significantly associated with outcome endpoints but UCP was observed 3 weeks after the beginning of ICIs in 7 out of 17 patients experiencing either early death (ED) or hyper-progression (HPD). UCP was never found among patients not experiencing ED or HPD after the start of ICIs.

Conclusions

Machine learning approach is able to define a binary classifier of somatic copy number alteration burden starting from a NGS test performed in plasma according to clinical practice. This classifier is potentially useful for clinical risk stratification during systemic treatment of NSCLC patients.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Istituto Oncologico Veneto IOV IRCCS.

Funding

5x1000 IOV.

Disclosure

L. Bonanno: Financial Interests, Institutional, Advisory Board: AstraZeneca, MSD, BMS, Roche; Financial Interests, Institutional, Invited Speaker: AstraZeneca, MSD, BMS, Roche; Financial Interests, Personal, Advisory Board: Novartis; Financial Interests, Personal, Invited Speaker: Lilly; Financial Interests, Institutional, Steering Committee Member: AstraZeneca; Non-Financial Interests, Principal Investigator: Roche, AstraZeneca, Boehringer Ingelheim, MSD, BMS, Janssen, PharmaMar, Arcus Biosciences. D. Rose: Financial Interests, Personal, Full or part-time Employment: Roche. G. Pasello: Financial Interests, Personal, Invited Speaker: Amgen, Eli Lilly, Novartis, MSD, Pfizer; Financial Interests, Personal, Advisory Board: AstraZeneca, Roche, Janssen; Financial Interests, Institutional, Research Grant: Roche; Financial Interests, Institutional, Other, unconditioned support: AstraZeneca, MSD; Non-Financial Interests, Principal Investigator: AstraZeneca, Roche, Novartis, Lilly, Janssen, PharmaMar. V. Guarneri: Financial Interests, Personal, Invited Speaker: Eli Lilly, Novartis, GSK, AstraZeneca, Gilead, Exact Sciences; Financial Interests, Personal, Advisory Board: Eli Lilly, Novartis, MSD, Gilead, Eli Lilly, Merck serono, Exact Sciences, Eisai, Olema Oncology, AstraZeneca, Daiichi Sankyo, Pfizer; Financial Interests, Institutional, Local PI: Eli Lilly, Roche, BMS, Novartis, AstraZeneca, MSD, Synton Biopharmaceuticals, Merck, GSK, Daiichi Sankyo, Nerviano, Pfizer; Non-Financial Interests, Member: ASCO. All other authors have declared no conflicts of interest.