Utility of liquid biopsy (LB)-based comprehensive genomic profiling (CGP) of circulating tumor DNA (ctDNA) in lymphomas and plasma cell neoplasms.

Abstract

e19543 Background: LB-based CGP of ctDNA can facilitate the molecular evaluation of lymphoid and plasma cell neoplasms (PCNs), especially when traditional CGP is infeasible due to insufficient tumor in the blood or tissue or when tissue biopsy itself cannot be performed. Methods: Retrospective study of LB samples from patients reported to have lymphomas or PCNs tested using hybrid-capture next-generation sequencing using the FoundationOne Liquid CDx (targeting 324 genes), FoundationOne Liquid (70 genes), or FoundationACT (62 genes) assays during routine clinical care between 5/2016 and 9/2021. Results: Among 48,459 unique patient samples submitted for CGP, 133 were from lymphomas or PCNs: 69 non-Hodgkin lymphoma (NHL), 32 PCN, 16 diffuse large B-cell lymphoma (DLBCL), and 16 classic Hodgkin lymphoma (cHL) cases. The median cell-free DNA yield was 72.7 ng (interquartile range [IQR]: 46.0-157.0), well above the minimum quantity needed for analysis. Reportable pathogenic genomic alterations were detected in 67.7% (90/133) of cases. The median maximum somatic allele frequency (MSAF) was 3.8% (IQR: 1.1-19.6%), with no statistically significant differences noted between DLBCL (8.2%), NHL (5.6%), cHL (1.7%), and PCNs (1.4%) ( P= 0.13, Kruskal-Wallis test). In all, 322 pathogenic short variants and 3 pathogenic rearrangements were detected (median variant allele frequency [VAF]: 0.6%, IQR: 0.3-2.6%, range: 0.09-81.8%). One hundred ninety-seven had VAFs < 1.0%, indicating excellent sensitivity. NHLs exhibited pathogenic alterations in TP53 (53.2% of cases in which pathogenic alterations were identified), KMT2D (14.9%), NF1 (10.6%), MYD88 (8.5%), BCL2 (6.4%), BRAF (6.4%), EZH2 (6.4%), NRAS (6.4%), BTK (4.3%, all of which were C481X ibrutinib resistance mutations in cases of chronic lymphocytic leukemia detected at VAFs ranging from 0.18-5.5%), KRAS (4.3%), RB1 (4.3%), and TNFAIP3 (4.3%), among others. PCNs exhibited pathogenic alterations in TP53 (54.6%), KRAS (18.2%), NF1 (13.6%), PTPN11 (13.6%), BRAF (9.1%), NRAS (9.1%), MAP2K1 (4.6%), and RB1 (4.6%), among others. DLBCLs exhibited pathogenic alterations in TP53 (63.6%), MYD88 (27.3%), CCND2/3 (9.1%), EZH2 (9.1%), and KMT2D (9.1%), among others. CD274 (PD-L1) and PDCD1LG2 (PD-L2) amplification and IGH- PAX5 and IGH- MYC fusions were also detected in individual cases of DLBCL. cHLs exhibited pathogenic alterations in TP53 (40.0%), ATM (20.0%), BCOR (10.0%), KRAS (10.0%), MYD88 (10.0%), and PDCD1LG2 (PD-L2) (10.0%), among others. Conclusions: LB identified clinically actionable genomic alterations in lymphomas and PCNs and may be used when traditional CGP of tumor biopsy specimens cannot be performed. Given its low limit of detection, LB may also allow for minimally invasive dynamic monitoring of response to therapy.