Applications of Circulating Tumor DNA in a Cohort of Phase I Solid Tumor Patients Treated With Immunotherapy

• Published in: JNCI cancer spectrum 2021-06, Vol.5 (3)
• Main Authors: Araujo, Daniel V, Wang, Ao, Torti, Dax, Leon, Alberto, Marsh, Kayla, McCarthy, Aoife, Berman, Hal, Spreafico, Anna, Hansen, Aaron R, Razak, Albiruni-Abdul, Bedard, Philippe L, Wang, Lisa, Plackmann, Eric, Chow, Helen, Bao, Hua, Wu, Xue, Pugh, Trevor J, Siu, Lillian L
• Format: Article
• Language: English
• Subjects: Adult; Aged; Circulating Tumor DNA - blood; Circulating Tumor DNA - genetics; Clinical Trials, Phase I as Topic; Cohort Studies; DNA Mutational Analysis - methods; Female; Humans; Immune Checkpoint Inhibitors - therapeutic use; Immunotherapy; Kaplan-Meier Estimate; Male; Middle Aged; Mutation Accumulation; Neoplasms - blood; Neoplasms - genetics; Neoplasms - pathology; Neoplasms - therapy; Paraffin Embedding; Pilot Projects; Prognosis; Proportional Hazards Models; Treatment Outcome; Young Adult
• ISSN: 2515-5091; 2515-5091
• DOI: 10.1093/jncics/pkaa122

Background The correlation between blood-based tumor mutation burden (bTMB) and tissue-based tumor mutation burden(tTMB) has not been broadly tested in a multicancer cohort. Here, we assess the correlation between bTMB with tTMB in phase I trial patients treated with immunotherapy. As an exploratory analysis, we evaluated circulating tumor DNA (ctDNA) dynamics in responders. Methods Patients treated with immunotherapy at the Princess Margaret phase I trials unit were enrolled. Pretreatment plasma ctDNA and matched normal blood controls were collected. Available archival tissue formalin-fixed paraffin-embedded (FFPE) samples were analyzed. A 425-gene panel was used to sequence both ctDNA and FFPE samples. Samples with TMB within the highest tertile were considered as high TMB. Results Thirty-eight patients were accrued from 25 different trials, 86.8% of which involved an anti-PD-1/PD-L1 agent. Thirty patients (78.9%) had detectable mutations in ctDNA, of which the median (range) bTMB was 5 (1-53) mutations per megabase (mut/Mb). Of the 22 patients with available FFPE samples, mutations were detected in 21 (95.4%); the median (range) tTMB was 6 (2-124) mut/Mb. Among the 16 patients with detectable mutations in both FFPE and ctDNA, a statistically significant correlation between bTMB and tTMB was observed (ρ = 0.71; P = .002). High TMB was not associated with better survival. All 3 responders had a decrease in the variant allele frequency of mutations detected in ctDNA at a second timepoint relative to baseline, indicating a potential early marker of response. Conclusions In this small series, bTMB correlated with tTMB. An on-treatment decrease in VAF of mutations detected in ctDNA at baseline was observed in responders. Larger studies to verify our findings are warranted.