Application of cfDNA-NGS Test in MRD Monitoring of Acute Leukemia Patients with CAR-T Cell Therapy

Background: The minimal residual disease (MRD) plays an important role in the comprehensive management of CAR-T cell therapy. However, some reasons hinder the application of MRD monitoring in CAR-T cell therapy. Dynamic monitoring of cfDNA in solid tumors and hematological malignancies can effectively evaluate the efficacy of treatment and predict long-term prognosis. However, there is no relevant researches on whether cfDNA can replace bone marrow for MRD monitoring after CAR-T cell therapy for acute leukemia.

Methods: 23 patients with acute leukemia who received CAR-T cell therapy from January 2021 to October 2022 in the First Affiliated Hospital of Suzhou University were collected retrospectively, including 14 AML patients and 9 ALL patients. The CAR-T cell targets included CD38, CD7, CLL-1 and CD19/CD22. Bone marrow and peripheral blood samples were collected before CAR-T cell therapy, +28d and every month after CAR-T cell therapy. We used QIAamp Circulating Nucleic Acid Kit to extract cfDNA, and then conducted high-throughput sequencing through Nextseq550 sequencing platform. We compared the test results of cfDNA-NGS and BM-NGS in terms of gene mutation types and variant allele frequency (VAF), and analyzed the efficacy of cfDNA-NGS, BM-NGS and FCM on the prediction of remission and replase of CAR-T cell therapy by the receiver operating characteristic curve (ROC curve).

Results: In this study, a total of 123 peripheral blood samples were collected and cfDNA was extracted from all samples. The results of 121 cfDNA-NGS and BM-NGS test were consistent, with a consistency rate of 98.4% (121/123), and the consistency rate with bone marrow FCM detection was 85.4% (105/123). cfDNA-NGS test was a good indicator for predicting the efficacy and replase after CAR-T cell therapy: compared with BM-NGS and FCM, cfDNA-NGS test before CAR-T cell therapy had the highest prediction efficiency for CR (AUC=0.767, 95%CI: 0.576-0.958, P=0.039). Survival analysis showed no difference in 1-year OS and LFS between cfDNA-NGS (+), BM-NGS (+) , FCM (+) patients at +28 days after CAR-T cell therapy (OS: 27.3% vs 33.3% vs 30%, P=0.949; LFS: 27.3% vs 33.3% vs 30%, P=0.970). At +28 days after CAR T treatment, there was no difference in 1-year OS and LFS between patients with cfDNA-NGS (-), BM-NGS (-), FCM (-) (OS: 91.7% vs 76.2% vs 83.3%, P=0.671; LFS: 73.3% vs 76.2% vs 82.1%, P=0.726). Then we assessed the long-term prognosis by cfDNA-NGS combined with traditional MRD detection methods after CAR-T cell therapy: Survival analysis showed that 1-year OS and LFS of patients with cfDNA (-), BM-MRD (-) and cfDNA1 (-), BM-MRD (+) were significantly better than those with cfDNA (+), BM-MRD (-) and cfDNA1 (+), BM-MRD (+) (OS: 90.9% vs 100% vs 50% vs 22.2%, P=0.008: LFS：68.2% vs 100% vs 50% vs 22.2%, P=0.017).

Conclusions: The cfDNA-NGS test was a good indicator for predicting the efficacy and replase after CAR-T cell therapy, and was not inferior to BM-NGS and FCM detection in evaluating efficacy, replase and long-term prognosis. Combination cfDNA-NGS test with traditional MRD detection methods could evaluate the long-term prognosis of CAR-T cell therapy more comprehensively.