Personalized Monitoring of Circulating Tumor DNA By a Specific Signature of Trackable Mutations after Chimeric Antigen Receptor t-Cell Therapy in Non-Hodgkin B Cell Lymphoma

Abstract

Introduction CART-cell therapy (CART) has produced a paradigm shift in the treatment of relapsing non-Hodgkin B-cell lymphoma (NHBcL). However, patients continue to relapse. Thus, developing strategies to optimize disease surveillance after this therapy is becoming increasingly important to attain disease-free survival and predict a potential relapse in advance. To facilitate its applicability, such strategies should also be minimally invasive. Hence, this study aims to explore the value of CloneSight, a circulating tumor DNA (ctDNA) monitoring approach based on a signature of personalized trackable mutations for the follow-up of NHBcL under CART. Methods A total of 30 NHBcL patients treated with CD19 CART-cell therapy were included in this work. Ten patients were diagnosed with Follicular Lymphoma (FL), and 20 with Diffuse Large B-Cell-Lymphoma (DLBCL). Genomic profiling was performed at relapse using FFPE biopsies and cfDNA liquid biopsies to detect somatic mutations by a custom capture enrichment panel (Twist, USA) covering 134 genes. Sequencing was conducted with the Illumina NextSeq platform. Then, the patient-specific somatic mutations were used as biomarkers for quantifying Minimal Residual Disease (MRD) on liquid biopsy follow-up samples with a sensitivity of 10 −4 (Jiménez-Ubieto A et al., Leukemia 2023). The amplicon-based, patient-specific MRD mini-panel also included primers to quantify the CART Phi construct relative to the albumin control gene. A total of 118 peripheral blood samples were collected. Plasma was isolated at day +7, +14, +30, +60, +90, and before progression (59 DLBCL and 59 FL samples). PET/CT examinations were performed on days +90, +180, +365 and every six months in FL, and the same for DLBCL but additionally on day +30. Results We found 185 trackable somatic mutations suitable for MRD monitoring (mean: 6.2 per patient). The most frequently mutated genes were CREBBP (80%), KMT2D (50%) and EP300 (30%) in FL, and KMT2D (45%), CREBBP (40%), and TP53 (35%) in DLBCL. The applicability of the test (detection of at least one somatic mutation) was 100%. Nineteen patients presented a cfDNA evaluation between days +30 and +60 ( Figure, left). Twelve progressed and ten of them were CloneSight-positive. Patient FL107, negative at day 28, presented a positive result at progression (day 180). For patient DLBCL16, the last sample available was at day +30, but the progression occurred at day 80, highlighting the need for sequential response assessment. Regarding the remaining nine cases in CR, five were negative and two positives by the test (SE 83%, SP 71%, PPV 83%, NPV 67%). Of note, FL101 becomes CloneSight-negative in the next Time-point (TP) available (day 112) and maintains MRD negativity in three more TPs (last at day 907). The remaining discordant case, DLBCL21, has a follow-up of only 60 days. Of note, patient DLBCL3 did not respond to CART but achieved CR and Clonesight-negativity under immune checkpoint inhibitor therapy. The ctDNA surveillance highly correlated with the PET/CT results ( Figure, right). Remarkably, 27 out of 32 PET/CT-negative TPs (84%) were consistent with a CloneSight-negative determination on belong two patients that did not progress. Four of five TPs with PET/CT-negative and CloneSight-positive status were classified as false negative PET/CTs, as those patients finally progressed. On the other hand, 13 out of 16 positive PET/CT samples (81%) were CloneSight-positive and the patient progressed. Of the remaining three PET/CT positive CloneSight-negative, two belong to patients that also progressed. Conclusions: In CAR-T therapy, monitoring with MRD by ctDNA is consolidated as a useful and necessary tool. Our CloneSight test can predict DLBCL and FL outcomes in advance and could be of high utility to detect false positive or negative PET/CT assessments during follow-ups. However, dynamic response assessment is crucial to draw meaningful conclusions. Further research and a larger cohort will be presented at the meeting.