Longitudinal clonotype tracking by next-generation sequencing (NGS) based T-cell receptor (TCR) rearrangement measurable residual disease (MRD) assay in frontline adult T-acute lymphoblastic leukemia (ALL)

Abstract

Abstract Introduction: Use of NGS assays based on IGH B-cell receptor (BCR) gene rearrangement has been a major advance in the treatment of B-ALL. (Short NJ et al, Blood Adv 2022). However, with the TCR sequences being less unique as compared to IGH, the utility of TCR NGS MRD in T-ALL remains unknown. We aimed to evaluate the longitudinal patterns of TCR clonotype tracking in adult T-ALL treated in a frontline setting at our center. Methods: TCR clonotype tracking was performed by the ClonoSEQ (Adaptive BioTechnologies, lnc.) NGS MRD assay, with a sensitivity of 10-6. Patients (pts) were included if they achieved CR on frontline therapy, had a trackable TCR clone, and underwent NGS MRD assessment in at least one time point. Results: 39 pts met inclusion criteria; median age was 39 years (range, 20-72). 8 pts (21%) had ETP-ALL, 11 (28%) T-LBL and 28 (72%) T-ALL. Most common treatment was combination of hyper-CVAD with nelarabine, pegasparaginase and/or venetoclax (n=33, 87%). Other therapies included mini-hyper-CVD + venetoclax (n=4, 10%), hyper-CVAD (n=1, 2.5%) and pediatric inspired (n=1, 2.5%). Overall, 4 pts (10%) proceeded to Allo SCT. The median follow-up of the whole cohort from diagnosis was 28.43 months (range, 2.17-82.8 months). There were a total of 5 relapses at a median remission duration of 14.4 months (range, 5.6-16.8 months). The number of trackable sequences available per pt was median 3 (range, 1-7). 26 pts (66%) had both TCRG and TCRB sequences, 11 (28%) pts had only TCRG, and 2 (5%) had only TCRB trackable. 26 of the 37 pts with TCRG sequence had >1 sequence trackable (2 sequences in 21, 3 sequences in 3, 4 sequences in 2). 13 of the 28 pts with TCRB sequence had >1 sequence trackable (2 sequences in 9, 3 sequences in 3, 4 sequences in 1). There were a total of 191 NGS MRD assays available during the follow-up duration with 145 in bone marrow (BM) and 46 in peripheral blood samples. The first NGS MRD assay was performed during intensive phase of treatment in 30 (77%) pts, maintenance phase in 6 (15%) pts and after completion of therapy in 3 (7%) pts. A total of 82 (43%) NGS assays in 30 pts were performed while on intensive phase of therapy, 76 (40%) assays in 22 pts were performed during maintenance phase, 11 (6%) assays in 4 pts were post Allo SCT and 22 (11%) assays in 7 pts were after completion of all therapy. 10 (7%) BM samples were positive for MRD by flow cytometry (FC), with disease quantification of a median 0.11% (range, 0.01%-3.11%). NGS MRD was detectable in all 10 FC positive samples. The median quantified NGS MRD in these samples was 1511 per million (range 4-21590/million). Three pts (8%) had persistent NGS MRD values >10000/million while the corresponding BM samples were in flow MRD-negative remission. Two pts had persistently elevated TCRB and TCRG sequences and one had TCRB only. These very high NGS values were thought to be tracking nonmalignant T cells rather than T-ALL and were disregarded for further analysis. Amongst 28 pts who had longitudinal NGS MRD tracking during intensive phase of therapy, the best NGS MRD response was negative in 16 (57%) pts and positive in 12 (42%) pts. There was one relapse event in the NGS negative group which was a myeloid lineage switch; cumulative incidence of relapse was 12% at 3 years. There were 4 relapse events in the NGS positive group; cumulative incidence of relapse was 38% at 3 years. (HR 6.36, 95% CI 1.05-38, P=0.0268) In 3 of the 5 pts to relapse, the NGS MRD after initially achieving best response of 54, 13 and 3/million respectively, started progressing while on intensive chemotherapy. The later two received an allo SCT consolidation but failed to clear NGS MRD. These 3 pts relapsed 3.7, 10.2 and 12.9 months respectively post best response. The other 2 of the 5 pts to relapse had best NGS MRD responses of 0 and 1/million. The first pt had a relapse with myeloid lineage switch 16.9 months post best response and the second pt had CNS relapse with BM NGS progression to 102/million 3.7 months post best response respectively. Conclusion: After adjusting for non-leukemic tracking sequences, pts with the persistent NGS MRD during the intensive phase of therapy had significantly higher risk of relapse than those who achieved NGS MRD negativity. Our findings suggest that NGS MRD is prognostic in T-ALL, although more pts with longer follow-up are needed to determine the prognostic impact of NGS MRD dynamics and outcomes in T-ALL.