Characterizing the T cell receptor (TCR) repertoire is essential for understanding antigen recognition and predicting clinical outcomes in patients with AML (pts). Disease phases and treatment regimens can significantly alter the TCR repertoire. However, most studies only analyze the TCR repertoire at a single time point, which may not adequately reflect its dynamic changes throughout leukemia treatment and disease progression. Here, we examined the temporal landscape of the TCR repertoire to assess how it evolves in response to different treatment regimens and disease stages in AML.

TCR sequencing was conducted on RNA samples from peripheral blood mononuclear cells of two AML pts. For pt USC5, who harbors FLT3-ITD, DNMT3A, and RUNX1 mutations, 9 samples were collected over 12.5 months. The pt underwent CLAG reinduction therapy (cladribine, cytarabine, G-CSF, mitoxantrone) for relapse, followed by decitabine and a regimen of fludarabine, cytarabine, and etoposide during the refractory stage, and subsequently a cycle of gemtuzumab ozogamycin, venetoclax, and 5-azacitidine. For pt USC6, with a FLT3-ITD mutation and post-second transplant complications including acute GVHD, CMV, and EBV infection, 6 samples were collected over 6.5 months; the pt received midostaurin treatment before relapsing at the last time point. TCR-seq libraries were prepared using the Takara 5'-RACE-like kit to target the complementarity determining region 3 (CDR3) and sequenced on the Illumina MiSeq platform. TCR-seq data were processed using MiXCR and visualized with the VDJTools and immunarch R package.

The sequencing results yielded an average of 1,270,356 ± 482,051 reads per sample, revealing an average of 8,279 ± 7,993 unique TRA and 5,095 ± 4,065 unique TRB clonotypes, covering most of the functional V and J exons (47 TRAV and 59 TRBV, 56 TRAJ and 13 TRBJ). For pt USC5, the lowest clonotype counts were observed at refractory disease stage for TRA (599) and at relapse for TRB (1,440) and the highest was post decitabine treatment (TRA: 22,570, TRB: 13,290). In pt USC6, the lowest clonotype counts were noted at 1-2 months post-CMV and EBV infection (TRA: 569, TRB: 670), and the highest occurred 2-3 months post-second transplant (TRA: 8,200, TRB: 3,265).

Hyperexpanded clones represented a significant portion of the TCR repertoire across disease progression and treatment in USC5 (TRA: 53 - 70%, TRB: 33 - 59%) and USC6 (TRA: 41% - 67%, TRB: 43 - 75%). The top 10 expanded TCR clones showed high conservation, with the dominant TRAV10 and TRBV4-3 genes capturing 11.3 - 33.9% and 10.9 - 22.4% of the entire repertoire in USC5. In USC6, the dominant TRAV4 and TRBV5-1 genes represented 5.52 - 14.64% and 10.61 - 28.37% of the TRA and TRB repertoires, respectively. The most abundant clonotypes fluctuated with treatment. In USC5, CVVSALNDYKLSF peaked at 31.1% of the TRA repertoire post-decitabine and dropped to 9.7% at the refractory period, while CASSPARNTEAFF ranged from 7.0% to 18.5% in the TRB repertoire. In USC6, CLHNFNKFYF and CASSLVVGQKNTEAFF showed similar variability, contributing significantly at specific treatment points and dropping at others.

The Morisita-Horn index was applied to quantify the overlap of TCR repertoires across various time points. In USC5, samples from the refractory disease stage had lower overlap indices, ranging from 0.49 - 0.75 for TRA and 0.63 - 0.86 for TRB, compared to 0.89 - 0.99 for TRA and 0.77 - 0.99 for TRB in other samples. For USC6, a sample at relapse stage showed notably lower overlap indices, ranging from 0.29 - 0.55 for TRA and 0.40 - 0.68 for TRB, and a sample collected post-CMV and EBV infection exhibited an overlap of 0.40 - 0.81 for TRB, compared to higher overlaps in other samples (0.63 - 0.97 for TRA and 0.62 - 0.99 for TRB). Hierarchical clustering revealed clusters and subclusters that closely correspond to specific treatments received and clinical events at the time of sampling, indicating a strong association between treatment, disease progression, and TCR profiles.

In conclusion, the TCR repertoire in AML pts is largely conserved across treatment, though marked deviations occur at clinical events such as relapse or viral infections. These findings highlight TCR profiling's value as a reliable and stable biomarker for monitoring dominant clonotypes, and they emphasize the necessity of multiple time-point analyses to fully capture the TCR repertoire's dynamic changes during critical clinical events.

Disclosures

Yaghmour:USC Keck School of Medicine: Current Employment; Pharmacyclics: Research Funding; Abbvie: Other: Advisory Board; Daiichi: Other: Advisory Board; Astazeneca: Other: Advisory Board; GSK: Speakers Bureau; Rigel: Speakers Bureau; Servier: Speakers Bureau; Jazz: Speakers Bureau; Bristol Myers Squibb: Speakers Bureau; Kite: Speakers Bureau; SOBI: Speakers Bureau; Incyte: Speakers Bureau; Secura Bio: Speakers Bureau; Astellas: Speakers Bureau; Blueprint: Speakers Bureau; ABBVie: Speakers Bureau; Stemline therapeutic: Speakers Bureau; Alexion: Other: consult. Alachkar:Servier: Consultancy.

This content is only available as a PDF.
2024
Sign in via your Institution