T-cell clones of uncertain significance are prevalent in cutaneous T-cell lymphomas Free

Background: T-cell receptor (TCR) sequencing is increasingly incorporated into the diagnosis and monitoring of T-cell lymphomas, including cutaneous T-cell lymphomas (CTCL). However, interpretation of TCR sequencing results is often confounded by the emergence of multiple dominant TCR clonotypes, whose etiology remains poorly understood. We aimed to determine the incidence of multiple dominant TCR clonotypes through serial TCR high-throughput sequencing (TCR-HTS), to immunophenotype and genetically characterize these dominant T-cell populations, and to assess their potential clonal relationships to CTCL.

Methods: We reviewed results from 1677 samples from 235 patients with CTCL (MF n=163, SS n=67, Other n= 23) sequenced using the clonoSEQ® platform (median 6 samples per patient) from 2012-2023. Dominant TCR clonotypes were defined per clonoSEQ® criteria, and only productive rearrangements were included. Clonotypes with identical VDJ rearrangements (e.g., single-nucleotide variants) were excluded.

Among patients with an emerging second dominant productive TCRβ clonotype, 11 had cryopreserved PBMCs available for analysis. These PBMCs, along with those from 10 age-matched healthy donors, underwent TCR-HTS using the LymphoTrack® assay. Multiparametric spectral flow cytometry (MFC), including antibodies to TCRBC1 and/or TRBV regions, enabled identification and characterization of clonal T-cell populations. Clonal and polyclonal T-cell populations were sorted by FACS and underwent whole-exome sequencing (WES) and hybrid-capture TCR sequencing. Sorted monocytes were used as germline for WES variant calling. To provide a reference for the mutational profile of benign T-cells, normal T-cells from one patient were singly sorting, ex-vivo expanded into colonies, and underwent WES.

Results: Among 235 CTCL patients, 109 (46%) had two or more unique dominant, productive TCRβ clonotypes identified in one or more specimens, and 41 (17%) had three or more. TCR-HTS using an alternative platform confirmed a high-frequency clonotype (>3%) in all 11 available CTCL PBMC samples, and in none of the 10 healthy donors tested. MFC identified one or more clonal T-cell populations per patient. These variably expressed CD4, CD8, or CD4-CD8- (double-negative), and most commonly displayed a terminal effector phenotype (CD45RA+, CD56+, and CD57+). These immunophenotypes were most consistent with T-cell clones of uncertain significance (T-CUS) and were distinct from typical CTCL profiles.

FACS-sorted T-CUS populations underwent TCR sequencing to confirm their clonal identity. In 10/11 patients, the dominant TCRβ clonotype matched that of TCR-HTS (median 91% TCR clonotype fraction, range 41-100%). WES was performed on 10 sorted T-CUS populations, 9 patient-matched CTCL biopsies, 8 polyclonal T-cell populations, and 15 single T-cell derived colonies. T-CUS harbored significantly fewer somatic variants than CTCL biopsies (mean 68 versus 781, p=0.004), but more than polyclonal T cells (mean 17, p=0.01). Single-cell-derived colonies of normal T-cells from a CTCL patient had a mutational burden (mean 41) comparable to T-CUS (p=0.24).

To evaluate whether T-CUS populations in CTCL patients shared a common lineage with the primary CTCL clone, we compared somatic variants in matched T-CUS and CTCL samples. No shared variants were found, indicating no common clonal origin. However, T-CUS harbored nonsynonymous variants involving CTCL-associated and/or cancer-associated genes including ARID1A, ATRX, FSIP2, GATA3, PTPRR, PTPRK, STAT1, TCF3, TP63, and TRRAP. STAT3 and STAT5B variants were not detected in the clonal sorted populations, and there were no unexplained cytopenias, arguing against T-cell large granular lymphocytic leukemia. Unlike CTCL, T-CUS lacked UV-associated mutational signatures.

Conclusions: Emergence of new dominant TCR clonotypes is common in CTCL and often arise within T-CUS populations that fluctuate in frequency in the blood and skin. While T-CUS harbor variants in genes associated with T-cell malignancies, there was no evidence of evolution of these populations into secondary malignancies, and T-CUS appear clonally unrelated to CTCL. The clinical significance of emerging dominant T-cell clonotypes in CTCL should be interpreted in the context of a high prevalence of T-CUS.