T-Prolymphocytic Leukemia Is Sensitive to Polyclonal T Cell-Derived Graft-Versus-Leukemia Effects: Evidence from Minimal Residual Disease Kinetics and TCR Repertoire Diversity Analyses

INTRODUCTION: T-prolymphocytic leukemia (T-PLL) is a rare T cell malignancy with an aggressive clinical course. Preliminary clinical data suggest that allogeneic stem cell transplantation (alloSCT) may provide long-term disease control in a proportion of patients. However, direct evidence that graft-versus-leukemia (GVL) activity is indeed effective in T-PLL is lacking. We sought to investigate GVL in T-PLL by correlating minimal residual disease (MRD) kinetics with immune modulatory interventions (immunosuppression tapering, donor lymphocyte infusions (DLI), chronic graft-versus-host disease (cGvHD)), and T cell receptor (TCR) repertoire diversity alterations after alloSCT.

METHODS: The study sample consisted of 10 consecutive patients who received alloSCT for T-PLL at the University of Heidelberg between 2007 and 2015. Quantitative MRD monitoring was performed using clone-specific real-time quantitative PCR (RQ-PCR) of clonal TCR beta (TRB) and/or gamma (TRG) gene rearrangements. Data interpretation followed EuroMRD guidelines. In selected patients, TCR repertoire diversity was analyzed longitudinally by next-generation sequencing (NGS). TRBV-TRBD-TRBJ gene rearrangements were amplified according to BIOMED2 protocol on genomic DNA. PCR products were sequenced on Illumina's MiSeq platform. NGS data were analyzed through a purpose-built bioinformatics immunoprofiler (EuroClonality-NGS consortium). Rearrangements with very similar junctional amino acid sequences and identical TRBV and TRBJ gene usage were defined as clonotypes.

RESULTS: Patients underwent alloSCT in remission after first-line (8) or salvage (2) alemtuzumab-based therapy. 5 patients were allografted with an unrelated donor, 4 with a related donor, and one received haploidentical alloSCT. Conditioning was fludarabine with cyclophosphamide and/or total body irradiation-based. All patients had a cytological complete response (CR) after alloSCT. 2 patients died early because of acute GvHD, and one had no MRD marker, leaving 7 patients for MRD monitoring. Of these, 3 were MRD- at alloSCT, whereas 5 patients remained or became MRD+ early after alloSCT. In all of these 5 patients, immunosuppression tapering (3) or DLI (2) resulted in significant reduction of MRD levels (range 1-3 log) and was accompanied by cGvHD in 3 patients. However, durable MRD- was obtained in only 2 patients (alive 81+ and 12+ months post transplant), whilst MRD re-increased in 3 patients after 5-28 months despite ongoing cGvHD in one of them. NGS of the TCR repertoire was performed longitudinally in 3 patients with the longest follow-up. A total of 104 samples (blood (n=91) and BM (n=10) plus 3 donor blood samples) were sequenced. The sample at diagnosis showed one or two major clonotypes in all 3 sequenced T-PLLs, predominantly reflecting a mono- or biallelic leukemic TRB gene rearrangement. Kinetics of this leukemic clonotype followed kinetics of RQ-PCR MRD measurement, demonstrating that NGS can be used to quantify MRD in T cell lymphoma. Immediately after transplantation, the TRB repertoire was heavily skewed in all 3 patients, but recovered over time. Also in all 3 patients, MRD responses were reproducibly associated with a shift from a clonal, T-PLL-driven profile to a polyclonal signature. This corresponded to the donor clonotype repertoire and disappeared with increasing MRD levels. In each of the samples after alloSCT, several expanded non-leukemicclonotypes were observed at a maximum frequency from 3% to 54% of total TRB sequences. However, during MRD response, novel dominant clonotypes that could explain a clonal GVL effect did not emerge. Figure 1 shows an example of a patient repeatedly showing MRD response to immune interventions. Currently, 5 patients are alive and in cytological CR (4-81 months after alloSCT), translating into a median relapse-free survival of 40 months.

CONCLUSIONS: The MRD responses to immune interventions observed here provide the first direct evidence for the efficacy of GVL in T-PLL. However, the GVL effect in T-PLL appears to be often only limited or transient. It also does not seem to be caused by the emergence of novel dominant T cell clones but is rather relying on a poly-/oligoclonal T cell response.