Molecular evidence for repertoire skewing of T large granular lymphocyte proliferation after allogeneic hematopoietic SCT: report of two cases

Large granular lymphocyte (LGL) expansion in patients after allo-HCT (poly- or monoclonal) can be either asymptomatic1 or accompanied by cytopenias and autoimmune manifestations.2 Strong association with viral infections in many cases implicates pathogenetically a chronic antigenic stimulation. We here present the results from a comprehensive evaluation, including detailed immunogenetic analysis, of two cases with benign oligoclonal expansion of CD3+ LGLs causing transient myelosuppression in the context of successive episodes of herpesvirus reactivation after allo-HSCT. Interestingly, both cases developed a highly restricted T-cell receptor β-chain (TRB) gene repertoire with common gene usage in the predominant clonotypes, strongly implicating Ag involvement in their development. A 27-year-old woman with AML in relapse underwent allo-HCT from a haploidentical donor in September 2004, receiving an anti-thymocyte globulin (ATG)-containing regimen. After successful engraftment, she attained complete donor chimerism (CDC) remaining in CDC ever since. CMV reactivation on day +45 prompted appropriate treatment with gancyclovir for 30 days. Fifteen days after documentation of the CMV infection, grade III (World Health Organization toxicity scale) neutropenia was noted (WBC: 1.6 × 109/L, ANC: 0.4 × 109/L, Hb: 11 g/L, plts: 70 × 109/L), persisting for 5 months despite CMV clearance. Hematopoiesis eventually recovered until September 2005, when EBV reactivation led to preemptive immunotherapy with Rituximab. Four days later, mild neutropenia developed (WBC: 5.3 × 109/L; ANC: 1 × 109/L). Further virologic and serologic screening was negative. BM aspiration revealed a hypocellular BM. PB lymphocyte subpopulation analysis with flow cytometry revealed: (i) inverted ratio of CD4/CD8 lymphocytes, (ii) absolute increase (>1.0 × 109/L) of CD3+/CD8+ lymphocytes and (iii) increased (>30%) CD3+/CD8+/CD57+ lymphocytes (Figure 1). Results from the immunogenetic analysis of the TRB repertoire are discussed below. After remaining uncomplicated for 7 months, neutropenia eventually resolved; however, PB lymphocyte imbalances persisted for 12 months after the recovery of hematopoiesis. Currently, the patient is alive in CR, without evidence of lymphocyte imbalances. A 14-year-old boy with bcr-abl (+) pre-B ALL in molecular remission underwent allo-HCT in October 2003, receiving a graft lymphodepleted in vivo with ATG from a matched unrelateddonor. CMV gastritis on day +30 prompted treatment with ganciclovir for 14 days. Grade III neutropenia was noted, eventually recovering in January 2004. Five months after the transplantation, he experienced EBV reactivation and was treated with one course of Rituximab (375 mg/m2). Fifteen days later, grade IV neutropenia was identified (WBC: 2.66 × 109/L, ANC: 0.27 × 109/L). As in case 1, no underlying pathology was identified. Examination of the BM aspirate revealed a hypercellular BM with hyperplastic myeloid series and a shift to the left. PB lymphocyte subpopulation analysis with flow cytometry revealed similar findings to those of case 1 (Figure 1). Immunogenetic analysis of the TRB gene repertoire is discussed below. Neutropenia persisted for 11 months. CD3+/CD8+/CD57+ lymphocyte expansion persisted for 3 years and eventually subsided in 2008. The patient is in continuous molecular remission and his blood counts have normalized. Immunogenetic analysis of the TRB gene repertoire was conducted on samples from two different timepoints (i) after CMV reactivation (+3 months for both patients) and (ii) after EBV reactivation (+7 and +2 months, for patients 1 and 2, respectively), revealing similar results for both patients (Figure 1). PCR amplification, subcloning of the PCR amplicons and sequencing analysis were performed as described previously. TRBV-TRBD-TRBJ gene rearrangement sequence data were interpreted using the IMGT database and tools. Patient 1: Overall, 41 productive TRBV-TRBD-TRBJ gene rearrangements were evaluated. Cluster analysis of the rearrangement sequences identified six clusters of identical rearrangements with three dominant clonotypes: (i) TRBV10-2/TRBD1/TRBJ2-7, n=8 (19.5%), (ii) TRBV27/TRBD2/TRBJ1-5, n=4 (9.7%) and (iii) TRBV6-5/TRBD1/TRBJ1-5, n=15 (36.6%). Different clonotypes were identified at different timepoints. Patient 2: Overall, 54 productive TRBV-TRBD-TRBJ gene rearrangements were evaluated. Three clusters of identical sequences were identified corresponding to the following dominant clonotypes: (i) TRBV2/TRBD2/TRBJ1-5, n=15 (29.4%), (ii) TRBV27/TRBD1/TRBJ1-6, n=15 (29.4%) and (iii) TRBV6-5/TRBD1/TRBJ2-7, n=14 (27.4%). Clonotypes (i) and (ii) were detected in both timepoints. T-LGL expansions, transient or persistent, have been reported in the setting of several clinical conditions, such as viral infections, autoimmune diseases, and after allo-HCT, solid organ transplantation and drug administration.3 In our patients, three of the most common underlying factors could be identified (allo-HCT, viral reactivation, Rituximab administration), thus implying a multifactorial origin for the observed T-LGL lymphoproliferation. The possibility that ATG could be implicated in the development of the observed lymphocyte imbalances cannot be ruled out; however, it is highly unlikely given the long intervening period between ATG administration and the detection of T-LGL expansion. Among patients undergoing allo-HCT, T-LGL lymphoproliferation can be either asymptomatic1, 4 or accompanied by cytopenias and autoimmune manifestations.2 Here, both patients experienced unexplained neutropenia persisting for several months; immunology and virology were negative, while the rapid onset of neutropenia after the first administration of Rituximab precluded the diagnosis of the recently recognized syndrome of Rituximab-associated late-onset neutropenia. T-LGL expansions in allo-HCT patients can be monoclonal,1, 4 however, with this monoclonality not translating to an aggressive (that is, malignant) clinical course. In the present study, immunogenetic analysis revealed oligoclonality in both cases with remarkable skewing of the TRB gene repertoire. Of note, both cases had experienced recent viral infection, prompting speculation about (super)Ag involvement in the selection and expansion of cytotoxic T cells. In healthy individuals, (super)antigenic drive may elicit the expansion of a specific cytotoxic clonotype among a polyclonal background, but usually the effect is short lived.5 To gain pathologic significance, immunodominance has to be sustained by an additional factor. In our case, the immunodominant clonotypes accounted for >30% of the analyzed sequences, suggesting additional factors that contributed towards perpetuating the initial response. This scenario is reminiscent of T-LGL leukemia, where the emergence of the clonal cytotoxic T cell expansion is thought to be a consequence of a nonrandom, Ag-driven process.6, 7 Along these lines, the finding of repertoire restrictions and similarities (that is, high frequency of the TRBV6-5 gene) in both cases reported here supports a dynamic process of cytotoxic T cell responses against auto- and exo-Ags as the initial event in T-LGL lymphoproliferation. The authors declare no conflict of interest.