HLA Class I Allele-Deficient Clones Regenerate in Immune Aplastic Anemia without Anti-Thymocyte Globulin

Background

The loss of specific HLA class I alleles on hematopoietic stem and progenitor cells (HSPCs) in aplastic anemia (AA) can facilitate immune escape from cytotoxic T lymphocyte (CTL)-mediated pathogenesis. However, HLA class I allele-deficient HSPCs are unable to sustain adequate hematopoiesis, necessitating treatments such as antithymocyte globulin (ATG)-based immunosuppressive therapy (IST) or hematopoietic cell transplantation (HCT). The outcomes of AA patients with HLA loss who do not receive ATG-based IST remain unclear.

Methods

We investigated the clinical and clonal outcomes of AA patients with HLA allelic loss who did not receive standard therapies. A complete response (CR) was defined as achieving all three of the following criteria: neutrophil counts ≥1000/µL, hemoglobin levels ≥10 g/dL, and platelet counts ≥100,000/µL. A partial response (PR) required at least two of the following criteria: neutrophil counts ≥500/µL, reticulocyte counts ≥40,000/µL, and platelet counts ≥20,000/µL, along with transfusion independence. HLA allele loss was assessed by flow cytometry using HLA allele-specific monoclonal antibodies. Targeted sequencing of HLA genes and myeloid neoplasia-related genes was performed on sorted HLA allele-deficient and -expressing cell populations.

Results

We identified 49 AA patients with HLA class I allele-deficient cells who were not treated with ATG-based IST or HCT. The median age at diagnosis was 65 years (range, 18-84 years); 24 had non-severe AA and 25 had severe AA. Of these, 39 (80%) were treated with cyclosporine, with or without thrombopoietin receptor agonists. Over a median follow-up of 5.7 years, most patients showed slow but significant hematologic recovery, regardless of cyclosporine use. Among the 25 severe AA patients, 21 (84%) achieved PR and 16 (64%) achieved CR; 19 (79%) of the 24 non-severe AA patients also achieved CR. The median time from diagnosis to CR was 1.9 years. Five patients died, including two elderly patients who declined any therapy and three who died from causes unrelated to cytopenia (two from cancer and one from renal failure). No patient showed clonal evolution to myelodysplastic syndrome (MDS), although progression to paroxysmal nocturnal hemoglobinuria (PNH) was observed in two patients. After hematologic recovery, the HLA allele-deficient cell population generally dominated myelopoiesis. Of 9 patients who were tested repeatedly, the percentage of HLA allele-deficient cells increased during recovery in 7 patients, while the other 2 patients exhibited an expansion of PNH clones. Targeted sequencing performed in 7 patients after achieving CR confirmed that almost all their hematopoiesis was supported by HSPC clones with HLA gene mutations and/or chromosome 6p loss of heterozygosity. Of these 7 patients, myeloid neoplasia-associated mutations were detected in 3 elderly patients (43%), aged 55, 74, and 76 years. The first patient had DNMT3A p.R882H (variant allele frequency, 15%); the second had DNMT3A p.F751V (46%) and TET2p.N1625fs (39%); and the third had U2AF1 p.Q157R (55%), ASXL1 p.G645fs (37%) and SETBP1 p.D868N (34%). These mutations were absent in the other four patients, all of whom were under the age of 55.

Discussion

Our study has revealed that a significant proportion of immune AA patients with HLA class I allele loss can achieve CR without undergoing ATG-based therapy or HCT. This was associated with an increase in HLA class I allele-deficient cells and a decrease in HLA-intact cells, likely due to ongoing myelosuppression by CTLs which do not recognize the HLA allele-deficient HSPCs, therefore allowing them to proliferate. This recovery pattern differs from ATG-based IST, which relies on the re-proliferation of HLA-intact HSPCs. The elimination of HLA-intact HSPCs by CTLs in patients who did not receive ATG therapy may explain the observed low incidence of clonal evolution to MDS and PNH, as these conditions typically arise from HLA-intact cells. Avoiding ATG and allowing hematologic regeneration by escape clones might be a feasible treatment strategy for immune AA patients with HLA loss, especially for those who cannot tolerate standard therapies. However, further research is needed to validate the prevalence, predictors, and long-term outcomes of this approach.