Long-term survival after HLA-haploidentical SCT from noninherited maternal antigen-mismatched family donors: impact of chronic GVHD

Allo-SCT from an HLA-haploidentical family donor has been the treatment of choice in patients with high-risk hematologic malignancies, who are expected to have a better prognosis with SCT but lack immediate access to a conventional stem cell source.1, 2 With intent to minimize the risk of severe GVHD, most haploidentical SCT protocols employ ex vivo or in vivo T-cell depletion, albeit at the expense of an increased risk of infection or relapse as a result of poor post-transplant immune reconstitution. To develop an alternative strategy to perform haploidentical SCT that confers improved immune recovery and acceptable risk of GVHD, we have explored the feasibility of T-cell-replete SCT from family donors mismatched for noninherited maternal HLA antigens (NIMA); NIMA-mismatched donor selection is based on the hypothesis that the detection of long-term maternal or fetal microchimerism in the donor's peripheral circulation is associated with immunological hyporesponsiveness against NIMA (in the case of NIMA mismatch in the graft-vs-host direction) or against inherited paternal HLA antigens (IPA) (in the case of NIMA mismatch in the host-vs-graft direction).3 According to this scenario, we and other groups showed that T-cell-replete HLA-haploidentical SCT from a NIMA-mismatched family donor is feasible in selected patients with poor-risk hematologic malignancies.3, 4 However, late complications and long-term outcomes in patients undergoing such transplantations have been so far largely unknown. Therefore, we retrospectively studied the severity of chronic GVHD, requirement for immunosupressive treatment, and status of primary disease in long-term survivors who received T-cell-replete NIMA-mismatched haploidentical SCT. We collected data on 16 consecutive patients who had survived more than 3 years after NIMA-mismatched SCT performed between January 2001 and July 2004 at 11 institutions that participated in our previous nationwide study (Table 1).3 At the time of SCT, they had a median age of 19 years (range, 2–56) and received BM (n=5) or G-CSF-mobilized peripheral blood (n=11) as treatment for acute myeloid leukemia (n=6), acute lymphoblastic leukemia (n=3), chronic myeloid leukemia (n=4) and other B-cell neoplasms (n=3); 6 patients had a chemosensitive disease and 10 had a refractory disease. Early outcomes of 14 of these transplantations have been described elsewhere.3, 4, 5 All patients received tacrolimus-based GVHD prophylaxis after myeloablative (n=10) or reduced-intensity conditioning (n=6). The type of donor was NIMA-mismatched sibling in 9 cases, mother in 6 and daughter in 1; all patient-donor pairs had two or three serologic mismatches at HLA-A, HLA-B and HLA-DR antigens in the graft-versus-host direction. The presence of long-term maternal or fetal microchimerism was detected in all donors through NIMA- or IPA-specific nested PCR as described earlier.3 Karnofsky score was employed to record the performance status for patients who were 16 years or older, whereas Lansky score was applied for those who were younger than 16 years of age. Organ-specific symptoms related to chronic GVHD and their severity were diagnosed and evaluated by consensus criteria proposed by the National Institutes of Health Chronic GVHD Diagnosis and Staging Working Group.6 At a median follow-up of 56 months (range, 38–74), 13 (81%) of 16 patients were alive and free of their primary disease. One patient was alive with relapsed disease, and two died from pulmonary complications at 51 and 52 months after transplantation in continuous remission. Fifteen (94%) patients developed classical chronic GVHD; type of onset was de novo in five, quiescent in six and progressive in four. According to the National Institutes of Health Clinical Scoring Systems the affected organs scored two or greater, which included lungs (n=5), skin (n=4), eyes (n=3), liver (n=3) and joints/fascia (n=2) (Table 2). Overall severity of chronic GVHD among these patients was classified as mild in three (20%) cases, moderate in seven (47%) and severe in five (33%). It is noted that immunosuppressive agents were successfully withdrawn from eight (50%) patients with de novo or quiescent onset of disease at a median of 19 months (range, 3–46) after transplantation, although two of four patients who experienced severe chronic GVHD (score 3) of lung eventually succumbed to bronchiolitis obliterans. Karnofsky or Lansky performance score at the time of last follow-up among the 14 surviving patients was 100% in 6 (43%), 80–90% in 5 (36%), 70% in 2 (14%) and less than 70% in 1 (7%). In the present study, we found that substantial proportion of long-term survivors after NIMA-mismatched haploidentical SCT could discontinue administration of immunosuppressive agents despite the frequent occurrence of moderate-to-severe chronic GVHD. This paradoxical observation contrasts sharply with the conventional assumption that the establishment of robust tolerance across multiple HLA disparities is hardly possible in the setting of marrow or peripheral blood SCTs without employing T-cell depletion. However, clinical significance of NIMA-mismatched donor selection has been controversial because the mechanisms underlying the tolerogenic effect against NIMA or IPA have not been fully elucidated.7 Using murine BM transplant models, Matsuoka et al.8 showed that transplant from donors exposed to NIMA in utero, but not from those exposed to IPA, reduced the morbidity and mortality associated with GVHD in an antigen-specific and a CD4+CD25+ T cells-dependent manner. In contrast, Opiela et al.9 showed that murine neonates exposed to low levels of NIMA can develop vigorous in vivo cytotoxic rather than tolerogenic responses against NIMA, which might explain the inconsistent severity of GVHD after NIMA-mismatched SCTs in humans. Intriguingly, Stern et al.10 recently reported that recipients of T cell-depleted haploidentical SCT using mother as the donor had the better overall survival than those using father as the donor, implying a mechanism by which previous exposure to paternally derived antigens in maternal donors would positively affect transplant outcomes. In conclusion, our observations in this study suggested that long-term survival without continuous immunosuppressive treatment is possible after T-cell-replete HLA-haploidentical SCT from a microchimeric NIMA-mismatched donor. Although the small number of patients limits the interpretation of our results, further studies are warranted to compare late sequelae after HLA-haploidentical SCTs with various protocols in larger cohorts.