Abstract 1945

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) from female donor to male recipient (F to M) has been reported to be a risk factor of poor prognosis. Although immune responses to male specific minor histocompatibility antigens (H-Y) are suggested to be involved, this hypothesis has not yet been well validated. To clarify the significance of exposure to H-Y, we retrospectively surveyed the medical records of 292 patients with hematological malignancies who underwent allo-HSCT from a related donor at National Cancer Center Hospital from 2001 to 2010. Data about children's sex and pregnancies of female donors were collected by chart review. The median follow-up of survivors was 69 months (range; 3–128) after HSCT. The median age of donors and recipients were 47 years (range; 12–69 years) and 51 years (range; 18–69 years), respectively. The distribution of sex combination of donors and recipients were as follows: 94 male to male, 57 female to female, 70 female to male and 71 male to female. About a half of female donors had pregnancy history (79 of 127; 62.2%, including abortion history). According to F to M sex-mismatch pattern and children's sex, patients were divided into three groups: 31 female donor with male child to male recipient (F to M with MC), 39 female donor without male child to male recipient (F to M without MC) and 222 non-F to M sex combination (non-F to M). The F to M sex-mismatch pattern did not significantly affect OS regardless of children's sex (5-year OS, F to M with MC; 52%, F to M without MC 40%; non-FtoM; 43%). In multivariate analysis, only disease status before HSCT significantly affected OS (hazard ratio [HR]; 1.64, 95% confidential interval [CI]; 1.17 to 2.28, P<0.01). However, non-relapse mortality (NRM) was significantly higher in F to M with MC than in the other groups (5-year NRM, F to M with MC; 34%, F to M without MC 24%; non-F to M; 20%, P=0.02), which was also confirmed in multivariate analysis (HR 2.00, 95% CI; 1.11 to 3.58, P=0.02). Furthermore, relapse incidence (RI) tended to be lower in F to M with MC than in the other groups (5-year RI, F to M with MC; 23%, F to M without MC 41%; non-F to M; 43%, P=0.12), which was of borderline significance in multivariate analysis (HR 0.48, 95% CI; 0.21 to 1.09, P=0.08). Because the cumulative incidences of acute and chronic GVHD in F to M with MC were higher than those of F to M without MC (26% v.s. 13%; P=0.12, 52% v.s. 37%; P=0.26, respectively), immune response to H-Y were suggested to be involved in the transplant outcome. In summary, F to M with MC had some special features, high NRM and low RI. These effects were counteracted by each other, resulting in similar OS to the other groups. These data suggest the possibility that poor outcome of allo-HSCT from F to M results from the immune response to H-Y and that the exposure of H-Y before donation is critical for the response. Further investigation in a large number of patients is warranted.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.