Graft-Versus-Host Disease Targets Granulosa Cell of Ovarian Follicle and Causes Infertility after Allogeneic Hematopoietic Stem Cell Transplantation

[Introduction] Infertility associated with ovarian failure is a serious late complication for female survivors of allogeneic hematopoietic stem cell transplantation (allo-SCT). While the role of pretransplant conditioning regimen has been well appreciated, it remains to be elucidated whether GVHD could impact female fertility. We have addressed this issue in murine models of allogeneic SCT.

[Methods] Female B6D2F1 (H-2^b/d) mice, remained unconditioned or conditioned with 12 mg/kg busulfan and 120 mg/kg cyclophosphamide on day -7, were injected with 80 x 10⁶ splenocytes from allogeneic B6 (H-2^b) or syngeneic B6D2F1 donors on day 0. In the indicated experiments, 10 mg/kg prednisolone (PSL) was administered daily from day 0 to 14 post SCT and every 3 days thereafter. The oocytes were enumerated after stimulation with pregnant mare's serum gonadotropin (PMSG) and human chorionic gonadotropin (hCG). The serum levels of anti-Müllerian hormone (AMH) were measured to evaluate functions of granulosa cells. Fertility was directly evaluated by mating recipient mice with naive B6D2F1 males 5-6 times between day 14 and day 136.

[Results] After unconditioned SCT, histological evaluation of the ovaries from allogeneic recipients demonstrated donor T-cell infiltration in the granulosa layers of follicles and apoptosis of granulosa cells surrounded by lymphocytes, whereas there was neither of these changes in syngeneic animals. Serum levels of AMH, which is produced by granulosa cells and regulates ovarian follicle development, were significantly decreased in allogeneic animals compared to syngeneic controls (Fig. 1). Upon superovulation, the numbers of oocytes recovered from allogeneic animals were significantly less than those from syngeneic animals, indicating an impairment of ovarian functions (Fig. 2). Remarkably, allogeneic animals delivered significantly less numbers of newborns by day 136 after SCT than syngeneic animals (Fig. 3). Next, we addressed whether GVHD impacts on the ovary functions in non-myeloablative SCT conditioned with busulfan and cyclophosphamide. Again, there were histological evidences of donor T cell infiltration and granulosa cell apoptosis, reduced oocytes, and impaired fertility in allogeneic mice when compared to syngeneic mice. Administration of PSL significantly reduced T cell infiltration in the granulosa cell layers and restored the serum levels of AMH (Fig. 1), numbers of oocytes (Fig. 2), and fertility (Fig. 3) in allogeneic mice.

[Conclusion] Our results demonstrate for the first time that GVHD induces infertility by targeting the granulosa cells in the ovaries and GVHD prevention could preserve ovarian functions. These results have important clinical implications in young female transplant recipients with nonmalignant diseases, where minimally toxic chemotherapies prior to SCT and conditioning regimen were used.