CD34-selected versus unmanipulated autologous haematopoietic stem cell transplantation in the treatment of severe systemic sclerosis: a post hoc analysis of a phase I/II clinical trial conducted in Japan

Background The effectiveness of autologous haematopoietic stem cell transplantation (auto-HSCT) in treating severe systemic sclerosis (SSc) is established; however, the necessity of purified CD34+ cell grafts and the appropriate conditioning regimen remain unclear. This study aimed to compare the efficacy and safety of CD34-selected auto-HSCT with unmanipulated auto-HSCT to treat severe SSc. Methods This study was a post hoc analysis of a phase I/II clinical trial conducted in Japan. Nineteen patients with severe SSc were enrolled. Peripheral blood stem cells (PBSCs) were mobilised with cyclophosphamide (4 g/m 2 ) and filgrastim (10 μg/kg/day). Following PBSC collection by apheresis, CD34+ cells were immunologically selected in 11 patients. All patients were treated with high-dose cyclophosphamide (200 mg/kg) monotherapy as a conditioning regimen and received CD34-selected ( n = 11) or unmanipulated auto-HSCT ( n = 8). Changes in skin sclerosis and pulmonary function were assessed over an 8-year follow-up period. Differences in the changes, toxicity, progression-free survival (PFS) and overall survival were compared between patients who had received CD34-selected auto-HSCT and those who had received unmanipulated auto-HSCT. Results Skin sclerosis progressively improved after transplantation over an 8-year follow-up period in both groups, and the improvement was significantly greater in the CD34-selected group than in the unmanipulated group. Forced vital capacity in the CD34-selected group continuously increased over 8 years, whereas in the unmanipulated group it returned to baseline 3 years after transplantation. Toxicity and viral infections, such as cytomegalovirus infection and herpes zoster, were more frequently found in the CD34-selected group than in the unmanipulated group. The frequency of severe adverse events, such as bacterial infections or organ toxicity, was similar between the two groups. No treatment-related deaths occurred in either treatment group. PFS of the CD34-selected group was greater than that of the unmanipulated group, and the 5-year PFS rates of the CD34-selected and unmanipulated group were 81.8% and 50% respectively. Conclusions CD34-selected auto-HSCT may produce favourable effects on improvement of skin sclerosis and pulmonary function compared with unmanipulated auto-HSCT. Use of CD34-selected auto-HSCT with high-dose cyclophosphamide monotherapy as a conditioning regimen may offer an excellent benefit-to-risk balance. Electronic supplementary material The online version of this article (10.1186/s13075-019-1823-0) contains supplementary material, which is available to authorized users.