Epidemiology of Second Allogeneic Hematopoietic Stem Cell Transplantation for Patients with Relapsed Myelodysplastic Syndrome

Introduction

Myelodysplastic syndrome (MDS) represent a group of clonal myeloid stem cell disorder with a heterogeneous spectrum of presentation. Although allogeneic hematopoietic stem cell transplantation (HSCT) was the well-studied and only curative treatment option for patients with MDS, relapse remains a major drawback like other hematological malignancies because it is associated with poor survival in general. The efficacy of second allogeneic HSCT was reported for patients with various hematological diseases. However, even fundamental epidemiological data of second allogenic HSCT after relapsed MDS is still lacking. The aim of this study was to investigate the epidemiology in patients who received second allogenic HSCT after relapsed MDS, and then to evaluate prognostic factor of second allogeneic HSCT by comparing and identifying the difference between first and second allogeneic HSCT in patients with MDS.

Methods

The present study included 3340 adult patients with MDS who received first allogeneic HSCT from January 1990 to December 2015 (Total population) from the Japanese Data Center for Hematopoietic Cell Transplantation. Among total population, 627 patients experienced recurrence of MDS and 101 patients underwent second allogenic HSCT for the treatment of relapsed MDS after allogenic HSCT (Recurrent population). We defined a relapse more than 18 months after allogeneic HSCT as a "late relapse" and the others as an "early relapse." Active disease was defined when the blast in bone marrow was greater than 5% and or presence of blast in peripheral blood. Performance Status (PS) was defined according to Eastern Cooperative Oncology Group and categorized into two group (i.e., good PS: 0-1 and poor PS: 2-4).

We set the primary endpoint as the 5-year overall survival (5yOS) which was evaluated by using log-rank test. We defined 5yOS as the time from first HSCT in total population and second HSCT in recurrent population to any cause of death. Other endpoints were evaluated by using Gray's method. To evaluate the risk factors for the primary endpoint, we employed multivariable Cox proportional hazards model. Adjusted covariates were selected clinically based on previous reports, and those for recurrent and total population were shown in Table.

Results

We analyzed 101 patients in recurrent population and 3340 patients in total population. In recurrent population, median age was 53 years old (interquartile: 45-59) and 58 patients (57.4%) were male. Eighty patients (82.5%) relapsed within 18 month after first allogeneic HSCT. The median duration between first and second HSCT was 281 days (interquartile: 145-610). Before second allogeneic HSCT, 71 (71.0%) patients had active disease and 31 patients (32.0%) had poor PS. About stem cell source, 35 (34.7%) received cord blood transplantation.

Regarding primary endpoint, 5yOS after allogeneic HSCT was 24.7% (15.8-34.7%) in recurrent population (Figure 1A) and 49.9% (95%CI: 48.0-51.7%) in total population (Figure 1B). Regarding secondary endpoints, 5-year cumulative incidence of relapse was 45.2% (35.0-54.9%) and 22.6% (21.1-24.1%), 5-year cumulative incidence of non-relapse mortality was 33.4 (23.2-43.4%) and 29.5 (27.9-31.2%), and 5-year cumulative incidence of tumor-related mortality was 41.5% (31.1-51.5%) and 20.0% (18.6-21.5%) in recurrent population and total population, respectively.

For analysis of prognostic factors influencing 5yOS in recurrent population, early relapse (HR: 2.82, 95%CI: 1.12-7.13, p=0.028) and poor PS (HR: 3.27, 95%CI: 1.79-5.97, p=0.001) were identified as significant independent predictors for 5yOS (Table). Ad-hoc subgroup analysis considering early relapse and poor PS demonstrated that the 5yOS were significantly worse in patients who had early relapse and poor PS as compared with other groups with median survival of only 57 days (p<0.001) (Figure 2). On the other hand, higher age, male gender, active disease, poor karyotype, poor PS and cord blood transplantation were associated with worse 5yOS in total population (Table).

Conclusion

In the present study, we revealed epidemiology of second allogeneic HSCT for relapsed MDS and prognostic factor that could aid the identification of patients who may benefit from second allogeneic HSCT. Patients who relapsed earlier than 18 months and had poor PS were recommended not to receive second allogeneic HSCT.

View largeDownload slide

View largeDownload slide

Close modal