Increased Mortality in Patients with Autoimmune Hemolytic Anemia after Allogeneic Hematopoietic Stem Cell Transplantation

Introduction

Autoimmune hemolytic anemia (AIHA) is the most commonly reported autoimmune complication following hematopoietic stem cell transplantation (HSCT) with incidence of 2-6%. The risk factors and pathogenesis remain poorly understood. Treatment often requires multiple therapeutic agents with variable efficacy and outcome. However no study to date has shown whether AIHA directly results in increased mortality. In order to better understand the risk factors, mortality and management of post-HSCT AIHA, we carried out a retrospective analysis of 533 allogeneic HSCTs in adult patients performed at King's College Hospital between 2005-2011.

Method

The median follow-up period after HSCT was 31 months (range 2.9 – 100 months). The primary endpoint was the onset of AIHA, defined by positive direct agglutinin test (DAT) arising after the HSCT, with biochemical markers of hemolysis (raised serum lactate dehydrogenase (LDH), reduced haptoglobin, or spherocytes on the blood film). Hemolysis was considered significant if the drop in hemoglobin was more than 20 g/l. Cases of DAT positivity due to ABO antibodies, as well as those with history of AIHA or positive DAT prior to HSCT were excluded. Potential risk factors for development of AIHA were calculated with univariate followed by multivariate analysis comparisons of incidence of AIHA for each clinical stratum. Kaplan-Meier method was used to compare mortality caused by AIHA against overall mortality (OM), and transplant-related mortality (TRM). AIHA was modelled as a time-dependent variable when estimating mortality. All patients alive at last follow-up who did not develop AIHA were censored.

Results

We identified 19 cases of AIHA following HSCT (overall incidence 3.6%). The median time to onset from HSCT to AIHA was 202 days. AIHA was associated with HSCT from unrelated donors (p = 0.026; Hazard Ratio = 5.28, 95% CI = 1.22 – 22.9), and concordant sex between HSCT recipients and donors (p = 0.045; Hazard Ratio = 3.52; 95% CI = 1.03 – 12.1). No significant association was observed between AIHA and the following eight factors: recipient gender; primary hematological disease; source of hematopoietic stem cells; conditioning regimen (alemtuzumab/ATG versus non-alemtuzumab/ATG, and reduced intensity versus myeloablative); HLA mismatch between donor/recipient; ABO mismatch; recipient CMV status; and concurrent chronic GvHD. AIHA patients also exhibited high frequency of simultaneous alloimmunization to Rh antigens coinciding with the onset of AIHA, which was not due to difference in transfusion rates, and not observed in the population who tested negative for DAT.

Majority of AIHA patients (14/19; 72%) required multiple agents for treatment, but only 9/19 (47%) cases achieved complete resolution of AIHA (1 with intravenous immunoglobulin; 2 with prednisolone alone; 6 with rituximab in combination with other agents). The median survival from onset of AIHA was 487 days (range 26 – 1977 days). We compared the mortality of the AIHA versus non-AIHA population that survived beyond the median time of onset for AIHA (202 days). Patients with post-transplant AIHA had a higher OM (p = 0.006, Hazard Ratio = 2.37, 95% CI = 1.28 – 4.39), 1-year OM of 22% versus 10% (p = 0.04) and 1-year TRM of 18% versus 4% (p = 0.001), respectively (Figure 1). 36% (4/11 cases) of deaths were attributable to AIHA.

Conclusion

The overall incidence of AIHA following allogeneic HSCT in our study was 3.6%. The risk factors associated with AIHA were receiving HSCT from unrelated donors, and matched gender between the donor/recipient, which has not been previously reported. We observed an association between allo and autoimmunization in the AIHA patients, suggesting a common immune defect underlying both phenomena. The most effective treatment was a combination regimen of rituximab with prednisolone or other immunosuppressive agents. The overall mortality rate for our AIHA patients was high at 53% (10/19 cases), with AIHA as a cause of death in 36% of deceased patients. We have shown that AIHA following HSCTs indeed leads to increased mortality with over 2 fold higher OM in the patients with AIHA.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal