Gvhd and Relapse Free Survival (GRFS) after Allogeneic Transplantation for Idiopathic Severe Aplastic Anemia: An Analysis from the Saawp Data Quality Initiative Program of EBMT

Background

Survival after Allo-HSCT for severe idiopathic aplastic anemia (SAA) has improved over past 20 years, approaching 75% at 5 years. However, beyond survival, a SAA-adapted composite endpoint GVHD and relapse free survival (GRFS) may more accurately assess patient outcomes, becoming a meaningful study endpoint. We analyzed GRFS aiming to identify risk factors and specific causes of GRFS failure.

Methods

This retrospective analysis from the SAAWP Data Quality Initiative (DQI registry database) program of EBMT included patients with: diagnosis of idiopathic SAA; first Allo-HSCT from 2005 to 2016; and matched related (MRD) or unrelated donor (UD) (no cord blood). Relevant events for Kaplan-Meier calculation of GRFS were: relapse (including primary and secondary graft failure); grade 3-4 acute GVHD; extensive chronic GVHD; and death. In addition, we used a competing-risk model to analyze cumulative incidences of specific causes of GRFS failure.

Results

We analyzed 580 patients (385 adults and 195 younger than 18 years), with a median age of 23 years (<0.1-77). Donor was matched related and unrelated in 337 and 243 patients, respectively. Median time from diagnosis to Allo-HSCT was 6 months (IQR: 2-15) and 310 (53%) patients underwent Allo-HSCT without prior treatment. GRFS at 5 years was 69% (65-73) in the whole cohort. Median follow up was 61 months (95%CI: 56-67). Multivariate cox model including age (continuous), graft source, conditioning intensity, sex mismatch, CMV-serostatus, donor type, time from diagnosis to Allo-HSCT (< vs. > 6 months) and previous treatment before Allo-HSCT showed that age (HR=1.02, [1.01-1.03], p<0.001) and CMV serostatus other than negative-donor to negative-recipient [D-/R-] (HR=1.51, [1.02-2.23], p=0.041) were the only independent factors associated with worse GRFS. Using cause specific cox model, we analyzed the risk of the different causes of GRFS failure and found that CMV-serostatus other than D-/R- was associated with higher risk of graft failure/relapse (HR=2.88, [1.12-7.38], p=0.028) while age influenced the risk of grade 3-4 acute GVHD (HR=1.03, [1.00-1.05], p=0.043), extensive chronic GVHD (HR=1.03, [1.01-1.06], p=0.008) and death without prior failure (HR=1.03, [1.01-1.04], p<0.001).

Among the 209 patients who underwent upfront Allo-HSCT from a MRD, 5-year GRFS was 77% (71-84). In multivariate analysis, time from diagnosis to Allo-HSCT (HR=2.64, [1.38-5.03], p=0.003) and age (HR=1.03, [1.00-1.05], p=0.039) independently influenced GRFS. When investigating the causes of GRFS failure in this subset of patients who underwent upfront MRD Allo-HSCT, time from diagnosis to Allo-HSCT was the only remaining factors significantly associated with the risk of death without prior failure (HR=0.29, [0.10-0.84], p=0.022). No factor was found specifically associated with any other causes of GRFS failure.

Conclusions

We observed 5-year GRFS of 69%, meaning that most of patients who underwent Allo-HSCT for idiopathic SAA are cured without experiencing severe forms of acute and chronic GVHD. In the context of upfront MRD Allo-HSCT, GRFS was even more promising (77%). In this particular setting, time from diagnosis to Allo-HSCT was the most important factor influencing GRFS, suggesting the need to proceed to Allo-HSCT as quick as possible when a MRD is available.