Background: Anemia is commonly present in patients with malignancies and is associated with reduced survival times. Recently, we described that low (<110 g/L) pre-transplant hemoglobin levels (PT-Hb) are associated with decreased early survival after allogeneic hematopoietic stem cell transplantation (alloHSCT)(

Xenocostas et al.
). We are now presenting data on long-term survival and causes of death in BMT recipients with and without anemia prior to alloHSCT.

Study Design and Methods: A retrospective analysis of 511 patients consecutively transplanted between January 1995 and March 2000 was performed to evaluate survival, cause of death, and PT-Hb. The end date for follow-up was June 2002. The median follow-up time was 993 days. Causes of death were categorized either as relapse, treatment-related mortality (TRM), or other. PT-Hb levels were determined within 2 weeks prior to transplantation, after commencing conditioning chemotherapy. Comparisons between groups were done using chi-squared tests.

Results: The 180-day survival of patients with low PT-Hb levels (<110 g/L) was significantly worse than that of patients with PT-Hb levels ≥110 g/L (57.1% versus 83.1%, p<0.0001). The survival difference remained significant at 5 years (36.2% versus 59.4%, p<0.0001). The difference in 180-day survival was contributed to by an increase in TRM (36.1% versus 14.9%, p<0.0001) as well as a higher relapse rate (4.4% versus 0.3%, p=0.019 by Fisher’s exact test). For patients surviving more than 180 days, there was no difference in TRM (16.7% versus 16.7%, p=0.987) or relapse rate (12.0% versus 7.3%, p=0.150). No difference in the rate of other causes of death was found between the groups at either the 180-day or 5-year time points.

Conclusions: Pre-transplant anemia is an independent risk factor for increased mortality following alloHSCT. Relapse and treatment-related deaths are both more likely to occur early in the post-transplant course of patients experiencing pre-transplant anemia. Differences in long-term survival are predominantly related to treatment-related deaths rather than relapse.

Author notes

Corresponding author