Abstract
Infection is a leading cause of morbidity following allogeneic hematopoietic stem cell transplantation (allo-HCT). In a previous retrospective study of related, HLA-matched myeloablative transplants, we identified associations between common inherited polymorphisms in the MBL2 gene encoding the mediator of innate immunity, mannose-binding lectin (MBL), and risk of major infection. Missense mutations resulting in low circulating MBL levels were associated with increased infection, and high-producing promoter haplotypes were protective. However, subsequent studies have been conflicting, there are no data examining non-myeloablative transplants, and the relationship between MBL2 genotype and serum MBL levels peri-transplant has not been studied. This is important as MBL is an acute phase reactant primarily synthesized by the liver, and many conditioning regimens are hepatotoxic. Here we report a prospective study examining MBL2 genotype, MBL levels and risk of major infection following HLA-matched sibling myeloablative and non-myeloablative allo-HCT. Data is available for 138 transplants, 81 myeloablative and 57 non-myeloablative. 49 of the myeloablative transplants utilized total body irradiation (TBI) based conditioning regimens. Five promoter and missense MBL2 polymorphisms were genotyped, and plasma MBL mannan-binding and C4-deposition levels were measured pretransplant for donor and recipient, and at days 0, 14 and 28 post-transplant. Major infection was defined as invasive or systemic episodes of microbiologically confirmed sepsis. MBL levels were higher in recipients than donors at baseline (t test P<0.0001), reflecting an acute phase response induced by disease and prior treatment. Recipient MBL levels increased during the peritransplant period (ANOVA P=0.0002), most strikingly in individuals without MBL2 coding mutations undergoing TBI. 59 of 138 (42.8%) recipients experienced at least one episode of major infection (myeloablative 38/81, 46.9%, non-myeloablative 21/57 (36.8%), P=NS). The most significant clinical risk factor for infection was the use of myeloablative TBI (HR 2.8, CI 1.39–5.8, p=0.004). Analyzing all transplants, MBL2 genotype (recipient P=0.07, donor p=0.08), and recipient mannan-binding MBL levels (P=0.10) were weak risk factors for infection, however the association was highly dependent upon the type and intensity of conditioning. No association was observed in non-myeloablative transplants, but a significant interaction was observed between recipient MBL2 coding mutations and the use of TBI in myeloablative transplants (Cox P=0.002). For example, 70.6% of recipients with a coding MBL2 mutation receiving TBI developed major sepsis, compared with 20% of those without mutations not receiving TBI. Our results confirm the association of MBL status with risk of infection risk post allo-HCT. Importantly, the association is restricted to myeloablative, TBI-conditioned transplants. Further studies examining the role of MBL replacement in this setting are warranted.
Disclosures: The Cooperative Research Centre for Vaccine Technology at the Australian Red Cross Blood Service funded the study.; Single Honorarium from Enzon to CGM in December 2005.; CGM is on the Scientific Advisory Board for Enzon, who have licensed recombinant mannose-binding lectin.
Author notes
Corresponding author