Abstract

Despite improvement in diagnosis and therapy, invasive fungal infections (IFI) remains one of the most important barriers to the success of allogeneic stem cell transplantation (allo-SCT). Mannan-binding lectin (MBL) gene polymorphisms have been associated with an increased risk of infections, particularly in immunocompromised patients. We analyzed the association of MBL gene polymorphisms with the incidence of IFI in 138 HLA identical donor-patient pairs, undergoing allo-SCT in our institution between March 1995 and December 2003. Median age at time of transplantation was 38 years (range 17 – 63); 73 (53%) patients were male. Diagnoses were AML (n=43), CML (n=23), ALL (n=19), NHL (n=18), and other hematological malignancies (n=35). 72 (52%) patients were in early phase of the disease and 66 (48%) in advanced phase. 129 (93%) donors were siblings and 9 (7%) unrelated. The source of stem cells was peripheral blood in 133 (96%) patients and bone marrow in 5 (4%). Conditioning regimen was myeloablative in 126 (91%) and non-myeloablative in 12 (9%). 71 patients (51%) received a T cell depleted graft by CD34+ selection, and in 67 (49%) cases the graft was unmanipulated. Six single nucleotide polymorphisms (−550 G/C, −221 C/G, 4 C/T, 52 CGT/TGT, 54 GGC/GAC and 57 GGA/GAA) in the MBL2 gene were genotyped using PCR and sequence specific primers. 22 (16%) donors and 21 (16%) patients of the 138 pairs carried O/O, LXA/LXA or LXA/O genotypes, which are related to a low production of MBL protein. After a median follow up of 21 moths, 25 (18%) patients developed an IFI, with an overall actuarial probability for this complication of 22%. In the univariate analysis, factors associated with a higher probability of IFI were: unmanipulated grafts (10% in CD34+ selected vs. 27% in unmanipulated grafts, p=0.001), aGVHD II–IV (9% vs. 30%, p=0.0007), and “low-producing” MBL donor genotype (14% vs. 41%, p=0.01). The group of patients with aGVHD II–IV and with a “low-producing” MBL donor genotype had an actuarial probability of IFI of 78% vs. 18% for the remaining patients (p=0.0003). No association between recipient’s MBL polymorphism and IFI was found. All variables with a p value < 0.2 in the univariate analysis were introduced in the multivariate analysis. Only two independent factors were associated with a higher incidence of IFI in the multivariate analysis: “low-producing” MBL donor genotype (RR 7.2, p=0.007) and aGVHD II–IV (RR 5.5, p=0.02). In conclusion, donor’s MBL gene polymorphism possesses strong predictive value for IFI. This finding could be useful for selecting the most appropriate donor. In addition, studies investigating primary anti-fungal prophylaxis in patients developing aGVHD II–IV and with a “low-producing” MBL donor genotype are warranted.

Author notes

Corresponding author