Reply to Smithson et al.

[Extract] To the Editor—Mannose-binding lectin (MBL) deficiency has been associated with a predisposition to numerous infectious diseases, and our recent data indicate that patients with pneumococcal infection who have MBL levels <0.5 μg/mL are more likely to die. The data included in our recent meta-analysis excluded the initial study that showed a strong association between MBL2 variant allele homozygosity and invasive pneumococcal infection, because these patients' MBL blood levels were not measured. If these patients could have been included, the association we ascertained between MBL deficiency and death in pneumococcal infection would very likely have been even stronger than our finding of an OR of 5.62 (95% CI, 1.27–24.92). In their letter, Smithson et al. indicate that conclusions regarding the significance of MBL deficiency need to be viewed in the context of in vitro data that describe the binding of this protein to bacterial cells and the consequent deposition of complement. The earliest data relating to Streptococcus pneumoniae demonstrated binding of the bacteria to MBL, but more-recent data indicate that encapsulation of pneumococci abrogates binding. One crucial aspect of the contribution of MBL to the killing of pneumococci that has not been studied to date is the contribution of neutrophils. Until these data are available, it is premature to ignore the strong association between MBL deficiency and poor outcomes of pneumococcal infection