We appreciate Dr Yende's interest in our article1 and the opportunity to respond to Dr Yende's concerns. It appears that Dr Yende is uneasy about 2 issues: (1) the use of a 1-sided statistical test and (2) the inclusion of only 73 patients for the control study group. Dr Yende states that "1-sided statistical tests should be limited to conditions in which overwhelming evidence suggests that the effect will occur only in 1 direction." Our decision to use a 1-sided statistical test was based on our specific hypothesis that the mutations of human TLR4 associated with hyporesponsiveness to inhaled endotoxin would also enhance the susceptibility of humans to gram-negative sepsis. This hypothesis is supported by our previous study that demonstrated that the mutations in human TLR4 under investigation in this study were associated with a decreased response to inhaled lipopolysaccharide,2 and studies in mice that have demonstrated that mice that contain mutations in TLR4 are both hyporesponsive to lipopolysaccharide3,4 and are more susceptible to gram-negative bacterial challenge.5 Thus, in our study under question,1 we were only investigating one side of the hypothesis that individuals with mutations of TLR4 would be more susceptible to gram-negative sepsis. The second concern raised by Dr Yende was also addressed in the article; the control population had been previously established by Dr Mira through the blood collection bank at the Cochin University Hospital (Paris, France) and, in part, was successfully used in a previous study.6 Moreover, the control population in the current study1 had the same prevalence of the lone Asp299Gly polymorphism as had been observed previously in an independently collected study population,2 providing further support for the representativeness of the control study group. Finally, while we believe that our findings are consistent with the biological and pathophysiological importance of TLR4, we recognize that our study is limited by the size of our study population and the generalizability of our findings. Given these limitations, we strongly endorse the need to have our hypothesis retested in an independently collected population.
Mira JP, Schwartz DA, Frees KL. Role of Toll-like Receptor 4 Mutations in Gram-Negative Septic Shock—Reply. Arch Intern Med. 2002;162(21):2496. doi: