Reply to “Bacterial Proliferation May Be the Key Component of Sepsis Mortality”

REPLY

In response to the letter by de Tymowski et al. (1) regarding our publication (2), we are happy that our study provides an open, fruitful discussion that is important to the scientific community.

Sepsis is a condition that affects many people worldwide every year. Its consequences are devastating, not just in the number of mortalities but also in the morbidities that in some cases extend beyond discharge from the hospital. There are no suitable predictors for sepsis except clinical symptoms, including positive bacterial cultures. Therapy is only supportive, particularly antibiotics, which have without any doubts saved many lives. However, the use of antibiotics in the clinical setting is not totally successful. Otherwise, sepsis would not be a major health problem (3).

Sepsis is a complex condition that is modulated by many factors, including genetics, age, sex, and the environment (4). Therefore, the treatment of this condition must also be complex. Since human beings are not ideal experimental models, our understanding of the cellular and molecular mechanisms undergone during sepsis has been greatly improved by studies using experimental animal models in which disease conditions could be regulated to the best of our abilities. Therefore, our study was directed toward evaluating the contribution of antibiotics in the gold-standard mouse model of sepsis, cecum ligation and puncture (CLP) (5, 6). The rationale for doing this study is that many laboratories have used antibiotic treatment in this experimental model. Although an improvement in survival has been observed, 80 to 100% protection has never been achieved, suggesting that other factors besides live bacteria could contribute to the final outcome. Our protocol for the use of imipenem, dosage and administration, was selected based on prior studies by many laboratories (7–12). We must admit that we were surprised when we did not observe any antibiotic protection in our model. Consequently, we tried different variations of the antibiotic protocol, including direct administration into the cecum, where bacteria reside. Based on these observations, we predicted that, indeed, factors other than live bacteria might be in play. Experiments in which the cecum content was sterilized by two different methods still resulted in 100% mortality, sustaining our initial assumption. Therefore, the spirit of our publication is not to propose that antibiotics are not necessary but rather to suggest that other unknown factors need to be considered for the development of highly needed new therapies.

de Tymowski et al. claimed that our antibiotic protocol does not resemble the one used in humans. We agree, but a mouse is not a human. So, it is difficult to reproduce the effectiveness of human treatment in mouse models. However, the human and mouse experiences have one thing in common: the use of antibiotics does not necessarily result in 100% recovery from sepsis, supporting our initial argument. de Tymowski et al. also claimed that the statistical analysis of our studies is not sound because we used a very small number of mice. Overall, we used 52 mice for antibiotic treatment and a control for four variations of treatment, rather than using a large number for a single protocol. Even combining all experiments, we did not observe a statistical difference between the two groups. de Tymowski et al. proposed that if the number of mice for each group is doubled theoretically, it would reach statistical significance. de Tymowski et al. are supposing that the results of an experiment with more animals will yield the same differences between groups, which may not be experimentally the case. However, assuming that this possibility is true, which I do not believe, the question is whether this is a “statistical artifact” or a real “biological phenomenon.” Even if statistical significance is observed, the second question is, Why is there not 100% protection? The answer is clear. It is because other factors are contributing to the outcome.

Another claim in de Tymowski et al.'s letter suggests that mortality after the administration of sterilized cecum may still be caused by bacterial proliferation due to contamination during the procedure. The sterilized cecum was analyzed for the presence of microorganisms, and we did not find live bacteria in the sample. Moreover, we did not detect a significant bacterial load in the blood after the injection of the sterilized cecum content. In addition, if mortality after two sterilization protocols is due to contaminating bacteria, why is the inflammatory response so different between the two treatments?

In summary, we are happy that our publication has opened a healthy debate on approaches to combat sepsis. Indeed, more studies are needed to address this tremendous health problem. We believe that research such as the type that we have conducted contributes, at least a little, to increasing our understanding of the complexity of sepsis. It suggests that any successful therapeutic intervention is likely to require more than one neutralizing agent as has been shown for other diseases. We believe that our contribution may open the possibility for further studies to find conditions to mitigate sepsis. In addition, it illustrates that a mouse is not a human.