Editorial comments by Dr. Daren Heyland:

This is a landmark article in that it demonstrates the feasibility and safety of enteral pharmaconutrition in the critically ill patient. The investigators are to be commended on this novel and innovative approach to providing key pharmaconutrients dissociated from the nutrition strategy. In this manner, one can initiate and deliver, in low volumes, the total amount of key nutrients without waiting for the patient to tolerate enteral nutrition (which is a problem with currently designed immunonutrition formulations where these key nutrients are mixed in with ready-to-feed solutions).

Unfortunately, there are too many other design flaws to enable us to make any clinical inferences as to the efficacy of this enteral pharmaconutrition approach. Specifically, the fact that patients in the experimental group also received an arginine containing solution disables us from making any inferences about the efficacy of either arginine-supplmented diets or the enteral pharmaconutrition. A much better design would have been to give both groups a standard feeding solution (not enrinched with pharmaconutrients) and then randomized them to the enteral glutamine and antioxidant solution vs placebo.

Another main concern with this paper, that limits the inferences we can make from it, is the competing nature of the primary endpoint with mortality. Patients who die initially have high SOFA scores and then the day of death, those high scores disappear, because the data are no longer available. Artificially then, the SOFA scores resolve quicker in groups where more patients die. There were 2 more deaths in the experimental group and how do we know that the drop outs due to death did not account for the observed differences in slopes. Of equal importance is the question, “What is the clinical importance of a difference of 0.18 of a regression (slope) coefficient?” The answer to this is ‘unknown’ and since the study was too small and underpowered, we really cannot comment on the impact of this experimental strategy on clinical outcomes.

It is unfortunate that the investigators stopped their study early. We are not given to understand whether there were clear stopping rules or whether the multiple tests of significance were incorporated into the final analysis.

Of final note, it is interesting that the authors postulate that there is a resolution of organ function with the enteral pharmaconutrition strategy and yet, there was no differences in markers of oxidative stress between the two groups. Glutamine, zinc, and Selenium levels were also the same between the two groups. Wonder what the mechanisms of benefit might be then?

In summary, based on this study, would neither conclude that arginine is safe in sepsis nor that enteral pharmaconutrition is beneficial. Clearly, more studies evaluating this pharmaconutrient strategy are warranted. Larger, more rigorously designed studies will hopefully illuminate subsequent clinical recommendations.