Reply.

Abstract

Potential conflict of interest: Nothing to report. We appreciate Kalal et al.'s comments to our article “Continuous Renal Replacement Therapy Is Associated With Reduced Serum Ammonia Levels and Mortality in Acute Liver Failure.”1 In acute liver failure, the lack of liver capacity to metabolize ammonia into urea and glutamine leads to its accumulation in the bloodstream. Uptake of part of this ammonia by the brain, muscle, and other tissues results in a gradient between arterial and venous ammonia levels.2 Therefore, the arterial ammonia level has been found to be a better surrogate for the risk of cerebral edema, intracranial hypertension, and mortality in acute liver failure.3 Nevertheless, venous ammonia has been shown to correlate well with arterial ammonia and to be associated with such clinical endpoints in acute liver failure.2 Taking this into account and to maximize sample size and the internal validity of our results, we decided to use both arterial and venous ammonia measurements. Furthermore, we were especially interested in the ammonia variation from days 1‐3 post–study inclusion with and without renal replacement therapy (RRT). Finally, in our cohort (n = 1,186), while the serum ammonia level (arterial or venous) at study inclusion was associated with 21‐day transplant‐free mortality (P < 0.001), the type of ammonia (arterial versus venous) at study inclusion was not associated with 21‐day transplant‐free mortality (P = 0.49). Moreover, median arterial and venous ammonia levels were similar for each of the continuous RRT, intermittent RRT, and no RRT subgroups on days 1, 2, and 3 (P ≥ 0.08 for all comparisons). Thus, we believe that our methodology was robust to extract our conclusions.