To the Editor:—
We read with interest the recent article by Molenaar et al. 1regarding aprotinin potentially reducing the need for vasopressors during orthotopic liver transplantation. We commend the authors for their work, but have several comments and questions. Although we accept the finding of a statistically significant difference in epinephrine requirements in the aprotinin treated groups, we question the clinical significance of a 50-μg difference in a liver transplant population. We also disagree that the authors demonstrate this reduced vasopressor requirement to be independent of the decreased transfusion requirements in these groups. This article reports a subgroup of a previously published, larger study. 2Perusal of the original article shows that the time of the greatest difference in blood transfusion requirements was the postreperfusion period, the time at which the current authors found their greatest intergroup differences in epinephrine requirements. The authors contend that because central venous pressures 5 min before and 30 min after reperfusion and hemoglobin values 5 min after reperfusion were equivalent between the groups, intravascular volume and fluid resuscitation were also similar between groups. We believe that during liver transplantation, central venous pressures may not accurately reflect fluid status, particularly after reperfusion. Pulmonary capillary wedge pressure might be a more appropriate measure in this setting. Regarding hemoglobin, it is well-recognized that blood loss may not be reflected in the hemoglobin concentration. Finally, the equivalence of values at a few time points does not exclude short-term fluctuations, such as blood loss requiring pressor support, before transfusion.
Regarding the design of the study and the analysis of the results, we would be interested to know what selection criteria were used to identify this subgroup from the original study population. Second, we would be interested to know which variables were log transformed to ensure normality. Third, we note from table 3 of Molenaar et al. 1that the requirements for epinephrine in the placebo group are extremely skewed, suggesting one or two high-requirement outliers, and wonder if this may have influenced the final analysis.
Finally, the authors postulate that their results are explained by aprotinin-mediated inhibition of the kallikrein-induced release of bradykinin, despite achieving adequate doses for inhibition only in the high-dose group. Ironically, the high-dose group experienced the same magnitude of hypotension at recirculation as the patients who received placebo. Observing that the low-dose group was normotensive immediately after reperfusion, the authors suggest that kallikrein inhibition can be obtained at lower concentrations than previously reported. This may be excessively speculative in the absence of appropriate measures of kallikrein activity. Other potential mechanisms of action for aprotinin 3include actions on ischemia–reperfusion injury, which may be of relevance during liver transplantation.