To the Editors:
We read with great interest the paper by Hajjar et al., “Vasopressin versus Norepinephrine in Patients with Vasoplegic Shock after Cardiac Surgery: The VANCS Randomized Controlled Trial,” and the accompanying editorial.1,2 There are a number of limitations inadequately addressed by the authors and in the editorial that limit the generalizability of VANCS findings to clinical practice.
First and most importantly, the VANCS treatment protocol used doses of vasopressin and norepinephrine that are not equivalent. Vasopressin was dosed at 0.01 to 0.06 U/min, which is similar to dosing in common clinical practice and to the regimens used in large randomized controlled trials of vasopressin versus norepinephrine in sepsis (Vasopressin and Septic Shock Trial [VASST] and the Vasopressin vs. Norepinephrine as Initial Therapy in Septic Shock [VANISH]).3,4 On the other hand, norepinephrine was dosed at 10 to 60 μg/min (~0.14 to 0.86 μg · kg–1 · min–1), which is approximately five times higher than the dose typically used following cardiopulmonary bypass at our center and many others. This dose also far exceeds that used in VASST (5 to 15 μg/min) and VANISH (maximum of 12μg/min). Using lower doses, neither VASST nor VANISH found significant differences in acute kidney injury or other outcomes between vasopressors.
Second, other than management of vasopressin and norepinephrine, clinical treatment was not protocolized. Balanced use of fluids, pressors, inotropes, and vasodilators is needed to assure adequate blood pressure and organ perfusion after cardiopulmonary bypass. VANCS patients randomized to the norepinephrine group subsequently required longer duration of treatment with the intervention vasopressor, greater use of dobutamine (P = 0.007), and greater use of open-label norepinephrine (19% vs. 11%; P = 0.06). Although intraoperative epinephrine was used in more than 25% of patients in both treatment groups, its postoperative use was not described. Similarly, the authors note that there were no differences in fluid administration between groups, but data on urine output were not provided. We suspect that greater use of dobutamine was required to offset intense vasoconstriction from the high dosage of norepinephrine.
Third, the unequal dosing regimens for vasopressin and norepinephrine had potential to unmask the clinical care teams to treatment allocation. In our experience, initiating norepinephrine at 10 μg/min causes a more dramatic rise in blood pressure than vasopressin at 0.01 U/min. Differences between treatment groups in duration and alternative pressor/inotrope usage would also have been noticeable to clinicians. Successful prediction of treatment allocation by clinicians could have introduced bias in the observed outcomes.
Fourth, the overall complication rates reported in this study are much higher than expected. With a baseline EuroSCORE of 5, one might expect a mortality of 3 to 5% instead of the 15% rate observed.5 Similarly, the 60 to 80% incidence of atrial fibrillation is much higher than the 26 to 32% rate reported in other studies.6,7 These outcome differences suggest a systematic difference in care that limits generalizability of findings.
VANCS investigators conclude that vasopressin causes lesser kidney injury and atrial fibrillation and should be considered for first-line treatment of vasoplegia after cardiopulmonary bypass. We caution against generalizing these results across centers, and bias against specific drug classes. Vasopressin and norepinephrine have distinct pharmacologic properties associated with benefits and risks. Given the fact that hypotension after cardiopulmonary bypass can be caused by a number of different pathologic processes, including ventricular dysfunction, hypovolemia, and vasodilation, it remains unclear which pressor(s) and dose provide optimal outcome for treatment of shock after cardiopulmonary bypass, or if a single regimen can be applied with equal effectiveness to all patients.
Research Support
Funded by the Department of Anesthesiology, Johns Hopkins University School of Medicine, Baltimore, Maryland.
Competing Interests
The authors declare no competing interests.