Portal de Periódicos da CAPES

PRO: The Food and Drug Administration Black Box Warning on Droperidol Is Not Justified

2008; Lippincott Williams & Wilkins; Volume: 106; Issue: 5 Linguagem: Inglês

10.1213/ane.0b013e31816ba463

1526-7598

Ashraf S. Habib, Tong J. Gan,

Anesthesia and Sedative Agents

Droperidol is a cost-effective antiemetic that has been used extensively in anesthetic practice. IV doses of 0.625–1.25 mg have been recommended as first-line therapy for the prophylaxis of postoperative nausea and vomiting (PONV).1,2 However, in December 2001, the United States Food and Drug Administration (FDA) issued a new black box warning due to the potential for serious proarrhythmic effects, and even death, after the administration of droperidol. The FDA noted that cases of QT prolongation and/or torsades de pointes (TdP) were reported, even when doses less than the 2.5 mg labeled dose of droperidol were used. The warning was based on 273 cases reported over a 4-yr period. A possible cardiac event occurred in 74 cases. There were 89 deaths reported, but the dose of droperidol was 2.5 mg or less in only two deaths. The majority of deaths involved droperidol doses that ranged from 25 to 250 mg. Five patients receiving droperidol 2.5 mg or less experienced either ventricular tachycardia or TdP.3 Furthermore, of all the cases reported, cardiac adverse events or death occurred in 10 patients when the recommended 0.625–1.25 mg antiemetic doses of droperidol were used. Careful review of those cases did not reveal evidence of a cause-and-effect relationship.4 As a result of this black box warning, many clinicians became reluctant to use droperidol, and many hospitals removed the drug from their formulary. There was a 10-fold decrease in the sales of droperidol in the year after the black box warning.4 This is despite the fact that, in one survey, 92% of responders did not believe that the black box warning was justified.5 Many experts in the field also challenged the basis of this warning.6–8 Interestingly, a number of studies have recently investigated the use of haloperidol as an alternative to droperidol for PONV prophylaxis.9–13 Given that haloperidol has the same effects on QTc as droperidol, this reflects the fact that anesthesiologists appreciate the antiemetic efficacy of droperidol, but do not believe that the black box warning is justified. So does the available evidence support the black box waning? First, does the literature suggest an increased risk of serious arrhythmias with droperidol? There are abundant data on the use of antiemetic doses of droperidol. A large, randomized, controlled trial involving 1028 patients who received droperidol 0.625 mg and 1.25 mg did not report any cardiovascular side effects or sudden deaths associated with its use.14 A subsequent meta-analysis of 76 trials involving 5351 patients who received different doses of droperidol ranging from 0.25 to 5 mg also did not suggest an increased risk of arrhythmias or sudden death with droperidol.15 Other systematic reviews involving about 3440 patients did not report an increased risk of cardiac events when regimens involving a combination of droperidol with 5-HT3 receptor antagonists were used.16,17 Similarly, in a large European multicenter study, 2030 patients were treated with droperidol with no reported arrhythmias or serious adverse effects.18 In a recent, large retrospective study, the authors estimated that approximately 16,791 patients were exposed to droperidol with no documented TdP being experienced by any patient.19 Since its introduction in 1970, there was only one case report of arrhythmias with the use of antiemetic doses of droperidol.20 It was one of the 10 cases in the FDA database in which adverse cardiac events were possibly related to the administration of droperidol.4 This event occurred intraoperatively in a patient who had baseline prolongation of the QTc (497 ms), secondary to chronic cyclobenzaprine therapy.20 The patient was also receiving fluoxetine, a known inhibitor of several cytochrome P450 isoenzymes, involved in the metabolism of cyclobenzaprine. Droperidol 0.625 mg was administered preoperatively, and the patient developed polymorphic ventricular tachycardia progressing into ventricular fibrillation intraoperatively approximately 2 h after the administration of droperidol. The authors suggested that prolongation of QT and subsequent arrhythmia was caused by inhibition of cyclobenzaprine metabolism by fluoxetine, complicated by administration of droperidol. They did not discuss the possibility that the inhaled drug used to maintain anesthesia might have had a more important contribution to this event, or that the event occurred about 2 h after the administration of droperidol, making it extremely unlikely that it was related to the antiemetic drug.4 Furthermore, several cases of TdP have been reported during inhaled drug-based anesthesia.21–23 Second, what is the evidence regarding the occurrence of QT prolongation with droperidol? There is no doubt that droperidol causes dose-dependent prolongation of the QT interval. Litschke et al. demonstrated rapid onset dose-related prolongation of the QTc after administration of 0.1–0.25 mg/kg droperidol to surgical patients. This prolongation occurred within 60 s after injection of droperidol, did not further increase, and started to decrease during the 10 min duration of the study.24 Similarly, Guy et al. reported a significant prolongation of the QT interval in 70% of patients receiving 0.25 mg/kg droperidol during general anesthesia. The course of this prolongation over the 10 min duration of the study was similar to Litschke et al.'s study. No arrhythmias occurred in any subjects in both studies.25 These studies used doses of droperidol that are much higher than what is recommended for the management of PONV. More recently, White et al. reported that the maximum prolongation of the QT interval occurred 3–6 min after the prophylactic administration of droperidol 0.625 mg and 1.25 mg during desflurane/remifentanil-based anesthesia. The average prolongation was not different compared with patients who received placebo, although a patient receiving 0.625 mg had a maximum QT prolongation of 120 ms, and a patient receiving 1.25 mg had a maximum QT prolongation of 133 ms. By 2 h after arrival in the postanesthesia care unit, the QTc was almost back to baseline. Again, no arrhythmias occurred in any patient in this study.26 Charbit et al. studied QTc prolongation after the administration of droperidol 0.75 mg or ondansetron 4 mg for the treatment of established PONV in the recovery room. They reported that QTc was prolonged in 21% of patients even before administration of the rescue antiemetic. The QTc was significantly prolonged after administration of both antiemetics with no significant differences between the groups. The maximal prolongation occurred 2–3 min after rescue antiemetic use, and was significantly increased from baseline until 5 min after administration. At 90 min, the QTc was significantly lower compared with predrug administration. No arrhythmias occurred in any patient except for supraventricular tachycardia in one patient who received ondansetron.27 Chan et al. also measured the QTc at 5 min and 2–3 h after the administration of droperidol 1.25 mg, ondansetron 4 mg, and the combination of both drugs. There was a moderate prolongation of the QTc at 5 min with return to baseline values in the 2–3 h electrocardiogram (ECG) recording, with no significant differences between the groups and no arrhythmias in any patient.28 The above studies suggest that while a temporal and dose-dependent relationship appears to be present with regard to QT prolongation, this is not the case with regard to the occurrence of arrhythmias or TdP. It also appears that maximum QT prolongation occurs within the first few minutes after administration of droperidol, then decreases towards baseline values. Although none of the studies continuously monitored the ECG to detect the exact time of return of QTc to baseline, ECGs performed between 90 min and 2 h after administration confirmed the return of QTc to baseline. Furthermore, QTc prolongation occurs commonly after anesthesia, even without the administration of droperidol.27,29 In fact, a number of drugs routinely used in anesthesia, prolong the QTc, including thiopental, isoflurane, sevoflurane, desfluane, succinylcholine, neostigmine, atropine, and glycopyrrolate.2–6 More interesting is the fact that cardiac arrhythmias and QTc prolongation have also been reported after administration of the 5-HT3 receptor antagonists (ondansetron, dolasetron, and granisetron),7 the class of drugs that have largely replaced droperidol for the management of PONV. In fact, there are a number of case reports documenting the occurrence of life-threatening arrhythmias and myocardial ischemia after administration of these drugs.30–32 Of note, the Canadian equivalent of the FDA has issued a black box warning on dolasetron causing QTc prolongation. It is important to note that the QT interval is only a surrogate measure for TdP, which is the primary concern for physicians. There is also a lack of direct relationship between the QTc interval and the potential for developing arrhythmias. Some drugs cause significant QTc prolongation but rarely cause TdP.33 Furthermore, there are problems and inconsistencies associated with measurement of the QT interval. For instance, there is no consensus on the way to correct QT interval to account for changes in heart rate.34 The threshold for defining a normal QT interval is also greatly dependent on the correction formula used.35 Charbit et al. reported that the Bazett formula over-corrects QT at heart rates more than 60/min.36 This results in false prolongation of the QTc at faster heart rates. Of note, the Bazett formula is the most commonly used formula with automatic reporting of ECGs. The QT interval is also subject to variation within the myocardium, and differs between ECG leads, a phenomenon referred to as QT dispersion. Furthermore, difficulty in identifying the end of the T-wave can introduce further variability in measurement of the QTc interval. In fact, manual and automatic measurements may not correlate well. As a result of the marked intraindividual variability in the QT interval, a strict distinction between normal and abnormal QTc values becomes problematic. These problems associated with measuring the QTc have actually led to continuing research into alternative measures of cardiac repolarization.33 In summary, droperidol has a long safety record when used in antiemetic doses. There is no convincing evidence to suggest increased incidence of arrhythmias with its use for the management of PONV. QTc is prolonged in a dose-related manner after the administration of droperidol. However, when given at doses <2.5 mg, this prolongation is modest and transient. Although a warning would have been appropriate, the current black box warning is excessive. Furthermore, the manufacturer was unwilling to supply the FDA with data to support a labeling indicating droperidol for treatment and prophylaxis of PONV at doses <2.5 mg, even though such a labeling change could be supported entirely from the published literature without additional studies. Therefore, as a result of this position by the FDA, the use of a very effective antiemetic is continuing to decline.