Three studies—two new this year—have shown that ketamine with its NMDA antagonist properties may be an effective therapy for ethanol withdrawal and delirium tremens, providing a useful adjunct to GABA agents like benzodiazepines, phenobarbital, and propofol.
Ethanol withdrawal is a complex disease. GABA, an inhibitory neurotransmitter, and glutamate, an excitatory transmitter that can act on NMDA receptors, are two of the main players here. Simplistically, chronic ethanol use leads to a down-regulation of GABA receptors and an up-regulation of glutaminergic receptors, such as NMDA. We are left, when ethanol is abruptly discontinued, with a largely excitatory state with less ability for GABA-mediated inhibition and more capacity for NMDA/glutamate-mediated excitation. Much of the treatment of severe ethanol withdrawal is focused on GABA, though phenobarbital and propofol can suppress the glutaminergic response to some degree.
But a retrospective study from 2015 showed promise for using ketamine for ethanol withdrawal. (Ann Pharmacother 2015;49:14.) Twenty-three adult patients received ketamine at a median infusion rate of 0.20 mg/kg/hr. Mean time to initiation of ketamine from the first treatment for ethanol withdrawal was 33.6 hours, and that treatment lasted 55.8 hours. No change was seen in sedation or alcohol withdrawal scores within six hours of ketamine initiation, but fewer benzodiazepines were needed; the median change in benzodiazepine requirements at 12 and 24 hours post-ketamine initiation were -40.0 and -13.3 mg, respectively.
The authors documented one adverse reaction of oversedation, which required dose reduction, and they concluded that ketamine appeared to reduce benzodiazepine requirements and was well tolerated at low doses.
The second ketamine study published in May was a retrospective observational cohort study where a guideline incorporating ketamine with benzodiazepines (or phenobarbital) for severe ethanol withdrawal. (Crit Care Med 18 May 8. doi: 10.1097/CCM.0000000000003204.) Sixty-three ICU patients with delirium tremens (DTs) were diagnosed by a board-certified medical toxicologist (29 patients pre-guideline, 34 patients post-guideline), and the median duration of ketamine treatment was 47 hours. The mean ketamine infusion rate was 0.19 mg/kg/hr, with 19 patients receiving a loading dose.
Ketamine was associated with decreased ICU length of stay by three days (95% CI, -5.58 to -0.089), decreased likelihood of intubation compared with pre-guideline (odds ratio, 0.14, p < 0.01), and fewer benzodiazepines: mean dose of a diazepam equivalent was 2,525 mg pre-guideline vs. 1,508 mg post-guideline (p=0.02). A more pronounced difference was found in the subgroup of 32 intubated patients: 3,016 mg vs. 833 mg (p=0.01). One documented adverse reaction of oversedation, requiring dose reduction, was recorded.
Also published this past May, the third retrospective study looked at 30 ICU patients with severe ethanol withdrawal who received ketamine in addition to benzodiazepines and phenobarbital. (J Med Toxicol 2018 May 10. doi: 10.1007/s13181-018-0662-8.) The mean time to initiation of ketamine from initiation of lorazepam infusion was 41.4 hours, and the median duration of ketamine treatment was 54 hours. The mean initial ketamine infusion rate was 0.75 mg/kg/hr (average max daily infusion 1.6 mg/kg/hr).
Ketamine was associated with fewer benzodiazepines: Lorazepam infusion decreased by 4 mg/hr at 24 hours (p=0.01) after ketamine infusion. Initial symptom control was within one hour of ketamine initiation in all 30 patients. Two patients experienced hypertension (none had tachycardia) related to ketamine, and six patients were intubated after ketamine initiation, two for escalating symptoms despite ketamine.
This study used much higher doses of ketamine compared with the other two. Six patients ultimately were intubated after ketamine initiation, those two for worsening withdrawal. The question is: Is a reduction of lorazepam infusion by 4 mg/hr (from about 14 mg/hr to about 10 mg/hr) over 24 hours clinically significant? The authors reported that initial symptom control was obtained within one hour, but many remained on infusions (lorazepam, ketamine, or both) for up to three more days. Some institutions did not use benzodiazepine infusions and instead opted for symptom-triggered bolus therapy.
Application to Clinical Practice
Ketamine has the potential to serve as a great partner to GABA agents in managing severe ethanol withdrawal. Unlike dexmedetomidine, which does not target the underlying pathophysiology, ketamine does attack part of the root cause. A reduction in benzodiazepines may be an appropriate endpoint with ketamine but not for dexmedetomidine.
All three ketamine studies were retrospective and not controlled, which should be taken into account when considering ketamine. The Critical Care Medicine study, however, reported on meaningful, patient-oriented outcomes like ICU length of stay and intubations. Well-designed trials are needed, but ketamine seemed to be effective with few reported adverse effects. The infusion rate used in two studies was approximately 0.2 mg/kg/hr, and seems to be a good place to start if considering ketamine for this indication. A bolus up to 0.3 mg/kg may be considered for some patients. The Journal of Medical Toxicology study used a much higher infusion rate.
I am excited to see more ketamine literature for severe ethanol withdrawal. The mechanism makes sense, the results look promising, and this clinical entity can present a significant management challenge. More options in the armamentarium are always welcome. For now, ketamine may be considered an adjunct to GABA agents in managing severe ethanol withdrawal and DTs.
A teaching point (and a plea from toxicologists): The term DTs is often incorrectly used interchangeably with ethanol withdrawal. DTs is a severe manifestation of ethanol withdrawal, but not all ethanol withdrawal is DTs.
Read Dr. Hayes' handout from his lecture at the 2018 American Academy of Emergency Medicine Scientific Assembly, “Beyond Benzos for Severe Ethanol Withdrawal,” which includes his general treatment algorithm athttp://bit.ly/2Kxcz5b. A version of this article first appeared on Academic Life in Emergency Medicine:http://bit.ly/2MBEIFQ.