Remifentanil in the Intensive Care Unit: Tolerance and Acute Withdrawal Syndrome after Prolonged Sedation

A 69-yr-old man was admitted to the hospital for facial and hip trauma after a bike accident. He underwent intubation on day 3 for respiratory failure. He was sedated with propofol at 200 mg/h and remifentanil at 0.1 μg · kg⁻¹· min⁻¹and then 0.16 μg · kg⁻¹· min⁻¹to maintain a Ramsay score of 3. On day 5, the propofol was stopped, and the remifentanil infusion was decreased to 0.08 μg · kg⁻¹· min⁻¹for 1 h. The patient then safely underwent extubation, and the remifentanil was stopped 1 h later. Within 10 min, the patient became very anxious, reporting vague discomfort but no pain. He had myoclonus and uncontrollable shaking of all four limbs, and was noted to be sweating, hypertensive, and tachycardic. We suspected an opioid withdrawal syndrome, which was treated with 20 mg intravenous morphine, 150 μg clonidine, and 10 mg haloperidol. Despite this treatment, the symptomatology persisted. We restarted remifentanil at 0.1 μg · kg⁻¹· min⁻¹, and the symptoms disappeared within 5 min. Subsequently, we started a morphine infusion at 40 mg/day and then progressively tapered remifentanil over the next 48 h.

A 52-yr-old man was admitted to the hospital for head trauma after a motorbike accident. He underwent intubation and received remifentanil at 0.1 μg · kg⁻¹· min⁻¹and midazolam at 15 mg/h. After 24 h, the midazolam was stopped. During the ensuing 30 days, we had to increase the infusion rate of remifentanil to 0.725 μg · kg⁻¹· min⁻¹to maintain a Ramsay score of 3. On day 33, after neurologic improvement, we were able to decrease the infusion rate to 0.325 μg · kg⁻¹· min⁻¹, and the patient underwent extubation after further tapering over a 2-h period. Ten minutes after discontinuation, we observed severe agitation with tachycardia, hypertension, tachypnea, mydriasis, sweating, and myoclonus. The patient was given 10 mg morphine and 150 μg clonidine intravenously, without any change. We restarted remifentanil infusion at a rate of 0.1 μg · kg⁻¹· min⁻¹, and the symptoms improved dramatically. An intravenous infusion of morphine was started, followed by tapering of the remifentanil over 24 h.

A 32-yr-old woman was admitted to the hospital after an automobile accident with head trauma, myocardial contusion, and fracture of the wrist and ankle. She underwent intubation and was sedated for the next 4 days to maintain a Ramsay score of 4, with propofol at 120 mg/h and remifentanil at 0.12 μg · kg⁻¹· min⁻¹. The remifentanil infusion rate had to be increased to 0.72 μg · kg⁻¹· min⁻¹and the propofol rate had to be increased to 200 mg/h because of pneumonia. After resolution of the pneumonia, the remifentanil infusion rate was decreased to 0.48 μg · kg⁻¹· min⁻¹, and propofol was stopped and replaced by midazolam at 10 mg/h because of signs of major anxiety. On day 15, midazolam and remifentanil were discontinued, and 10 mg morphine was injected intravenously. The patient became extremely agitated, although not confused, with associated tachycardia, hypertension, tachypnea, and mydriasis. She received 150 μg intravenous clonidine and 20 mg ketamine, without any improvement. She improved immediately after reintroduction of remifentanil. The patient underwent extubation 2 days later, after remifentanil was tapered in conjunction with an intravenous infusion of morphine and clonidine.

We use remifentanil-based sedation to facilitate frequent awakening for neurologic and respiratory evaluation of our patients.4,5,12,13During a 6-month period, 4 of 40 patients receiving a similar sedation regimen in our department presented the same acute-onset withdrawal syndrome, in addition to development of acute tolerance to remifentanil.

Initial infusion rates complied with the manufacturer’s recommendations (0.10–0.15 μg · kg⁻¹· min⁻¹). Subsequently, they were adjusted, following an established protocol, to maintain a predefined Ramsay score. When sedation was no longer needed, the infusions were tapered and stopped over a 2-h period. This protocol was similar to protocols in place in other intensive care units in France.

We think that the observed symptoms are those of withdrawal syndrome because, first, the clinical presentation was typical of an opioid withdrawal syndrome, and the onset was approximately 10 min after discontinuation of remifentanil; second, the patients were not in pain at the time of symptoms; third, the patients had regained consciousness and did not show any confusion; and finally, the patients did not have any significant preexisting medical conditions or history of substance abuse (e.g. , alcohol) that could explain the symptoms.

We did not find any similar cases of acute-onset withdrawal syndrome in studies with adult patients9,10or children.14–17In adult studies, withdrawal syndrome occurs when there is a sudden discontinuation of an opioid drug (usually fentanyl).9,10We postulate that in our series of patients, the same phenomenon occurred after discontinuation of the short-acting remifentanil over a short period of time (2 h).

Acute tolerance to the analgesic effects of opioids has been shown to occur within a few hours after commencing administration of opioids,6,7including remifentanil,6,8but acute tolerance to their sedative properties has been studied little.16,17 

Our patients had no history of alcohol or substance abuse. However, other patients who received the same drug regimen, including some who were alcohol abusers, did not present such a withdrawal syndrome. We believe there may be a genetic component to these observations.18,19Pharmacogenetics is a gene-by-gene approach, and pharmacogenomics is an approach to the genome focusing on its expression.20The latter approach seems more appropriate in evaluating a genetic basis for tolerance and withdrawal, because these processes are complex and likely involve many pathways (e.g. , the cyclic adenosine monophosphate signal transduction pathway and its distal effectors).20However, much research remains to be done in this field before we can accurately evaluate the potential importance of these various mechanisms.21 

In conclusion, we hypothesize that remifentanil-based sedation may result in acute tolerance and a unique withdrawal syndrome, in that it can only be treated by remifentanil. The obvious precipitating factor seems to be the rate of infusion discontinuation. Because there were no medical risk factors identified, we postulate that this difference could be due to yet unknown genetic factors. Current guidelines recommend a systematic prevention of opioid withdrawal syndrome by gradually tapering the infusion rate.22However, further studies are necessary to prospectively evaluate the clinical importance of this phenomenon in relation to remifentanil. In the meantime, we recommend tapering a remifentanil-based sedation over 24–48 h, in conjunction with a concurrently running morphine infusion (evidence-based recommendation grade C).

The authors thank Francis Veyckemans, M.D. (Professor of Anesthesiology, Anesthesiology Department, Chairman of Acute Medicine Department, University Hospital Saint-Luc, Catholic University of Louvain, Brussels, Belgium), and Scott Nordstrom, M.D. (Staff Anesthesiologist, Montreal General Hospital, McGill University Health Center, Montreal, Canada).