Happy 50th? Sedation for Colonoscopy in HIV-Infected Patients
We asked three experts — two pharmacologists and a gastroenterologist — about their approaches to sedation in HIV-infected patients receiving ritonavir.
A 50-year-old man with stable HIV infection (undetectable viral load, CD4 count >500 cells/mm3) is referred for routine screening colonoscopy. However, he is concerned that he will not get sufficient anesthesia for the procedure because his partner, who is also HIV-positive and on the same antiretroviral regimen as he is (tenofovir/FTC and ritonavir-boosted atazanavir), underwent the procedure last year and did not receive the usual sedation (midazolam plus either meperidine or fentanyl) because of potential drug–drug interactions with his antiretrovirals. The patient also tells you that one of his friends was told to stop his HIV therapy the day of the procedure, which then allowed the usual sedatives to be used.
What is the proper approach to giving anesthesia to HIV-infected patients who require colonoscopy and similar procedures? Which medications should be used? Can drugs such as midazolam (which is contraindicated with ritonavir) be given at lower doses? Is stopping the HIV medications a safe strategy?
— Edward P. Acosta, PharmD
To deal with this case, we first need to understand the drug interactions that could occur and their significance. Midazolam, meperidine, and fentanyl are all metabolized to some extent by the cytochrome P450 enzyme 3A4 (CYP3A4). Ritonavir is an extremely potent inhibitor of this enzyme and can thus affect the pharmacokinetics of each of these drugs. Midazolam, particularly oral midazolam, is contraindicated with ritonavir use because of increased midazolam concentrations that result in excessive sedation and respiratory depression. Meperidine concentrations are decreased by ritonavir, but the active metabolite, normepiridine, is increased by approximately 50%. Fentanyl concentrations are also increased by ritonavir. Collectively, these data suggest that midazolam, meperidine, and fentanyl should not be used with ritonavir. These compounds can have serious side effects, which can then be exacerbated by ritonavir.
What are the alternatives? One option would be for the patient to undergo an alternative colorectal cancer screening test that does not require sedation (such as fecal occult blood testing, flexible sigmoidoscopy, or virtual colonoscopy); however, if any polyps or cancers were detected on such a test, the patient would be referred for colonoscopy and would be in the same position that hes in now. Sedation-free colonoscopy is also an option, but this patient seems unlikely to choose it, given his stated concerns about inadequate sedation. We are left then to deal with this case pharmacologically.
Stopping this patients antiretrovirals on the day of the procedure will do no good, because the drugs effects on the CYP3A4 enzymes will take much longer to reverse. Given that the patient has an undetectable viral load and a high CD4-cell count, the simplest option would be to stop his ritonavir at least 10 to 14 days before colonoscopy and instead prescribe unboosted atazanavir (400 mg once daily), still with tenofovir/FTC. This approach would allow enough time for ritonavirs induction and inhibitory effects on CYP450 to cease while still providing good antiretroviral coverage. Clearly, this dosage of unboosted atazanavir would not result in the same plasma concentrations as the boosted dosage (300 mg, plus 100 mg of ritonavir, both once daily). Nonetheless, it is approved by the FDA and has been shown to be an effective component in antiretroviral regimens. For clinicians or patients who may be leery of this dosage, another option would be unboosted atazanavir at 200 or 300 mg twice daily. Pharmacokinetic studies have shown that this dosage produces atazanavir trough concentrations well in excess of those observed with 400 mg once daily but still below those seen with ritonavir boosting. The switch to unboosted atazanavir, whether at 400 mg once daily or at 200 to 300 mg twice daily, would be brief, so the risk of virologic failure is small. The risk of new toxicities is also small, given that the patient has already been receiving atazanavir.
— David W. Kubiak, PharmD, BCPS
Fentanyl and midazolam are the preferred agents for procedural sedation in patients who require colonoscopy and similar procedures. However, these medications are contraindicated in HIV-infected patients taking ritonavir and efavirenz. Current guidelines for procedural sedation do not address the management of these patients (Anesthesiology 2002; 96:1004), and thus practice varies widely among physicians.
Alternative strategies that have been explored raise some significant concerns. For example, instructing patients to hold their antiretrovirals the day before the procedure will not eliminate the interaction; despite ritonavirs relatively short half-life, the drugs effect on CYP3A4 inhibition can persist for 48 hours or longer after administration (Clin Pharmacol Ther 2003; 73:406). Similarly, instructing patients to hold their antiretrovirals for several days could potentially cause drug resistance, especially in patients whose virus is not well controlled.
Although several analgesics and sedatives exist that do not interact with antiretrovirals, their pharmacodynamic and pharmacokinetic properties do not seem to be as favorable as those of midazolam and fentanyl. For example, lorazepam has a slower onset and much longer duration of action than midazolam, which may prolong induction and recovery times; propofol requires more-careful hemodynamic and respiratory monitoring, which might not be available in all practice settings; and morphine has a less predictable dose-response effect, especially in older patients, and tends to have a slower onset, longer duration of action, and higher incidence of nausea and vomiting than fentanyl (Anesth Analg 1997; 84:509). Meperidine has similar pharmacologic properties to fentanyl with respect to dose response, onset, and duration of action, but it is limited by interactions with commonly prescribed selective serotonin reuptake inhibitors, such as fluoxetine.
Given that the alternative regimens for procedural sedation are less desirable than fentanyl and midazolam, these two drugs should still be administered to HIV-infected patients, but with careful monitoring for potential increased effects. Lower doses can be used initially, but sedation should ultimately be titrated to patient comfort, and patients should be allowed extra time for recovery. Alternatively, in practice settings where clinicians are trained in using propofol for procedural sedation, fentanyl and propofol could be an option. In either situation, fentanyl may be best prescribed in small single doses rather than continuously, given the results of a pharmacokinetic study showing that ritonavir significantly increased the AUC and decreased the clearance of fentanyl when the sedative was given as a continuous infusion but not when it was administered in small bolus doses (Anesthesiology 1999; 91:681). Last, reversal agents (naloxone and flumazenil) and respiratory support should always be available for use if necessary.
In my practice, we sedate our HIV-infected endoscopy patients the same as we do all our endoscopy patients: We titrate midazolam for effect. In my mind, there is little reason to view midazolam as "contraindicated" with ritonavir. Use of the two drugs together does significantly increase blood levels of midazolam, but this effect should lead to adequate sedation at lower doses without the potential for oversedation, assuming the midazolam is used appropriately — that is, infused at a low dose with a reassessment of effect and level of sedation before additional doses are given. I have performed hundreds of endoscopies using midazolam in patients on ritonavir or efavirenz (another antiretroviral contraindicated with midazolam) and have never witnessed a significant adverse effect. Until interaction between these sedatives and antiretrovirals is proven to be clinically relevant (versus pharmacologic or theoretical, as it is now), deviating from a safe and time-tested sedation strategy does not seem justified.
Tell us what you would do
How would you manage this patient? Submit your comments below.
Dr. Acosta is Professor, Division of Clinical Pharmacology, and Director, Antiviral Pharmacology Laboratory, University of Alabama at Birmingham School of Medicine.
Dr. Kubiak is Infectious Disease Clinical Specialist, Department of Pharmacy Services/Division of Infectious Diseases, Brigham and Womens Hospital, Boston.
Published in Journal Watch HIV/AIDS Clinical Care April 6, 2009
- Unboosted ATV dosing
Rob Sandmann, Pharm.D., AAHIVE, 8 Apr 2009 2:17 PM EST
As a pharmacist, I agree with the last two responses which essentially use the drugs at reduced dosing to compensate... [more]
- Use propofol without the fentanyl
Ravpreet S. Gill, 9 Apr 2009 9:07 AM EST
During my anesthesiology residency I did hundreds of colonoscopies with propofol alone -- you don't necessarily need the fentanyl.
- What Dr. Fennerty said...
James E. Carroll, Critical Access Hospital, 9 Apr 2009 9:07 AM EST
Dr. Fennerty's astute comment captures the clinician's dilemma. Titration of sedation drugs to effect is at the core of what... [more]
- titrated to effect, modazolam-ketamine is a good alternative
Pinny Halpern, Tel Aviv Medical Center, 9 Apr 2009 9:07 AM EST
We use midazolam titration until spontaneous eye closure and thenadd ketamine 0.5-1.0 mg/kg to excellent effect in such pts, with... [more]
- Use sedatives with caution in any patient
A. J. Lee, 9 Apr 2009 9:07 AM EST
Speaking as an anesthesiologist, these sedative drugs are always titrated to effect for individual patients. We are aware that responses... [more]
- Use common sense
Karen V Williams, Pharm D, Health Partners Specialty Center, St Paul, MN, 9 Apr 2009 9:07 AM EST
I have queried several gastroenterologists treating HIV patients in Minneapolis and their practice follows Response 3. They feel reduced doses... [more]
- Ask the sailor how to sail not the professor.
Jonathan C Hausheer, 9 Apr 2009 11:41 AM EST
Why do you ask people who don't give sedation for colonoscopy how do do it. Clearly the first two responders... [more]
- common sense
Robert C. Barker, Sparks Hosptal, Ft. Smith, AR, 14 Apr 2009 10:29 AM EST
Common sense would would seem consistent with Dr. Fennerty's approach.
- Fentanyl and Midazolam
Kevin Parent, Geisinger Clinic, 14 Apr 2009 10:29 AM EST
I would use fentanyl and Midazolam at low and slow doses to achieve desired effect.
- no nfusions for short procedures
Ramanathan Moorthy, Manipal Hospital, Bangalore, 14 Apr 2009 10:29 AM EST
By logical thinking, when you are talking about lack of availability for prolonged monitoring, do not use infusions, especially for... [more]
- Similar concern as previously with azoles
George R Thompson, UTHSCSA, 14 Apr 2009 10:29 AM EST
Although admittedly ritonavir is a more potent CYP450 inhibitor, this concern received attention previously early in the AIDS epidemic due... [more]
- PK principles
Cari Brackett, 14 Apr 2009 10:29 AM EST
An important thing to remember is that IV sedation for a procedure constitutes a SINGLE rather than repeated dosing model.... [more]
- Titration and Post-Procedure Monitoring Are Key
John K. DesMarteau, MD, 22 Apr 2009 6:27 PM EST
As someone who happens to be on efavirenz I received midazolam and fentanyl for a colonoscopy without incident at age... [more]
To ensure that your Reader Remark is not formatted as one long paragraph, precede new paragraphs with either a blank line or an indentation.