Antiretroviral agents have frequent side effects and numerous potential drug interactions [1–3]. Although the effects of antiretroviral agents on other drugs used to manage HIV infection are relatively well studied, less is known about their effects on other drug classes [4,5].
Many HIV-infected individuals are also receiving medications for psychiatric disorders, including depression. Depressive disorders are more common in HIV- infected individuals than the general adult population, occurring in 22–45%[5,6]. Injection drug users with HIV infection have been reported to have a rate of major depression that is five times that of the general population . The potential for clinically significant drug interactions resulting from antiretrovirals and antidepressants has been identified [6,8,9], but actual cases are rarely described. We report a series of cases of HIV infected patients receiving highly-active antiretroviral therapy and antidepressants who developed serotonin syndrome due to drug–drug or drug–food interactions.
A 42-year-old African American male with a history of AIDS, depression, and Kaposi's sarcoma was receiving a stable regimen of 40 mg fluoxetine daily, 40 mg stavudine twice daily, 150 mg lamivudine twice daily, one double-strength cotrimoxazole tablet daily, 1.2 g azithromycin weekly, 200 mg fluconazole daily, 400 mg acyclovir twice daily, and 2.5 mg dronabinol twice daily. In April 1997, 600 mg ritonavir twice daily was added to his regimen in accordance with a study protocol. Within 1 week he noted severe anxiety, diarrhea, fever to 103.5°F, nausea, vomiting, headache, and confusion. He was admitted to the hospital with dehydration, hypokalemia and other electrolyte abnormalities. AFB blood cultures were negative. The treatment with ritonavir was discontinued and his symptoms resolved.
One year later, he was receiving a regimen containing saquinavir and ritonavir, 400 mg twice daily. Fluoxetine was empirically reduced to 20 mg daily. He tolerated these medications well.
A 49-year-old white male with HIV who had been diagnosed in 1989, Grave's disease (1989), major depression (1990), and non-Hodgkin's lymphoma (1992) had been treated with a variety of antiretroviral regimen since 1993. Other medications included 40 mg fluoxetine daily (since September 1997), 0.5 mg clonazepam twice daily, 200 mg levothyroxine daily, and one double-strength cotrimoxazole tablet daily. On 9 March 1998 he complained of increasing depression and a desire to quit smoking. His psychiatrist prescribed 150 mg bupropion SR twice daily.
On 13 March 1998, he received a new antiretroviral treatment regimen consisting of 400 mg didanosine daily, 500 mg hydroxyurea twice daily, 400 mg ritonavir twice daily, and 400 mg saquinavir twice daily. After 2 weeks on this regimen, the patient complained of nausea, vomiting, diarrhea, diaphoresis, paranoia, ‘flight of ideas', ‘fitful’ sleeping and feeling disconnected.
His thyroid stimulating hormone level 14.85 IU/ml was high (normal range 0.3–5) and his levothyroxine was increased to 0.3 mg daily. On April 4, 1998 ritonavir was discontinued and nelfinavir was initiated. One week later his symptoms had resolved.
A 32-year-old African American female with a history of HIV infection diagnosed in 1992 and major depression, had been receiving 40 mg fluoxetine daily since November 1997 and one double-strength cotrimoxazole tablet daily. On 9 June 1998 she began a new antiretroviral regimen of didanosine, stavudine, hydroxyurea, and efavirenz. After 2 days on this regimen, she complained that she was ‘crawling out of her skin', restless, anxious, and sweating. After 4 weeks she was advised to decrease her daily dose of fluoxetine to 20 mg, which resulted in a resolution of her symptoms.
A 57 year old Hispanic male, who was a former intravenous drug abuser was found to be HIV positive in October 1990. His past medical history included hypertension, depression and post-traumatic stress disorder. Since March 1997 he had been receiving his first protease-containing regimen of indinavir, stavudine, and lamivudine.
The patient did well on this combination regimen until June 1998 when he presented to the ambulatory clinic with complaints of dizziness, mild confusion, diarrhea, visual changes and a general feeling of being ‘out of sorts’ for approximately 1 month. His symptoms had significantly worsened over the previous week. He described a syncopal episode without a clear triggering event while waiting for a bus. His other medications included daily doses of 20 mg fluoxetine, 200 mg trazodone, 10 mg benazepril, and one double-strength cotrimoxazole tablet. On physical examination, his sitting and standing blood pressures were 140/85 and 144/90 mmHg, respectively. His heart rates were 59 and 96 beats/min, respectively. Routine blood chemistry tests and blood counts were all within normal limits. His electrocardiogram showed normal sinus rhythm with normal intervals and no acute changes.
Upon further interview, it was learned that the patient had traveled to Florida in the previous month and returned with five crates of grapefruit. He would normally have one grapefruit each morning but had recently increased his consumption to three per day so that the fruit would not be wasted. The patient was advised to discontinue eating grapefruit. His symptoms had resolved at follow-up 1 month later.
A 40-year-old HIV-infected African American with a history of bipolar disease and substance abuse had been receiving a stable treatment regimen of 300 mg zidovudine twice daily, 150 mg lamivudine twice daily, and 800 mg indinavir every 8 h since May 1998. In March 2000, his antiretroviral regimen was changed to 300 mg zidovudine, 150 mg lamivudine, 800 mg indinavir, and 200 mg ritonavir, all taken twice daily. Other medications included 20 mg fluoxetine daily, 200 mg trazodone at bedtime, 1200 mg lithium daily, 750 mg valproic acid daily, and 30 mg temazepam at bedtime.
Eight weeks later he returned to the clinic complaining of nausea, abdominal pain, diarrhea, insomnia, tachycardia, mania, and muscle twitching for approximately 6 to 8 weeks. The trazodone was discontinued and promethazine was given for nausea and sleep. After 4 weeks the mania, tachycardia, and myoclonus had resolved.
Serotonin syndrome is an important and potentially fatal complication of psychopharmacologic therapy resulting from serotonin hyperstimulation [10,11]. Sternbach's definition of the serotonin syndrome requires the development of three or more of the following clinical features coincident with the addition of or increase in a known serotonergic agent to an established medication regimen: mental status changes, agitation, myoclonus, hyper-reflexia, diaphoresis, shivering, tremor, diarrhea, lack of co-ordination, and fever (Table 1) . Other symptoms identified include nausea, abdominal pain, anxiety, dyspnea, hyper- and hypotension, and seizures. The syndrome may appear with different degrees of severity and is probably under-reported because it is not recognized.
Serotonin syndrome is generally the result of a pharmacodynamic interaction from concurrent administration of two or more medications that increase the central nervous system (CNS) serotonin levels by different mechanisms [e.g. selective serotonin reuptake inhibitors (SSRIs), tricyclic antidepressants, and monoamine oxidase inhibitors][10,12]. It also occurs as a result of a pharmacokinetic drug interaction, when drugs that inhibit SSRI metabolism are added to SSRI-containing regimens [13,14]. The cases described here were probably the result of inhibition of SSRI metabolism by the protease inhibitors, non-nucleoside reverse transcriptase inhibitors, and grapefruit juice (Table 2).
Fluoxetine (Prozac) is metabolized primarily by P450 2D6 to an active metabolite, norfluoxetine, which is further metabolized by 2D6. Other isoenzymes involved in the metabolism of fluoxetine include 2C9, 2C19, and 3A4. A deficiency in 2D6 has been identified in 8.5% of African Americans, 5–8% of white people, and 2–10% of Asians.  In such cases, metabolic shunting would increase the role of these alternative isoenzymes and prolong the half-lives of fluoxetine and norfluoxetine .
The P450 inhibitors involved in this series are known to inhibit isoenzyme 3A4, but only ritonavir is known to also markedly inhibit 2D6 at therapeutic doses. Although Ouellet et al. found that fluoxetine and norfluoxetine increase the ‘area under the curve’ of ritonavir, the effect of ritonavir on fluoxetine levels was not studied.  Efavirenz may also inhibit isoenzymes 2C9 and 2C19, but inhibits 2D6 only at concentrations greater than those achieved clinically. Grapefruit and grapefruit juice inhibit cytochrome P450 CYP3A4 in the wall of the small intestine. A large amount must be ingested for it to affect other drugs, as was the case with our patient. We agree with Michalets who concluded that there is a need for definitive in vivo drug interaction studies and continued patient reporting by clinicians, as in vitro data are not always consistent with in vivo experience and many variables influence patient outcomes .
All of the cases described involved fluoxetine. Fluoxetine may be more prone to cause the serotonin syndrome than other SSRIs because of its long half-life, widespread use, and its effects on dopamine reuptake [11,19].
Certain issues in the cases should be discussed. In the first case the patient was prescribed fluconazole, which is a 2C9 substrate, but he did not manifest signs of the serotonin syndrome until the introduction of ritonavir. The symptoms demonstrated by Case 2 could be the result of both serotonin syndrome and bupropion toxicity. Serotonin excess may have resulted from a pharmacokinetic interaction between fluoxetine and ritonavir, and a pharmacodynamic interaction between fluoxetine and bupropion. Moreover, bupropion toxicity can manifest as mania. The case emphasizes the necessity of obtaining a complete medication history from all patients, particularly when multiple prescribers are involved. In Case 3, one of the drugs implicated in the drug–drug interaction is efavirenz. CNS effects are commonly experienced early in the course of efavirenz treatment, but .the duration of symptoms (4 weeks) and the presence of diaphoresis distinguished the serotonin syndrome from early adverse efavirenz effects. In addition, the inhibition of fluoxetine and norfluoxetine metabolism by efavirenz is plausible if the patient is deficient in 2D6.
Serotonin syndrome is a clinical diagnosis. Therapeutic efficacy and toxicity of SSRIs are not related to serum concentrations, so levels are not monitored. As noted earlier, Sternbach's criteria for the serotonin syndrome include a temporal relationship between an increase in a known serotonergic agent to an established medication regimen. We would emphasize that an increase in levels of certain serotonergic agents occurs after a P450-inhibiting compound is ingested.
When adding P450 inhibitors to SSRI-containing regimens, recommendations vary whether to empirically reduce the SSRI dose or simply monitor for signs of serotonin excess. Penzak and colleagues recommend monitoring for antidepressant side effects when initiating ritonavir-containing regimens . Others suggest that when adding ritonavir the SSRI dose should be reduced by one-half to two-thirds, and then adjusted again as necessary, depending on response [8,9]. No SSRI dose adjustments are suggested when adding antiretrovirals other than ritonavir.
HIV and mental health disorders commonly occur concurrently. Treatment of depressive disorders in those with HIV infection can improve adherence and limit disease progression. All together, medication regimens for those with HIV disease are becoming more complex. Interactions between the HIV medications, mental health medications, and food or nutrients are likely. Proper identification of the adverse effects caused by drug interactions should allow appropriate management via dose reduction of the altered drug rather than incorrectly attributing the symptoms to the antiretroviral itself or pursuing an unnecessary diagnostic work-up. This should ultimately lead to better tolerability, medication adherence, and optimal patient outcomes. In vivo drug interaction studies between antiretrovirals and antidepressants are needed.
1. United States Department of Health and Human Resources. Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. February 1, 2001.http://www.hivatis.org/guidelines/adult
2. Carpenter CC, Cooper DA, Fischl MA. et al
. Antiretroviral therapy in adults: updated recommendations of the International AIDS Society–USA Panel. JAMA 2000, 283: 381–390.
3. Flexner C. HIV-protease inhibitors. N Engl J Med 1998, 338: 1281–1292.
4. Piscitelli SC, Flexner C, Minor JR. et al
. Drug interactions
in patients infected with the human immunodeficiency virus. Clin Infect Dis 1996, 23: 685–693.
5. Tseng AL, Foisy MM. Management of drug interactions
in patients with HIV. Ann Pharmacother 1997, 31: 1040–1058.
6. Penzak SR, Reddy YS, Grimsley SR. Depression
in patients with HIV infection. Am J Health-Syst Pharm 2000, 57: 376–389.
7. Friedland GH. Issues in HIV therapy in ‘triple-diagnosed’ patients. Int AIDS Soc – USA 2000, 8: 14–17.
9. Tseng AL, Foisy MM. Significant interactions
with new antiretrovirals and psychotropic drugs. Ann Pharmacother 1999, 33: 461–473.
10. Sternbach H. The serotonin syndrome
. Am J Psych 1991, 148: 705–713.
11. Mills KC. Serotonin syndrome
. Am Fam Phys 1995, 52: 1475–1482.
12. Mitchell PB. Drug interactions
of clinical significance with selective serotonin reuptake inhibitors. Drug Saf 1997, 17: 390–406.
13. Lee DO, Lee CD. Serotonin syndrome
in a child associated with erythromycin and sertraline. Pharmacotherapy 1999, 19: 894–896.
14. Pollack PT, Sketris IS, MacKenzie SL. et al
. Delirium probably induced by clarithromycin in a patient receiving fluoxetine
. Ann Pharmacother 1995, 29: 486–488.
15. Cohen LJ, DeVane CL. Clinical implications of antidepressant pharmacokinetics and pharmacogenetics. Ann Pharmacother 1996, 30: 1471–1480.
16. Fjordside L, Jeppesen U, Eap CB, Powell K, Baumann P, Brosen K. The steroselective metabolism of fluoxetine
in poor and extensive metabolizers of sparteine. Pharmacogenetics 1999, 9: 55–60.
17. Ouellet D, Hsu A, Qian J. et al
. Effect of fluoxetine
on pharmacokinetics of ritonavir. Antimicrob Agents Chemother 1998, 42: 3107–3112.
18. Michalets EL. Update: clinically significant cytochrome P-450 drug interactions
. Pharmacotherapy 1998, 18: 84–112.
19. Finley PR. Selective serotonin reuptake inhibitors: pharmacologic profiles and potential therapeutic distinctions. Ann Pharmacother 1994, 28: 1359–1369.