Sex Differences in Nevirapine Rash and Hepatitis

Last year, the AIDS world was rocked when news came that the South African Medicines Control Commission had halted a study of an experimental HIV drug after two women in the trial died and several others developed severe liver disease. Coming in the months leading up to the International AIDS Conference in Durban, this episode catapulted the problems of bringing HIV treatments to South Africa into the international spotlight.

The experimental drug in the trial was FTC, a compound in the same therapeutic class as AZT and ddI. But blame for the women's deaths came to rest on one of the other drugs the women received in their combination cocktail: Nevirapine, an approved, non-nucleoside reverse transcriptase inhibitor already at the center of another raging controversy about access to treatment in South Africa.

A Breakthrough for Treatment in Africa?

A widely hailed study (HIVNET 012) reported in 1999 that, for only four U.S. dollars, a single dose of nevirapine (Viramune) given to a pregnant woman about to give birth and another dose to her baby could dramatically reduce the chances that the infant would be infected during labor and delivery. For the first time it seemed a grip had been gained on the intractable problem of treating HIV in poor countries. Nevirapine for prevention of mother-to-child HIV transmission was effective, it was practical and it was relatively affordable. But was it safe? South African government officials, reluctant to spend scarce resources on saving the babies of HIV-infected mothers, claimed there was too little known about the side effects of nevirapine to make it widely available. Ominously, the FTC trial deaths seemed to back them up.

A close look at the experience with single-dose nevirapine finds little cause for concern. In the HIVNET 012 study, 310 pregnant women received single-dose nevirapine and another 308 women received AZT just before giving birth. The rates of adverse events and drug-related toxicity were observed to be similar in the groups. The incidence of laboratory abnormalities was also similar in each group. Mild to moderate rash was reported for four mothers in the nevirapine group and five in the AZT group. Among the babies, nine in each group had a mild or moderate rash. No serious rashes were reported for any participants.1 In another large trial of nevirapine involving 652 pregnant mothers in South Africa who each received two doses of the drug, there were no episodes of liver toxicity or serious rash reported.2

Because no placebos were used in these trials, one cannot conclude that the risks of adverse events are any greater or lesser for mothers who do not receive antiretroviral drugs just before birth. And the long-term effect upon the development of treated children can only be determined by years of follow-up. But the low rates of serious adverse events observed suggest that these treatment regimens are safe, at least in the short term. So safe, in fact, that the HIVNET authors proposed that presumptive treatment of all pregnant women with single-dose nevirapine may be justified in settings where HIV testing and pre-test counseling are not affordable.

But what caused the deaths and liver disease in the FTC trial and was nevirapine responsible? In a complicated stratified trial design, participants received FTC or lamivudine (3TC, Epivir) plus, depending on their baseline viral load, either nevirapine or efavirenz (Sustiva/Stocrin, another drug in the same class as nevirapine). Participants also received stavudine (d4T, Zerit) as a background nucleoside analog to round out the cocktail. All participants were monitored for elevated liver enzyme levels, the standard laboratory markers for the development of hepatitis. There were no reported differences in the incidence of severe (Grade 4) liver enzyme elevations between those who received 3TC and those who received FTC. Nor were any differences observed in these enzyme levels between blacks and non-blacks in the study. But of 385 patients receiving nevirapine in the trial, 36 (9%) had Grade 4 liver enzyme elevations while no such liver enzyme abnormalities were observed in 83 patients receiving efavirenz. This clearly pointed the finger at nevirapine. And, as we later learned from a report at the Fifth International Congress on Drug Therapy in HIV Infection in Glasgow, Scotland last November, women in the trial were twice as likely as men to experience a Grade 4 elevation of liver enzymes during the first four weeks of taking nevirapine. This is compelling evidence from a randomized trial that there may be a sex difference in the likelihood of liver disease for those who receive nevirapine.3

PEP Tragedies

About the same time as the nevirapine crisis was developing in South Africa, in the United States an HIV-negative African-American healthcare worker suffered a needle-stick accident as she was drawing blood from an HIV-infected patient. She was placed on a post-exposure prophylaxis (PEP) protocol containing AZT, 3TC and nevirapine. Fourteen days after starting the drugs, she developed malaise, fatigue, fever and chills. At 20 days, the HIV drugs were halted after her liver enzyme levels increased dramatically above the normal range. A week later she developed liver failure and went into a coma. She subsequently received a liver transplant, recovered and remained HIV-negative six months after the accident.4

Two more tragic case reports involving post-exposure prophylaxis were presented at the Glasgow conference. Two women who had been raped were treated with a PEP regimen containing nevirapine. Both women required liver transplants after liver disease developed within a month on the regimen.5 And in yet another needle-stick accident, a female health-care worker in Chicago complained of nausea and anorexia eight days after starting a PEP regimen containing nevirapine. She subsequently developed a severe rash despite discontinuing the drug after the first symptoms occurred.6 Due to these and other cases, the U.S. Centers for Disease Control (CDC) in January 2001 warned against the routine use of nevirapine in post-exposure prophylaxis regimens.7

Background to the Warnings

January 2001 marked the tenth anniversary of the first use of nevirapine in humans. The problems with skin rash and liver enzyme abnormalities were detected during the earliest trials of the drug. These eventually led to manufacturer recommendations that patients start nevirapine treatment at a lower dose for two weeks until it was clear the full dose could be safely tolerated. After several years of clinical experience with nevirapine and a mounting number of serious adverse reactions, culminating in the South African deaths, the European Medicines Evaluation Agency (EMEA) issued a strong public statement in April of 2000 drawing sharp attention to the risk of liver disease. The statement stressed the importance of following the two-week lead-in period, intensive monitoring during the first eight weeks of therapy and immediate discontinuation of nevirapine if certain symptoms occur. Of particular concern are symptoms of severe rash, hypersensitivity reaction or hepatitis, including elevated liver enzymes.8

These warnings were strengthened six months later when the manufacturer sent a letter to U.S. health care professionals extending the critical initial period of close monitoring to 12 weeks. Yet, as noted in that warning, approximately one-third of adverse events have been reported to occur after 12 weeks.9

Despite the strength of these warnings to the medical community, despite the doubled rate of hepatitis among women in the FTC trial, and despite an apparent preponderance of case reports about severe and fatal adverse events involving women, there have been no official cautions about any particular danger for women starting nevirapine.

Sex Differences in Nevirapine Rash

Recently, additional evidence was published that says women may have more frequent skin-related adverse events when starting nevirapine. An article in Clinical Infectious Diseases (January 2001) reports on a retrospective cohort study that finds likely significant differences in the incidence of nevirapine-associated rash between men and women.10

The researchers examined the medical records of all patients from three large U.S. HIV clinics who received nevirapine for at least 30 days during the period from 1993 through September 1998. Patients who discontinued nevirapine due to severe skin rash prior to 30 days on treatment were also included. The patients studied had been treated in clinics at Washington University in St. Louis, the University of North Carolina in Chapel Hill and at UCSF in San Francisco. The primary outcome of interest was a report of severe (Grade 3 or 4) rash during the first 90 days of nevirapine treatment. A secondary outcome was discontinuation of nevirapine due to any degree of rash.

The search of medical records found 358 individuals (95 women; 263 men) from the three sites that met the entry criteria for nevirapine use. The overall incidence of severe rash during the first 90 days after starting nevirapine was 3.4 percent. But women were eight times more likely to have had severe rash than men, with an incidence of 9.5 percent compared to 1.1 percent for the men. Also, women were five times more likely to have discontinued nevirapine due to any degree of rash than were men.

When any rash (from mild to severe) was counted, about 16 percent of women and 8 percent of men were affected. Two women and two men experienced Stevens-Johnson syndrome, a life-threatening form of severe rash in which large sheets of skin die and peel off. Because of the smaller number of women treated with nevirapine, the two cases represent a larger proportion of women with Stevens-Johnson syndrome than men. There were apparently no deaths reported due to nevirapine in this sample. The incidences of elevated liver enzymes and liver disease were not collected.

While the sex of the patient was strongly associated with the development of severe rash in this study, other possible factors were also identified. Individuals with higher CD4 counts when starting nevirapine tended to have more episodes of severe rash and had to stop therapy due to rash more often. Patients younger than 35 years tended to discontinue nevirapine more often than older patients but did not have significantly more cases of severe rash. Of 23 women who took birth control hormones, 4 (17%) experienced severe rash compared to 5 of 72 (7%) women not taking contraceptive hormones. In this population, 76 percent of the women were African-American compared with 43 percent of the men. The authors report, however, that no association was observed between race and the incidence of rash in the study.

Rash Decisions

Although the key clinical trials that led to the approval of nevirapine uncovered a risk of having skin reactions when starting the drug, none of those studies examined the differential risk between men and women. One company-sponsored report also presented at the Glasgow conference last year described an early, large (2,249 patient) trial comparing nevirapine with placebo. The authors reported an equal incidence of elevated liver enzymes in each comparison arm and a greater number of liver-related deaths in the placebo arm. Overall, women comprised 20 percent of that study population although, once again, the adverse event rates for men and women were not reported separately.11

There is a small amount of provocative evidence that women may experience more adverse reactions to medications in general. Though not HIV-specific, two surveys of hospital inpatients performed in the 1970s and 1980s found a generally increased risk of drug-related skin reactions among women, reporting an incidence in women 35 to 50 percent higher than in men.12, 13 But the authors of the nevirapine sex differences study caution that, while the differences they found in the incidence of severe rash between men and women are statistically significant, the overall number of episodes of rash in their study are small and could possibly be due to other, unexamined, factors. Another limitation of the study is that the observations were not pre-planned but were collected from medical records after the fact. Retrospective studies of this sort can contain undetected biases if, for example, different staging systems for rash were used at different times or at different sites. One limitation in particular is a lack of information about any additional medications patients may have been taking while they were receiving nevirapine.

Despite these caveats, the authors conclude, it is at least safe to say that the risk of nevirapine rash cannot simply be extrapolated between sexes, as is commonly done. This is a particular problem since our knowledge of adverse reaction rates is drawn from study populations with small proportions of women. As demonstrated by the sex-differences study, overall incidence rates of rash may be similar to the rates for men, but rates for women may appear dramatically higher when they are isolated. Since no previously published study of nevirapine has differentiated adverse reaction rates by sex, clinicians and patients have no choice but to judge expected rash episodes by published overall incidence rates or by anecdotal experience.

Although most nevirapine-associated deaths have been liver-related, the sex-differences study only reported on the relative incidence of rash. Yet, mild to moderate rash may be one of the warning symptoms of liver disease. In the South African FTC trial, rash was temporally associated with Grade 4 liver enzyme elevation in 25 percent of cases. If this association is common, then an increased incidence of rash may possibly warn of an increased risk for liver disease among women starting nevirapine. Compounding the risk for black women, it's been suggested that the appearance of rash on darker skin may be more difficult to diagnose than on lighter skin.14 Add to this an increasing trend among physicians to "treat through" mild nevirapine rash when initiating therapy, and the need for close clinical management and frequent monitoring of liver enzymes becomes crucial.

The evidence of sex differences in the incidence of drug-associated rash and liver disease suggests that women, with whom most U.S. clinicians probably have less overall experience, need vigilant monitoring and extra counseling about possible treatment-associated adverse events -- including rash -- when initiating an antiretroviral regimen containing nevirapine.


  1. Guay LA, et al. Intrapartum and neonatal single-dose nevirapine compared with zidovudine for prevention of mother-to-child transmission of HIV-1 in Kampala, Uganda: HIVNET 012 randomized trial. The Lancet. 1999 254:795-802.

  2. Moodley D, et al. Evaluation of safety and efficacy of two simple regimens for the prevention of mother to child transmission (MTCT) of HIV infection: nevirapine vs lamivudine and zidovudine used in a randomised clinical trial (the SAINT study). (Abstract TuOrB356) XIII International AIDS Conference, Durban, South Africa, July 9-14, 2000.

  3. Sanne I, et al. Severe liver toxicity in patients receiving two nucleoside analogues and a non-nucleoside reverse transcriptase inhibitor. (Abstract) Fifth International Congress on Drug Therapy in HIV Infection, Glasgow, UK. AIDS. 2000;14(supplement):PL 9.3.

  4. Sha BE, et al. Adverse effects associated with use of nevirapine in HIV postexposure prophylaxis for 2 health care workers. (Research Letter) JAMA. 2000;284:21.

  5. Henderson D. Post-exposure prophylaxis. (Abstract) Fifth International Congress on Drug Therapy in HIV Infection, Glasgow, UK. AIDS. 2000;14(supplement):PL 6.1.

  6. Johnson S, et al. Adverse effects associated with use of nevirapine in HIV postexposure prophylaxis for 2 health care workers. (Research Letter) JAMA. 2000;284:21.

  7. Centers for Disease Control and Prevention (CDC). Serious adverse events attributed to nevirapine regimens for postexposure prophylaxis after HIV exposures -- Worldwide, 1997-2000. MMWR. 2001 Jan 5.

  8. European Agency for the Evaluation of Medicinal Products. EMEA Public Statement on Viramune. April 2000; EMEA/11260/00.

  9. Roxane Laboratories. Important Drug Warning Re: Severe, life-threatening and fatal cases of hepatotoxicity with Viramune. November 2000.

  10. Bersoff-Matcha SJ, et al. Sex Differences in Nevirapine Rash. Clinical Infectious Diseases. 2001;32:124-9. (This study was first presented as a poster at the 6th Conference on Retroviruses and Opportunistic Infections in 1999).

  11. Cahn P, et al. Hepatic safety with nevirapine (NVP) and two nucleosides in patients with advanced HIV infection, from a placebo (PBO) controlled clinical endpoint trial (1090). (Abstract) Fifth International Congress on Drug Therapy in HIV Infection, Glasgow, UK. AIDS. 2000;14(supplement):PL 8.6

  12. Arndt KA, Jick H. Rates of cutaneous reactions to drugs: a report from the Boston Collaborative Drug Surveillance Program. JAMA. 1976;235:918-22.

  13. Bugby M, et al. Drug induced cutaneous reactions: a report from the Boston Collaborative Drug Surveillance Program on 15,438 consecutive inpatients. JAMA. 1986;256:3358-63.

  14. Gilden D. Liver Failures Spark Nevirapine Warning. (Statement by Franck Rosseau, V.P. for Medical Affairs, Triangle Pharmaceuticals, sponsor of the FTC trial). The amfAR Treatment Insider. August 2000;1(4).

But Is It Really, Really Safe?

Although no severe rashes or liver problems have been reported for single-dose nevirapine given to pregnant women, the original Phase I study of nevirapine in pregnant women may be worth taking a closer look at. In a trial conducted at seven hospitals in the United States and Puerto Rico, seventeen women in active labor were given single doses of nevirapine (either 100mg or 200mg) when they arrived at the hospital. The women had previously been enrolled in the study and were receiving a standard course of zidovudine to prevent HIV transmission to their infants. Blood samples were drawn at frequent intervals after the nevirapine dose and daily after delivery. The levels of nevirapine in the blood samples were charted to find the peak concentration and the rate of clearance from the body.

Nevirapine has a relatively long life in the blood (ideal for preventing mother-to-child transmission) but the amount of time nevirapine remained in the bodies of these pregnant women varied dramatically -- both between the women on the study and from historical data in men and non-pregnant women. One woman, who received a 100mg dose, still had more than half the nevirapine in her blood after seven days. Considering that one of the needle-stick victims who started prophylactic nevirapine (see PEP Tragedies above) developed symptoms within 8 days, it's not unimaginable that there may eventually be a rare adverse event resulting from single-dose nevirapine. Yet, this scenario is highly unlikely because, as in the needle-stick case, symptoms have only developed after several consecutive days of daily dosing.

If nevirapine is slower to clear from women's bodies, whether pregnant or not, then the increased incidence of rash and liver problems among women on daily-dosing regimens may be due to a gradual accumulation of the drug to toxic levels. The wide variability of drug concentrations seen in pregnant women and reports of slower clearance of nevirapine from non-pregnant women point to the need for additional pharmacokinetic studies of nevirapine in women.


Mirochnick M, et al. Pharmacokinetics of Nevirapine in Human Immnodeficiency Virus Type 1-Infected Pregnant Women and Their Neonates. Journal of Infectious Diseases. 1998;178:368-74.

Lamson MJ, et al. Effects of gender on the single and multiple dose pharmacokinetics of nevirapine. Pharm Res. 1995; 12:S-101.

Representation of Women in Key Clinical Trials of Nevirapine
Clinical Trial Investigator, Journal, YearTotal NVP EnrollmentNumber of Women Enrolled
De Jong, JID 1994100
Havlir, JID 1995210
Havlir, JID 1995215
Cheeseman, JAIDS 1995622
D'Aquila, Annals 199619934
De Jong, JID 1997200
Henry, JAIDS 199833041
Montaner, JAMA 1998988
Floridia, JAIDS 1999275
Barreiro, AIDS 200010414
Podzamczer, 7th Retro 2000~72~19
Cahn, Glasgow 2000~1125~234
~ denotes estimation

Back to the GMHC Treatment Issues January 2001 contents page.