Share this article on:

Efficacy and Durability of Nevirapine in Antiretroviral-Experienced Patients

Harris, Marianne

Section Editor(s): Cooper, David; Lange, Joep M. A.; Montaner, Julio S. G.

JAIDS Journal of Acquired Immune Deficiency Syndromes: September 2003 - Volume 34 - Issue - p S53-S58
Optimising HIV Treatment - Building on Experience; Based on A Symposium Held Friday 15 March 2002, Seville, Spain

The non-nucleoside reverse transcriptase inhibitor (NNRTI) nevirapine (NVP) has been used as a component of salvage therapy for patients who have experienced virologic failure while taking nucleosides and protease inhibitors (PIs). In spite of broad cross-resistance within the NNRTI class, NVP may also play a role in salvage therapy for patients who have experienced failure while taking NNRTIs. Another role for NVP in treatment-experienced patients is in so-called "PI switch" strategies, in which NVP is substituted for a PI within a virologically successful combination regimen. This paper reviews the clinical data that support the use of NVP in salvage regimens and PI switch strategies.

(Australia) (Cooper)

(The Netherlands) (Lange)

(Canada) (Montaner)

B. C. Centre for Excellence in HIV/AIDS and the Canadian HIV Trials Network, St Paul's Hospital, Providence Health Care, University of British Columbia, 1081 Burrard Street, Vancouver, V6Z 1 Y6, Canada

Address correspondence and reprint requests to Marianne Harris, B.C. Centre for Excellence in HIV/AIDS and the Canadian HIV Trials Network, St Paul's Hospital, Providence Health Care, University of British Columbia, 1081 Burrard Street, Vancouver, V6Z 1 Y6, Canada; e-mail:

Back to Top | Article Outline


Nevirapine (NVP) has been used as a component of salvage therapy for patients who have experienced virologic failure while taking nucleosides and protease inhibitors (PIs). In spite of broad cross-resistance within the non-nucleoside reverse transcriptase inhibitor (NNRTI) class, NVP may also play a role in salvage therapy for patients who have experienced failure while taking NNRTIs. Another role for NVP in treatment-experienced patients is in so-called "switch" strategies, where NVP is substituted for a PI within a virologically successful combination regimen.

Back to Top | Article Outline


Nevirapine after nucleoside failure

When NVP and PIs first became available, a combination including these two agents was an attractive option for salvage of patients with advanced HIV and previous exposure to nucleoside mono- and dual therapies. In one such salvage study, 22 patients with extensive prior nucleoside exposure and CD4 cell counts of less than 50 cells/mm3 (median 30 cells/mm3) were treated with NVP, indinavir (IDV) (800 mg every eight hours), and lamivudine (3TC) (1). One patient had previously received a PI (ritonavir) and two had received an experimental NNRTI (loviride), but the remaining 19 patients were naïve to PIs and NNRTIs. Fifteen patients had a previous AIDS-defining illness. The median plasma viral load (pVL) at baseline was 5.16 log10 copies/mL (range 2.58 to 5.84 log10 copies/mL). Four patients withdrew due to adverse drug-related events, two attributable to NVP (one nausea/vomiting and one rash) and two to IDV (one urinary frequency/nocturia and one nausea/vomiting). Three other patients withdrew for personal reasons unrelated to toxicity or lack of efficacy. Of the 15 patients still taking study medications at 24 weeks, 11 had pVLs <500 copies/mL and six had pVLs <20 copies/mL. CD4 cell counts increased by a median of 95 cells/mm3 after 24 weeks. In conclusion, standard doses of NVP, IDV, and 3TC in combination were generally safe and well-tolerated, with substantial virologic and immunologic effects in nucleoside-pretreated patients with advanced HIV infection.

Back to Top | Article Outline

NVP after nucleoside and PI failure in NNRTI-naïve patients

A number of studies have examined the use of NVP in salvage therapy after failure of initial nucleoside and PI combinations. One such study included 20 patients failing an IDV- or ritonavir-containing regimen (2). Of note, patients had had a detectable pVL for a median of 12.4 months while taking their PI regimen. The first 10 subjects were treated with nelfinavir, saquinavir soft gel capsule (SGC), abacavir, and four of these also received another new nucleoside. The second 10 patients received the same PIs, abacavir, and NVP. One patient in the first group but none in the NVP group had previously received an NNRTI (efavirenz). Median baseline CD4 cell counts were 293 and 288 cells/mm3 and pVLs were 4.50 and 4.24 log10 copies/mL in the nucleoside and NVP groups, respectively. Three subjects in the nucleoside group and one in the NVP group discontinued study drugs, the latter after 3 days due to diarrhea. After 24 weeks, the group which received NVP had significantly greater reduction in pVL (2.67 versus 0.39 log10 copies/mL, p = .02) and a significantly greater proportion of patients with pVL <500 copies/mL (seven out of nine versus one out of seven, p = .04). The proportions of patients with pVL <50 copies/mL at 24 weeks were not significantly different between groups in this small study (five out of nine for NVP versus one out of seven for nucleoside, p = .15). The investigators concluded that virologic suppression was possible after failure of an IDV- or ritonavir-containing regimen, and was more likely if NVP was included in the salvage regimen of NNRTI-naïve patients.

Patients failing nucleosides and one PI, but naïve to NNRTIs, were treated with a salvage regimen including NVP in a study primarily designed to examine the use of lopinavir/ritonavir (3,4). The 70 patients enrolled had median baseline pVLs of 4.0 log10 copies/mL (range 2.9-5.8) and CD4 cell counts of 349 cells/mm3 (range 72-807). They were randomized to receive one of two doses of lopinavir/ritonavir twice daily: 400/100 mg (n = 36) or 400/200 mg (n = 34). Lopinavir/ritonavir was substituted for the PI in the failing regimen for the first two weeks of the study, after which the regimen was changed and at least one new nucleoside and NVP were added. By week 144, 22 patients (31.4%) had discontinued study therapy, but only six of these discontinuations (8.6%) were due to drug-related adverse events (diarrhea in two; flatulence, rash, asthenia, and depression in one each). Overall, proportions of patients with pVL <50 copies/mL at 96 weeks were 63% by on-treatment (OT) and 49% by intent to treat (missing = failure) (ITT, M = F) analyses. At 144 weeks, the proportions of patients achieving the same outcome were 76% by OT and 44% by ITT analyses. The mean CD4 cell count at 144 weeks was 587 cells/mm3, an increase of 211 cells/mm3 from baseline. In summary, in PI-experienced, NNRTI-naïve patients, a combination of NVP, lopinavir/ritonavir and nucleosides was well-tolerated and had a potent and durable antiviral effect, with pVL suppression in approximately half of patients after 2 and 3 years.

Nevirapine was also studied as salvage therapy of patients failing a single PI in Study ACTG 373 (5). Fifty-six patients who had been treated with amprenavir monotherapy (n = 36) or amprenavir, zidovudine, and 3TC (n = 20) in a previous study received four-drug salvage therapy with NVP, IDV (1000 mg every 8 hours), stavudine (d4T), and 3TC. One patient had experienced intolerance while taking the amprenavir-containing regimen and the remainder had virologic failure. All but five were NNRTI naïve. Median baseline pVL before four-drug salvage was 4.2 log10 copies/mL (range <2.3-5.8) and baseline CD4 count was 346 cells/mm3 (range 36-975). Overall, the proportions of patients with pVL <500 copies/mL at 48 weeks were 78% as-treated (AT) and 59% by ITT analysis. All of the patients with previous NNRTI experience had virologic failure while taking the four-drug regimen. The investigators concluded that most subjects in this study who had taken amprenavir-based regimens achieved durable virologic suppression (1 year) while taking a four-drug regimen of NVP, IDV, d4T, and 3TC.

Back to Top | Article Outline

NVP after triple class failure

Patients who have been exposed to multiple combination regimens, often including drugs from all three available classes, present an increasing clinical challenge to the treating physician. One approach is multiple drug rescue therapy (MDRT) with five or more agents, also known as mega-HAART. This strategy has had remarkable success in treating patients for whom complete viral suppression is thought by some to be an unrealistic goal (6).

This MDRT strategy was studied in an observational fashion in 106 patients with prior failure of at least two prior regimens (7). Patients had a median prior exposure to antiretrovirals of 43 months and 38% had been exposed to NNRTIs. The rescue regimen, commenced between August 1997 and June 1998, consisted of five to nine drugs: up to four nucleosides, up to two PIs and up to two NNRTIs, with or without hydroxyurea. The median number of drugs in the MDRT regimen was five and most patients received NVP as a component of MDRT (79/106, 75%). The median baseline (pre-MDRT) pVL was 4.79 log10 copies/mL (interquartile range 4.18-5.34) and the CD4 cell count was 180 cells/mm3 (interquartile range 90-260). Fifty-six patients (59%) had four-fold or greater decreased susceptibility to seven or more antiretroviral drugs before starting MDRT. Six patients discontinued MDRT due to severe adverse drug events or laboratory toxicities. After 1 year, 48% of patients (AT analysis) and 40% of patients (ITT analysis) had pVLs <400 copies/mL, and 49% (52/106) had pVLs <400 copies/mL on at least two consecutive visits during the study. The pVL of <50 copies/mL on at least two consecutive visits was achieved in 33% (35/106) of patients. As expected, being NNRTI-naïve increased the likelihood of a response to MDRT (defined as pVL <400 copies/mL on two consecutive visits): 2.83× overall and 3.06× for patients who were prescribed an NNRTI as part of their regimen.

The same strategy has been applied more recently, but using newer drugs such as lopinavir/ritonavir (8). Sixty-eight patients commenced MDRT with a median of six drugs (lopinavir/ritonavir = two drugs) during the first 6 months of 2000. Median pre-MDRT pVL was 5.12 log10 copies/mL (interquartile range 4.27-5.66) and CD4 cell count was 130 cells/mm3 (interquartile range 40-200). Median prior antiviral exposure time was 32 months, and 56% of patients (38/68) were NNRTI-experienced. The MDRT regimen included NVP in 84% (57/68) of cases. After 1 year, the pVL was <50 copies/mL in 55% (AT analysis) and 44% (ITT analysis) of patients. Thus, it appears that availability of some of the newer antiviral drugs has improved the rate of response to MDRT, even when using the more sensitive HIV RNA assay (limit of detection = 50 copies/mL). These studies demonstrate that relying exclusively on previous treatment experience, and even resistance testing, is not sufficient to rule out an antiviral response to multiple drug regimens in patients who are able to tolerate them. Drugs such as NVP can be successfully recycled in heavily pretreated patients and undetectable pVL remains a viable therapeutic goal for many (9).

Back to Top | Article Outline

NVP-PI drug interactions

When NVP is combined with PIs, most commonly in the context of salvage therapy, drug interactions need to be considered. PIs do not significantly affect the pharmacokinetics of NVP (10). On the other hand, NVP, a cytochrome P450 (CYP) inducer, has the potential to decrease exposure to PIs. For example, the area under the plasma concentration curve (AUC) for IDV (given as 800 mg every 8 hours) is decreased by 28% when co-administered with NVP (11). This has led some to implement a dose increase of IDV to 1000 mg every 8 hours in this situation (5), although this may not be necessary (1,11).

The current practice of boosting PI levels with ritonavir, a potent CYP inhibitor, may overcome any clinically significant lowering of PI levels by NVP. Pharmacokinetic data for the effect of NVP on ritonavir-boosted PIs are limited, but for lopinavir 400 mg/ritonavir 100 mg twice daily, the mean ratio of AUC with NVP/AUC to without NVP is 0.99 (no effect = 1.00) (12). However, the 90% confidence intervals around this mean value are 0.74 and 1.32, leading to a recommendation to increase lopinavir/ritonavir dosing to 533/133 mg twice daily when given with NVP. Taking this precaution would ensure that patients with the greatest pharmacokinetic impact of NVP still achieve adequate lopinavir levels. This magnitude of pharmacokinetic variability between subjects is typical, and demonstrates the need for therapeutic drug monitoring to direct rational drug dosing in patients receiving these complex multiple drug regimens.

Back to Top | Article Outline

Summary of NVP in salvage therapy

Combinations including NVP have been used successfully in the treatment of patients experiencing virologic failure while taking nucleosides and protease inhibitors (Table 1). Success rates are lower in patients who have previously been exposed to NNRTIs, but even in the latter group, NVP may play a role as a component of multiple drug rescue therapy.



Back to Top | Article Outline


A number of groups have studied replacing NVP for the PI in the regimens of patients who have achieved pVL suppression while taking PI-based triple therapy. Potential advantages of such a switch strategy are treatment of PI-related side effects, prevention of PI-related long-term metabolic effects and simplification of the dosing regimen. Virologic outcomes and effects on dyslipidemias and abnormal body fat distribution have been examined in both uncontrolled and controlled studies; the latter studies have compared patients who have switched to NVP with those who have continued to take the same PI-based regimen.

Back to Top | Article Outline

Uncontrolled studies

Martinez et al. (13) followed 23 NNRTI naïve patients (11 women and 12 men) who switched to NVP because of perceived fat redistribution while taking one or more PIs. Patients' pVLs had been <200 copies/mL for at least 3 months (median 9 months) prior to the switch. Six patients were taking their first-line therapy and the remaining 17 had received nucleosides prior to receiving PIs. Virologic failure (pVL >200 copies/mL) was observed in only one patient (1/23, 4%) after 6 months of NVP therapy. Significant improvements were observed in fasting cholesterol, triglyceride and glucose plasma levels, as well as in the insulin resistance index at 6 months as compared with baseline. Subjective partial improvements in body fat distribution were reported by 21/23 patients (91%) after 6 months, particularly with regard to peripheral fat wasting. Waist-hip ratios decreased significantly (0.91 to 0.85, p = .048) but body mass index did not change over a 6-month period in the NVP-treated patients. Nevirapine was well-tolerated by all patients. The investigators concluded that PI-associated metabolic abnormalities may revert at least partially, while virologic suppression was preserved in most patients for 6 months after replacing the PI with NVP.

In another study by Raffi et al. (14), 73 NNRTI-naïve patients who had received a PI-based regimen for at least 12 months (mean 22 months), and had pVLs <400 copies/mL for at least 6 months (mean 18 months), had the PI replaced by NVP (n = 63) or efavirenz (n = 10). The nucleoside component of each patient's regimen was not altered. The PI regimen was first-line therapy in 31 patients, while the remaining 42 had received other antiretrovirals prior to their PI regimen. Patients were switched to an NNRTI because of one or more reasons: PI intolerance (gastrointestinal symptoms or renal colic, n = 21), lipodystrophy (n = 18), or adherence issues/wish to simplify (n = 43). Of note, CD4 cell counts were relatively high at the time of the switch to an NNRTI (median 473 cells/mm3, range 111-1504). After a median follow-up time of 52 weeks, 63/73 patients (86%) had undetectable pVL (<200 copies/mL in 8 and <80 copies/mL in 55). Virologic breakthrough at 12 months tended to be more common in patients who had received antiretroviral therapy prior to their PI regimen than in those who had not (19.2% versus 6.5%), but this difference was not significant (p = .10). Nevirapine was discontinued due to adverse effects in 4/63 patients (6%) (two myalgias, one localized cutaneous bullous lesions, one transaminase elevation). The authors concluded that switching from a PI to an NNRTI was virologically successful in the majority of patients who had a good immunologic response and undetectable pVL for at least 6 months during first-line treatment with highly active antiretroviral therapy.

Back to Top | Article Outline

Controlled studies

The safety and efficacy of a switch from a PI to NVP has been prospectively evaluated in two controlled studies. Barreiro et al. (15) randomized 138 NNRTI-naïve patients who had pVLs <400 copies/mL for at least 6 months (mean 9 months) to either switch to a NVP-based (n = 104) or continue to take their PI-based regimen (n = 34). In an ITT analysis, pVL rebound occurred during the first 6 months after switching to NVP in 11% of patients as compared with 29% of those who continued to take a PI (p = .007). Fasting lipid profiles tended to normalize in both the PI- and NVP-treated patients, but significant changes were not observed between baseline and month 6 in either arm. Half of the patients who switched to NVP subjectively reported at least partial improvement in body fat distribution, while those taking their PI-based regimen reported no changes. Body fat distribution was not assessed objectively. Tolerance of the switch to NVP was not described. Most patients who switched to NVP reported an improvement in their quality of life and their scores on the McGill Quality of Life questionnaire were significantly better than the PI-treated patients (9.1/10 versus 4.4/10, p < .01). The investigators concluded the replacement of a PI by NVP was virologically and immunologically safe over a 6-month period, with significant improvements in quality of life and subjective improvements in lipodystrophic body shape abnormalities (in half of the patients studied).

A randomized PI switch study involving 106 NNRTI-naïve patients was conducted by Ruiz et al. (16). Patients with clinical lipodystrophy, who had had pVLs <400 copies/mL for at least 6 months while taking a PI-based regimen, were enrolled. Viral load results at 48 weeks were equivalent in the 52 patients who were randomized to switch to NVP and the 54 who were randomized to continue PI-based therapy: 79% versus 77% <400 copies/mL, and 74% versus 72% <50 copies/mL, for the NVP and PI arms, respectively. Treatment discontinuation rates were also similar: 17% in the NVP arm and 19% in the PI arm. Nevirapine was discontinued due to adverse events in six patients: two out of three who had rash and four out of six who had hepatitis (all of whom were co-infected with hepatitis C).

Fasting cholesterol and triglyceride plasma levels improved at week 48 as compared with baseline in the NVP-treated but not in the PI-treated group. However, there were no significant differences in the fasting cholesterol and triglyceride plasma levels between the two groups at week 48. Lipodystrophy-related body shape changes were measured by anthropometric measurements (skinfolds and waist, hip, leg and mid-arm circumference) and DEXA scans at baseline and 48 weeks. Abdominal skinfold decreased in the NVP- but not in the PI-treated group. Otherwise, no significant changes or differences between groups were observed in any of these parameters over the 48-week study period. Quality of life reports were better in the NVP group than in the controls at 48 weeks, which was mainly attributed to the greater simplicity of the NVP-based drug regimen. The conclusions of the study were that a switch from PI- to NVP-based triple therapy was as effective as continuing a PI regimen in sustaining virologic suppression and immunologic responses at 48 weeks of follow-up. Metabolic parameters improved in patients receiving NVP, but significant changes in lipodystrophic body shape abnormalities were not objectively demonstrated 1 year after substituting NVP for the PI. The simpler dosing schedule of the NVP regimen facilitated adherence and improved quality of life in the patients studied.

Back to Top | Article Outline

Rash rates in PI switch studies

Rash is the most common side effect of NVP, occurring with a frequency of up to 17% in clinical trials and resulting in drug discontinuation in up to 7% (10). Interestingly, observed rash rates are much lower in studies which involved PI replacement with NVP when patients had undetectable pVLs (17). In switch studies that reported adverse events, rash occurred in 10/240 patients, or 4.2%. In switch studies describing reasons for discontinuation, there were six discontinuations due to rash in 353 patients, representing a frequency of 1.7%. The reasons for this lower rash rate are unknown, but it appears that NVP-related rash may occur less frequently in patients who start the drug when their pVL is already suppressed or who have already had induction of their CYP system.

Back to Top | Article Outline

Summary of PI switch studies

Replacing a PI with NVP in patients who have stable undetectable pVL while taking a triple combination regimen results in sustained virologic suppression in 74-86% of patients a year after the switch. In randomized, controlled trials, switching to NVP has been shown to be equivalent (Ruiz et al. [16], at 1 year) or superior (Barreiro et al. [15], at 6 months) to continuing to take a PI-based regimen in terms of sustained virologic suppression. Metabolic indices including fasting lipid profiles have been shown to improve 6 or 12 months after the switch to NVP in some studies, but remained unchanged and equivalent to those observed in patients who continued to take their PI-based regimens. Body fat distribution abnormalities may improve subjectively but objective improvements have not been demonstrated. Rates of rash (4.2%) and discontinuations due to rash (1.7%) are relatively low in patients who switch to NVP when their pVL is undetectable. The strategy of replacing a PI with NVP is most likely to be safe and effective in patients with prolonged suppression of viral replication and good CD4 cell responses while taking first-line, PI-based therapy.

Back to Top | Article Outline


1. Harris M, Durakovic C, Rae S, et al. A pilot study of nevirapine, indinavir, and lamivudine among patients with advanced Human Immunodeficiency Virus disease who had previously failed combination nucleoside therapy. J Infect Dis 1998;177:1514-1520.
2. Deeks SG, Hellmann NS, Grant RM, et al. Novel four-drug salvage treatment regimens after failure of a Human Immunodeficiency Virus Type 1 protease inhibitor-containing regimen: antiviral activity and correlation of baseline phenotypic drug susceptibility with virologic outcome. J Infect Dis 1999;179:1375-1381.
3. Feinberg J, Brun S, Xu Y, et al. Durable suppression of HIV+ RNA after two years of therapy with ABT-378/ritonavir (ABT-378/r) in single protease inhibitor experienced patients [abstract P101]. AIDS 2000;14(suppl 4):S46.
4. Hicks C, Brun S, King M, et al. Lopinavir/ritonavir therapy in single protease inhibitor experienced patients: 144-week followup [abstract 220]. Presented at the 8th European Conference on Clinical Aspects and Treatment of HIV Infections, Athens, Greece, 2001.
5. Gulick RM, Smeaton LM, D'Aquila RT, et al. Indinavir, nevirapine, stavudine, and lamivudine for Human Immunodeficiency Virus-infected, amprenavir-experienced subjects: AIDS Clinical Trial Group Protocol 373. J Infect Dis 2001;183:715-721.
6. Deeks SG, Martin JN. Reassessing the goal of antiretroviral therapy in the heavily pre-treated HIV-infected patient. AIDS 2001;15:117-119.
7. Montaner JSG, Harrigan PR, Jahnke N, et al. Multiple drug rescue therapy for HIV-infected individuals with prior virologic failure to multiple regimens. AIDS 2001;15:61-69.
8. Harris M, Yip B, Hogg R, et al. Multiple drug rescue therapy (MDRT) with and without lopinavir/r (LPV/r) [abstract 114]. Presented at the 1st IAS Conference on HIV Pathogenesis and Treatment, Buenos Aires, Argentina, 2001.
9. Montaner JSG, Harris M, Harrigan R, Hogg R, Wood E. A compromise strategy for patients with multiple drug failure (letter). AIDS 2001;15:2470.
10. Viramune Product Monograph, Version 3.0, Boehringer Ingelheim International GmbH.
11. Murphy RL, Sommadossi J-P, Lamson M, et al. Antiviral effect and pharmacokinetic interaction between nevirapine and indinavir in persons infected with Human Immunodeficiency Virus Type 1. J Infect Dis 1999;179:1116-1123.
12. Kaletra Product Information, Abbott Laboratories, North Chicago, IL, U.S.A., 2000.
13. Martinez E, Conget I, Lozano L, Casamitjana R, Gatell JM. Reversion of metabolic abnormalities after switching from HIV-1 protease inhibitors to nevirapine. AIDS 1999;13:805-810.
14. Raffi F, Bonnet B, Ferre V, et al. Substitution of a nonnucleoside reverse transcriptase inhibitor for a protease inhibitor in the treatment of patients with undetectable plasma Human Immunodeficiency Virus Type 1 RNA. Clin Infect Dis 2000;31:1274-1278.
15. Barreiro P, Soriano V, Blanco F, et al. Risks and benefits of replacing protease inhibitors by nevirapine in HIV-infected subjects under long-term successful triple combination therapy. AIDS 2000;14:807-812.
16. Ruiz L, Negredo E, Domingo P, et al. Antiretroviral treatment simplification with nevirapine in protease inhibitor-experienced patients with HIV-associated lipodystrophy. JAIDS 2001;27:229-236.
17. Cotton G. Sustained virologic suppression in subjects switched from protease inhibitors to nevirapine [abstract WePeB4198]. Presented at the XIII International AIDS Conference, Durban, South Africa, 2000.
Back to Top | Article Outline

Section Description

This publication has been made possible by an educational grant from Boehringer Ingelheim.


Nevirapine; Salvage regimen; PI switch strategy; Antiretroviral therapy

Copyright © 2003 Wolters Kluwer Health, Inc. All rights reserved.