Share this article on:

The Interaction of CD4 T-Cell Count and Nevirapine Hepatotoxicity in China: A Change in National Treatment Guidelines May Be Warranted

Zhang, Chengda*; Wang, Wei MD*; Zhou, Mengyu*; Han, Yang MD*; Xie, Jing MD, PhD*; Qiu, Zhifeng MD, PhD*; Guo, Fuping MD, PhD*; Li, Yanling MD*; Wang, Huanling MD, PhD*; Ghanem, Khalil G. MD; Li, Taisheng MD, PhD*

JAIDS Journal of Acquired Immune Deficiency Syndromes: April 15th, 2013 - Volume 62 - Issue 5 - p 540–545
doi: 10.1097/QAI.0b013e3182845cc7
Clinical Science

Objectives: Nevirapine (NVP), a still widely used nonnucleoside reverse transcriptase inhibitor, can cause severe hepatotoxicity. Previous studies suggest that CD4 cell counts more than 250 cells per microliter in women and more than 400 cells per microliter in men are risk factors for NVP-related hepatotoxicity. These studies have informed Chinese national treatment guidelines. We evaluate whether current Chinese guidelines for NVP use are appropriate.

Methods: Longitudinal data were pooled from 2 clinical trials between 2005 and 2009 across mainland China. Five hundred sixty-six antiretroviral therapy–naive Chinese patients were given NVP-containing antiretroviral therapy for 24 weeks. Hepatotoxicity was defined as alanine aminotransferase, aspartate transaminase, or total bilirubin level greater than 1.25 times the upper limit of normal range. Severe hepatotoxicity was defined as greater than 5 times the upper limit of normal range.

Results: One hundred ninety-seven (36.1%) patients developed hepatotoxicity during treatment, including 42 (7.7%) patients with severe hepatotoxicity. CD4 cell count more than 250 cells per microliter was an independent predictor for hepatotoxicity both in men [relative risk = 1.22 (95% confidence interval: 1.04 to 1.44)] and in women [relative risk = 1.72 (95% confidence interval: 1.20 to 2.46)]. Severe hepatotoxicity was also more common among all persons with CD4 >250 cells per microliter.

Conclusions: Hepatotoxicity was a common adverse effect of NVP among men and women with CD4 >250 cells per microliter. Chinese treatment guidelines should be considered to reflect this risk.

*Department of Infectious Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Beijing, P. R. China; and

Division of Infectious Diseases, Johns Hopkins University School of Medicine, Baltimore, MD.

Correspondence to: Taisheng Li, MD, PhD, Department of Infectious Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, No.1 Shuaifuyuan, Wangfujing Street, Beijing, China 100730 (e-mail:

Supported by the National Natural Science Foundation of China (grant 81071372 to T.L.); National Key Technologies R&D Program for the 12th Five-Year Plan (grant 2012ZX10001003-001); and Key Clinical Program of the Ministry of Health (2010–2012). J.X. was supported by the grant of Chinese Ministry of Human Resources and Social Security (2011).H.W. was supported by National Key Technologies R&D Program for the 12th Five-Year Plan (Grant 2012ZX09303013).

The authors T.L. designed the study. T.L., H.W., and K.G.G. provided critical review of this article. Y.H., J.X., Z.Q., F.G., Y.L., and H.W. preserved the sample and collected the data. W.W. and Y.H. did the laboratory test. C.Z., W.W., and M.Z. analyzed the data. C.Z., W.W., and M.Z. wrote the article.

The authors have no conflicts of interest to disclose.

Received October 11, 2012

Accepted December 20, 2012

Back to Top | Article Outline


Nevirapine (NVP), due to its cost-effectiveness,1 is currently recommended by the Chinese national treatment guidelines as a first-line drug for antiretroviral therapy (ART) –naive HIV-infected patients. Women with a CD4 cell count above 250 cells per microliter and men with a CD4 cell count greater than 400 cells per milliliter are excluded from this recommendation due to increased risk for NVP-induced hepatotoxicity.2 These exclusions are similar to those found in US treatment guidelines3 and are mainly derived from studies conducted in blacks,4 whites,4 hepatitis C virus (HCV)–coinfected patients5 and pregnant women.6

Recent studies from resource-limited areas have demonstrated conflicting results. A study among women who took stavudine (D4T) or zidovudine (AZT) plus lamivudine (3TC) with NVP in Zambia, Thailand, and Kenya found that CD4 cell count was not a reliable predictor of NVP-related hepatotoxicity.7 A retrospective study in 1110 Argentinian patients did not identify high CD4 cell count as a risk factor for NVP-related hepatotoxicity.8 A study from South Africa revealed a low incidence of NVP-related hepatotoxicity, and it suggested that age, gender, and baseline CD4 count were not associated with hepatotoxicity.9 A study from Thailand showed that patients' sex and baseline CD4 cell count were not associated with the risk of hepatotoxicity or rash, but pregnant women with CD4 cell counts more than 250 cells per microliter were likely to develop NVP-related hepatotoxicity.10 A study among pregnant women in Kenya showed that baseline CD4 cell count more than 250 cells per microliter was not associated with severe hepatotoxicity.11 Finally, a retrospective study of 703 treatment-naive pregnant African women concluded that it was safe to prescribe NVP for the prophylaxis of vertical transmission in women independent of CD4 cell counts.12 These findings suggest that the relationship between CD4 cell count and NVP-associated hepatotoxicity is complex and may be multifactorial. We know little about the risks of NVP toxicity in Chinese patients. To better inform current treatment guidelines in China, we examined the predictors of NVP-associated hepatotoxicity within our clinical trials network.

Back to Top | Article Outline


Study Population

Between 2005 and 2009, we recruited 728 ART-naive patients from 12 clinical trial units located across mainland China, including Beijing, Shanghai, Guangzhou, Shenzhen, Fuzhou, Kunming, Zhengzhou, and Xi’an. Exclusion criteria were as follows: (1) pregnancy or breastfeeding, (2) AIDS-defining illness within 2 weeks of entry, (3) white blood cell count less than 2.0 × 109/L, (4) hemoglobin level less than 90 g/dL, (5) platelet count less than 75 × 109/L, (6) baseline alanine aminotransferase (ALT) level greater than 3 times the upper limit of normal (ULN) range, (7) baseline total bilirubin level >2.5 × ULN (8) baseline serum creatinine level >1.5 × ULN.

The initial goal of these 2 randomized controlled trials was to evaluate the safety and efficacy of 3 antiretroviral regimens that were used in China. We discontinued the AZT + didanosine (ddI) + NVP arm prematurely due to significant adverse effects observed in that group compared with the other 2 groups.1,13 So the earlier trial randomized patients into 3 groups, whereas the later trial randomized patients into only 2 groups. As a whole, patients were randomly assigned to receive 3 different NVP-containing antiretroviral regimens: (1) AZT (300 mg twice a day) + ddI (200 mg twice a day or 125 mg twice a day when body weight less than 60 kg) + NVP (200 mg every day for 2 weeks and 200 mg twice a day thereafter); (2) 3TC (150 mg twice a day) + NVP (200 mg every day for 2 weeks and 200 mg twice a day thereafter) + D4T (30 mg twice a day or 20 mg twice a day when less than 60 kg) or (3) 3TC (150 mg twice a day) + NVP (200 mg every day for 2 weeks and 200 mg twice a day thereafter) + AZT (300 mg twice a day). AZT, ddI, D4T, and NVP are generic drugs manufactured in China. All patients were followed up at 2, 4, 12, and 24 weeks after treatment initiation. Immunological and virological responses to ART and liver function tests were evaluated at each follow-up visit. All patients with available CD4 and liver function test results were eligible to participate in this study.

Back to Top | Article Outline

Hepatotoxicity Grading

Hepatotoxicity was graded according to the criteria described in the WHO Toxicity Grading Scale for Determining The Severity of Adverse Events.14 Briefly, these criteria include standard elevations in the serum ALT level, serum aspartate transaminase (AST) level, and total bilirubin level. Exceeding 1.25 times of ULN range in any of the above criteria was considered mild hepatotoxicity, while higher than 5 × ULN was considered to be severe. All patients were then classified into 3 groups by the presence and severity of hepatotoxicity as follows: (1) normal; (2) mild; and (3) severe.

Back to Top | Article Outline

Statistical Analyses

The χ2 test was used to compare the prevalence of hepatotoxicity among groups. We used log-binomial regression to analyze the relationship between baseline CD4 cell count and hepatotoxicity, and to calculate the relative risk (RR). Goodness-of-fit was assessed by deviance/DF. The χ2 test was carried out by IBM SPSS v19.0.0. Log-binomial regression was carried out by SAS v 9.13. An α value for all hypothesis testing of 0.05 was considered to represent statistical significance.

Back to Top | Article Outline


From 2005 to 2009, 914 antiretroviral treatment–naive HIV-infected patients were screened for the 2 trials. A total of 566 patients were included in this analysis. The detailed patient flow chart and baseline characteristics are shown in Figure 1 and Table 1. There was no significant difference in age, body mass index, and baseline ALT and AST among patient groups. No significant differences in baseline characteristics were noted for the 162 participants excluded from the analysis.





Overall, 217 [38.3%, 95% confidence interval (CI): 34.4 to 42.4] participants developed hepatotoxicity, with 61 (10.8%, 95% CI: 8.5 to 13.6) participants experiencing severe hepatotoxicity. Among men, 144 (36.8%, 95% CI: 32.2 to 41.7) participants experienced hepatotoxicity; of those, 31 (7.9%, 95% CI: 5.6 to 11.0) were judged to be severe cases. Among women, 73 (41.7%, 95% CI: 34 to 49.1) participants developed hepatotoxicity; of those 30 (17.1%, 95% CI: 12.3 to 23.4) were judged to be severe. Among those who experienced hepatotoxicity, the peak ALT and AST levels occurred at week 4, compared with other follow-ups. Women had higher prevalence of hepatotoxicity (P = 0.002). More women developed severe hepatotoxicity than men (P = 0.001).

As shown in Table 2, in the multivariable model, CD4 cell count and hepatitis C viral antibody serostatus were independent predictors of hepatotoxicity. Baseline CD4 cell count more than 250 cells per microliter was an independent risk factor for hepatotoxicity in Chinese men [adjusted RR = 1.22 (95% CI:, 1.04 to 2.44)] and women [adjusted RR = 1.72 (95% CI: 1.20 to 2.46)]. Regimen 1 (ddI + AZT + NVP) was associated with more risk of hepatotoxicity compared with the other 2 regimens [adjusted RR = 1.36 (95% CI: 1.09 to 1.69)].



Severe hepatotoxicity was more likely in both men and women with CD4 >250 cells per microliter. About 13.7% men and 24.6% women with CD4 >250 cells per microliter developed severe hepatotoxicity, compared with 5.9% men (P = 0.01) and 13.2% women (P = 0.05) with CD4 <250 cells per microliter developed severe hepatotoxicity. Among 61 patients with severe hepatotoxicity, 49 recovered after discontinuation of NVP and the rest were lost to follow-up after 12 weeks.

Twenty-one patients developed rash and 14 patients developed fever during our study. Nine (42.9%) patients with rash had concomitant hepatotoxicity, 4 (19.0%) with severe hepatotoxicity. Nine (64.3%) patients with fever had hepatotoxicity, 2 (14.3%) were severe. There were no reported deaths due to liver failure in our population.

Back to Top | Article Outline


Current Chinese guidelines recommend that women with a CD4 cell count more than 250 cells per microliter and men with a CD4 cell count more than 400 cells per microliter avoid NVP-containing ART regimens. In this study, we found that CD4 cell count more than 250 cells per microliter was an independent risk factor for hepatotoxicity in both men and women. Our findings suggest that we may need to reconsider prescribing NVP to all Chinese patients with a CD4 cell count more than 250 cells per microliter regardless of sex. Given that NVP is recommended as a first-line antiretroviral agent by official Chinese treatment guidelines, the impact on the population at large is likely to be significant.

Racial differences, study methodology, and other differences in study populations may each play a role in explaining these findings. The exact mechanism of NVP-related hepatotoxicity and its relationship with CD4 cell count is still not fully understood. NVP-related hepatotoxicity may be immune mediated. Antunes et al15 showed that metabolic activation of NVP into a form called 2-hydroxy NVP could induce the formation of DNA protein covalent adducts in vivo, which might play a role in NVP-related hepatotoxicity. Clinically, however, patients did not benefit from anti-inflammatory therapy.16 There may also be racial differences in drug metabolism. Finally, the impact of drug–drug interactions must be accounted for. In our study, the combination of ddI, AZT, and NVP was associated with a higher rate of hepatotoxicity compared with other drug combinations.

The prevalence of hepatotoxicity in our study is consistent with findings from most other reports. In an US study which included patients from different racial/ethnic groups (including white, African American, Hispanic, and Asian), the prevalence of hepatotoxicity (defined as ALT or AST higher than 5 × ULN) from 200 mg twice daily oral doses of NVP was 8.8% overall.4 That study was underpowered to detect significant racial differences. A comprehensive summary by Dieterich et al17 on clinical trials from the United States, Italy, the Netherlands, Australia, Canada, Argentina, and South Africa found a prevalence of NVP-related hepatotoxicity of 10%. The prevalence in another study from South Africa was 17%. The higher prevalence in the latter study may reflect a difference in study population as patients with baseline CD4 cell count <200 cells per microliter were excluded.18 At the same time, several studies reported a lower rate of hepatotoxicity. One study from Boehringer Ingelheim Company, although disputed by several peers,19 reported a prevalence of 5.8%.4 Another study from Africa showed prevalence of grade 3–4 hepatotoxicity at 1.3%, which might be due to a lower baseline CD4 cell count and normal baseline AST/ALT of the studies population.20

Other predictors of hepatotoxicity found in our analysis are consistent with published studies. An increased prevalence of hepatotoxicity and severe hepatotoxicity in women is consistent with findings from several other studies.19,21 Similarly, HCV coinfection were previously reported to be independent predictors of hepatotoxicity with NVP use.22 The lack of association between hepatitis B virus coinfection and hepatotoxicity was observed in our previous study13 and several others.23–25

This study has several strengths. The data were derived from a randomized controlled trial, which tends to minimize the potential for confounding. The trial was a multicenter study that included a broad representation of the Chinese population thereby enhancing external validity. Our study, however, has several limitations. First, we used a limited number of tests to define liver toxicity. Although these tests are sensitive measures of liver damage, we did not assess other measures of hepatic function, such as serum alkaline phosphatase, gamma-glutamyl transpeptidase, and albumin levels, and we did not confirm these findings histologically. Second, we could not assess for all causes of liver damage, but given that the parent study was randomized, it is unlikely that this bias was differential among groups. Third, we excluded pregnant women from enrollment to this study, and we could not generalize our finding to pregnant women directly. However, based on our findings, and what we know about the risks of NVP-induced hepatotoxicity in pregnant women whose CD4 counts are greater than 250 cells per microliter, it seems reasonable to assume that our recommendations apply to that population as well. Finally, the sample size was relatively small. As such, lack of associations may reflect limited power rather than a true null finding.

In conclusion, CD4 cell count more than 250 cells per microliter was an independent risk factor for both hepatotoxicity and severe hepatotoxicity when using NVP-containing antiretroviral regimens in Chinese men and women. Additional studies are needed to confirm these findings. Current Chinese national treatment guidelines may need to reflect this increased risk in both sexes. Based on our findings, we suggested the following changes be considered as a way to minimize hepatoxicity: (1) efavirenz, the alternate nonnucleoside reverse transcriptase inhibitor, may be used as the first-line agent for ART-naive patients with CD4 cell counts >250 cells per microliter and in those who are coinfected with hepatitis B virus and/ or HCV; (2) The combination regimen of ddI + AZT + NVP seems to be associated with greater risk of hepatotoxicity, and consideration should be made to no longer recommend it as a first-line regimen.

Back to Top | Article Outline


The authors thank the study participants for their cooperation. The following clinical institutions or hospitals participated in this study: Peking Union Medical College Hospital, Beijing Youan Hospital, Beijing Ditan Hospital, Zhengzhou Sixth People’s Hospital, Tangdu Hospital, Guangzhou Eighth People’s Hospital, Shanghai Public Health Clinical Center, Shenzhen Third People’s Hospital, Fuzhou Infectious Diseases Hospital, Yunnan AIDS Care Center, Kunming Third People’s Hospital, and Honghe First People’s Hospital. The authors are grateful to Ang Li, from Columbia University College of Physicians and Surgeons for his insightful comments and editorial support; and to Weidong Zhang, MD, PhD, director of the department of epidemiology, School of Public Health, Zhengzhou University for his kind assistance in data analysis.

Back to Top | Article Outline


1. Li T, Dai Y, Kuang J, et al.. Three generic nevirapine-based antiretroviral treatments in Chinese HIV/AIDS patients: multicentric observation cohort. PLoS One. 2008;3:e3918.
2. Zhang F, Li T, Wang Y, et al.. National Guide for Free Antiviral Treatment Of HIV Patients. Beijing, China: People's Medical Publishing House; 2007.
3. Viramune tablets and oral suspension, US [package insert]. Ingelheim, German: Boehringer Ingelheim; 2006.
4. Stern JO, Robinson PA, Love J, et al.. A comprehensive hepatic safety analysis of nevirapine in different populations of HIV infected patients. J Acquir Immune Defic Syndr. 2003;34(suppl 1):S21–S33.
5. Torti C, Costarelli S, De Silvestri A, et al.. Analysis of severe hepatic events associated with nevirapine-containing regimens: CD4+ T-cell count and gender in hepatitis C seropositive and seronegative patients. Drug Saf. 2007;30:1161–1169.
6. Lyons F, Hopkins S, Kelleher B, et al.. Maternal hepatotoxicity with nevirapine as part of combination antiretroviral therapy in pregnancy. HIV Med. 2006;7:255–260.
7. Peters PJ, Stringer J, McConnell MS, et al.. Nevirapine-associated hepatotoxicity was not predicted by CD4 count >/=250 cells/μL among women in Zambia, Thailand and Kenya. HIV Med. 2010;11:650–660.
8. Bottaro EG, Huberman MJ, Iannella MC, et al.. Nevirapine-associated toxicity in clinical practice in Buenos Aires, Argentina. J Int Assoc Physicians AIDS Care (Chic). 2010;9:306–312.
9. Chu KM, Boulle AM, Ford N, et al.. Nevirapine-associated early hepatotoxicity: incidence, risk factors, and associated mortality in a primary care ART programme in South Africa. PLoS One 2010;5:e9183.
10. Phanuphak N, Apornpong T, Teeratakulpisarn S, et al.. Nevirapine-associated toxicity in HIV-infected Thai men and women, including pregnant women. HIV Med. 2007;8:357–366.
11. Peters PJ, Polle N, Zeh C, et al.. Nevirapine-associated hepatotoxicity and rash among HIV-infected pregnant women in Kenya. J Int Assoc Physicians AIDS Care (Chic). 2012;11:142–149.
12. Marazzi MC, Germano P, Liotta G, et al.. Safety of nevirapine-containing antiretroviral triple therapy regimens to prevent vertical transmission in an African cohort of HIV-1-infected pregnant women. HIV Med. 2006;7:338–344.
13. Wang H, Li Y, Zhang C, et al.. Immunological and virological responses to cART in HIV/HBV co-infected patients from a multicenter cohort. AIDS. 2012;26:1755–1763.
14. World Health Organization. Toxicity grading scale for determining the severity of adverse events. Available at: Accessed December 9, 2011.
15. Antunes AM, Novais DA, Da SJ, et al.. Synthesis and oxidation of 2-hydroxynevirapine, a metabolite of the HIV reverse transcriptase inhibitor nevirapine. Org Biomol Chem. 2011;9:7822–7835.
16. Bonnet F, Lawson-Ayayi S, Thiebaut R, et al.. A cohort study of nevirapine tolerance in clinical practice: French aquitaine cohort, 1997–1999. Clin Infect Dis. 2002;35:1231–1237.
17. Dieterich DT, Robinson PA, Love J, et al.. Drug-induced liver injury associated with the use of nonnucleoside reverse-transcriptase inhibitors. Clin Infect Dis. 2004;38(suppl 2):S80–S89.
18. Sanne I, Mommeja-Marin H, Hinkle J, et al.. Severe hepatotoxicity associated with nevirapine use in HIV-infected subjects. J Infect Dis. 2005;191:825–829.
19. Baylor MS, Johann-Liang R, Hepatotoxicity associated with nevirapine use. J Acquir Immune Defic Syndr. 2004;35:538–539.
20. Kalyesubula R, et al.. Hepatotoxicity from first line antiretroviral therapy: an experience from a resource limited setting. Afr Health Sci. 2011;11:16–23.
21. Wit FW, Weverling GJ, Weel J, et al.. Incidence of and risk factors for severe hepatotoxicity associated with antiretroviral combination therapy. J Infect Dis. 2002;186:23–31.
22. Labarga P, Soriano V, Vispo ME, et al.. Hepatotoxicity of antiretroviral drugs is reduced after successful treatment of chronic hepatitis C in HIV-infected patients. J Infect Dis. 2007;196:670–676.
23. Ocama P, Castelnuovo B, Kamya MR, et al.. Low frequency of liver enzyme elevation in HIV-infected patients attending a large urban treatment centre in Uganda. Int J STD AIDS. 2010;21:553–557.
24. Palmon R, Koo BC, Shoultz DA, et al.. Lack of hepatotoxicity associated with nonnucleoside reverse transcriptase inhibitors. J Acquir Immune Defic Syndr. 2002;29:340–345.
25. Mbougua JB, Laurent C, Kouanfack C, et al.. Hepatotoxicity and effectiveness of a nevirapine-based antiretroviral therapy in HIV-infected patients with or without viral hepatitis B or C infection in Cameroon. BMC Public Health. 2010;10:105.

nevirapine; CD4 lymphocyte count; drug toxicity; hepatotoxicity

© 2013 Lippincott Williams & Wilkins, Inc.