The number of patients who will start highly active antiretroviral therapy (HAART) in sub-Saharan Africa is likely to increase dramatically over the next few years. This scaling up entails numerous technical and scientific challenges and obstacles, including the choice of first-line HAART regimens to be recommended in large-scale public health programs to thousands of people without the benefit or possibility of individual case reviews. The choice of first-line regimens should take into account not only international knowledge on the efficacy and tolerance of HIV drugs, but also local specificities in HIV disease (eg, frequency of tuberculosis, frequency of HIV-2 infection, and frequency of hepatitis B virus [HBV] coinfection) and drug tolerance.1 Furthermore, a given drug regimen in a given type of patient can also show variable efficacy depending on the conditions under which the patient can be followed (eg, type of health care facilities and access to them) or even on the patient's socioeconomic and family conditions of life. Until recently, most of the existing knowledge pertaining to HAART came from industrialized countries. An increasing number of HAART experiences have now been reported from low-resource settings.2-7 Most of these reports were from "pioneer" cohorts, from the era when access to HAART was limited and no large-scale standardized recommendations were available. On the one hand, these studies clearly showed that HAART was globally feasible and efficient in the context of low-resource settings. On the other hand, they mostly include either numerous patients following multiple HAART regimens or a reduced number of patients on a single regimen. The efficacy and tolerance of each specific HAART regimen in the African context remain to be detailed in standardized cohorts with large number of patients. We report here the 6-month efficacy and tolerance of zidovudine, lamivudine, and efavirenz in a standardized prospective cohort of 740 HIV-infected adults living in Abidjan, the capital city of Côte d'Ivoire.
In December 2002, a multicenter randomized trial (Trivacan Agence Nationale de Recherches sur le SIDA [ANRS] 1269 trial) was launched in 5 outpatient clinics of Abidjan.8 The main objective of this trial was to assess various structured treatment interruption strategies of HAART. The trial was designed in 2 phases. Patients were included in the first phase ("prerandomization phase") if they met the following criteria: 18 years or older, naive to curative antiretroviral therapy, CD4 T-cell (CD4) count between 150 and 350 cells/μL (or CD4 percentage between 12.5% and 20%), absence of pregnancy, absence of severe renal or hepatic failure, and written informed consent. In this prerandomization phase, all patients received continuous HAART. The first-line regimen was zidovudine-lamivudine in combination with (i) preferably efavirenz, for HIV-1-infected or HIV-1 and HIV-2 dually infected men and women with effective contraception and no history of nevirapine prophylactic treatment; (ii) indinavir/ritonavir, for HIV-2-infected patients, women not desiring contraception or with a history of nevirapine prophylaxis. After at least 6 months in the prerandomization phase, patients were randomized into 1 of the 3 arms of the structured treatment interruption trial. We present here the data concerning the first 6 months on HAART in patients who were included in the prerandomization phase of the Trivacan trial and who were prescribed zidovudine-lamivudine-efavirenz at enrollment.
The protocol of the Trivacan trial was approved by the ethics committee of the Ivorian Ministry of Health and the institutional review board of the ANRS.
At enrollment, a standardized questionnaire was used to record baseline characteristics. Patients were asked to return to their study center at day 7, at 1 month, and every month thereafter. During these visits, counseling was given which covered compliance with HAART and with contraception, and questionnaires were administered to record self-reported symptoms since last visit, adherence to treatment during the previous 4 days, and, for women, adherence to contraception and date of last menstrual period. Pregnancy testing was only done in case of clinical signs of pregnancy or delayed date of menstruation. Between these scheduled visits, patients had free access to the study clinics whenever they had medical problem.
At enrollment and then at each monthly visits, subjects were given numbered boxes containing the quantity of pills for 35 days, that is, 70 pills of zidovudine 300 mg-lamivudine 150 mg (Duovir, Cipla Ltd, Mumbai, India) with 105 pills of 200 mg efavirenz (Stocrin; Merck Sharp & Dohme BV, Amsterdam, The Netherlands) before May 1, 2003, or 35 pills of efavirenz 600 mg after this date. Patients were asked to return monthly with their box from the previous month, including unused tablets, and to exchange it for a new one. Cotrimoxazole prophylaxis (800/160 mg daily) was also systematically prescribed.
CD4 count (True Count technique on FACScan; Becton Dickinson, Erenbodegen, Belgium) and plasmatic HIV-1 RNA (real-time polymerase chain reaction on Taq Man technology Abi Prism 7000, Applied Biosystems, Rotkreuz, Switzerland, threshold of detectability at 300 copies/mL)9 were measured at months 3 and 6. Blood cell count (MaxM Coulter, Beckman, Fullerton CA), serum liver enzymes, and serum creatinine were measured at months 1, 3, and 6.
At each scheduled and unscheduled visit, signs and symptoms were managed following standardized algorithms of investigation and treatment.10,11 All clinical events were reviewed by an event documentation committee. The diagnostic criteria were the same as those used in a cohort which was followed in Abidjan during the same period, under the supervision of the same team and whose procedures were described elsewhere.10,12 All care was free of charge, including transport, consultations, investigations, hospitalization, and drugs.
Baseline was the date of enrolment in the prerandomization phase. The incidence rate of a given disease was defined as the number of patients with at least 1 episode per 100 patient-years of at-risk follow-up. The at-risk period began on the day of enrollment and continued to the date of the 6-month visit, death, default, or the date of the first episode of the disease. At study termination, patients were defined as lost to follow-up if (i) their last contact was before the sixth month visit, (ii) they were not known to be dead, and (iii) no further information on their vital status was available on March 15, 2005. The monthly follow-up rate was the percentage of patients who appeared for monthly visits among those who had survived the interval.
Patients and Follow-up
From December 2002 to March 2004, 740 patients were prescribed zidovudine-lamivudine-efavirenz. Their main baseline and follow-up characteristics are shown in Table 1. The schooling was less than or equal to primary school level in 50% of patients, 78% of patients had running water at home, 41% had a refrigerator at home, and 44% had a personal regular source of income. For 80% of patients, the overall monthly income of the family was less than US$112. One hundred eighty-six patients (25%) had informed no one of their HIV seropositivity. Of the remaining 554, 41% had informed their partner, 39% had informed somebody else living in the same home, and 63% had informed somebody else not living in the same home. Of the 548 women, 10% had a history of short-course zidovudine regimen for preventing mother-to-child transmission of HIV.
During the first 6 months of HAART, the incidence of death was 2.5/100 patient-years (95% confidence interval [CI], 2.5-4.1).
The attendance rate for months 1 to 6 visits was 98.8%, 98.2%, 98.6%, 97.7%, 97.9%, and 98.2%, respectively. At these monthly visits, 11%, 11%, 10%, 8%, 9%, and 8% of patients declared that they had missed at least 1 dose of medication during the previous 4 days, respectively. Only 1 self-reported reason for missing at least 1 dose significantly changed over time, the "fear of side effects," which was reported by 22% of patients who declared to have missed 1 dose at month 1, versus 11%, 7%, 8%, 10%, and 6%, at months 2, 3, 4, 5, and 6, respectively.
CD4 Count and Viral Load Evolution
Figure 1A shows the evolution of the mean CD4 count between baseline, month 3, and month 6 by group of baseline CD4 count. The mean gain in CD4 count at month 6 was +156 cells/μL overall and was not significantly different between groups of baseline CD4 count (P > 0.9 when comparing all groups to each other, except the group of patients with baseline CD4 count of more than 350 cells/μL; for comparison between patients with baseline CD4 count of more than 350 cells/μL and other groups, P = 0.23-0.28).
Figure 1B shows the evolution of plasma HIV-1 viral load between baseline and month 6, in terms of percentage of patients with undetectable viral load, detectable viral load with difference [baseline to month 3] or [baseline to month 6] of 1 log10/mL or more, and detectable viral load with [baseline to month 3] or [baseline to month 6] less than 1 log10/mL. As shown in Figure 1B, 87% of patients had undetectable viral load at month 6. Prior short-course zidovudine prophylaxis was not associated with virologic or immunologic failure at 6 months (data not shown).
During the first semester, there were 198 episodes of severe morbidity in 147 patients (incidence of severe morbidity, 3.7/100 person-months; 95% CI, 3.1-4.3), including 35 episodes of WHO stage 3 or 4 defining diseases in 33 patients (incidence, 0.7/100 person-months; 95% CI, 0.5-1.0) and 163 episodes of nonclassifying diseases with at least 1 day at the day-care hospitals of the study clinics and/or with HAART regimen modification in 121 patients (incidence, 3.0/100 person-months; 95% CI, 2.4-3.5). As shown in Table 2, the most frequent causes of WHO staging system classifiable morbidity were tuberculosis and bacterial diseases.
Adverse Effects Leading to Treatment Modification
During follow-up, 37 patients interrupted their initial HAART regimen temporarily or definitely, all due to serious side effects, yielding an overall incidence of serious side effects of 0.9/100 person-months (95% CI, 0.6-1.2). Of these 37 episodes, (i) 28 were zidovudine-attributable side effects, including 25 severe anemia, 2 severe neutropenia, and 1 nonobstructive cardiomyopathy. In all these 28 cases, zidovudine was replaced by stavudine, an incidence of zidovudine definitive discontinuation for side effects of 0.6/100 person-months (95% CI, 0.4-0.9); (ii) 9 were efavirenz-attributable side effects, including 5 neurologic (severe dizziness, n = 4; acute delirium, n = 1), 3 cutaneous (febrile rash, n = 1; intolerable pruritus, n = 2), and 1 grade 3 cytolytic hepatitis. Efavirenz was definitively interrupted in 6 of these 9 cases, an incidence of efavirenz definitive discontinuation for side effects of 0.1/100 person-months (95% CI, 0.01-0.3). In the remaining 3 cases (the 2 severe pruritus and the acute delirium), efavirenz was successfully reintroduced, and the HAART regimen was not modified. Of the 6 patients who permanently interrupted efavirenz, 3 received lopinavir-ritonavir, and 2 received indinavir-ritonavir in combination with zidovudine and lamivudine. The last patient did not come back to the study center and was lost to follow-up.
Other Adverse Effects
Contrasting with this rarity of serious neurologic side effects, the rate of mild self-declared neurologic symptoms which did not lead to efavirenz discontinuation was high in the first month and decreased rapidly over time. At months 1, 2, 3, 4, 5, and 6, 73%, 36%, 29%, 25%, 21%, and 17% of patients reported insomnia, vertigo, or nightmare within the previous month, respectively (Fig. 2).
At baseline and month 6, 21% and 6.6% of patients had a transaminase value over 1.25 times the upper limit of normal (ULN), respectively. This included 17% versus 5.8% patients with 1.25 ULN < transaminases ≤ 2.5 ULN, 3% versus 0.4% patients with 2.5 ULN < transaminases ≤ 5 ULN, and 1% versus 0.4% patients with > 5 ULN. Table 3 shows the higher transaminases value during follow-up by stratum of baseline value in patients with negative and positive serum HBs antigen at baseline.
The percentage of patients with creatinine clearance of less than 50 mL/min was 3% at baseline and 1% at month 6.
Finally, at month 6, 11% of patients self-declared that they felt body changes since baseline. These modifications were declared to be abdominal hypertrophy in 88% of cases, breast hypertrophy in 4%, arm atrophy in 4%, and breast atrophy in 4%. The measured mean waist circumference/hip circumference (WC/HC) ratio was 0.87 (median, 0.86; interquartile range [IQR], 0.80-0.93) at baseline and 0.86 (median, 0.86; IQR, 0.85-0.92) at month 6, with no significant difference between both measures. The median change in WC/HC ratio between baseline and month 6 was +0.00 (IQR, −0.04 to +0.04).
Contraception and Pregnancy
At months 1, 2, 3, 4, 5, and 6, 58%, 70%, 77%, 79%, 80%, and 80% of women declared that they were actually currently using a contraceptive method, respectively. Contraceptive drugs were intramuscular progesterone in 65% of women and oral estrogen/progesterone in 35%, with no significant change in the type of contraceptive drug during the period. There were 7 pregnancies and overall incidence of 2.6/100 person-years (95% CI, 0.67-4.51). Six women voluntary interrupted the pregnancy. For the sole woman who chose to continue the pregnancy, the initial HAART regimen was modified to zidovudine-lamivudine-nelfinavir on the eighth week of pregnancy. In this woman, pregnancy and delivery were normal. The child was not HIV infected (as determined by HIV polymerase chain reaction) at week 8. No malformations were detected. There was no examination of fetuses that did not go to term for malformations.
In resource-limited settings, experts are currently recommending the combination of 2 nucleoside reverse transcriptase inhibitors (NRTIs) and a nonnucleoside reverse transcriptase inhibitor (NNRTI) as a first-line antiretroviral treatment for HIV-1-infected adults and adolescents.1,13
In this large cohort study of 740 patients, we provide further evidence that the combination zidovudine-lamivudine-efavirenz is highly effective at 6 months among sub-Saharan adults, and we generate further arguments in the debate concerning the respective positions of efavirenz and nevirapine as first-line NNRTI.
With respect to efficacy, the percentage of persons found with a viral load of less than 300 copies/mL (87%) and the mean rise in CD4 at 6 months (+156 cells/μL) were both in the upper bracket of what has previously been described not only in Africa, but also in Europe or the United States.2-7,14 The truly original contribution of this study, however, concerns the question of drug tolerance and precautions for use. As long as an HIV drug is not tested on a large scale, one can never be sure that its tolerance will be identical to what has been found previously in industrialized countries. On the one hand, the HIV-positive population in Africa is not the same as in Europe or in the United States, in terms of sex ratio (higher percentage of women in Africa), of coinfections (more persons are coinfected with HBV in Africa),15 and of the prevalence of risk factors associated with a poorer prognosis (more anemia and more hemoglobinopathies in Africa), but in terms of hypersensitivity to certain medications (fewer cutaneous reactions to cotrimoxazole in sub-Saharan Africa than in industrialized countries).16 On the other hand, follow-up conditions and the possibility of ensuring a high level of compliance through access to a large cadre of health providers, incentives, and/or counseling are not the same; thus, painful symptoms occurring at the onset of treatment can have potentially significant effects on compliance.
In this context, the high neurologic, cutaneous, and hepatic tolerance to efavirenz in sub-Saharan Africa in a large cohort of several hundred persons is a welcome finding.
As far as neurologic symptoms with efavirenz, those that were self-reported seemed to be as frequent in Cote d'Ivoire as that which has been described elsewhere in the world,14,17,18 with a clear peak during the first month, yet frequency did not seem to jeopardize compliance, as both our data on compliance and 6-month efficacy suggest. Severe neurologic intolerance leading to replacement of efavirenz with another drug was rare.
As far as cutaneous and hepatic events, the frequency of rash or severe drug-related hepatitis was low, both within the lower bracket of what has been described for efavirenz in industrialized countries,14,17,19-21 and quite lower than figures that have been reported for nevirapine both in industrialized and low-income countries.14,17,19-22 For hepatic tolerance, this was even more remarkable than it would otherwise be; as in our population, 14% of subjects were carriers of serum HBs antigen, and 19% had abnormal levels of serum transaminases before starting treatment.19,21
The strongest argument in favor of nevirapine as first-line NNRTI is its absence of teratogenicity, in a context where more than half of adults infected with HIV are women of childbearing age.1,23 On this latter point, our study provides conflicting evidence. On the one hand, all the women participating in this study agreed to use contraception, with an overall compliance rate reaching 80% at 6 months. The incidence of pregnancy was low, and all pregnancies, except one, were terminated at the request of the subject (without intervention on our part). Our experience shows that effective contraception is possible and that efavirenz can be prescribed for women of childbearing age without a disastrous outcome. The current wisdom, "in Africa, the NNRTI of choice is efavirenz for men and nevirapine for women," can thus be further nuanced.23 On the other hand, this study also illustrates that even if one maximally facilitates access to contraception, we will never attain the goal of zero live births among women taking efavirenz. Furthermore, in our study with limited number of women, we did not have the power to estimate the true risk of efavirenz-induced teratogenicity. In a large number of women taking efavirenz through a population-based implementation program, even a high adherence to contraception would result in a significant number of pregnancies and therefore in a risk of teratogenicity. Further studies should estimate the extent of this risk.
Our study has several other limitations. First, results from a clinical trial may not translate exactly into results on a population basis. In our study, patients were followed up free of charge in a dedicated center with more procedures available than in most of the other HIV outpatient clinics in Abidjan. Under these cohort conditions, HAART efficacy and adherence to contraception were likely to be higher in our patients than in most patients receiving HIV care in Abidjan. That said, in Côte d'Ivoire, where constant efforts have been made for the last 6 years to reduce the cost of HIV drugs, free access to antiretroviral drugs is likely to become the standard of care in the near future.
Second, we cannot extrapolate that the low rate of pregnancy that we found during the first 6 months would remain so with continued treatment. In Abidjan before the HAART era, women with more advanced disease had lower pregnancy rates.24 With effective treatment, immune and clinical status will improve, and hence, fertility may increase over time. As women improve in their health status, they may be more likely to desire to become pregnant
Finally, a major argument in favor of NNRTIs in Africa is the possibility of combining it successfully with rifampicin, in a context where tuberculosis is the first cause of severe morbidity under antiretroviral treatment. This was the case in our study, as well as others previously conducted in Africa.2,25
The combination zidovudine-lamivudine used in our study was 1 of the 2 first-line NRTI combinations recommended by experts in 2003 in Africa, the other being the combination of stavudine-lamivudine.1 In our study, zidovudine-related anemia was by far the major cause of modification in treatment regimen during the first semester of treatment. Nevertheless, its frequency was acceptable and comparable with what has been described previously in industrialized countries.26 It is a well-known adverse effect that does not raise major issues concerning its use even in a difficult context. Our study did not, in fact, produce any new information on the risk of lipoatrophy and lipohypertrophy. At 6 months, self-reported body changes were mainly nonspecific and compatible with a process of gaining back weight that had been lost during treatment. The evolution of the WC/HC ratio also did not show any large pathologic phenomena. Nevertheless, more targeted and longer duration studies are now clearly necessary to estimate the incidence of lipoatrophy and lipohypertrophy among sub-Saharan African patients, especially those receiving protease inhibitors or specific NRTIs such as stavudine.27
In conclusion, the regimen zidovudine-lamivudine-efavirenz was efficacious and adhered to over the first semester among adults infected with HIV-1 in Côte d Ivoire. Zidovudine required replacement because of intolerance among 3.8% of patients, and efavirenz required replacement by another drug among 0.8% of patients. Hepatic and cutaneous tolerance to efavirenz was very high, thus conferring advantages on efavirenz as the first-line NNRTI. Considering these advantages in terms of drug tolerance, efavirenz could thus be a valuable first-line drug for women of childbearing age who agree to use contraception, provided that the risk of teratogenicity should be closely monitored.
2. Seyler C, Anglaret X, Dakoury-Dogbo N, et al. Medium-term survival, morbidity and immunovirological evolution in HIV-infected adults receiving antiretroviral therapy, Abidjan, Cote d'Ivoire. Antivir Ther
. 2003; 8:385-393.
3. Landman R, Schiemann R, Thiam S, et al. Once-a-day highly active antiretroviral therapy in treatment-naive HIV-1-infected adults in Senegal. AIDS
. 2003; 17:1017-1022.
4. Laurent C, Kouanfack C, Koulla-Shiro S, et al. Effectiveness and safety of a generic fixed-dose combination of nevirapine, stavudine, and lamivudine in HIV-1-infected adults in Cameroon: open-label multicentre trial. Lancet
. 2004; 364:29-34.
5. Laurent C, Diakhaté N, Ngom Gueye N, et al. The Senegalese government's highly active antiretroviral therapy initiative: an 18-month follow-up study. AIDS
. 2002; 16:1363-1370.
6. Jack C, Lalloo U, Karim QA, et al. A pilot study of once-daily antiretroviral therapy integrated with tuberculosis directly observed therapy in a resource-limited setting. J Acquir Immune Defic Syndr
. 2004; 36:929-934.
7. Coetzee D, Hildebrand K, Boulle A, et al. Outcomes after two years of providing antiretroviral treatment in Khayelitsha, South Africa. AIDS
. 2004; 18:887-895.
8. Danel C, Moh R, Sorho S, et al. Progress report on a structured treatment interruption (STI) trial: Trivacan ANRS 1269 trial, Abidjan, Côte d'Ivoire [Abstract WeOrB1284 2004]. Paper presented at: XV International AIDS Conference; July 11-16, 2004; Bangkok, Thailand.
9. Rouet F, Ekouevi D, Chaix M, et al. Real-time PCR test for human immunodeficiency virus type-1 infection in an African resource-limited setting. J Clin Microbiol
. 2005; 43:2709-2717.
10. Anglaret X, Messou E, Ouassa T, et al. Pattern of bacterial diseases in a cohort of HIV-1 infected adults receiving cotrimoxazole prophylaxis in Abidjan, Cote d'Ivoire. AIDS
. 2003; 17:575-584.
11. Anglaret X, Dakoury-Dogbo N, Bonard D, et al. Causes and empirical treatment of fever in HIV-infected adult outpatients, Abidjan, Cote d'Ivoire. AIDS
. 2002; 16:909-918.
12. Attia A, Huet C, Anglaret X, et al. HIV-1-related morbidity in adults, Abidjan, Cote d'Ivoire: a nidus for bacterial diseases. J Acquir Immune Defic Syndr
. 2001; 28:478-486.
13. WHO. Antiretroviral drugs for the treatment of HIV infection in adults and adolescents in resource-limited settings: recommendations for a public health approach (2005-2006 revisions). Presented at: Guidelines Development Group, Brief Meeting Report; June 22-23, 2005; Montreux, Switzerland. Available at: http://www.who.int/3by5/ARVmeetingreport_June2005.pdf
. Accessed April 13, 2006.
14. Staszewski S, Morales-Ramirez J, Tashima KT, et al. Efavirenz plus zidovudine and lamivudine, efavirenz plus indinavir, and indinavir plus zidovudine and lamivudine in the treatment of HIV-1 infection in adults. Study 006 Team. N Engl J Med
. 1999; 341:1865-1873.
15. Burnett RJ, Francois G, Kew MC, et al. Hepatitis B virus and human immunodeficiency virus co-infection in sub-Saharan Africa: a call for further investigation. Liver Int
. 2005; 25:201-213.
16. Colebunders R, Izaley L, Bila K, et al. Cutaneous reactions to trimethoprim-sulfamethoxazole in African patients with the acquired immunodeficiency syndrome. Ann Intern Med
. 1987; 107:599-600.
17. Perez-Molina JA. Safety and tolerance of efavirenz in different antiretroviral regimens: results from a national multicenter prospective study in 1,033 HIV-infected patients. HIV Clin Trials
. 2002; 3:279-286.
18. Fumaz CR, Tuldra A, Ferrer MJ, et al. Quality of life, emotional status, and adherence of HIV-1-infected patients treated with efavirenz versus protease inhibitor-containing regimens. J Acquir Immune Defic Syndr
. 2002; 29:244-253.
19. Sulkowski MS, Thomas DL, Mehta SH, et al. Hepatotoxicity associated with nevirapine or efavirenz-containing antiretroviral therapy: role of hepatitis C and B infections. Hepatology
. 2002; 35:182-189.
20. van Leth F, Phanuphak P, Ruxrungtham K, et al. Comparison of first-line antiretroviral therapy with regimens including nevirapine, efavirenz, or both drugs, plus stavudine and lamivudine: a randomised open-label trial, the 2NN Study. Lancet
. 2004; 363:1253-1263.
21. Law WP, Dore GJ, Duncombe CJ, et al. Risk of severe hepatotoxicity associated with antiretroviral therapy in the HIV-NAT Cohort, Thailand, 1996-2001. AIDS
. 2003; 17:2191-2199.
22. Ekouevi D, Tonwe-Gold B, Touré H, et al. Incidence and risk factors of severe adverse events with HAART in adult patients in the MTCT Plus program in Abidjan, Côte d'Ivoire [Abstract TuPe7.3C02]. Presented at: 3rd IAS Conference on HIV Pathogenesis and Treatment; July 24-27, 2005; Rio de Janeiro, Brazil.
23. Thorne C, Newell ML. The safety of antiretroviral drugs in pregnancy. Expert Opin Drug Saf
. 2005; 4:323-335.
24. Loko MA, Toure S, Dakoury-Dogbo N, et al. Decreasing incidence of pregnancy by decreasing CD4 cell count in HIV-infected women in Cote d'Ivoire: a 7-year cohort study. AIDS
. 2005; 19:443-445.
25. Kwara A, Flanigan TP, Carter EJ. Highly active antiretroviral therapy (HAART) in adults with tuberculosis: current status. Int J Tuberc Lung Dis
. 2005; 9:248-257.
26. Moh R, Danel C, Sorho S, et al. Haematological changes in adults receiving a zidovudine-containing HAART regimen in combination with cotrimoxazole in Cote d'Ivoire. Antivir Ther
. 2005; 10:615-624.
27. Lichtenstein KA. Redefining lipodystrophy syndrome: risks and impact on clinical decision making. J Acquir Immune Defic Syndr
. 2005; 39:395-400.