Malaria in pregnancy is responsible for mother and infant morbidity, including maternal anemia and low birth weight.1–3 In HIV-infected women, consequences are worsened as follows: peripheral and placental parasitemias are higher than in HIV-negative patients,4,5 and anemia and low birth weight are more frequent.6,7 Moreover, placental malaria could increase the HIV mother-to-child transmission, by altering the placental barrier and the cytokine inflammatory profile.8,9
In HIV-negative women, prevention of malaria in pregnancy resides in the use of insecticide-treated bednets and intermittent preventive treatment during pregnancy (IPTp), with 2 intakes of a curative dose of sulfadoxine-pyrimethamine (SP), from the second trimester of pregnancy.10 In HIV-infected women, the World Health Organization (WHO) advocates 2 possible strategies11 as follows: (1) if the woman is eligible to cotrimoxazole prophylaxis for the prevention of opportunistic infections (low CD4 cell count or advanced HIV disease), no additional drug intake is recommended as cotrimoxazole is assumed to protect her from malaria as well. Alternatively, an optimized IPTp, using 3 intakes of SP instead of 2, is recommended.
As resistance to sulfa drugs spreads everywhere in sub-Saharan Africa,12 other molecules than SP should urgently be tested for IPTp.13 Mefloquine (MQ) is the best candidate because of its long half-life and low resistance rates in Africa.14,15 MQ is presently allowed during all the course of pregnancy. However, it has been rarely used as IPTp, possibly because frequent, although moderate, adverse reactions (ARs) have been reported.16 A trial recently evaluated mefloquine intermittent preventive treatment (MQ IPTp) in 1600 HIV-negative pregnant women in Ouidah, South Benin, and showed comparable efficacy to SP IPTp on low birth weight and superiority on placental infection.14 Minor ARs were however frequently reported.
MQ tolerability in HIV-infected pregnant women is a major concern. Indeed, as these women already receive many medications [including antiretrovirals (ARVs) and cotrimoxazole], in vitro mechanisms suggest that potential interactions could favor MQ-associated ARs.17–19 Moreover, because of poor health conditions related to HIV immune deficiency, these women might be more susceptible to develop ARs in general.
In HIV-infected pregnant women, the ongoing PACOME trial is comparing MQ IPTp to cotrimoxazole prophylaxis for the prevention of placental malaria. We investigated the tolerability of MQ IPTp in HIV-infected pregnant women in this trial during the first year of recruitment and compared it with that observed in HIV-negative women from the Ouidah trial.
Study Site and Population
In southern Benin, Plasmodium falciparum malaria transmission is perennial with peaks during rainy seasons.20 HIV overall prevalence in Benin is around 2%21 and can reach 5% in pregnant women attending antenatal visits in Cotonou.
For the present study aiming to compare the prevalence of ARs between HIV-infected and HIV-negative women, we pooled 2 samples of pregnant women treated with MQ from a past trial in HIV-negative women (the Ouidah trial) and from the ongoing PACOME trial.
The PACOME trial (NCT00970879) compares the efficacy of cotrimoxazole and MQ IPTp for malaria prevention in HIV-infected pregnant women. It was started in December 2009 in 5 hospitals in Cotonou and Porto-Novo. Women are randomized to receive either cotrimoxazole prophylaxis daily during all the course of pregnancy, or 3 intakes of MQ IPTp, between 16 and 28 weeks, 24 and 32 weeks, then 28 and 36 weeks of gestation. In case of low CD4 cell count or advanced HIV disease, cotrimoxazole is associated with MQ for the prevention of opportunistic infections. Combined ARV therapy is prescribed for the prevention of mother-to-child transmission of HIV, according to national guidelines. The primary endpoint of this trial is the prevalence of placental malaria. The trial will last until mid-2012 to reach the target sample size of 250 pregnant women per treatment arm. Data concerning all the women enrolled in PACOME during the first year (until January 31, 2011) and allocated to MQ IPTp were selected for the present study (103 women, Fig.1).
The Ouidah trial (NCT00274235)14 was conducted in Benin between 2005 and 2008. It compared the efficacy of 2 IPTp regimens, SP and MQ, in 3 maternity clinics in the Ouidah region, 40 kilometers West of Cotonou. Women were randomized to receive either SP or MQ twice during pregnancy, between 16 and 28 weeks, and after 30 weeks. The primary endpoint was the proportion of low birth weight infants. HIV infection was an exclusion criterion. However, HIV screening tests were not available in all the maternity clinics at the beginning of the enrollment period, so only 878 of the 1601 women enrolled in the Ouidah trial (55%) were tested for HIV. Data concerning all the women of the Ouidah trial allocated to MQ and with a negative result to the HIV screening test were selected for the present analysis (421 women, Fig. 1), those with no HIV test performed were excluded.
In both trials, women were eligible after having given signed informed consent, if they were between 16 and 28 weeks of gestation, had no history of a neuropsychiatric disorder or severe kidney or liver disease, and had no history of severe ARs with MQ or sulfa drugs. MQ IPTp was administered with an interval of at least 4 weeks between intakes, at the dose of 15 mg/kg concomitantly with a rich-fat collation, under observation of the investigator. In case of vomiting within 30 minutes after the intake, an extra dose was administered. Women also received insecticide-treated bed nets and ferrous and folic acid supplementation, according to the national malaria control program guidelines.
Socioeconomic and Medical Data
In both studies, socioeconomic data and medical history were recorded at enrollment. On each scheduled visit, clinical information and medications were recorded. In HIV-infected women, information about HIV WHO clinical stage, ARV and cotrimoxazole intakes were collected additionally.
Thick blood smears were performed on each scheduled visit to detect malaria parasites. HIV rapid diagnostic screening tests were performed first by the Determine test, further confirmed by the Bioline test. Women with 2 positive results were tested by Elisa for confirmation.
In HIV-infected women, HIV viral load at enrolment was assessed using the real-time polymerase chain reaction thermocycler M2000 RP Abott, with a 40 copies per milliliter detection threshold. Absolute CD4 cell count was assessed at each scheduled visit by a Cyflow or Facscount cytometer.
ARs occurring after the MQ intakes were collected using open-labeled questionnaires. Early ARs were collected within 30 minutes after MQ intake in both trials. Late ARs were collected by phone with a median collection time of 2 days for HIV-infected women (PACOME) or during visits at home within 1 week after the intake for HIV-negative women (Ouidah). If the woman reported at least 1 AR, the investigator called her again (PACOME) or paid a visit to her (Ouidah) every day until complete recovery. She received free care when needed. Later ARs were recorded at the following monthly visits, as were ARs for the very few women who had not been reached by phone calls or visits at home.
The severity of an AR was classified as minor if it was not disturbing for daily activity, moderate if it was disturbing for daily activity (outpatient symptomatic medication when adequate but no need for other specific care), and severe if it needed more intensive treatment like parenteral treatment, surgical intervention, or hospital admission. Spontaneous abortions, stillbirths, and congenital malformations were also recorded. In each trial, an independent data and safety monitoring board reviewed all serious adverse events and had the possibility to recommend the trial's discontinuation.
Data capture was performed using Access 2003 (PACOME) and Epidata (Ouidah). Data analysis was performed using Stata version 11.
Women's medical and socioeconomic characteristics were described in each sample and compared using χ2 test for proportions and Student test for means.
Descriptive analysis compared the proportion of subjects reporting at least 1 AR during follow-up in the 2 samples in women who completed at least 2 MQ intakes. Analysis was restricted to the first 2 intakes in HIV-infected women to harmonize the exposure to MQ in the 2 cohorts (ie, 2 MQ intakes were considered for each woman). A sample size of 100 HIV-infected and 400 HIV-negative women allowed to obtain a power of 80% for a difference in proportion of ARs between both populations exceeding 15% (alpha risk = 0.05, observed prevalence in HIV-negative women = 78%).14
To verify accurately if the ARs were reported more frequently in HIV-infected women, we performed a multilevel logistic regression, taking into account the correlations within hospitals (clustered data) and within subjects (repeated measures). The probability of reporting an AR after each MQ intake (level 1) was modeled using fixed effects for the HIV status and relevant covariates and random effects for the patient (level 2) and the hospital (level 3). The first step of the analysis tested covariates separately. To investigate the specific effect of HIV-related covariates on the reporting of ARs, we tested one by one these covariates while adjusting for HIV status in the whole population (both HIV-infected and HIV-negative) of pregnant women. In the second step, relevant covariates (those with P <0.20 in first step analysis and those which differed between the populations of HIV-infected and HIV-negative women) were tested simultaneously in multivariate models. Stepwise procedures were used to select the strongest parsimonious model. For sensitivity analysis, we tested in the final model the reporting of the most frequent types of ARs and of “moderate and severe” ARs.
The Ouidah and PACOME trials have been ethically approved in France by “Comité consultatif de déontologie et d'éthique de l'IRD” and in Benin by “Comité d'éthique de l'université d'Abomey-Calavi” and “Comité national provisoire d'éthique pour la recherche en santé”. The procedures followed were in accordance with the Helsinki Declaration of 1975, as revised in 2000.
A total of 524 pregnant women were enrolled. Women's general and HIV-related characteristics are detailed in Tables 1 and 2, respectively. HIV-infected women were older and less likely to be primigravid than HIV-negative women, and they were seen earlier in pregnancy (21 weeks vs 24 weeks of gestation at the first MQ intake, in means). HIV-infected women, enrolled in big cities, had higher socioeconomic status compared with HIV-negative women enrolled in semi-rural areas.
In HIV-infected women, median duration since HIV diagnosis was 12.5 months, and mean CD4 cell count at enrollment was 374 cells per cubic millimeter; 55 (53%) had already started ARVs before enrollment; 12 (12%) initiated a new ARV regimen concomitantly with MQ intake (ARVs started during the week preceding the MQ intake), 9 with efavirenz. More than 95% of the ARV regimens included nevirapine or efavirenz. Seventy women (68%) also received cotrimoxazole prophylaxis.
As shown in Figure 1, 9 of the 103 HIV-infected women (8.6%) and 36 of the 421 HIV-negative women (8.7%) did not receive a second dose of MQ. In particular, 2 HIV-infected women (1.9%) discontinued MQ because of an AR at the first intake as follows: one with moderate dizziness and the other with severe dizziness, vomiting, and sleeplessness. The latter was admitted at the hospital for 24 hours parenteral antiemetic treatment. There were 12 HIV-negative patients who stopped the trial or discontinued MQ IPTp because of an AR (2.9%, P = 0.61 for the comparison between HIV-negative and HIV-infected women, χ2 test): All these women presented with vomiting, associated with dizziness in 4 women, with fatigue in 2 women, and with both in 5 women. Abortions and stillbirths occurred in 5 HIV-infected (4.9%) and 15 (3.6%) HIV-negative women, respectively (P = 0.57). In both trials, they were not significantly more frequent in the MQ arm than in the control arm. The data and safety monitoring board reviewed all cases and did not estimate them imputable to MQ.
Table 3 describes the ARs by type and according to HIV status at the woman level. Analysis was restricted to women who had received at least 2 MQ intakes during follow-up. The most frequently reported ARs were vomiting and dizziness. HIV-infected women reported less ARs than HIV-negative women (65% versus 78%, P = 0.009), especially vomiting (33% vs. 56%) and fatigue (15% vs. 42%). They also reported less moderate and severe ARs (22% vs. 41% of women, P = 0.002). ARs were much more frequent at the first MQ intake than at the second MQ intake, especially in HIV-negative patients as follows: 49% of HIV-infected women reported ARs after the first intake, versus 72% in HIV-negative women (P < 0.0001). This proportion decreased to 36% in HIV-infected women and 37% in HIV-negative women at the second intake (P = 0.90). The mean duration of ARs was 1.5 days (ranging from 1 hour for vomiting to 10 days for headache).
Factors associated with the reporting of an AR on each intake have been identified by multilevel logistic regression (Table 4). The first step of the analysis investigated the effect of covariates tested one by one. The risk for reporting an AR was markedly increased in the first MQ intake as compared with further intakes. HIV infection was associated with a decreased risk for reporting an AR [odds ratio (OR) = 0.45, P = 0.002], whereas an older maternal age, an older gestational age at MQ intake, and a higher education level increased the risk. Concerning HIV-related covariates other than HIV infection itself, ARs were more frequent when HIV viral load was detectable (OR = 2.2, P = 0.03) and when a new ARV regimen was initiated concomitantly with a MQ intake (OR = 2.17, P = 0.02). ARs were less frequent with ongoing ARVs (OR = 0.4, P = 0.01). In the second step (multivariate analysis), the effect of gestational age at MQ intake did not remain significant after adjustment for the timing of MQ intake. The effect of ongoing ARVs also vanished after adjustment for other HIV-related covariates. The factors which remained associated with the reporting of an AR were HIV infection (decreasing the risk), first (compared with second and third) MQ intake, detectable viral load, older age, and higher education level (increasing the risk). The effect of an advanced HIV disease was borderline. Similarly, the effect of the initiation of a new ARV regimen concomitantly with a MQ intake did not reach statistical significance (only 12 women were involved), though it was marked (OR > 3). In a complementary model considering the reporting of moderate and severe ARs only, the initiation of a new ARV regimen concomitantly with a MQ intake significantly increased the risk (OR = 9.5, P = 0.002; for whole sensitivity analysis, see Table, Supplemental Digital Content 1, http://links.lww.com/QAI/A336).
TABLE 4-a Factors As...Image Tools
TABLE 4-b Factors As...Image Tools
This study is to our knowledge the first to investigate the tolerability of MQ IPTp in HIV-infected pregnant women. It shows that in Beninese women, ARs were frequently reported (by more than 2/3 of the women), although these were mostly minor and did not impair adherence to MQ IPTp. Indeed, only 2% of the women discontinued MQ because of an AR. The most frequently reported ARs were dizziness and nausea or vomiting. No serious neuropsychiatric disorder like seizures or delirium, which are the most feared reactions,22 occurred. The prevalence of spontaneous abortions and stillbirths observed with MQ (5% altogether) was consistent with the prevalence observed in other cohorts of HIV-infected pregnant women receiving ARVs in Malawi and Mozambique (5%)23 or in Ivory Coast (5.2% and 6.7% prevalence for miscarriages and stillbirths, respectively).24
As compared with HIV-negative women, reported tolerability of MQ was better in HIV-infected pregnant women. This result was unexpected, as we first believed that an impaired health condition and polymedications would increase the occurrence of ARs in these women. In a trial testing SP IPTp in Kenya, ARs were as frequent in HIV-infected women as in HIV-negative women.25 To explain our results, we hypothesize that HIV-infected women, especially those with longer HIV disease and ARV course, might be more familiar with experiencing adverse events of any kind, so less prone to systematically report them. The most frequently reported ARs were dizziness and nausea, which are subjective symptoms, thus very reliant upon perception. It has been shown that patients treated for chronic diseases did not have the same feeling about ARs as patients treated for acute diseases,26 the latter being more concerned when an AR occurred. Accordingly, healthy (HIV negative) pregnant women who receive the antimalarial drug only for prevention are probably more bothered when an AR occurs than chronically sick HIV-infected women. Indeed, HIV-infected women with undetectable viral load at enrollment, who had experienced a longer HIV duration and ARV course, were at lower risk to report ARs. ARs also tended to be less reported by women with an advanced HIV disease (stage 3, 4), who had already experienced visible symptoms. Whatever the cause, such under-reporting in HIV-infected women is reassuring as it suggests that adherence to MQ IPTp would not be compromised by the occurence of ARs.
ARs were most frequently reported after the first MQ intake. We first thought that it could be due to a younger gestational age at first intake, as nausea and vomiting of early pregnancy could be mistaken for MQ ARs. However, only the effect of the timing of MQ intakes remained after adjustment in the multivariate model, rather suggesting a habituation phenomenon after the first IPTp dose.
In this study, moderate neuropsychiatric ARs to MQ were observed, such as dizziness and sleep disorders. Interestingly, adverse neuropsychiatric reactions to efavirenz seem comparable to those observed with MQ in type, frequency, and timing. They occur during the first weeks of treatment in more than 50% of treated patients and are dominated by dizziness, feeling of drunkenness, nightmares, and sleep disorders.27 Incidence and severity are also major during the first hours and days of treatment, then they gradually decrease and vanish by weeks 4–6 in most patients.27–30 Some studies have pointed out ARs linked with efavirenz drug plasma levels,31 but others found no correlation.28 According to this, an increased frequency of neuropsychiatric ARs could be feared when efavirenz and MQ are associated, especially at the beginning of the treatment. From our results, it seemed that women initiating a new ARV regimen less than 1 week before the intake of MQ had a higher risk for moderate and severe ARs. This finding relies only on 12 observations, however. Also, we cannot assess whether ARs were specifically attributable to MQ, to ARVs, or to the drug interactions between MQ and ARVs in these women.27 Finally, women initiating a new ARV regimen all attended specific pretherapeutic counseling concerning drug toxicity and schemes of administration to improve their adherence. This counseling might have increased their vigilance and ability to report ARs to any drug (ARVs and MQ).
Other covariates associated with the reporting of ARs were older age and higher level of education. The first has been classically pointed out as a determinant of drug tolerability,32 whereas the second is more original. Educated women could be more aware of their health condition. Older age and higher education have also been associated with improved adherence to ARVs.33
Finally, the study design did not allow a perfect comparability between HIV-infected and HIV-uninfected pregnant women because they were taken from different studies, different hospitals, and different observation periods. The collection of ARs also differed between the 2 studies, one being performed by home visits and the other by telephone calls. In mixed models, the site effect (hospital) was taken into account as a random effect, which allowed accounting for disparities between the populations of pregnant women in the different hospitals and for the effect of different investigators collecting ARs. Adjustments for socioeconomic and pregnancy-related characteristics were also performed to take into account the effect of all covariates that seemed different in the populations of HIV-negative and HIV-infected women, like age, rank of gestation, scholarship, and socioeconomic status. Finally, HIV infection remained associated with an under-reporting of ARs after all these adjustments and in sensitivity analyses. These results strengthen our findings that HIV-infected women are not at higher risk for ARs to MQ IPTp than HIV-negative pregnant women.
In conclusion, increasing resistance to SP is urging its replacement by other molecules in IPTp. The PACOME trial aims at assessing adequately the efficacy of cotrimoxazole for the prevention of malaria in HIV-infected pregnant women, which has never been done before in this population. The effectiveness of this policy could be compromised by the spreading of resistances to sulfa drugs and possible poor adherence due to its daily administration. MQ IPTp becomes a major alternative in this high-risk population, as our study provides very reassuring data on its use in HIV-infected pregnant women. If MQ IPTp had to be implemented for malaria prevention in HIV-infected women, we would encourage at least 1 week interval be observed between the initiation of a new ARV regimen and MQ intakes to improve the tolerability and adherence of both medications.
The authors kindly acknowledge all the study participants, the midwives, and physicians in the study sites and the Ouidah and PACOME study teams, especially Natacha Aïhonnou, Jeanne Amassé, Clarisse Briga, Claire Degnonvi, Florelle Kottin, and Rodolphe Ladekpo. The authors thank the Programme National de Lutte contre le Sida for technical support in biological investigations.
1. Dreyfuss ML, Msamanga GI, Spiegelman D, et al.. Determinants of low birth weight among HIV-infected pregnant women in Tanzania. Am J Clin Nutr. 2001;74:814–826.
2. Shulman CE, Dorman EK. Importance and prevention of malaria in pregnancy. Trans R Soc Trop Med Hyg. 2003;97:30–35.
3. Desai M, ter Kuile FO, Nosten F, et al.. Epidemiology and burden of malaria in pregnancy. Lancet Infect Dis. 2007;7:93–104.
4. Verhoeff FH, Brabin BJ, Hart CA, et al.. Increased prevalence of malaria in HIV-infected pregnant women and its implications for malaria control. Trop Med Int Health. 1999;4:5–12.
5. ter Kuile FO, Parise ME, Verhoeff FH, et al.. The burden of co-infection with human immunodeficiency virus type 1 and malaria in pregnant women in sub-saharan Africa. Am J Trop Med Hyg. 2004;71:41–54.
6. van Eijk AM, Ayisi JG, ter Kuile FO, et al.. Human immunodeficiency virus seropositivity and malaria as risk factors for third-trimester anemia in asymptomatic pregnant women in western Kenya. Am J Trop Med Hyg. 2001;65:623–630.
7. Ayisi JG, van Eijk AM, ter Kuile FO, et al.. The effect of dual infection with HIV and malaria on pregnancy outcome in western Kenya. AIDS. 2003;17:585–594.
8. Kfutwah A, Mary JY, Lemen B, et al.. Plasmodium falciparum infection significantly impairs placental cytokine profile in HIV infected Cameroonian women. PLoS One. 2009;4:e8114.
9. Ayisi JG, van Eijk AM, Newman RD, et al.. Maternal malaria and perinatal HIV transmission, western Kenya. Emerg Infect Dis. 2004;10:643–652.
10. WHO. WHO Expert Committee on Malaria. World Health Organ Tech Rep Ser. 2000;892:1–74.
11. WHO. Malaria and HIV/AIDS Interactions and Their Implications for Public Health Policies. Report of a Technical Consultation. Geneva, Switzerland: WHO; 2004.
12. ter Kuile FO, van Eijk AM, Filler SJ. Effect of sulfadoxine-pyrimethamine resistance on the efficacy of intermittent preventive therapy for malaria control during pregnancy: a systematic review. JAMA. 2007;297:2603–2616.
13. WHO. A Strategic Framework for Malaria Prevention and Control During Pregnancy in the African Region. Brazzaville, Democratic Republic of Congo: WHO regional office for Africa; 2004.
14. Briand V, Bottero J, Noel H, et al.. Intermittent treatment for the prevention of malaria during pregnancy in Benin: a randomized, open-label equivalence trial comparing sulfadoxine-pyrimethamine with mefloquine. J Infect Dis. 2009;200:991–1001.
15. Aubouy A, Fievet N, Bertin G, et al.. Dramatically decreased therapeutic efficacy of chloroquine and sulfadoxine-pyrimethamine, but not mefloquine, in southern Benin. Trop Med Int Health. 2007;12:886–894.
16. Nosten F, ter Kuile F, Maelankiri L, et al.. Mefloquine prophylaxis prevents malaria during pregnancy: a double-blind, placebo-controlled study. J Infect Dis. 1994;169:595–603.
17. Dooley KE, Flexner C, Andrade AS. Drug interactions involving combination antiretroviral therapy and other anti-infective agents: repercussions for resource-limited countries. J Infect Dis. 2008;198:948–961.
18. Skinner-Adams TS, McCarthy JS, Gardiner DL, et al.. HIV and malaria co-infection: interactions and consequences of chemotherapy. Trends Parasitol. 2008;24:264–271.
19. Uneke CJ, Ogbonna A. Malaria and HIV co-infection in pregnancy in sub-Saharan Africa: impact of treatment using antimalarial and antiretroviral agents. Trans R Soc Trop Med Hyg. 2009;103:761–767.
20. Akogbeto M, Modiano D, Bosman A. Malaria transmission in the lagoon area of Cotonou, Benin. Parassitologia. 1992;34:147–154.
21. UNAIDS. UNAIDS Report on the Global AIDS Epidemic 2010. Geneva, Switzerland: UNAIDS; 2010.
22. Chen LH, Wilson ME, Schlagenhauf P. Controversies and misconceptions in malaria chemoprophylaxis for travelers. JAMA. 2007;297:2251–2263.
23. Marazzi MC, Palombi L, Nielsen-Saines K, et al.. Extended antenatal use of triple antiretroviral therapy for prevention of mother-to-child transmission of HIV-1 correlates with favorable pregnancy outcomes. AIDS. 2011;25:1611–1618.
24. Ekouevi DK, Coffie PA, Ouattara E, et al.. Pregnancy outcomes in women exposed to efavirenz and nevirapine: an appraisal of the IeDEA West Africa and ANRS Databases, Abidjan, Cote d'Ivoire. J Acquir Immune Defic Syndr. 2011;56:183–187.
25. Parise ME, Ayisi JG, Nahlen BL, et al.. Efficacy of sulfadoxine-pyrimethamine for prevention of placental malaria in an area of Kenya with a high prevalence of malaria and human immunodeficiency virus infection. Am J Trop Med Hyg. 1998;59:813–822.
26. Lorimer S, Cox A, Langford NJ. A patient's perspective: the impact of adverse drug reactions on patients and their views on reporting. J Clin Pharm Ther. 2012;37:148–152.
27. Gutierrez-Valencia A, Viciana P, Palacios R, et al.. Stepped-dose versus full-dose efavirenz for HIV infection and neuropsychiatric adverse events: a randomized trial. Ann Intern Med. 2009;151:149–156.
28. Clifford DB, Evans S, Yang Y, et al.. Impact of efavirenz on neuropsychological performance and symptoms in HIV-infected individuals. Ann Intern Med. 2005;143:714–721.
29. Staszewski S, Miller V, Sabin C, et al.. Virological response to protease inhibitor therapy in an HIV clinic cohort. AIDS. 1999;13:367–373.
30. Molina JM, Ferchal F, Rancinan C, et al.. Once-daily combination therapy with emtricitabine, didanosine, and efavirenz in human immunodeficiency virus-infected patients. J Infect Dis. 2000;182:599–602.
31. Marzolini C, Telenti A, Decosterd LA, et al.. Efavirenz plasma levels can predict treatment failure and central nervous system side effects in HIV-1-infected patients. AIDS. 2001;15:71–75.
32. Marzolini C, Back D, Weber R, et al.. Ageing with HIV: medication use and risk for potential drug-drug interactions. J Antimicrob Chemother. 2011;66:2107–2111.
33. El-Khatib Z, Ekstrom AM, Coovadia A, et al.. Adherence and virologic suppression during the first 24 weeks on antiretroviral therapy among women in Johannesburg, South Africa—a prospective cohort study. BMC Public Health. 2011;11:88.
© 2012 Lippincott Williams & Wilkins, Inc.