Share this article on:

Low Prevalence of Detectable HIV Plasma Viremia in Patients Treated With Antiretroviral Therapy in Burkina Faso and Mali

Boileau, Catherine PhD*; Nguyen, Vinh-Kim MD, PhD; Sylla, Mohamed MSc; Machouf, Nima PhD§; Chamberland, Annie PhD; Traoré, Hamar A MD; Niamba, Pascal A MD, PhD; Diallo, Ismaël MD#; Maïga, Moussa MD**; Cissé, Mamadou MD††; Rashed, Sélim MD, PhD‡‡; Tremblay, Cécile MD, PhD

JAIDS Journal of Acquired Immune Deficiency Syndromes: August 1st, 2008 - Volume 48 - Issue 4 - p 476-484
doi: 10.1097/QAI.0b013e31817dc416
Epidemiology and Social Science

Background: Sub-Saharan Africa has seen dramatic increases in the numbers of people treated with antiretroviral therapy (ART). Although standard ART regimens are now universally applied, viral load measurement is not currently part of standard monitoring protocols in sub-Saharan Africa.

Methods: We describe the prevalence of inadequate virological response (IVR) to ART (viral load ≥ 500 copies/mL) and identify factors associated with this outcome in 606 HIV-positive patients treated for at least 6 months. Recruitment took place in 7 hospitals and community-based sites in Bamako and Ouagadougou, and information was collected using medical charts and interviews.

Results: The overall prevalence of IVR in treatment-naive patients was 12.3% and 24.4% for pretreated patients. There were no differences in rates of IVR according to ART delivery sites and time on treatment. Patients living farther away [odds ratio (OR) = 2.48; 95% confidence interval (CI) 1.40 to 4.39], those on protease inhibitor or nucleoside reverse transcriptase inhibitor regimens (OR = 3.23; 95% CI 1.79 to 5.82) and those reporting treatment interruptions (OR = 2.36; 95% CI 1.35 to 4.15), had increased odds of IVR. Immune suppression (OR = 3.32, 95% CI 1.94 to 5.70) and poor self-rated health (OR = 2.00; 95% CI 1.17 to 3.41) were also associated with IVR.

Conclusions: Sufficient expertise and dedication exist in public hospital and community-based programs to achieve rates of treatment success comparable to better-resourced settings.

From the *Institute for Health and Social Policy-McGill University, Montreal, Quebec, Canada; †Département de Médecine Sociale et Préventive-Université de Montréal, Montreal, Canada; ‡Département de Microbiologie et Immunologie-Université de Montréal, Montreal, Canada; §Clinique Médicale L'Actuel, Montréal, Quebec, Canada; Hôpital National du Point G-Unité de Médecine Interne, Bamako, Mali; ¶Centre Universitaire Hospitalier Yalgado-Ouédraogo-Unité de Dermatologie, Ouagadougou, Burkina Faso; #Centre Universitaire Hospitalier Yalgado-Ouédraogo-Unité de Médecine Interne, Ouagadougou, Burkina Faso; **Hôpital Gabriel Touré, Bamako, Mali; ††Centre de soins, d'animation et de conseils pour les personnes atteintes du VIH/SIDA (CESAC), Bamako, Mali; and ‡‡Unité de Santé Internationale-Université de Montréal, Montreal, Canada.

Received for publication January 3, 2008; accepted April 24, 2008.

Supported by the Canadian Institute of Health Research (Canadian Institute for Health Research-Recherche in Institut de Recherche en Santé du Canada).

Presented at the XVI International AIDS conference, August 13-18, 2006, Toronto.

Correspondence to: Catherine Boileau, PhD, Institute for Health and Social Policy-McGill University, 1130 Pine Avenue West, Montréal, Québec, Canada H3A 1A3 (e-mail:

Back to Top | Article Outline


Sub-Saharan Africa has seen dramatic increases in the numbers of people treated with antiretroviral therapy (ART). ART is now considered an integral part of the comprehensive response to HIV prevention, care, and support.1 Highly active antiretroviral therapy (HAART) requires that patients take at least 3 different antiretrovirals (ARVs) daily for life. Factors that are predictors of response to ART include adherence,2-4 stage of disease with immune status at baseline,5,6 treatment history,7,8 and sociodemographic factors such as age and gender.9 But, by far, the most important determinant of therapeutic success is adherence to treatment.

HIV infection has management requirements that are especially difficult to meet in sub-Saharan countries where resources are scarce, such as long-term follow-up and support of patients and frequent laboratory monitoring. Despite considerable challenges, ART treatment programs in resource-poor settings have success rates similar to those reported for developed countries despite low standards of monitoring.10 Early results from Senegal and Cote d'Ivoire,11-13 Uganda,14,15 and from South Africa16 were particularly encouraging. These studies show remarkably high rates of adherence and excellent virological and immunological outcomes, allaying initial concerns about difficulties with adherence and inadequate infrastructure. Less commented upon, however, has been the fact that these initial pilot projects were better resourced than what can be expected in more routine settings. As demonstration projects, many received technical assistance by expatriate physicians, had selected for the most motivated patients, and had direct links to academic resources in the North. However, as treatment access is expanded and decentralized, it will be difficult to maintain this intensity of resources, follow-up, and motivation. Moreover, free ARVs do not release the patient from the burden of additional expenditures. Laboratory monitoring may not be free, and even if it is, patients must often incur transportation costs to travel to appointments and often find that their food expenditures increase with improved health and appetite. This has been cited as a potential barrier to adherence.17

Standard HAART comprises 2 nucleoside reverse transcriptase inhibitors (NRTIs) and a nonnucleoside reverse transcriptase inhibitor (NNRTI) for first-line therapy and a protease inhibitor (PI)-based regimen for second-line treatment.1 Although these regimens are now universally applied, standards for laboratory monitoring of patients greatly depend on availability of basic laboratory services and appropriate training and quality control. These continue to be limited, and many regions in developing countries still rely on clinical status alone for treatment decisions. CD4 cell count remains the strongest predictor of HIV-related complications18 and is widely advocated by World Health Organization (WHO) for treatment initiation and monitoring. Viral load (VL) measurements1 and drug resistance surveillance19 are not recommended for wide use in public health management of ART and remain restricted because of cost. Nonetheless, a growing number of urban centers are equipped with facilities able to perform VL measurement, and this is used to monitor therapy for selected patients. In some major cities, even genotypic resistance testing is now available. Both VL monitoring20 and genotypic resistance testing at diagnosis21 and after treatment failure22,23 have been shown to improve clinical outcomes. This is because they permit early detection of a failing regimen and guide switch to a second line24 before a broad cross-class resistance has evolved.

Mali has an estimated 120,000 people living with HIV25 and a national program which is now treating over 10,000 with free ART. Laboratory monitoring is not covered by the program. Burkina Faso has an estimated 300,000 people living with HIV.26 There, most patients are receiving ART through nongovernmental mechanisms27; however, the public health system is rapidly developing ART delivery programs for HIV patients. Scale-up is occurring rapidly from a baseline of 4446 patients in April 2005 to a target of over 30,000 by 2008.

In this study, we focus on virological outcomes (concentrations of plasma HIV RNA) of HIV-positive patients treated with HAART in 7 hospitals and community-based sites in Bamako (Mali) and Ougadougou (Burkina Faso) to describe treatment efficacy and identify main factors associated with inadequate virological response (IVR), defined as HIV RNA levels higher than 500 copies/mL in observational settings.

Back to Top | Article Outline



A sample of 636 HIV-infected patients treated with HAART for more than 6 months voluntarily enrolled in this study between January and October 2005 in a pilot study to assess adherence, clinical, immunological, and virological outcomes. A convenience sample of patients were recruited in 3 public hospitals (n = 406) and 3 community-based organizations (CBOs) (n = 230) in Bamako (Mali) (n = 314) and Ouagadougou (Burkina Faso) (n = 322) to provide a representative mix of public, nongovernmental, community, and hospital-based ART delivery sites. As of 2005, sites participating in our study treated from 450 to 2372 patients.

To evaluate adherence and potential causes of treatment failure, patients were interviewed independently by trained staff including medical personnel or another person living with HIV using a standard questionnaire. The questionnaire covered demographic and relational characteristics, psychosocial factors, self-reported adherence, and serostatus disclosure. VL was obtained in all patients except for 30. We report the results of the remaining 606 patients for whom questionnaire and VL data were available.

HAART regimens were generally combinations of 2 NRTIs with 1 NNRTI or with 1 PI. NRTIs mostly used were lamivudine (3TC) (57.4%), D-drugs [didanosine (ddI) and stavudine (d4T)] (38.6%), and zidovudine (AZT) (25.1%). NNRTIs were efavirenz (43.9%) and nevirapine (30.7%). PIs included indinavir-unboosted (15.8%) and ritonavir-unbooster (1.0%) as sole PI. Combination ART including dual combination ART (Combivir, Duovir, or Avocomb) (30.2%) and triple combination ART (Triomune) (6.8%) were also available. Triple combination ART that was used was Triomune (Cipla), and dual therapies were a combination of 3TC + AZT drugs: Duovir-N (Cipla), Combivir (GlaxoSmithKline), and Avocomb (Ranbaxy).

Consent for participation was obtained from all patients, and the study was approved by the Ethics Committees in Bamako, Ouagadougou, and at University of Montreal.

Back to Top | Article Outline

Clinical Data

Clinical and immunological responses to HAART were measured as part of routine checkup and medical examination but varied slightly according to site. Start and stop dates for each ARV and the most recent clinical evaluation by physician and CD4 cell count were obtained from patients' medical files. HIV-1 RNA plasma viral load (PVL) was measured in all subjects. Immunological and virological failures (VFs) were defined according to WHO guidelines. Immunological suppression (IS) was defined as CD4 count below 200 cells/mm3 and virological failure as HIV-1 RNA more than 10,000 copies/mL. VL persisting ≥500 c/mL after more than 6 months is a clinically significant outcome as it is prognostic for treatment failure, we defined this as inadequate virological response. Body mass index (BMI) was calculated as weight (kilograms) divided by square of height (meters). Self-rated health (poor, neither good nor poor, and good) is a good predictor of health outcomes including mortality,28 hence, it was also included as a proxy for patients' general health.

Back to Top | Article Outline


Variables that were selected from the standardized questionnaire were grouped into 4 categories as follows: (1) sociodemographic characteristics including age, gender, food insecurity (having often skipped a meal or reduced the amount of food intake in the last 6 months because of money), literacy (being able to read and write in French), and distance from ART delivery site (≤1 hour or >1 hour); (2) context of ART care including country and type of site (public hospital or CBO); (3) treatment factors including prescribed ART, ART history (treatment naive or experienced), drug formulation (3 different pills or fixed-dose ART), and patient-reported drug-related adverse events (digestive, skin rashes, arthralgia, neurological, and fatigue); and (4) behavioral factors including adherence (patient-reported number of doses missed, yesterday, the day before yesterday, and in the last 7 days, and treatment interruption) and alcohol use (yes or no).

Back to Top | Article Outline

Laboratory Methods

CD4 cells were quantified by flow cytometry (Becton Dickinson FACSCaliburTM system in Ouagadougou and FACScan in Bamako). Plasma HIV-1 RNA loads were measured in the Centre Hospitalier de l'Universite de Montreal laboratory using a quantitative reverse transcriptase polymerase chain reaction (Ultrasensitive Assay version 1.5; Roche Molecular Systems, Branchburg, NJ), which has a lower limit of detection of 50 c/mL.

Back to Top | Article Outline


The proportions of patients with IVR were compared with patients with low (<500 c/mL) or undetectable PVL (<50 c/mL) using the χ2 statistic. The level of significance was 0.05. Factors associated with PVL ≥500 copies/mL were analyzed in a logistic regression model using multivariate approach with staggered entry. We modeled the associations with IVR by including sociodemographic factors (country, type or ART delivery site, distance from ART delivery site, age, gender, literacy, and food insecurity), treatment factors (prescribed HAART, previous HAART usage, use of fixed-dose ART, and ART-related adverse events), behavioral factors (previous treatment interruption, 7-day adherence, and alcohol use), and clinical characteristics (BMI, CD4 cell count, and self-rated health) in 4 sequential steps. Variables from each step were selected by using stepwise backward strategy in which statistical criteria for entry and retention of variables in each model were P ≤ 0.05 and P ≤ 0.10, respectively. Fifty-seven cases were removed from the final model due to missing information on any of the variables included in the analysis. This resulted in 549 cases selected in the final model. Results are presented as odds ratios (ORs) with 95% confidence intervals (95% CIs).

Back to Top | Article Outline



This analysis is based on 606 patients treated with HAART for at least 6 months for whom VL and questionnaire data were available. Mean age of the study population was 38.0 ± 7.9 years, and 66.8% were females. Patients were equally distributed in each country, and more than two-thirds (64%; n = 388) were treated in public hospitals as compared with CBOs (Fig. 1). Patients were divided by time on treatment into 3 groups: 6-12 months (22.4% of patients), 12-24 months (48.5%), and >24 months (29.2%). The majority of patients were on NNRTI-based regimens (83.7%), 15.4% were on PI-based regimens, and 5 patients were on a triple NRTI regimen. Fixed-dose medications with double (Duovir-N, Combivir, Avocomb) or triple (Triomune, Trizivir) ARVs were prescribed in, respectively, 30.2% and 6.8% of the patients (Table 1).







Eighty-two (15.8%) patients had previously received ARVs. Pretreated patients were more likely to be females (15.9% versus 9.0%; P = 0.021) to be treated in Burkina Faso (17.3% versus 9.3%; P = 0.004) and in CBOs (19.3% versus 10.4%; P = 0.002). For 52 of the 82 patients who had previously received an ART, we were able to ascertain their first regimen. The majority (90.4%) had received triple ARV as an initial regimen; only 4 patients had been previously treated with NRTI dual therapy (both NRTIs), and 1 woman reported having taken zidovudine (AZT) monotherapy during pregnancy. For their second ART, 23 patients (44.2%) were receiving an NNRTI-based regimen, 23 (44.2%) were on a PI-based regimen, and 6 (11.5%) were on a triple NRTI regimen. In half the patients (26), their regimen was a fixed-dose ART (eg, Combivir). Changes in treatment were based on clinical or immunological criteria because VL quantification was not available at the time of the study. The majority of treatment switches (n = 34, 65.4%) had resulted in the substitution of only one drug. Substitutions of 2 or all drugs were observed in 15 (28.8%) and 3 (5.8%) patients, respectively. Forty-four percent of the treatment changes (23 cases) involved switching ARVs without switching classes either for toxicity or supply reasons. Among the 29 patients who switched classes, 18 switched from a PI-based regimen to an NNRTI-based regimen, 4 switched from an NRTI-only regimen to an NNRTI-based regimen, 4 switched from an NNRTI-based regimen to a PI-based regimen, and 1 patient switched from an NRTI-only regimen to a PI-based regimen. In all but one case, patients were on unboosted PIs.

Treatment adherence in the entire sample was generally good with 391(64.5%) patients reporting perfect adherence in the last week and 471(77.7%) never having interrupted their treatment regimen. Of the 493 patients for whom BMI was measured, 287(58.2%) were considered to have a normal BMI, 65(13.2%) patients were considered underweight, and 141(29.6%) were overweight (Table 1).

Back to Top | Article Outline

Virological and Immunological Outcomes

More than three quarters (77.2%) of the patients had suppressed PVL below the detection level (50 c/mL), 8.7% had low detectable VL (50-499 c/mL), 14.0% had IVR, of which 70.6% (60 patients) had VF (Fig. 1). Prevalence of IVR ranged from 7.9% to 20.7% across ART delivery sites, but the differences in the proportion of patients in each VL category between countries, types of sites, and actual ART delivery sites were not statistically significant (Fig. 1). Proportions of patients with VL ≥500 c/mL (IVR) were not different between time-on-treatment groups. However, differences were observed between treatment regimens and between pretreated and naive patients. Overall prevalence of virological suppression was 79% in naive patients and 68% in pretreated patients. Virological suppression was also more frequent in NNRTI-treated patients (79.6%) as compared with patients treated with other regimens (62.9%; P < 0.001). In Figure 2, rates of IVR are disaggregated by time on treatment, type of regimen, and history of ART use. Higher proportions of naive patients had IVR in the first year if they were on PI as compared with NNRTI. More so, a previous ART use was significantly associated with higher rates of IVR. Pretreated patients taking an NNRTI regimen were significantly more likely to have IVR in the first 2 years of treatment whereas those taking a PI regimen were significantly more likely to have IVR after their second year of treatment.



One hundred forty-four (25.9%) patients did not increase their CD4 counts over the threshold of 200 cells/mm3 (IS), even though two thirds of these patients had undetectable VL. IS suppression was more common in women (31.2% versus 23.2%; P = 0.004) and among those who reported food insecurity in the last 6 months (33.1% versus 19.7%; P ≤ 0.001). There were no significant differences in rates of IS according to time on treatment, treatment history, and type of HAART regimen, but patients on fixed-dose ART (Duovir + NNRTI or Triomune) were less likely to have IS (19.4% versus 30.0%; P = 0.007). Furthermore, rates of IS correlated well with BMI and self-rated health.

Back to Top | Article Outline

Correlates of IVR

Bivariate analysis (Table 1) indicates that there are no differences in the rates of IVR between men and women or between literate and illiterate patients. However, patients who reported living more than one hour away from their ART delivery site were more likely to have an IVR. IVR was also more frequent in patients treated with PI- and NRTI-only regimens than in patients taking NNRTIs, in patients taking more pills, and in treatment-experienced patients. None of the specific drug-related events or cumulative number of adverse events (not shown) was associated with IVR. Treatment interruptions were associated with IVR whereas partial adherence in the last week was not. As expected, clinical factors such as low BMI, self-rated health, and immunosuppression were also associated with IVR. Although only marginally significant, the relationship between BMI and HIV RNA seemed curvilinear with overweight patients having the best virological outcomes (6.1% IVR; P = 0.061) as compared with others with higher or lower BMI.

Results of the multivariable logistic regression models are reported in Table 2. Regression models suggest that factors related to patients' sociodemographic characteristics, treatment regimen, adherence, and clinical status may all be independently associated with high HIV RNA concentration. Model 1 describes the effect of patients' sociodemographic characteristics on virological outcomes. Age, gender, or literacy and food insecurity were selected into this model; however, living more than one hour away from ART delivery site increased the odds of having an IVR (OR = 1.66; 95% CI 1.01 to 2.73). In model 2, treatment-related factors are included. In this model, PI- or NRTI-only regimens (OR = 3.12; 95% CI 1.81 to 5.38) and previous history of ART (OR = 2.32; 95% CI 1.27 to 4.22) were associated with increased odds of having high PVL when compared with NNRTI regimens and treatment-naive patients, respectively. Once history of treatment interruption was included in model 3, the odds of having an IVR in pretreated patients slightly dropped and the association lost its significance, but the effect of other variables remained relatively unchanged. Contrary to our expectations, this finding indicates that the mechanism by which distance from site affects virological outcome is not explained by its effect on ART interruption. ART interruption was a significant predictor of poor virological outcome (OR = 2.63; 95% CI 1.57 to 4.41) once adjusted for model 1 and 2 variables. Finally, IS (3.32; 95% CI 1.94 to 5.70) and poor self-rated health (OR = 2.00; 95% CI 1.17 to 3.41) significantly increased the odds of having an IVR, and their effects were independent of distance to ART delivery sites, treatment regimen, previous ART use, and treatment interruptions. Surprisingly, the odds of IVR for patients living farther away from ART delivery sites increased with this final adjustment, suggesting that this variable interacts with either of the 2 clinical variables.



Back to Top | Article Outline


This study describes virological and immunological responses to HAART in routine care in a large group of patients receiving treatment in a variety of hospitals and CBOs in Burkina Faso29 and Mali. This multicenter 2-country study is significant as it describes treatment outcomes in representative health care settings. Overall, 76.6% and 74.6% of patients in the first, second year, and later had adequate immunological and virological outcomes, respectively.

The overall prevalence of undetectable VL in treatment-naive patients was 79% and 68% in pretreated patients. Rates of virological suppression in naive and pretreated patients remained stable across ART delivery sites and time on treatment. Results are encouraging given that they were neither obtained from randomized trials nor obtained from highly monitored and supervised programs. Moreover, results reported herein are superior to some of the results obtained in resource-limited contexts and are quite comparable to those obtained from middle- or high-income countries. For example, rates of undetectable VL at 12 months of 40% and 50% have been reported in Uganda30 and Côte d'Ivoire.13 High rates of virological suppression (80%) were observed in the paying ART clinic in Blantyre (Malawi), but patient population is unlikely to be representative of HIV-positive people in need of treatment because more than half were lost to follow-up after initiating treatment.31 According to the number provided by the national HIV program in Mali (December 2005), 69.78% of the patients who have initiated treatment are still being routinely followed in the 3 HAART delivery sites in Bamako. Although similar information was not available for Burkina Faso, we hypothesize that patients may be more representative of HIV-infected people in our setting because the free access to treatment is likely to improve patient retention. In Brazil, 25% of treatment-naive patients had IVR at 6 months after initiation of treatment with nonbrand name drugs.32 In a heterogeneous clinic population in the United States, one quarter of patients had IVR over a period of 2 years.33 Strikingly similar results were observed in patients initiating HAART in North America and Europe, where 83% had attained HIV RNA <500 c/mL at 6 months.34 However, slightly higher virological outcomes were obtained from the Swiss cohort, where 90% of the naive patients and 79% of the pretreated patients had suppressed their viral RNA at 12 months.8

The appropriate threshold for switching regimens remains to be determined and is a research priority; in high-income countries, a confirmed detectable VL at any level after being suppressed is sufficient to switch therapy. In resource-limited settings, WHO guidelines recommend that 10,000 c/mL may be an appropriate threshold for switching.1 Resistance testing was conducted on a subsample of 46 patients who had high viremia (858-160,000 c/mL); 33 (71%) had VL >10,000 consistent with WHO guidelines of VF and an indication for treatment switch. However, in this subgroup, 39 patients (85%) already harbored HIV with major resistance mutations.35 These results suggest that an important proportion of patients with high viremia already have future treatment options compromised and would require access to second-line treatments.

Immunological outcomes are a concern in these patients because 1 of 4 patients had CD4 cell count below 200 cells/μL. Immune reconstitution is challenging; some studies have reported that some patients may never have CD4 counts that exceed 200 cells/mm3 and thus never leave the zone of AIDS-defining immunosuppression.36 Such low counts are associated with significantly increased odds of contracting opportunistic infections or clinical progression to AIDS.37,38 Our data also show that food insecurity was associated with IS, as reported by others.39 This supports recommendations for beginning treatment earlier in contexts where immune reconstitution may be a challenge, for instance in contexts where malnutrition, poor hygiene, and infectious diseases are prevalent.40

As would be expected, treatment interruptions, low CD4 count, and poor self-rated health were associated with treatment failure, consistent with the SMART study on treatment interruption.41 The results of studies evaluating differences in responses based on gender or socioeconomic background have been inconsistent. Nevertheless, poor socioeconomic conditions and low literacy have not been reported to be major obstacles for proper adherence and therapeutic success in resource-limited settings42 or in more marginalized populations in high-income countries.43 This is in line with our findings showing a lack of association between literacy or demographic factors and treatment failure. However, geographic distance from ART delivery site was identified as predictor for poor virological outcome in our sample, and contrary to our expectations, the effect of this factor was not explained by the fact that it may be a barrier for proper adherence.17 Because ART delivery mechanisms in this study varied greatly with regard to infrastructure, human resources, management culture, services to person living with HIV, and patient characteristics, neither selection bias nor site-specific factors would explain these results. We believe these results to be generalizable to the whole population of people living with HIV in these 2 countries.

A rate of 14.7% of IVR response to ART suggests that, in these patients, the risk of treatment failure and resistance is significant. IVR was associated with treatment regimen, history of ART, and previous treatment interruption. We measured adherence with 2 questions: having ever interrupted the current treatment regimen and having forgotten one or more doses in the last week. The former was associated with IVR whereas the latter was not. Part of the reason for the absence of association between self-reported last week adherence and virological response may be that this measure is more susceptible to social desirability bias than reporting events that date further back. Another potential explanation for this is that patients may have been more likely to report treatments interruptions if these events consisted of longer or more significant numbers of pills missed and less likely to report missing 1 pill in the last week. Nevertheless, our results suggest that events outside the patients' control leading to treatment interruptions may also contribute to IVR as does actual patient behavior. This underscores the importance of developing strategies to support adherence over the long term.

PI-based regimen and NRTI-only regimens have poorer efficacy than NNRTI-based regimens-even when adjusted for ART history and adherence. NRTI-only regimens are known to be less effective than HAART regimens with a PI or an NNRTI,33 and PIs have also been reported to be less effective than NNRTI regimens.44 The association between IVR and treatment with a PI-containing regimen is explained by the fact that the majority (93.8%) of these patients were taking unboosted indinavir, which is known to have low genetic barrier and to be less effective than boosted PIs.45 Indinavir is a challenging drug to use as it requires strict dosing and, in our experience, is particularly poorly tolerated with patients often experiencing recurrent nausea, vomiting, and diarrhea. This is reflected in a higher switch rate from PIs than from NNRTIs. As treatment switches were not guided by VL response but by clinical factors, the majority of switches were for tolerability reasons. Only 2 patients had a complete regimen switch for presumed treatment failure.

This study was limited by its cross-sectional design. Single time-point measurement of VL made it impossible to determine whether a VL of more than 500 c/mL was a single “blip” or represented persistent viremia, which selects for resistance and is associated with treatment failure.46 Similarly, it was not possible to distinguish whether patients with CD4 <200 cells/mm3 were true immunological failures as these patients may still be able to reconstitute their immune systems to above 200 cells/mm3 with more time. This was a convenience sample of patients recruited at ART delivery sites during routine examinations; none of the respondents were hospitalized at the time of the study. Although refusal rates were low, patients volunteering for this study might have better health.

Back to Top | Article Outline


These results suggest that, even without significant external inputs, sufficient expertise and dedication exist in public hospital and community-based programs to achieve rates of treatment success comparable to better-resourced settings. That self-reported treatment interruptions were associated with IVR whereas self-reported partial adherence in the past week suggests that structural factors outside the control of patients may be more important than behavioral factors in this resource-poor setting. In this study, structural factors outside the control of patients included geographic distance from care sites, availability and accessibility of ART including type of regimen (eg, use of regimens containing indinavir as an unboosted PI), and absence of VL monitoring. Because IVR is not detected in the absence of VL monitoring, in these minority of patients with treatment failures, future salvage options may be compromised as continued treatment with a suboptimal regimen that will lead to broad cross-class resistance. Further research is needed on determinants of treatment success and IVR. Adherence must be measured at the level of ART availability and accessibility and individual behavior. Even in the absence of accessible VL monitoring, ongoing operations research using VL as a gold standard will be needed to develop easily usable screening tools to predict IVR. This suggests that ARV rollout programs need to include strengthening of health care delivery capacity, adherence support programs, and monitoring and drug distribution.

Back to Top | Article Outline


We would like to acknowledge the work done by every researcher and clinician collaborating in the ATARAO project. These include Prof. Flabou Bougoudogo (INRSP-Bamako), Prof. Hamar Alassane Traoré (Hôpital National du Point G, Service de Médecine Interne Bamako), Dr Souleymane Ag Aboubacrine (Hôpital National du Point G, Service de Médecine Interne Bamako), Dr Daouda Minta (Hôpital National du Point G, Service des Maladies Infectieuses, Bamako), Dr Seydou Coulibaly (Hôpital National du Point G, Pharmacie, Bamako), Prof. Moussa Maïga (Hôpital Gabriel Touré, Bamako), Dr Abdoulaye Kallé (Hôpital Gabriel Touré, Bamako), Dr Dramane Koné (Hôpital Gabriel Touré, Pharmacie, Bamako), Dr Mamadou Cissé (CESAC, Bamako), Dr Ousmane Traoré (CESAC, Pharmacie, Bamako), Prof. Somita Keïta (CNAM, Bamako), Dr Mohamed Sylla (Projet SIDA-3, Bamako), Issoufou Tiendrébéogo (AAS, Ouagadougou), Dr Samuel Koala (AAS, Ouagadougou), Laure Salembéré (ALAVI, Ouagadougou), Dr Alain Ouermi (ALAVI, Ouagadougou), Prof. Joseph Drabo (CHU-Yalgado-Ouédraogo, Service de Médecine Interne, Ouagadougou), Dr Ismaël Diallo (CHU-Yalgado-Ouédraogo, Service de Médecine Interne, Ouagadougou), Prof. Adama Traoré (CHU-Yalgado-Ouédraogo, Service de Dermatologie, Ouagadougou), Dr Pascal Niamba (CHU-Yalgado-Ouédraogo, Service de Dermatologie, Ouagadougou), Dr Maria-Victoria Zunzunegui (Université de Montréal, Unité de Santé Internationale). We would also like to thank the interviewers (Ali Djerma, Idrissa Coulibaly, Cheickné Touré, Saliou Mahamadou), laboratory technicians (Issa Cissé; INRSP and Leon Sawadogo; CNLAT), and pharmacists for their contribution to this work. Finally, we express our appreciation to all the participants who generously volunteered to give a blood sample and be interviewed for this study.

Back to Top | Article Outline


1. World Health Organization. Antiretroviral Therapy for HIV in Adults and Adolescents: Recommendations for a Public Health Approach. WHO Press; Geneva, Switzerland. 2006:1-134.
2. Phillips AN, Miller V, Sabin C, et al. Durability of HIV-1 viral suppression over 3.3 years with multi-drug antiretroviral therapy in previously drug-naive individuals. AIDS. 2001;15:2379-2384.
3. Bartlett J, DeMasi R, Quinn J, et al. overview of the effectiveness of triple combination therapy in antiretroviral-naïve HIV-1 infected adults. AIDS. 2001;15:1369-1377.
4. Harrigan PR, Hogg RS, Dong WW, et al. Predictors of HIV drug-resistance mutations in a large antiretroviral-naive cohort initiating triple antiretroviral therapy. J Infect Dis. 2005;191:339-347.
5. Wood E, Hogg RS, Yip B, et al. Impact of baseline viral load and adherence on survival of HIV-infected adults with baseline CD4 cell counts > or = 200 cells/microl. AIDS. 2006;20:1117-1123.
6. Skowron G, Street JC, Obee EM. Baseline CD4(+) cell count, not viral load, correlates with virologic suppression induced by potent antiretroviral therapy. J Acquir Immune Defic Syndr. 2001;28:313-319.
7. Abgrall S, Duval X, Joly V, et al. Clinical and immunologic outcome in patients with human immunodeficiency virus infection, according to virologic efficacy in the year after virus undetectability, during antiretroviral therapy. Clin Infect Dis. 2003;37:1517-1526.
8. Ledergerber B, Egger M, Opravil M, et al. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Swiss HIV Cohort Study. Lancet. 1999;353:863-868.
9. Anastos K, Gange SJ, Lau B, et al. Association of race and gender with HIV-1 RNA levels and immunologic progression. J Acquir Immune Defic Syndr. 2000;24:218-226.
10. Ivers LC, Kendrick D, Doucette K. Efficacy of antiretroviral therapy programs in resource-poor settings: a meta-analysis of the published literature. Clin Infect Dis. 2005;41:217-224.
11. Laurent C, Diakhate N, Gueye NF, et al. The Senegalese government's highly active antiretroviral therapy initiative: an 18-month follow-up study. AIDS. 2002;16:1363-1370.
12. 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.
13. Djomand G, Roels T, Ellerbrock T, et al. Virologic and immunologic outcomes and programmatic challenges of an antiretroviral treatment pilot project in Abidjan, Cote d'Ivoire. AIDS. 2003;17 (Suppl 3):S5-S15.
14. Kebba A, Atwine D, Mwebaze R, et al. Therapeutic responses to AZT + 3TC + EFV in advanced antiretroviral naive HIV type 1-infected Ugandan patients. AIDS Res Hum Retroviruses. 2002;18:1181-1187.
15. Byakika-Tusiime J, Oyugi JH, Tumwikirize WA, et al. Adherence to HIV antiretroviral therapy in HIV+ Ugandan patients purchasing therapy. Int J STD AIDS. 2005;16:38-41.
16. 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.
17. Hardon AP, Akurut D, Comoro C, et al. Hunger, waiting time and transport costs: time to confront challenges to ART adherence in Africa. AIDS Care. 2007;19:658-665.
18. Garcia F, de Lazzari E, Plana M, et al. Long-term CD4+ T-cell response to highly active antiretroviral therapy according to baseline CD4+ T-cell count. J Acquir Immune Defic Syndr. 2004;36:702-713.
19. Walensky RP, Weinstein MC, Yazdanpanah Y, et al. HIV drug resistance surveillance for prioritizing treatment in resource-limited settings. AIDS. 2007;21:973-982.
20. Hughes MD, Johnson VA, Hirsch MS, et al. Monitoring plasma HIV-1 RNA levels in addition to CD4+ lymphocyte count improves assessment of antiretroviral therapeutic response. ACTG 241 Protocol Virology Substudy Team. Ann Intern Med. 1997;126:929-938.
21. Boden D, Hurley A, Zhang L, et al. HIV-1 drug resistance in newly infected individuals. JAMA. 1999;282:1135-1141.
22. Weinstein MC, Goldie SJ, Losina E, et al. Use of genotypic resistance testing to guide HIV therapy: clinical impact and cost-effectiveness. Ann Intern Med. 2001;134:440-450.
23. Torre D, Tambini R. Antiretroviral drug resistance testing in patients with HIV-1 infection: a meta-analysis study. HIV Clin Trials. 2002;3:1-8.
24. Jones S, Klotman ME. Impact of genotypic resistance testing on physician selection of antiretroviral therapy. J Hum Virol. 2001;4:214-216.
25. UNAIDS/WHO. Epidemiological Fact Sheets on HIV/AIDS and Sexually Transmitted Infections. Mali; August 2006.
26. UNAIDS/WHO. Epidemiological Fact Sheets on HIV/AIDS and Sexually Transmitted Infections. Burkina Faso: UNAIDS/WHO; Geneva, Switzerland. August 2006.
27. Nguyen VK, Grennan T, Peschard K, et al. Antiretroviral use in Ouagadougou, Burkina Faso. AIDS. 2003;17:S109-S111.
28. Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav. 1997;38:21-37.
29. Lagarde E, Congo Z, Meda N, et al. Epidemiology of HIV infection in urban Burkina Faso. Int J STD AIDS. 2004;15:395-402.
30. Weidle PJ, Malamba S, Mwebaze R, et al. Assessment of a pilot antiretroviral drug therapy programme in Uganda: patients' response, survival, and drug resistance. Lancet. 2002;360:34-40.
31. van Oosterhout JJ, Bodasing N, Kumwenda JJ, et al. Evaluation of antiretroviral therapy results in a resource-poor setting in Blantyre, Malawi. Trop Med Int Health. 2005;10:464-470.
32. May SB, Barroso PF, Nunes EP, et al. Effectiveness of highly active antiretroviral therapy using non-brand name drugs in Brazil. Braz J Med Biol Res. 2007;40:551-555.
33. Robbins GK, Daniels B, Zheng H, et al. Predictors of antiretroviral treatment failure in an urban HIV clinic. J Acquir Immune Defic Syndr. 2007;1:30-37.
34. May MT, Sterne JA, Costagliola D, et al. HIV treatment response and prognosis in Europe and North America in the first decade of highly active antiretroviral therapy: a collaborative analysis. Lancet. 2006;368:451-458.
35. Sylla M, Chamberland A, Boileau C, et al. ATARAO Group. Characterization of drug resistance in antiretroviral-treated patients infected with HIV-1 CRF02_AG and AGK subtypes in Mali and Burkina Faso. Antivir Ther. 2008;13:141-148.
36. Dronda F, Moreno S, Moreno A, et al. Long-term outcomes among antiretroviral-naive human immunodeficiency virus-infected patients with small increases in CD4+ cell counts after successful virologic suppression. Clin Infect Dis. 2002;35:1005-1009.
37. Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet. 2002;360:119-129.
38. Ledergerber B, Egger M, Erard V, et al. AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy: the Swiss HIV Cohort Study. JAMA. 1999;282:2220-2226.
39. Paton NI, Sangeetha S, Earnest A, et al. The impact of malnutrition on survival and the CD4 count response in HIV-infected patients starting antiretroviral therapy. HIV Med. 2006;7:323-330.
40. Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1-infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet. 2006;367:817-824.
41. Julg B, Goebel FD. Treatment interruption in HIV therapy: a SMART strategy? Infection. 2006;34:186-188.
42. Mills EJ, Nachega JB, Buchan I, et al. Adherence to antiretroviral therapy in sub-Saharan Africa and North America: a meta-analysis. JAMA. 2006;296:679-690.
43. Paasche-Orlow MK, Cheng DM, Palepu A, et al. Health literacy, antiretroviral adherence, and HIV-RNA suppression: a longitudinal perspective. J Gen Intern Med. 2006;21:835-840.
44. Weiser SD, Guzman D, Riley ED, et al. Higher rates of viral suppression with nonnucleoside reverse transcriptase inhibitors compared to single protease inhibitors are not explained by better adherence. HIV Clin Trials. 2004;5:278-287.
45. Boyd MA, Srasuebkul P, Khongphattanayothin M, et al. Boosted versus unboosted indinavir with zidovudine and lamivudine in nucleoside pre-treated patients: a randomized, open-label trial with 112 weeks of follow-up (HIV-NAT 005). Antivir Ther. 2006;11:223-232.
46. Sungkanuparph S, Groger RK, Overton ET, et al. Persistent low-level viraemia and virological failure in HIV-1-infected patients treated with highly active antiretroviral therapy. HIV Med. 2006;7:437-441.

viral load; antiretroviral treatment monitoring; sub-Saharan Africa; immune suppression; public hospital; community-based organization

© 2008 Lippincott Williams & Wilkins, Inc.