Patterns and Predictors of First-Line Antiretroviral Therapy Modification in HIV-1-Infected Adults in a Large Urban Outpatient Cohort in Nigeria

Abstract

Objective:

We described the magnitude, type, and factors associated with first-line antiretroviral therapy (ART) modification in HIV-1-infected adults on ART in Jos, Nigeria.

Method:

Data on 6309 patients initiated on first-line ART between January 2004 and December 2006 were analyzed retrospectively. Factors predictive of modification to initial ART were assessed by chi-square and multivariable logistic regression analysis.

Results:

Overall, 5212 (83%) included patients incurred a modification (73.3% drug substitution and 9.7% drug switch) to their initial first-line ARV regimen during a median (interquartile range) follow-up period of 7 (3-8) years. Drug substitutions of zidovudine (ZDV) were less likely than of tenofovir (TDF; adjusted odd ratio [AOR] 0.6; 95% confidence interval [CI]: 0.51-0.71), and Drug substitutions of efavirenz (EFV) were more likely than of nevirapine (NVP)-containing (AOR 1.82; 95% CI: 1.42-2.33) regimens. Predictors of switch to second-line regimen include older age (AOR 2.05; 95% CI: 1.68-2.51), CD4 count ≤100 cells/mm³ (AOR 1.89; 95% CI: 1.49-2.37), EFV compared to NVP (AOR 1.38; 95% CI: 1.02-1.88), and drug toxicity (AOR 1.90; 95% CI: 1.48-2.43).

Conclusion:

Modification to initial ART was common in this study. Further evaluation of the contribution of guideline changes on regimen modification and treatment outcomes is recommended.

Keywords

antiretroviral therapy regimen durability switch substitution

Introduction

The benefits of treating HIV infection with combination antiretroviral therapy (cART) have been well described, including durable viral suppression, quantitative and qualitative immune recovery, reductions in morbidity and mortality, and prevention of HIV transmission.^1

–6 As survival of HIV-infected patients continues to improve, more attention is being focused on factors that ensure long-term success of ART and ART programs.

Maximal regimen durability with regard to safety and efficacy is a critical factor for long-term success of ART.⁷ Modification to ART presents a number of challenges such as limited therapeutic alternatives, viral cross-resistance between possible alternatives, and overlapping of toxicity between potential alternatives, in addition to adherence and quality-of-life challenges.^8

–11 In sub-Saharan Africa (SSA), where limited drug options are common due to limited resources, regimen durability takes on an even more important role than in industrialized countries. Most public sector ART programs, which are donor driven, often adopt treatment guidelines that promote a public health approach to the management of HIV, and providers are constrained to operate with a limited formulary.¹² It becomes imperative to carefully select ARV regimen at initiation and understand the determinants of regimen change in order to ensure treatment success.

There are limited data on the magnitude and predictors of regimen durability in ART programs in Nigeria, where there has been rapid scaling up of ART services over the past decade.¹³ In this study, which spans 8 years, we aimed to determine the magnitude and type of modification of first-line ART among a large cohort of HIV-1-infected patients who initiated ART at the Jos University Teaching Hospital (JUTH) as well as to identify baseline demographic and clinical characteristics that predict first-line regimen modification.

Methods

Setting

The comprehensive HIV care and treatment program at JUTH located in North Central Nigeria, started in 2002 with support from the Federal Government of Nigeria. In 2004, there was a rapid scaling up of ART services with further support from the Harvard School of Public Health (HSPH) and AIDS Prevention Initiative in Nigeria (APIN), with funding from the US President’s Emergency Plan for AIDS Relief (PEPFAR). The JUTH HIV clinic provides comprehensive HIV care and treatment services for the city of Jos and serves as a referral center for health facilities in Plateau State and other states in the region. Currently, the JUTH HIV clinic provides HIV care and treatment services to over 14 000 HIV-infected adults and children.

Study Population

The study population included HIV-1-infected, treatment-naive adults (aged 15 years and older) who initiated their first ART between January 2004 and December 2006. All patients included in the analysis initiated nonnucleoside reverse transcriptase inhibitor (NNRTI)-based ART, which included an NNRTI (nevirapine [NVP] or efavirenz [EFV]) plus 2 nucleoside reverse transcriptase inhibitors (NRTIs), often stavudine (d4T), didanosine (DDI), zidovudine (ZDV), tenofovir (TDF), or abacavir (ABC) plus lamivudine (3TC). Patients were excluded from the analysis if they made only 1 pharmacy pickup for their antiretroviral (ARV) medications, if they initiated first-line ART with a triple NRTI regimen or a protease inhibitor (PI)-based regimen, or if they were ART experienced.

Standard of Care

Per the standard care at the JUTH HIV clinic, all patients had their HIV-1 infection status confirmed positive by Western blot. The decision to initiate ART was based on the Nigerian National Adult ART guidelines.¹⁴ At the time of the study, ART was initiated based on a CD4 count of 350 cells/mm³ with HIV-related symptoms or a CD4 count <200 cells/mm³ regardless of the symptoms.¹⁴ Selection of the initial first-line ARV regimen was also based on the national ART guidelines. An electronic database (FileMaker Pro, v10; FileMaker, Inc, Santa Clara, California, USA) of patients’ demographic, clinical, pharmacy, and laboratory data designed by the Harvard PEPFAR/APIN program was utilized in the clinic. Dispensing of ART occurred monthly and at scheduled clinical visits, and each time a medication was dispensed at the clinic (pharmacy pickup), data were documented in the electronic pharmacy database. The ART efficacy and toxicity was monitored regularly based on the standard clinical procedures. Free-of-charge ARV drugs and laboratory monitoring were provided to the patients.

Definition of Terms and Outcome Variable

The outcome variable included the modification of a first-line ARV regimen, which is described as drug substitution or drug switch. This was evaluated between the first documented pharmacy pickup at the time of ART initiation (baseline) and the last pharmacy pickup occurring within the end of the study period (December 31, 2011). Drug substitution was defined as the replacement of one or more drugs in the first-line ARV regimen (NRTI or NNRTI) with another drug from the same ARV class (NRTI or NNRTI), for example, an NNRTI substitution from NVP to EFV or an NRTI substitution from d4T to ZDV. However, a change between 3TC and emtricitabine (or vice versa) was not classified as a drug substitution because these 2 NRTIs are considered therapeutically interchangeable.¹⁵ Drug switch was defined as a change from the first-line NNRTI-based ART to the second-line PI-based ARV regimen. Per treatment guidelines,¹⁴ drug switches are usually done after confirmation of virologic failure. Virologic failure was defined as plasma viral load (VL) of above 1000 copies/mL on 2 consecutive VL measurements after 6 to 12 months of cART initiation, with documented optimal ART adherence.

Variables Collected

All data were extracted from the Harvard PEPFAR/APIN Plus program database. The patient data assessed pre-ART (baseline) included demographics (sex and age), World Health Organization (WHO) clinical stage of HIV, coinfection with hepatitis B virus (HBV), coinfection with active tuberculosis (TB) at baseline, and CD4 count. Regimen data assessed included initial first-line ARV regimen (by NNRTI and NRTI backbone) and last ART regimen dispensed during the study period (by NNRTI or PI and NRTI backbone). Data on reported drug toxicity and the date of documentation were also collected. The HIV-1 VL assay was performed using Roche COBAS Ampiclor HIV-1 Monitor, version 1.5 (Roche Diagnostics GmbH, Mannheim, Germany), while the CD4 count was determined using CyFlow (Partec, Munster, Germany). Hepatitis B surface antigen (HBsAg) was determined using enzyme immunoassay (Monolisa HBsAg Ultra3; Bio-Rad, Hercules, USA).

Ethical Considerations

All patients enrolled in the clinic provided written informed consent for the use of their data for research. The ethical committee of JUTH approved retrospective analysis of existing data as part of the program treatment and study protocol, and the use of secondary data was approved by HSPH.

Statistical Analysis

Continuous variables were described by means and standard deviation or median and interquartile range (IQR, presented as the upper value of the 25% and 75% quartiles), as appropriate based on data distribution, and categorical variables were described using frequencies and proportions. Univariate analysis to evaluate the baseline patient or regimen factors associated with the outcome variable (drug substitution or drug switch) was performed using chi-square. For the multivariate analysis, the following predictor variables were stratified: age (<35 years versus ≥35 years, which was the median age for the population), WHO clinical stage (stage 3 or 4 versus stage 1 or 2), and CD4 count (≤100 cells/mm³ versus >100 cells/mm³). World Health Organization clinical stages 3 and 4 and CD4 count ≤100 cell/mm³ were representative of severely immunocompromised patients. Multivariate logistic regression analysis was performed to evaluate independent predictors of drug substitution or drug switch (dependent variables); results are presented using an adjusted odds ratio (AOR) with 95% confidence intervals (CIs). All variables that were significant in the univariate analysis to a P value <.25 were included in a multivariate model as well as age and sex to account for any residual confounders. All analyses were performed using IBM SPSS Statistics for Windows, Version 20 (IBM Corp, Armonk, New York, USA).

Results

Baseline Patient Characteristics

A total of 6309 patients were eligible for inclusion in the study and were followed for a median (IQR) of 7 (3-8) years. Forty-one patients initiated ART during the specified period but were excluded from the analysis, that is, 2 were ART experienced and 39 initiated a triple NRTI regimen. Table 1 summarizes patient demographics and baseline clinical and regimen characteristics overall and for those with drug substitution or drug switch. In all, the study population was largely female (66%) with a median (IQR) age of 35 (30-42) years and a median (IQR) baseline CD4 count of 154 (76-259) cells/mm³. Reflective of ART guidelines and ARV drug availability at the time of inclusion, most patients initiated ART with NVP as the NNRTI (91%) and either ZDV/3TC (39%) or d4T/3TC (34%) as the NRTI backbone.

Table 1.

Univariate Analysis of Demographic and Clinical Characteristics Associated with a Drug Substitution or Drug Switch among Patients on ART in Jos, Nigeria.^a

Characteristics	Group	N = 6309 (Percentage of total)	Substitution		Switch
Characteristics	Group	N = 6309 (Percentage of total)	N = 4623/6309 (Percentage of Row)	P Value	N = 589/6309 (Percentage of Row)	P Value
Sex	Male	2153 (34)	1604 (75)	.11	225 (11)	.03
Sex	Female	4156 (66)	3019 (73)		364 (9)
Age, years	≤35	3055 (48)	2262 (74)	.29	194 (6.4)	<.001
Age, years	>35	3254 (52)	2361 (73)		392 (12.1)
Log viral load, copies/mL	<5	816 (12.9)	560 (68.6)	<.001	72 (8.8)	.56
	≥5	312 (4.9)	200 (64.1)		34 (10.9)
	Unknown	5181 (82.1)	3863 (74.6)		480 (9.3)
HBV positive	No	4410 (70)	3169 (72)	<.001	417 (10)	.86
	Unknown	812 (13)	596 (73)		75 (9)
	Yes	1087 (17)	858 (79)		97 (9)
WHO clinical stage^b	1	2038 (36)	1482 (73)	.03	158 (8)	.01
	2	1777 (31)	1262 (71)		183 (10)
	3	1523 (27)	1147 (75)		158 (10)
	4	357 (6)	271 (76)		42 (12)
CD4 count^b, cells/mm³	≤50	1016 (16)	734 (72)	.25	145 (14)	<.001
	51-100	1046 (17)	752 (72)		121 (12)
	101-200	1872 (30)	1366 (73)		166 (9)
	>200	2326 (37)	1738 (75)		151 (7)
NNRTI	Efavirenz	558 (9)	443 (79)	<.001	76 (14)	<.001
NNRTI	Nevirapine	5751 (91)	4183 (73)		513 (9)
NRTI	Abacavir	307 (5)	277 (90)	<.001	25 (8)	.89
	Zidovudine	2426 (39)	1382 (57)		221 (9)
	Stavudine	2137 (34)	1879 (88)		206 (10)
	Didanosine	322 (5)	292 (91)		29 (9)
	Tenofovir	1117 (18)	793 (71)		108 (10)
Drug toxicity documented	No	5684 (90)	4177 (74)	.25	486 (9)	<.001
Drug toxicity documented	Yes	625 (10)	446 (71)		103 (17)

Abbreviations: NNRTI, nonnucleoside reverse transcriptase inhibitor, NRTI, nucleoside reverse transcriptase inhibitor; ART, antiretroviral therapy; HBV, hepatitis B virus; WHO, World Health Organization.

^a All characteristics are at ART initiation except drug toxicity.

^b Totals may not sum to 6309 due to missing data.

Initial First-Line ART Modification

Overall, 5212 (83%) included patients incurred a modification to their initial first-line ARV regimen, that is, 4623 (73.3%) experienced an NRTI and/or NNRTI drug substitution and 589 (9.3%) experienced a drug switch from NNRTI-based to PI-based ART.

Drug substitution

Of the 4623 patients who incurred a drug substitution, 2789 (44%) involved a single NRTI substitution, 484 (8%) involved a single NNRTI substitution, and 1350 (22%) involved both an NRTI and an NNRTI substitution. Tables 1 and 2 summarize factors associated with NRTI and/or NNRTI drug substitutions on bivariate and multivariate analyses, respectively. In bivariate analysis (Table 1), drug substitution rates were significantly higher among patients who initiated EFV-based ART (79% versus 73% for NVP-based ART, P < .001) and were lowest among those who initiated ZDV/3TC (57%, P < .001) across NRTI backbones. Differences were also noted by HBV and across WHO clinical stage groups. After adjustment for confounding factors (Table 2), several regimen and patient baseline factors were identified as predictors of drug substitution. Patients who initiated EFV-based ART were at an 82% greater odds of experiencing a drug substitution than those who initiated NVP-based ART (AOR 1.82; 95% CI: 1.42-2.33). Across NRTIs, patients who received ABC, d4T, or DDI were at 4.4-, 3.5-, and 3.8-fold, respectively, greater odds to experience a drug substitution than those who received TDF, whereas those who received ZDV were 40% less likely to have a drug substitution (AOR 0.6; 95% CI: 0.51-0.71). Baseline patient characteristics identified as independent predictors of drug substitution included HBV coinfection, and more advanced HIV disease, evidenced by CD4 count ≤100 cells/mm³ or WHO clinical stage 3 or 4. Documentation of drug toxicity throughout the study period was not associated with drug substitution.

Table 2.

Adjusted Logistics Regression Analysis of Characteristics Associated with Drug Substitution or Drug Switch after Commencing First-Line Antiretroviral Therapy.

		Substitution		Switch
		AOR (95% CI)	P Value	AOR (95% CI)	P Value
Sex	Male	1.06 (.92-1.22)	.42	.91 (.74-1.11)	.35
Sex	Female	Reference		Reference
Age, years	>35	.79 (.68-.89)	<.001	2.05 (1.68-2.51)	<.001
Age, years	≤35	Reference		Reference
HBV positive	Yes	1.57 (1.32-1.87)	<.001	NE
HBV positive	No/unknown	Reference
CD4 count^a	≤100	1.37 (1.89-1.59)	<.001	1.87 (1.49-2.37)	<.001
CD4 count^a	>100	Reference		Reference
WHO stage^b	3 and 4	1.25 (1.09-1.42)	.01	1.21 (.99-1.46)	.05
WHO stage^b	1 and 2	Reference		Reference
NNRTI	Efavirenz	1.82 (1.42-2.33)	<.001	1.38 (1.02-1.88)	.04
NNRTI	Nevirapine	Reference		Reference
NRTI	Abacavir	4.39 (2.88-6.71)	<.001	.82 (.51-1.34)	.44
	Zidovudine	.60 (.51-.71)	<.001	.97 (.74-1.25)	.78
	Stavudine	3.48 (2.86-4.24)	<.001	.99 (.75-1.29)	.92
	Didanosine	3.82 (2.52-5.79)	<.001	.82 (.52-1.29)	.39
	Tenofovir	Reference	<.001	Reference	.86
Drug toxicity documented	Yes	.88 (.72-1.08)	.21	1.90 (1.48-2.43)	<.001
Drug toxicity documented	No	Reference		Reference

Abbreviations: AOR, adjusted odd ratio; CI, confidence interval; HBV, hepatitis B virus; NE, not included in the model; NNRTI, nonnucleoside reverse transcriptase inhibitor; NRTI, nucleoside/nucleotide reverse transcriptase inhibitor; WHO, World Health Organization.

^a Forty-nine patients without data were excluded.

^bSix hundred and fourteen patients without data were excluded.

Drug switch

Tables 1 and 2 summarize factors associated with drug switch for the 589 (9.4%) patients who switched from NNRTI-based to PI-based ART. In the univariate analysis (Table 1), drug switch was significantly higher in male than in female patients (11% versus 9%; P = .03) and more common in patients >45 years (13%; P < .001) than across the different age-groups. Although there was no significant difference in the occurrence of drug switch across the NRTI groups, it was significantly higher in those on EFV than those on NVP-based ART (14% versus 9%; P < .001). There were more drug switches in patients who had documentation of drug toxicity than those with no documented drug toxicity (17% versus 9%; P < .001). After adjustment for confounders (Table 2), the odds of drug switch were doubled in patients aged >35 compared to those ≤35 years (AOR 2.05; 95% CI: 1.68-2.51). Other independent predictors of drug switch were advanced HIV disease (CD4 count ≤100 cells/mm³ and WHO clinical stage 3 or 4 disease) and EFV compared to NVP-based ART (AOR 1.38; 95% CI: 1.02-1.88). The odd of drug switch was 90% higher in patients with documented drug toxicity than those without.

Discussion

This study evaluated a large clinical cohort of patients initiating ART, who were followed for an 8-year time period. A high incidence of ART modification was observed among patients included in our analysis, wherein the within-class drug substitutions of the NRTI and/or NNTRI accounted for the majority, while drug switch from NNRTI-based to PI-based second-line ART was low. This study, to the best of our knowledge, is the first report of ARV regimen durability in Nigeria.

In terms of frequency, the high rates of first-line ART modification observed in this study are similar to that reported in other cohort studies in industrialized countries^16,17 but lower than a recent study in Kenya.¹⁸ Although we did not quantify the impact of treatment guideline changes on the magnitude of first-line ART modification in our study, the contribution of changes in policy was likely a significant contributor to the high rates of drug substitution observed in our study. Between 2004 and 2006, a wide variety of ARV regimens were available for use within the Harvard PEPFAR/APIN Plus program. However, with time, regimen options were streamlined based on the changing programmatic ART guidelines, which were largely based on WHO and national treatment guidelines, program experience, and new scientific evidence. According to the WHO 2003 treatment guideline, d4T/3TC or AZT/3TC in combination with NVP or EFV was the preferred first-line regimen, although EFV was largely avoided in women of childbearing potential or in pregnant women.¹⁹ In 2006, however, AZT/3TC or TDF/3TC in combination with NVP or EFV became the preferred first-line regimen.²⁰ In the 2010 guideline, the use of d4T was discouraged, given the high rates of lactic acidosis, lipodystrophy, peripheral neuropathy, and other mitochondrial toxicity–related adverse effects observed with its use.²¹ As a result, about 88% of patients in our cohort who initiated a d4T-containing regimen during the study period experienced a drug substitution as their regimen was modified to a less toxic NRTI backbone. Similarly, the use of DDI, that has a similar toxicity profile as d4T, was discontinued in 91% of all patients initiated on DDI. Additionally, programmatic ARV quantification assumptions gave priority to extant treatment guidelines and limited the choice of regimens available for treatment. Hence the ARV regimens of stable patients were modified in some cases based on product availability, moving toward regimens with more favorable toxicity profiles or more convenient dosage formulations. The adherence, quality of life, and virologic implication of regimen modification in a stable patient in this setting are unknown and needs to be further investigated.

Interestingly, in our study, we observed that patients who initiated EFV-based ART were at significantly greater odds of experiencing a drug substitution and drug switch as compared to those who initiated NVP-based ART. Prior studies have reported conflicting findings to this point, either in agreement²² or in disagreement with ours.^23,24 Our finding is important, considering that the fixed-dose combination consisting of EFV, 3TC + emtricitabine, and TDF is now the preferred first-line ART regimen.²⁵ However, given the limitations of our study, we are unable to characterize the reason for drug substitution, which is important in evaluating whether the higher rate of modification noted with EFV is a result of efficacy or toxicity. Further investigation without our cohort is required.

Considering the NRTI background, the higher rates of drug substitution noted among patients who initiated d4T-, DDI-, or ABC-containing therapy are well supported by changes to ART guidelines in favor of safer and now more affordable options,²⁵ such as TDF. However, we also observed a lower risk of drug substitution among patients who initiated first-line ART with an AZT-containing regimen compared to TDF, which differs from prior studies,^26

–29 and is contrary to the most recent ART guidelines that recommend TDF as the preferred NRTI.²⁵ From our study, we are unable to discern the rationale for drug substitution and, therefore, unable to evaluate whether differences in toxicity, programmatic recommendations, ARV drug availability, or other factors impacted this finding. However, given the shift toward TDF-containing regimen as the preferred first-line ART in SSA,²⁵ it remains important to further evaluate these patterns of drug usage.

Furthermore, we observed that patients with severe immunodeficiency at the start of ART (CD4 count ≤100 cells/mm³) and advanced HIV disease (WHO clinical stages 3 and 4) had a higher likelihood of drug substitution and drug switch. This association has been reported in a recent study in industrialized countries.²⁶ It lends credence to the findings which suggest that treatment initiation at higher CD4 count thresholds is advantageous in achieving better treatment outcomes, which is also favored by current treatment guidelines.²⁵ We also identified HBV coinfection as an important factor for drug substitution. This evolved over the course of the study, in that clinician became more aware of preferred treatment practices for HBV and simultaneously, TDF availability became more widespread.^27
–29 Both tending toward the expected shift in NRTI backbone for many of these patients who were (hopefully) moved to more optimal regimens.

Unlike the Kenya study, we found a lower likelihood of drug substitution in older patients (>35 years),¹⁸ whereas the odds of drug switch almost doubled in older patients in our study, suggesting increased risk of virologic failure in older patients at ART initiation, synonymous with the finding of Fong et al.³⁰

Unlike a previous study,³¹ surprisingly we did not find a significant association between drug substitution and drug toxicity; however, drug toxicity was predictive of drug switch. Drug side effects are associated with poor adherence to therapy,³² virologic failures, and higher drug switch rates,^33,34 as observed in this study.

Most patients in our study cohort were retained on NNRTI-based first-line ART during the study period, suggestive of high NNRTI-based regimen durability in our patient cohort. This pattern has been reported in other African settings.³⁵ Based on the treatment guidelines, PI-based ART is reserved for patients who failed the first-line, NNRTI-based regimen and who, therefore, require second-line ART. Although the observed drug switch rates of less than 10% in our cohort is suggestive of highly effective and durable NNRTI-based ART, it is worth noting that the requirement for documented adherence and a repeat VL >1000 copies/mL may have resulted in a lag of drug switching relative to virologic failure. To the latter point, maintaining patients on a failing regimen has implications for the accumulation of more resistant mutant strains.^36
–38 Promptly identifying patients with virological failure who are in need of a drug switch is important to the preservation of treatment options and the achievement of optimal treatment outcomes.

Our study had some limitations that must be carefully considered when interpreting our findings. First, while the criteria for drug switch were established in the clinic, the reasons for drug substitutions were poorly documented. Thus, it remains unclear whether there were additional clinical (eg, drug interactions or opportunistic coinfections), laboratory (eg, laboratory values associated with drug toxicities), or programmatic factors (eg, treatment guideline changes or ARV supply chain considerations) that were independent predictors of drug substitution. Second, the impact of TB coinfection on drug-substitution or drug-switch rates was not evaluated as data were only available for 18% of our cohort, which is not a representative sample.

Conclusion

Among patients initiating first-line ART between 2004 and 2006 in Nigeria, over 80% experienced a regimen modification within a period of 8 years, commonly an NRTI and/or NNRTI drug substitution. Drug-switch rates from NNRTI- to second-line PI-based regimens were low (<10%), suggesting high durability of NNRTI-based regimen in our setting. Significant predictors of switch to second-line regimen were older age, greater immunosuppression, EFV compared to NVP, and drug toxicity. Our study highlights the dynamic nature of first-line ARV regimens used in our setting over time, which were likely driven by the dynamic nature of ART guidelines over this period. The need to evaluate the contribution of guideline changes on regimen modification and treatment outcomes is desirable in this setting.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded in part by the US Department of Health and Human Services, Health Resources and Services Administration (U51HA02522), and the Centers for Disease Control and Prevention (CDC) through a cooperative agreement with APIN (PS 001058). We are grateful to the Fogarty International Center, Northwestern University AIDS International Training and Research Program (Grant Numbers R24TW008878 & D43TW007995) for training support and mentorship toward this manuscript. The contents are solely the responsibility of the authors and do not represent the official views of the funding institutions.

References

Palella

Jr Deloria-Knoll

Chmiel

. Survival benefit of initiating antiretroviral therapy in HIV-infected persons in different CD4+ cell strata. Ann Intern Med. 2003;38(8):620–626.

Palella

Jr Delaney

Moorman

. Declining mortality and morbidity in HIV-infected ambulatory patients. N Engl J Med. 1998;328:853–860.

Weidle

Malamba

Mwebaze

. Assessment of a pilot antiretroviral drug therapy programme in Uganda: patients’ response, survival, and drug resistance. Lancet. 2002;360(9326):34–40.

Cohen

Chen

McCauley

. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365(6):493–505.

Wester

Kim

Bussmann

Avalos

Ndwapi

Peter

. Initial response to highly active antiretroviral therapy in HIV-1C–infected adults in a public sector treatment program in Botswana. J Acquir Immune Defic Syndr. 2005;40(3):336–343.

Idigbe

Adewole

Eisen

. Management of HIV-1 infection with a combination of nevirapine, stavudine, and lamivudine: a preliminary report on the Nigerian antiretroviral program. J Acquir Immune Defic Syndr. 2005;40(1):65–69.

Palella

Jr Chmiel

Moormana

Holmberga

and the HIV Outpatient Study Investigators. Durability and predictors of success of highly active antiretroviral therapy for ambulatory HIV-infected patients. AIDS. 2002;16(12):1617–1626.

Drechsler

William

Powderly

. Switching effective antiretroviral therapy: a review. Clin Infect Dis. 2002;35(10):1219–1230.

Becker

Rachlis

Gill

. Successful substitution of protease inhibitors with SUSTIVA (efavirenz) in patients with undetectable plasma HIV-1 RNA levels: results of a prospective, randomized, multicenter, open label study (DMP 266-049) [Abstract 20]. Programs and Abstracts of the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy (Toronto). Washington, DC: American Society for Microbiology; 2000.

10.

Martinez

Romeu

Garcia-Viejo

. An open, randomized study on the replacement of HIV-1 protease inhibitors by efavirenz in chronically suppressed HIV-1-infected patients with lipodystrophy [Abstract 668]. Programs and Abstracts of the 8th Conference on Retroviruses and Opportunistic Infections (Chicago). Alexandria, VA: Foundation for Retroviruses and Human Health; 2001.

11.

Moyle

Baldwin

Mandalia

. Changes in metabolic parameters and body shape after replacement of protease inhibitor with efavirenz in virologically controlled HIV-1–positive persons: single arm observational cohort. J Acquir Immune Defic Syndr. 2001;28(4):399–401.

12.

The ART-LINC Collaboration of the International Databases to Evaluate AIDS (IeDEA). Antiretroviral therapy in resource-limited settings 1996 to 2006: patient characteristics, treatment regimens and monitoring in sub-Saharan Africa, Asia and Latin America. Trop Med Int Health. 2008;13(7):870–879.

13.

Federal Republic of NIGERIA: GLOBAL AIDS RESPONSE Country Progress Report Nigeria GARPR [Internet]. 2012. http://www.unaids.org/en/dataanalysis/knowyourresponse/countryprogressreports/2012countries/Nigeria%202012%20GARPR%20Report%20Revised.pdf. Accessed February 18, 2014.

14.

Federal Ministry of Health. National guidelines for HIV and AIDS treatment and care in adolescents and adults. Abuja, Nigeria [Internet]. 2007. http://www.who.int/hiv/amds/Nigeria_adult_2007.pdf. Accessed February 18, 2014.

15.

Ford

Shubbert

Hill

. Comparative efficacy of lamivudine and emtricitabine: a systematic review and meta-analysis of randomized trials. PLoS One. 2013;8(11):e79981.

16.

Elzi

Marzolini

Furrer

. Treatment modification in human immunodeficiency virus-infected individuals starting combination antiretroviral therapy between 2005 and 2008. Arch Intern Med. 2010;170(1):57–65.

17.

Lodwick

Smith

Youle

. Stability of antiretroviral regimens in patients with viral suppression. AIDS. 2008;22(9):1030–1046.

18.

Inzaule

Otieno

Kalyango

. Incidence and predictors of first line antiretroviral regimen modification in western Kenya. PLoS One. 2014;9(4):e93106. doi:10.1371/journal.pone.0093106.

19.

WHO. Scaling up antiretroviral therapy in resource-limited settings: treatment guidelines for a public health approach 2003 revision [Internet]. 2003. http://www.who.int/hiv/pub/prev_care/en/arvrevision2003en.pdf?ua=1. Accessed May 6, 2014.

20.

WHO. Antiretroviral therapy for HIV infection in adults and adolescents: recommendations for a public health approach, 2006 revision [Internet]. 2006. http://www.who.int/hiv/pub/guidelines/artadultguidelines.pdf?ua=1. Accessed May 6, 2014.

21.

WHO. Antiretroviral drugs for treating pregnant women and preventing HIV infection in infants: recommendations for a public health approach, 2010 version [Internet]. 2010. http://whqlibdoc.who.int/publications/2010/9789241599818_eng.pdf?ua=1. Accessed May 6, 2014.

22.

Reekie

Reiss

Ledergerber

. Comparison of the Long-term Durability of Nevirapine, Efavirenz and Lopinavir in Routine Clinical Practice in Europe. HIV Med. 2011;12(5):259–268.

23.

Takuva

Evans

Zuma

Okello

Louwagie

. Comparative durability of nevirapine versus efavirenz in first-line regimens during the first year of initiating antiretroviral therapy among Swaziland HIV-infected adults. Pan Afr Med J. 2013;15:5.

24.

Pillay

Ford

Shubber

Ferrand

. Outcomes for Efavirenz versus Nevirapine-containing regimens for treatment of HIV-1 infection: a systematic reviewand meta-analysis. PLoS One. 2013;8(7):e68995.

25.

WHO. Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection: Recommendations for a public health approach, June 2013 [Internet]. 2013. http://apps.who.int/iris/bitstream/10665/85321/1/9789241505727_eng.pdf. Accessed May 6, 2014.

26.

The Antiretroviral Therapy Cohort Collaboration (ART-CC). Durability of first ART regimen and risk factors for modification, interruption or death in HIV-positive patients starting ART in Europe and North America 2002–2009. AIDS. 2013;27(5):803–813.

27.

Arribas

Pozniak

Gallant

. Tenofovirdisoproxilfumarate, emtricitabine, and efavirenz compared with zidovudine/lamivudine and efavirenz in treatment-naive patients: 144-week analysis. J Acquir Immune Defic Syndr. 2008;47(1):74–78.

28.

Gallant

DeJesus

Arribas

, for the Study 934 Group. Tenofovir DF, Emtricitabine, and Efavirenz vs. Zidovudine, Lamivudine, and Efavirenz for HIV. N Engl J Med. 2006;354(3):251–260.

29.

Njuguna

Orrell

Kaplan

Bekker

Wood

Lawn

. Rates of switching antiretroviral drugs in a primary care service in South Africa before and after introduction of tenofovir. PLoS One. 2013;8(5):e63596.

30.

Fong

Cheng

Vujovic

Hoy

. Factors associated with virological failure in a cohort of combination antiretroviral therapy-treated patients managed at a tertiary referral centre. Sex Health. 2013;10(5):442–447.

31.

Njuguna

Orrell

Kaplan

Bekker

Wood

Lawn

. Rates of switching antiretroviral drugs in a primary care service in South Africa before and after introduction of tenofovir. PLoS One. 2013;8(5):e63596.

32.

Friedland

. Adherence: the Achilles heel of highly active antiretroviral therapy. Improving Manage HIV Dis. 1997;5:13–15.

33.

Van Roon

Verzijl

Juttmann

. Incidence of discontinuation of highly active antiretroviral therapy (HAART) and its determinants. J Acquir Immune Defic Syndr Hum Retrovirol. 1999;20(3):290–294.

34.

d'Arminio

Monforte A

Testa

Adorni

. Clinical outcome and predictive factors of failure of highly active antiretroviral therapy in antiretroviral-experienced patients in advanced stages of HIV-1 infection. AIDS. 1998;12(13):1631–1637.

35.

Pujades-Rodriguez

O’Brien

Humblet

Calmy

. Second-line antiretroviral therapy in resource-limited settings: the experience of Médecins Sans Frontières. AIDS. 2008;22(11):1305–1312.

36.

Rawizza

Chaplin

Meloni

. Accumulation of protease mutations among patients failing second-line antiretroviral therapy and response to salvage therapy in Nigeria. PLoS One. 2013;8(9):e73582.

37.

Tenorio

Smith

Kuritzkes

. HIV-1-infected antiretroviral-treated patients with prolonged partial viral suppression. J Acquir Immune Defic Syndr. 2003;34(5):491–496.

38.

Nettles

Kieffer

Simmons

. Genotypic resistance in HIV-1-infected patients with persistently detectable low-level viremia while receiving highly active antiretroviral therapy. Clin Infect Dis. 2004;39(7):1030–1037.