Virological, Weight, and Drug Resistance Outcomes among Patients Initiating a Dolutegravir-based First Line ART Regimen in Zimbabwe

Abstract

Objective:

Dolutegravir (DTG)-based ART is being scaled up in Africa. However, clinical experience with DTG and patterns of HIV drug resistance (HIVDR) are sparse in Zimbabwe. We assessed virological, weight, and HIVDR outcomes among individuals initiating on a DTG-based ART.

Design:

We conducted a prospective cohort study among HIV infected adult (≥18 years old) individuals attending care at Parirenyatwa hospital, Harare, Zimbabwe between October 2021 and April 2023.

Methods:

Viral load and weight were assessed at both baseline and follow-up (≥24weeks) visits. HIVDR genotyping was performed by Sanger sequencing among participants with virological failure (VL≥1000 copies/ml) at follow up visit. Factors associated with weight gain were determined using logistic regression analysis on STATA 17.0.

Results:

172 participants were enrolled in the study. The median (IQR) age was 39(29-48) years whilst the median (IQR) CD4 cell count and log₁₀ VL at enrolment was 175(58-328) cells/mm³ and 5.41(4.80-5.74) respectively. After a median (IQR) duration of 27 (25-30) weeks on DTG, of the 131 participants with follow-up VL data available, 129 (98%) had VL<1000 copies/ml and among the 2 (2%) participants with VL≥1000 copies/ml , no emergent HIVDR was detected. We observed a significant increase in weight among the participants. The average weight gain was 5.25 kgs (p<0.0001). Baseline CD4 cell count ≥200 cells/mm³ was significantly associated with at a smaller weight gain (OR= 0.26; 95% CI: 0.12-0.58, p=0.001).

Conclusion:

We found high virological suppression and an increased weight among people initiating on DTG in a resource-limited setting. Encouragingly, HIVDR to DTG remains rare.

Introduction

Due to the wide availability of effective antiretroviral therapy (ART), the effect of the HIV/AIDS epidemic has dramatically reduced in numerous low-and middle-income countries (LMICs). In these regions, public health approaches to ART had been used to simplify and streamline care for people living with HIV (PLHIV) given the prevailing limited therapeutic options. Following the current WHO ART guidelines, several LMICs are scaling up the single-tablet regimen of tenofovir disoproxil fumarate (TDF), lamivudine and dolutegravir (TLD) as part of their preferred first-line ART for all PLHIV [1]. The efficacy of dolutegravir (DTG)-based ART has previously been reported in randomized clinical trials [2–5], showing superiority compared to protease inhibitor (PI) and non-nucleotide reverse transcriptase inhibitor (NNRTI)-based regimens in ART-naïve people [6–8]. DTG with its high potency, tolerability and high genetic barrier to resistance is expected to address the existing problems of virological failure and the emergence of HIV drug resistance (HIVDR) in LMICs [9]. TLD has become available in LMICs since 2018 and more than 90% of PLHIV were anticipated to be on DTG by the end of 2022 in this region [10,11].

Zimbabwe has likewise adopted the use of DTG as the preferred first-line ART as of July 2019 [12] and up to 78% of PLHIV have been on TLD since March 2021[13]. However, clinical experience with DTG in the country has not been documented. Despite the optimal safety profile of DTG demonstrated in clinical trials, concerns including metabolic disorders and weight gain associated with DTG in real-world settings have recently arisen [14–16]. As DTG-based ART becomes widely available in LMICs, there is a need for continued monitoring and surveillance of HIVDR. Clinical experience with DTG in Zimbabwe is lacking and little is known about the patterns of emerging resistance to DTG to inform public health approach to effective ART optimization. Growing data in routine care settings are needed. Therefore, we investigated virological outcome, weight profile and pattern of resistance among adult PLHIV on TLD in Zimbabwe.

Materials and Methods

Study design, setting and population

This was a prospective cohort study investigating the clinical outcomes of adult PLHIV initiating on DTG-based ART between October 2021 and April 2023. Consenting participants were ART-naive adults (aged ≥18 years) initiating on DTG or patients re-initiating first-line DTG-based ART after defaulting for at least 3 months. Participants were attending care at Parirenyatwa Hospital Family Care Centre (PHFCC), a tertiary level setting, in Harare, Zimbabwe.

Participant enrolment procedure

Individuals were consecutively invited to participate in the study and then screened for eligibility. Written informed consent was obtained from eligible participants prior to recruitment. Socio-demographic and clinical data (age, sex, marital status, height, weight, education and occupation) and ART history data (prior ART exposure) were extracted via an interview-based questionnaire. Missing data and baseline CD4 cell count data were extracted from the clinic’s electronic medical records. Upon enrolment, whole blood was collected in an ethylenediamine tetra acetic acid (EDTA) tubes for viral load (VL) measurement.

Participant follow up at ≥ 24 weeks after ART initiation

All participants had their phone number and next of kin phone number recorded in the clinic’s electronic medical records at enrolment as per the clinic protocol for future clinical and refill visits. As per the Zimbabwe ART guidelines, individuals initiating or re-initiating on DTG-based first-line ART should have a clinical and refill visit at 24 weeks. All enrolled participants were seen at or after 24 weeks follow up where besides other routine clinical tests, weight was measured and whole blood collected in a EDTA tube for VL measurement and genotypic resistance testing in case of virological failure (VL≥1000 copies/ml).

Laboratory tests

Preparation and storage of plasma samples

The freshly collected blood specimen was centrifuged at 3,000 revolution per minute (rpm) for 15 mins in a SorVall Legend X1R Centrifuge (ThermoFisher Scientific, Waltham, MA, USA). The harvested plasma was aliquoted into cryotubes for each participant and immediately stored at −80°C prior to VL quantification and genotyping respectively.

Viral load quantification

The quantitation of HIV-1 ribonucleic acid (RNA) in human plasma was done using the Cobas Ampliprep/Cobas TaqMan HIV-1 test v2.0 (Roche diagnostics, Indianapolis, USA) in the Infectious Diseases Research Laboratory, at the University of Zimbabwe. The linear range of the assay is 20-10⁷ HIV-1 RNA copies/mL and the specificity is 100%. Briefly,1050μl of plasma sample was allowed to thaw at room temperature, centrifuged and loaded onto the Cobas Ampliprep instrument together with the negative and positive controls for the nucleic acid purification. This was followed by the simultaneous amplification and detection of this nucleic acid on the TaqMan instrument.

Genotypic resistance testing

HIV viral RNA was extracted from 140 ul plasma sample using the QIAamp Viral RNA Mini kit protocol (Qiagen, Hilden, Germany) from only samples with VL ≥1000 copies/ml. The eluted viral RNA was reversed transcribed and amplified using the Applied Biosystems^™ TaqPath^™ Seq HIV-1 Genotyping Assay Kit (ThermoFisher Scientific, Waltham, MA, USA). The amplicon (HIV HXB2: 2253-5096) covered the protease (PR, HXB2: 2253-2550), reverse transcriptase (RT, HXB2: 2582-3366) and integrase (IN, HXB2: 4229-5093) regions of the HIV pol gene. The quality of the PCR product was assessed on a 2% agarose gel with a molecular weight marker of known DNA size (O’GeneRuler 1kb DNA ladder) and purified prior to Sanger sequencing. The amplicons were sequenced by Sanger sequencing at Molecular Cloning Laboratories, San Francisco, California, USA. The sequences generated were cleaned and assembled to create a consensus on Geneious software version 11.0 (Biomatters, Ltd. Auckland, New Zealand). HIV subtypes and mutations were analyzed on the REGA HIV-1 subtyping tool and the online Stanford HIV drug resistance database.

Statistical Analysis

Descriptive statistics were used to summarize the baseline demographic and clinical characteristics and were presented as proportions and medians (interquartile range-IQR). At baseline, these characteristics were compared between ART-naive and prior ART exposed participants using the Student’s T-test for parametric variables and the Wilcoxon rank- sum test for non-parametric variables as well as the Chi-square test for categorical data. Virological failure was described as VL ≥1000 copies/ml (as a single snapshot algorithm measurement) at follow up as per the WHO guidelines [1]. The time points used for analysis were baseline (defined as the time at which DTG was initiated for each patient), and ≥ 24 weeks. Proportion of participants with virological suppression (VL<1000 copies/ml) was described.

Body mass index (BMI) which is a measure of body fat based on height and weight was categorized as: underweight ≤18.5, normal weight = 18.5–24.9, overweight = 25-29.9 and obesity = BMI of 30 or greater as per the National Institute of Health. Wilcoxon rank-sum test was used to assess changes in weight and body mass index (BMI) at ART initiation and at ≥24 weeks follow up. Univariable and multivariable logistic regression analyses were performed to explore associations between baseline variables and weight gain whilst on DTG. Participants’ characteristics with P less than 0.1 in the univariable analyses were included in the multivariable logistic regression analysis. A sensitivity analysis was conducted to assess the robustness of any association between baseline variables and weight gain, whilst accounting for the weight missing data at follow-up. All missing weight values at follow-up were imputed in 2 different scenarios as follows: 0 for (no weight gain) and 1 for (weight gain). All statistical analyses were performed using Stata version 17.0 (StataCorp LP, College Station, Texas, USA; 800-STATA-PC).

Ethics

The study was reviewed and approved by the local institutional review board of the Joint Research and Ethics Committee of the University of Zimbabwe and the Parirenyatwa Group of Hospitals (JREC/352/2021) and by the Medical Research Council of Zimbabwe (MRCZ/A/2826). Consented forms were obtained from all study participants.

Sequence data

The sequence data for the 2 participants with VL≥1000 copies/ml at follow up visit have been provided as supplementary data.

Results

Baseline characteristics of the participants

A total number of 172 participants were enrolled in the study (See Figure 1). The median (IQR) age of the participants was 39(29-48) years and slightly above half (54%) of the participants were females. See Table 1. The median (IQR) CD4 cell count and log₁₀ VL at enrolment was 175(58-328) cells/mm³ and 5.41(4.80-5.74) respectively. Most of the participants were ART naïve (83%; n=142/172) at enrolment and among ART experienced, 43% (13/30) had a DTG-based regimen as previous ART, 40% (12/30) were on an efavirenz-based regimen and the remaining 17% (5/30) had an atazanavir/ritonavir-based regimen. Among the defaulted, patients were more frequently males than females (67% vs 33%, p=0.012; Chi-square test). The majority of the participants were married (54%), unemployed (52%) and had gone through a secondary education (78%).There were more single and divorced individuals among those who defaulted ART compared to ART naïve (p <0.001).

Figure 1:

Flowchart of the study population, VL-Viral load

Table 1:

Baseline sociodemographic and clinical characteristic of the participants.

Characteristics	All participants, N=172	ART naïve n=142	ART defaulted n=30	P value
Age, years, median (IQR)	39 (29-48)	40 (31-49)	37 (28-44)	0.114
Gender, n(%)
Female	93 (54%)	83 (58%)	10 (33%)	0.012
Male	79 (46%)	59 (42%)	20 (67%)
Plasma VL in log10 copies/ml, median (IQR)	5.41 (4.80-5.74)	5.42 (4.78-5.76)	5.32 (4.80-5.74)	0.637
CD4 cell count in cells/mm3, median (IQR)	175 (58-328)	165 (55-357)	213 (77-320)	0.795
Weight, mean (SD)	60.2 (11.6)	60.5 (12.2)	58.9 (7.9)	0.509
BMI, median (IQR)	21.8 (19.1-25.9)	21.6 (19.0-25.5)	23.3 (20.3-27.6)	0.159
Marital status, n(%)
Married	93 (54%)	86 (61%)	7 (23%)
Divorced	23 (13%)	15 (10%)	8 (27%)	< 0.001
Single	41 (24%)	28 (20%)	13 (43%)
widowed	15 (9%)	13 (9%)	2 (7%)
Education, n(%)
None	1 (1%)	1 (1%)	0 (0%)
Primary	14 (8%)	11 (8%)	3 (10%)	0.847
Secondary	135 (78%)	111 (78%)	24 (80%)
Tertiary	22 (13%)	19 (13%)	3 (10%)
Occupation, n(%)
Employed	79 (46%)	69 (48.5%)	10 (33%)
Unemployed	89 (52%)	69 (48.5%)	20 (67%)	0.364
Student	4 (2%)	4 (3%)	0 (0%)

IQR=Interquartile range, VL=Viral load, BMI=Body mass index, ART=Antiretroviral therapy, SD=Standard deviation

At baseline, participants had a normal BMI with a median (IQR) of 21.9 (19.1-26.0) kgs/m² and 8% of the participants were obese with BMI≥30 kgs/m². Females had a significant higher median (IQR) BMI compared to males (23(20-27) vs 20(19-24), p=0.009, two-sample Wilcoxon rank-sum test).

Clinical outcomes of the participants at ≥ 24 weeks follow up

After a median (IQR) duration of 27 (25-30) weeks on DTG, a total of 137 (79.7%) were still in care and followed up at the clinic, 10(5.8%) participants were loss to follow up, 10(5.8%) transferred out and 15(8.7%) died. The 15 participants who died had severe advanced HIV disease (baseline CD4<100 cells/mm³) at enrolment into the study. Fourteen (93%) out of the 15 were ART naive. There was a statistically significant difference between their baseline CD4 cell count compared to those in care (57(22-227) vs 180(71-351), p=0.035).

Virological outcomes of the participants at ≥ 24 weeks follow up

From the 137 participants in care, 131 had follow-up VL data available. Of these 131, 129 (98%) participants were suppressed with VL<1000 copies/ml. Using the cut off 50 copies/ml, 110 (84%) had VL<50 copies/ml and up to 21(16%) had VL≥50 copies/ml. Of these 21 participants, the majority (76%,16/21) were ART naïve and males (67%, 14/21).

HIV drug resistance among the participants with virological failure on DTG

The 2 participants with VL≥1000 copies/ml (1 participant had a VL of 23946 copies/ml and the other had 2640 copies/ml copies/ml) at follow-up underwent genotypic resistance testing. Unsurprisingly, no emerging resistance to DTG was found in these two participants.

Weight outcome at ≥ 24 weeks follow-up and factors associated with weight gain

A total number of 120 participants among those in care had weight data available at follow up. The mean(SD) weight at follow-up was 66.9 kgs (12.8) from 60.2 (11.6) at baseline and the median (IQR) BMI was 24.1(21.1-26.7) kgs/m² from 21.8 (19.1-25.9) at baseline. There were significant increases in weight and BMI among participants at follow-up. The average weight gain was 5.25 kgs (p<0.0001) with the lowest weight gain being observed among ART naive and the largest in patients with baseline CD4 cell count <200 cells/mm³. The average BMI (p<0.001) increase was 2.2 kgs/m². The cut-off used for multivariable logistic regression analysis of on-treatment weight gain was at least 5.25kg (50^th percentile). Baseline CD4 cell count ≥200 cells/mm³ (OR= 0.26; 95% CI: 0.12-0.58, p=0.001) was significantly associated with at a smaller weight gain. See Table 2. When the sensitivity analysis was performed whilst accounting for the weight missing values at follow-up, baseline CD4 cell count ≥200 cells/mm³ was still significantly related with at a smaller weight gain. (p<0.05).

Table 2:

Factors associated with weight gain

Variables		Univariate				Multivariate
	Proportion	OR	95%CI	P	OR	95%CI	P
Age (n=172), in years
<40	0.51	1	0.89-3.79	0.098	1	0.61-3.04	0.449
≥40	0.49	1.84			1.36
BMI (n=147) in kgs/m2
<30	0.92	1	0.23-4.13	0.979
≥30	0.08	0.98
Gender (n=172)
Male	0.46	1	0.39-1.65	0.552
Female	0.54	0.80
ART status (n=172)
ART naïve	0.83	1	0.21-1.51	0.256
ART defaulter	0.17	0.57
Baseline CD4 (n=161)
<200	0.53	1	0.11-0.51	<0.001	1	0.12-0.58	0.001
≥200	0.47	0.23			0.26
Viral load (n=170)
<10000	0.07	1	0.38-4.58	0.664
≥10000	0.93	1.32
Marital Status (n=172)
Married	0.54	1	0.82-1.63	0.402
Divorced	0.13	1.16
Single	0.24
Widowed	0.09
Occupation (n=169)
Employed	0.45	1	0.24-1.06	0.069	1	0.28-1.42	0.268
unemployed	0.55	0.51			0.63

OR=Odds ratio, CI=Confidence interval, ART=Antiretroviral therapy, BMI=Body mass index. Participants’ characteristics with P less than 0.1 in the univariable analyses were included in the multivariable logistic regression analysis.

Discussion

Current WHO recommendations suggest the use of second-generation integrase strand transfer inhibitors (INSTIs) such as DTG for initial ART in many LMICs. DTG plays a crucial role in the WHO’s efforts to control the HIV pandemic. In this study, we found high level of virological suppression (98%) among the participants, significant increases in weight and BMI following a median duration of 24 weeks post-DTG initiation and no evidence of acquired drug resistance to DTG among those with virological failure.

We found a high rate (98%) of virological suppression (VL<1000 copies/ml) among ART naïve individuals following DTG initiation in a real-world setting and up to 84% had VL<50 copies/ml. The high potency of DTG has previously been established in clinical trials.[2–5] Our finding is consistent with previous findings from large clinical trials on virological efficacy of ART-naïve individuals receiving DTG plus two NRTIs.[6,17,18] The SINGLE (multicenter) clinical trials conducted among adults initiating on a DTG-based ART reported virological suppression (VL<50 copies/ml) of 88% by week 48.[8] Furthermore, the high virological suppression rate found in our study is comparable to findings from two landmark clinical trials conducted exclusively in sub-Saharan Africa (NAMSAL and ADVANCE).[19,20] Virological suppression rates (VL<50 copies/ml at week 48) of 75% and 84% were seen among ART naïve individuals initiating on DTG in the NAMSAL and ADVANCE studies respectively. In these two studies, the effectiveness of DTG was compared to efavirenz (EFV) among the participants. Higher rates of virological failure and emergent drug resistance were found on EFV-based regimens compared to DTG-based regimens. In the NAMSAL study, virologic failure (VL≥50 copies/ml) at week 48 was observed in 31% in the EFV group compared to 25% in the DTG group. Additionally, among those with VL≥1000 copies/ml as per the WHO definition, none of the 3 participants in the DTG group had drug resistance whereas 6 of 16 participants in the EFV group had drug resistance. In the ADVANCE study, virologic failure (VL≥50 copies/ml) at week 48 was observed in 21% in the EFV group compared to 16% in the DTG group. Although in these 2 studies, virological suppression rate was reported after a longer period (by week 48) than this current study (by week 24), our finding on the effectiveness of DTG in ART naïve individuals is reassuring and just confirms previous studies in this population.

Additionally, Jacobson et al (2017) in a study on viral suppression among treatment naïve HIV-Infected adults initiating ART, showed that individuals on INSTI-based ART regimens experienced higher rates of virological suppression (93.2%) at 6 months compared to those on NNRTI (69.7%).[21]

On the other hand, clinical experience with DTG is growing in Africa and consistent findings from cohort studies among ART naïve and ART-experienced individuals are also being reported in real-world settings [22–25]. We recently showed that virological suppression increased from 76% to 95% following a switch to TLD among adolescents in rural Zimbabwe [26]. Similarly Brown et al (2022) showed that virological suppression (VL<1000 copies/ml) of 99% was achieved at 4 months following the transition to DTG among individuals in Lesotho [27].

Despite the high effectiveness, excellent safety profiles and minimal side effect rates of DTG reported so far, concerns about subsequent weight gain observed with INSTIs particular DTG has emerged [28–30]. In this study, we found significant increases in weight and BMI among individuals following 24 weeks post-DTG initiation. The mean weight among participants increased to 66.90 kgs from 60.2 kgs at baseline.

This is consistent with previous studies that have associated body weight and BMI gain with the use of INSTI among treatment naïve and experienced individuals [31,32]. Menard et al (2017) showed an abnormal weight gain ranging between 4 and 12 kgs on DTG after a mean time from baseline to weight assessment of 276 ± 79 days.[32] Similarly, the ADVANCE study in South Africa showed that naive individuals receiving DTG-based ART (with either TDF or TAF) had a greater increased weight at 48 weeks compared with efavirenz-based ART.[20] In the NAMSAL study, more weight gain was also seen in the DTG than in the efavirenz group (median weight gain, 5.0 kg vs. 3.0 kg; incidence of obesity, 12.3% vs. 5.4%, respectively).[19] In contrast to these studies, no other drug was used as a comparator to link the observed weight gain to the use of DTG in our study. Hence, the weight gain observed in our study is more likely due to the ‘return to health’ phenomenon following ART initiation. This ‘return to health’ phenomenon may also be supported by our described population (ART naïve individuals).

However, we found in our study a significant association between baseline CD4 cell count <200 cells/mm³ and increased weight gain. Factors including being black and Hispanic women and having a BMI ≥30 kg/m² have previously been associated with weight gain.[15] The association between low baseline CD4 cell count and weight gain in our study may also be explained by the return to health phenomenon that occurs following ART initiation among ART naïve individuals [33,34]. Additionally, the weight gain phenomenon has been hypothesized to be multifactorial in nature. These factors all contributing include demographic factors, genetic predisposition and HIV-related factors, the ‘return to health’ weight gain.

We also found that among the defaulted, patients were more frequently males than females (67% vs 33%, p=0.012). Additionally, there were more single and divorced individuals among those who defaulted ART compared to ART naïve (p <0.001). Previous studies have showed that ART defaulters had an increased rate of pre-treatment HIVDR which predisposed them to an increased risk of further developing HIVDR and experienced treatment failure on NNRTI-based regimens [35,36]. Although the high potency and high genetic barrier to resistance of DTG has been well established, strategies to identify vulnerable population (single and divorced men) may be needed so as to preserve DTG activity and mitigate resistance to DTG in many resource limited settings.

In our study, the 2 participants with virological failure (VL≥1000 copies/ml) at follow-up did not develop any emerging HIV drug resistance on TLD. In line with this, no resistance to DTG was detected in large clinical trial studies (NAMSAL, FLAMINGO, ARIA and ADVANCE studies) of ART-naïve PLHIV receiving DTG plus 2 NRTIs.[6,18,19,37] Furthermore, no emerging resistance INSTI-associated DRMs was found among ART-Naïve PLHIV receiving DTG+ 2NRTIs in the SINGLE and SPRING-2 clinical trial studies [8,17]. On the contrary, higher rates of drug resistance to NNRTIs (e.g. efavirenz) have previously been reported in LMICs [38–42]. In line with this, treatment-emergent resistance to NNRTIs occurred in the efavirenz arms of both ADVANCE and NAMSAL studies. [19,20,43,44]

Furthermore, Malet et al (2017) previously showed that mutations outside the integrase gene, precisely in the 3’ polypurine tract region have been associated with virological failure and resistance to DTG. [45] Similarly, Hikichi et al (2023) also found that INSTIs can lose their efficacy not due to mutations arising in the integrase enzyme, but due to mutations in a completely different part of the virus, its envelope protein.[46] However, we did not assess this in our study.

Our prospective cohort study shows that virological suppression is high among ART naïve adults initiating on DTG and confirms the relationship between low CD4 cell counts and an increased weight following ART initiation. However, our study has some limitations. One limitation includes the fact that there was no ART comparator to assess any weight gain due to DTG. Similarly, we were unable to account for other factors including diet, genes and physical activities that have an effect on body weight. Finally, our follow-up time period was short (∼24 weeks from the time of initiation on DTG). More longitudinal studies are needed to assess the long-term virological outcomes and weight on DTG.

Conclusions:

We found that virological suppression was high among people initiating on TLD in a resource-limited setting and that DTG resistance remains scarce among ART naïve individuals.