To evaluate the impact of switching from efavirenz-based antiretroviral therapy (EFV) to dolutegravir (DTG) on virological suppression in people living with HIV followed at the Lubumbashi Centre of Excellence.

1) Describe the sociodemographic and clinical characteristics of the patients followed.

2) Compare the virological suppression rates between patients on EFV and on DTG.M2

3) Identify the factors associated with virological suppression (sex, CD4, medical follow-up, duration of treatment).

4) Estimate the independent effect of DTG on virological suppression by logistic regression.

5) Evaluate clinical outcomes (deaths, loss to follow-up) according to the treatment regimen.

This is a retrospective longitudinal intra-individual (before/after) cohort study conducted at the HIV/AIDS Centre of Excellence of the University of Lubumbashi (DRC).

Unitofanalysis:thepatient

All included patients were followed successively:

Under efavirenz-based (EFV) regimen (pre-switch period)Then under a dolutegravir-based (DTG) regimen (post-switch period)

Thus, each patient constitutes their own control, allowing for a matched intra-patient comparison.

The primary endpoint (virological suppression) was assessed between 6 and 12 months after the switch to DTG, in order to standardize the follow-up period between patients.

A further analysis taking into account the time until virological suppression could be carried out using a Cox proportional hazards model, although not implemented in the present study.

For each patient, the periods were defined as follows:

EFVperiod(pre-switch): 2014-2019

The last phase of treatment before the introduction of DTG

Switchdate:

DTG initiation date (2019-2024)

DTG(post-switch)period:

Period following the introduction of DTG

Viralloadmeasurement:

A viral load was recorded for each period.This was the first available viral load ≥ 6 months later:The exhibition at EFVAnd after the switch to DTG

An intention-to-treat (ITT) analysis was performed. All included patients were retained in the analysis, and patients without viral load measurement were considered virological failures.

The initial characteristics were evaluated at the time of the switch to DTG:

AgeSexWHO stageCD4 count: The CD4 count was categorized according to clinically relevant thresholds. A low CD4 count was defined as <200 cells/mm³, in accordance with WHO recommendations.This measurement was taken at the time of the switch to dolutegravir, in order to reflect the initial immunological status at the beginning of the post-switch period.Previous viral loadDuration of antiretroviral treatmentType of follow-up (doctor, nurse, etc.)

These variables were used for:

Describe the populationAdjust the multivariate analysis

The following variables were not available in the database:

Therapeutic complianceReason for the switch: decision of the national HIV control programExact duration under EFV before switchingStandardized initial virological status

These limitations were explicitly taken into account in the discussion.

1)Descriptiveanalysis

Averages, medians, proportions

2)Pairedanalysis(CRITICALCORRECTION)

Intra-patient comparison EFV vs DTGChi-square test for paired proportions (or McNemar test if specified)

3)Multivariateanalysis

Binary logistic regression:

Dependent variable:Virological suppression (yes/no)Independent variables:Treatment regimen (DTG vs EFV)SexCD4Type of monitoringDuration of treatment

Results expressed as adjusted OR (aOR) with 95% CI.

Patients without a viral load were excluded from the main analysisA descriptive analysis was carried out to quantify the missing data (~45%)No replacement (imputation) was made

The study was approved by the Ethics Committee of the University of Lubumbashi.

Approval number: ESP/CE/98B/2025Nature of the study: retrospective review of files

Individual informed consent was obtained, in accordance with ethical recommendations for retrospective studies. The data were anonymized.

A total of 4617 patients living with HIV were included in the study. 65.7% were male (n = 3033), compared to 34.3% female (n = 1584). The majority of patients (76.6%) were followed by physicians, while 21.9% were followed by nurses, and only 0.2% by community facilitators (Table 1).

Table 1. Sociodemographic and patient follow-up characteristics.

The gender distribution was similar between the two groups, with a male predominance (65.7%) (Table 2).

Table 2. Distribution of patients according to treatment regimen and sex.

The proportion of patients achieving virological suppression (<1000 copies/ml) in the intention-to-treat (ITT) analysis was 40.3% with DTG versus 32.7% with EFV. In both groups, 45% of patients did not undergo viral load testing. Death and loss-to-follow-up rates were similar between the two regimens (Table 3).

Table 3. Virological results and clinical outcomes according to the treatment regimen.

Dolutegravir was associated with a significantly higher probability of virological suppression (OR = 2.47; 95% CI [2.11 - 2.88]; p < 0.001) (Table 4).

Table 4. Summary of statistical associations.

No significant difference in mortality between the EFV and DTG regimens (p = 0.372; OR = 0.96; 95% CI: 0.80 - 1.17).

Follow-up by physician, compared to that by nurse, showed no significant difference after adjustment (p = 0.06; OR = 1.23; 95% CI: 0.93 - 1.63).

Our results show that patients receiving DTG are twice as likely to achieve virological suppression compared to those receiving EFV. However, the high proportion of patients without viral load testing (≈45%) is a major limitation. Deaths and loss to follow-up remained comparable between the two groups (Table 5).

Table 5. Logistics regression results.

Effect of the treatment regimen (DTG vs EFV)

Dolutegravir is strongly associated with virological suppression (ORa = 2.47; 95% CI [2.11 - 2.88]; p < 0.001). Patients on DTG are approximately 2.5 times more likely to achieve virological suppression than those on EFV, independent of other factors.

Sex (M vs F)

No significant association was observed (ORa = 1.02; p = 0.72). Sex did not influence the virological response in this cohort.

CD4 bas

Significant association (ORa = 1.85; 95% CI [1.40 - 2.30]; p < 0.01). Patients with impaired immune status have a modified probability of virological suppression, reflecting the importance of the immunological background in the response to treatment.

Follow-up by a doctor

Borderline association (ORa = 1.23; p = 0.06). A trend suggests better suppression under medical supervision, but without formal statistical evidence.

Treatment duration

Not significant (ORa = 0.95; p = 0.08). Treatment duration is not an indeK8pendent factor in this model, although a trend is observed (Figure 1).

Figure 1. Orest plot-logistic regression.

Intention-to-treat analysis helps to reduce bias related to missing data. However, the high proportion of patients without viral load measurements (~45%) remains an important limitation.

Our results show a higher probability of viral suppression with dolutegravir (DTG) compared with efavirenz (EFV). This observation is consistent with major randomized clinical trials. The SINGLE study [6] demonstrated faster and more durable viral suppression with DTG compared with EFV. In South Africa, Venter et al. [7] confirmed the superior virological efficacy of DTG. Similar findings were reported in the SPRING-2 and FLAMINGO studies [8][9]. The ADVANCE trial also demonstrated the high virological efficacy and favorable safety profile of dolutegravir-based first-line regimens in treatment-naïve adults [10].

In African settings, real-world data also show good virological outcomes with DTG. Dorward et al. [5] reported strong outcomes following implementation of DTG-based regimens in a national program. Okoli et al. [11] also described favorable clinical outcomes in routine care. These findings are consistent with WHO recommendations supporting the widespread use of dolutegravir because of its high genetic barrier to resistance and favorable treatment profile [12].

Evidence from African clinical studies also supports the effectiveness of dolutegravir in achieving virological suppression, including among treatment-experienced patients receiving second-line therapy [13]. Nicholas et al. [14] showed that point-of-care viral load testing improves the detection of virological failure and supports clinical decision-making, without being directly related to DTG efficacy.

Dolutegravir has a high genetic barrier to resistance [15]. In settings such as the DRC, this represents an important advantage in cases of suboptimal adherence.

Efavirenz is associated with neuropsychiatric adverse effects that may reduce adherence [16]. The GEMINI-1 and GEMINI-2 studies [17] showed that DTG + 3TC is non-inferior to standard triple therapy in terms of virological suppression.

Fonjungo et al. [18] highlighted structural limitations in viral load monitoring systems in sub-Saharan Africa, particularly regarding testing coverage and patient follow-up.

Cases of dolutegravir resistance remain rare but have been documented. Blanco-Arévalo et al. [19] described virological failure patterns associated with integrase inhibitors. Loosli et al. [20] also reported resistance in treatment-experienced patients.

Regarding pregnancy, Zash et al. [21] initially reported a potential signal for neural tube defects, but updated analyses showed a risk comparable to that of the general population [22]. These findings contributed to WHO recommendations supporting DTG use [3].

Weight gain under DTG has been reported in several studies. Sax et al. [16] and Norwood et al. [23] described weight gain associated with integrase inhibitor–based regimens. Kanters et al. [24] reported an association between DTG and weight gain without definitive conclusions on its clinical impact.

The high proportion of missing viral load data reflects structural limitations of the health system. Boeke et al. [25] showed that limited access to viral load testing remains a major challenge in sub-Saharan Africa.

Economic evaluations suggest that point-of-care viral load testing could represent a cost-effective strategy to improve access to virological monitoring in resource-limited settings [26]. Paton et al. [13] demonstrated the effectiveness of DTG in second-line treatment or treatment failure settings.

Regarding care models, Nicholas et al. [14] showed that good outcomes can be achieved in decentralized care when viral load monitoring is available. Ford et al. [27] and Eshun-Wilson et al. [28] support differentiated service delivery models adapted to local contexts.