Efficacy and safety profiles of dolutegravir plus lamivudine vs. bictegravir/emtricitabine/tenofovir alafenamide in therapy-naïve adults with HIV-1

Abstract

Background:

Dual regimen dolutegravir (DTG) plus lamivudine (3TC) has demonstrated non-inferior efficacy compared to DTG-based three-drug regimens (3DRs), yet directly comparative data regarding the efficacy and safety of DTG + 3TC and bictegravir/emtricitabine/tenofovir alafenamide (B/F/TAF) for therapy-naïve people with human immunodeficiency virus (HIV)-1 (PWH) are still limited. We aimed to assess the antiviral potency and safety profiles of DTG + 3TC vs. B/F/TAF based on antiretroviral therapy (ART)-naïve PWH in China.

Methods:

This retrospective multicenter study enrolled PWH initiating ART with DTG + 3TC or B/F/TAF from 2020 to 2022 in Guangdong and Guangxi. We analyzed response rates based on target not detected (TND) status using intention-to-treat (ITT) analysis. Subgroups were formed based on baseline viral load (VL) (<100,000 vs. ≥100,000 copies/mL) and CD4⁺ cell count (<200 vs. ≥200 cell/µL). Median time to TND VL was assessed by Kaplan–Meier method. We also measured changes from baseline in CD4⁺ cell counts, CD4/CD8 ratio, lipid parameters, weight, creatinine (Cr), estimated glomerular filtration rate (eGFR), and drug-related adverse effects (DRAEs).

Results:

We enrolled 280 participants, including 137 (48.9%) on DTG + 3TC and 143 (51.1%) on B/F/TAF. At week 48, 96.4% (132/137) on DTG+3TC and 100% (143/143) on B/F/TAF achieved TND (P = 0.064). At week 12, TND responses were higher with B/F/TAF (78.3% [112/143]) than DTG+3TC (30.7% [42/137]) (P <0.001). This trend held across subgroups. B/F/TAF achieved TND faster (12 weeks) than DTG+3TC (24 weeks) (P <0.001). No differences were seen in CD4⁺ cell count and CD4/CD8 ratio, except in the high-VL subgroup, where B/F/TAF showed better recovery. DRAEs were significantly lower with B/F/TAF (4.9% [7/143]) than with DTG + 3TC (13.1% [18/137]) (P = 0.016). Lipid parameters, body weight, and Cr increased in both groups over 48 weeks, with DTG+3TC showing a more favorable effect on triglycerides, high-density lipoprotein (HDL) cholesterol, and weight gain.

Conclusions:

In this real-life study, B/F/TAF led to a faster viral decline and fewer DRAEs compared to DTG+3TC. No significant difference was observed in the TND rate at week 48, regardless of baseline VL and CD4⁺ cell count. CD4⁺ recovery was superior for B/F/TAF in participants with high VL. The DTG + 3TC regimen had less impact on metabolic changes than B/F/TAF.

Introduction

The treatment of human immunodeficiency virus (HIV) has improved dramatically since the advent of new antiretroviral classes, particularly second-generation integrase strand transfer inhibitors (INSTIs) that are characterized by a higher barrier to resistance, better efficacy and safety profiles, and fewer drug–drug interactions (DDIs). These favorable properties are extremely important in an aging HIV population with cumulative antiretroviral therapy (ART)-related toxicities, high prevalence of multiple comorbidities, and concomitant polypharmacy. Thus, many guidelines recommend using dolutegravir (DTG)- or bictegravir (BIC)-based regimens with either one or two nucleoside reverse transcriptase inhibitors as a preferred first-line treatment for most people with HIV (PWH).^[1–3]

In comparison to other 2-drug or 3-drug INSTI-based regimens, bictegravir/emtricitabine/tenofovir alafenamide (B/F/TAF) and dolutegravir plus lamivudine (DTG + 3TC) are currently more often prescribed worldwide in ART-naïve and virologically suppressed individuals as demonstrated in clinical trials and real-life studies.^[4–9] However, in therapy-naïve PWH, these two mainstream regimens have not been directly compared in clinical trials. Results of an indirect treatment comparison have indicated that DTG + 3TC offered similar efficacy with fewer serious adverse events (AEs) than B/F/TAF and dolutegravir/abacavir/lamivudine at week 144 after ART initiation.^[10] This finding reflected no observed differences between 3TC + DTG and B/F/TAF in long-term efficacy among treatment naïve PWH. Nevertheless, it is still unknown whether B/F/TAF has a different impact on initial HIV-1 RNA decay kinetics compared to DTG + 3TC in real-life contexts. In the Observational Pharmacoepidemiology Research and Analysis (OPERA^®) cohort in United States, B/F/TAF initiation was linked to a faster virological control (viral load [VL] <200 copies/mL) than other INSTI-based regimens among individuals with advanced HIV infection.^[11] Both DTG + 3TC and B/F/TAF have strong antiviral properties and well-described fast VL decay.^[12] As a novel and potent inhibitor of HIV-1, BIC exhibits an improved virology profile relative to other INSTIs and demonstrates powerful synergistic antiviral effects in combination with emtricitabine (FTC) and TAF in vitro.^[13] Undoubtedly, DTG also presents high potency on HIV-1 suppression, yet relevant knowledge remains deficient on whether the exclusion of tenofovir/abacavir may confer a weakness on the rapid viral reduction capacity of DTG + 3TC. Hence, whether dual therapy DTG + 3TC is less potent than triple B/F/TAF in early viral suppression, as mirrored by a rapid drop decline in HIV-1 RNA and achievement of undetectable VL (target not detected [TND]), has aroused particular interests.

Therefore, we conducted a multicenter retrospective observational study to explore the differences between DTG + 3TC and B/F/TAF regimens in terms of clinical outcomes with a particular focus on initial HIV-1 RNA decay for treatment naïve PWH in China.

Methods

Ethical approval

The study received approval from the research ethics committee of the participating hospitals: Dongguan Ninth People's Hospital (No. 2022-Ethical Review-04) and the First Affiliated Hospital of Guangxi Medical University (No. 2023-E346-01). Written informed consent was obtained from participants at the Dongguan Ninth People's Hospital. The requirement for informed consent was waived by the research ethics committees of the First Affiliated Hospital of Guangxi Medical University.

Study population and design

We conducted a multicenter, retrospective observational study at the infectious disease unit of Dongguan Ninth People's Hospital and the First Affiliated Hospital of Guangxi Medical University. PWH who initiated ART with DTG + 3TC or B/F/TAF were enrolled between December 1, 2020 and May 31, 2022.

Participants who met the following criteria were included: ART-naïve, age ≥18 years, with complete baseline data available, and at least two data entries of plasma viral load (PVL), CD4 cell, and biochemical parameters measured during follow-up. Exclusion criteria included pre-existing mutations conferring resistance to any antiretroviral classes, co-infection with hepatitis B virus, drug abuse, pre-existing mental disorders, and pregnancy.

Here, we defined an achievement of undetectable HIV-RNA (TND) as complete virological suppression (CVS) to further assess the high potency of DTG + 3TC vs. B/F/TAF regimens. Safety was assessed throughout the study as laboratory abnormalities, incidence and severity of AEs, drug-related AEs, and regimen discontinuation due to AEs.

Data collection

Baseline demographics and clinical characteristics (age, sex, weight, HIV transmission category, comorbidities, opportunistic infections [OIs], and time from diagnosis to ART initiation) were abstracted from clinical charts. All laboratory results were conducted by the laboratories of each hospital. HIV-related data including HIV-RNA, CD4⁺ count, CD4/CD8 ratio, genotypic drug resistance testing, and other biochemical parameters such as fasting lipid profile, total bilirubin, liver enzymes, creatinine, and fasting plasma glucose were collected anonymously at initiation of ART and each follow-up visit. Study visits after therapy initiation were scheduled at weeks 4, 12, 24, and 48 (a 2-week window period was allowed for each visit). AEs were recorded at each follow-up visits. Subgroup analysis was stratified by baseline VL (<100,000 vs. ≥100,000 copies/mL) and CD4 cell count (<200 vs. ≥200 cell/μL). Plasma HIV-RNA was measured using a quantitative polymerase chain reaction (PCR) kit, with a detection threshold of 30 IU/mL (1 copy = 1.7 IU; Livzon Diagnostics, Zhuhai, China). Baseline genotype for HIV-1 drug-resistant mutations was assessed using HIV-1 genotyping kits (SUPBIO, Guangzhou, China).

Outcomes of interest

The primary objective of this study was to assess the proportion of participants with CVS (TND VL) under DTG + 3TC vs. B/F/TAF regimens at each visit time points, using an intention-to-treat (ITT) analysis, considering missed cases as failures (M = F). Moreover, the secondary objective was to compare the median time from ART initiation to CVS, changes from baseline in CD4⁺ cell counts and CD4/CD8 ratio, and safety at 48 weeks between the two regimens.

Statistical analysis

Categorical variables were presented as frequency distributions, whereas continuous variables were described in terms of median with interquartile range (IQR) or mean with standard deviation (SD). Between-regimen differences were analyzed using a chi-squared test or Fisher's exact tests for categorical variables and the Mann–Whitney U test for continuous variables as appropriate. Median time to CVS was estimated using the Kaplan–Meier method, and differences between the two groups were assessed using the log-rank test. The Wilcoxon signed-rank test was performed to assess changes in laboratory parameters from baseline. A two-tailed P-value <0.05 was considered statistically significant. Analyses were conducted using SPSS software, version 23.0 (IBM Corporation, Armonk, NY, USA).

Results

Out of the 295 participants screened from December 1, 2020 to May 31, 2022, 280 were finally enrolled in the study, and 15 were excluded due to missing baseline data or having three missing data entries of PVL, CD4⁺ cell, and biochemical parameters following ART initiation. None of them reported baseline HIV-1 drug-resistant mutations to currently available antiretroviral classes. In total, 137 (48.9%) participants were on DTG/3TC, and 143 (51.1%) participants were on B/F/TAF. Through week 48, four (2.9%) participants missed several scheduled appointments and three (2.2%) modified their ART regimen at week 36 but remained on study in the DTG/3TC group. Complete follow-up data were available for all participants in the B/F/TAF group.

Baseline characteristics

Demographics and baseline characteristics of the participants are summarized in Table 1. Out of 280 participants, 235 (83.9%) were male, and the median age was 34.0 (IQR: 26.0, 47.0) years. The median time from diagnosis to ART initiation was 14.0 (IQR: 8.0, 34.0) days. Further, of the 280 participants, 70 (25.0%) had at least one comorbidity, while 27 (9.6%) presented with OIs. The median baseline VL was 70,600 copies/mL, and the median baseline CD4⁺ cell count was 264 cells/μL. A total of 124 (44.3%) had HIV-1 RNA ≥100,000 copies/mL, 83 (29.6%) had a baseline VL ≥500,000 copies/mL, 156 (55.7%) had HIV-1 RNA <100,000 copies/mL, and 100 (35.7%) had a CD4⁺ T-cell count <200 cells/μL.

Table 1.

Baseline demographics and participant characteristics of B/F/TAF and DTG/3TC groups.

Characteristics	All (n = 280)	B/F/TAF (n = 143)	DTG/3TC (n = 137)	Statistics	P-value
Gender (male)	235 (83.9)	125 (87.4)	110 (80.3)	2.238*	0.091
Age (years)	34.0 (26.0, 47.0)	34.0 (27.0, 48.0)	33.0 (25.0, 45.0)	-1.505†	0.706
Weight (kg)	60.7 (52.6, 67.4)	60.7 (52.6, 67.4)	60.0 (50.1, 65.0)	-2.047†	0.041
HIV transmission				12.694*	0.002
Homosexual	159 (56.8)	95 (66.4)	64 (46.7)
Heterosexual	102 (36.4)	43 (30.1)	59 (43.1)
Others	19 (6.8)	5 (3.5)	14 (10.2)
OIs	27 (9.6)	18 (12.6)	9 (6.6)	2.908*	0.088
Comorbidities	70 (25.0)	30 (21.0)	40 (29.2)	2.520*	0.112
Interval from HIV diagnosis to ART initiation (days)	14.0 (8.0, 34.0)	14 (8.0, 36.0)	15 (9.0, 33.0)	-0.515†	0.607
HIV-RNA	70,600 (19,525, 251,500)	73,500 (19,500, 286,000)	65,900 (20,450, 216,500)	-0.609†	0.543
<105 copies/mL	156 (55.7)	74 (51.8)	82 (59.9)	1.863*	0.172
≥105 copies/mL	124 (44.3)	69 (48.3)	55 (40.2)	1.863*	0.172
CD4+ count	264 (136,394 )	230 (114, 367)	278 (149, 432)	-2.340†	0.019
<200 cells/μL	100 (35.7)	61 (42.7)	39 (28.5)	6.136*	0.013
≥200 cells/μL	180 (64.3)	82 (57.3)	98 (71.5)	6.136*	0.013
CD4/CD8	0.26 (0.14, 0.38)	0.24 (0.13, 0.40)	0.27 (0.15, 0.38)	-0.524*	0.600
Cr (μmol/L)	75.8 (65.0, 84.0)	77.0 (67.0, 83.0)	72.0 (62.3, 84.6)	1.159†	0.246
TG (mmol/L)	1.35 (0.95, 1.93)	1.32 (1.00, 1.80)	1.41 (0.91, 2.14)	1.533†	0.125
CHOL (mmol/L)	4.03 (3.47, 4.62)	4.03 (3.51, 4.66)	4.00 (3.43, 4.57)	-1.195†	0.233
HDL-C (mmol/L)	1.04 (0.88, 1.29)	1.00 (0.87, 1.23)	1.07 (0.89, 1.40)	-2.131†	0.033
LDL-C (mmol/L)	2.53 (2.07, 3.07)	2.58 (2.15, 3.11)	2.49 (1.94, 3.04)	-1.885†	0.059

Data were shown as n (%) or median (IQR). ^*χ² values; ^†Z values; ART: Antiretroviral therapy; B/F/TAF: Bictegravir/emtricitabine/tenofovir alafenamide; CHOL: Total cholesterol; Cr: Creatinine; DTG/3TC: Dolutegravir/lamivudine; HDL-C: High-density lipoprotein cholesterol; HIV: Human immunodeficiency virus; IQR: Interquartile range; LDL-C: Low-density lipoprotein cholesterol; OIs: Opportunistic infections; TG: Triglyceride.

Overall, main baseline characteristics were generally similar between the two groups, with the exception that participants in the B/F/TAF group were more likely to be infected with HIV through homosexual transmission among men who have sex with men (MSM), and to have a relatively higher body weight and lower CD4⁺ cell count than those in the DTG/3TC group.

Comparison of virological outcomes of DTG + 3TC vs. B/F/TAF

Overall, 98.2% (275/280) of the participants achieved CVS (TND) by week 48; 96.4% (132/137) of participants in the DTG + 3TC group and 100% (143/143) of those in the B/F/TAF group obtained TND VL values at week 48 by ITT (M = F) analysis (χ² = 3.437, P = 0.064). For the remaining five participants in the DTG + 3TC group, three had the initial DTG + 3TC regimen modified before week 48, and two had missing virological data. All subjects with available HIV-1 RNA data at week 48 met the criteria of CVS without virological rebound. Similar results were observed at week 24, as 95.6% (131/137) of participants in the DTG + 3TC group and 99.3% (142/143) in the B/F/TAF group achieved undetectable HIV-1 RNA (χ² = 2.525, P = 0.112). Notably, the rate of CVS was significantly higher in the B/F/TAF group (78.3% [112/143]) than in the DTG + 3TC group (30.7% [42/137]) at week 12 (χ² = 64.227, P <0.001), though no between-regimen difference was found at week 4 (0% [0/137] with DTG + 3TC vs. 0.7% [1/143] with B/F/TAF; Figure 1).

Figure 1.

Overall proportion of participants with TND in B/F/TAF and DTG/3TC groups by visit (ITT analysis). B/F/TAF: Bictegravir/emtricitabine/tenofovir alafenamide; cART: Combination antiretroviral therapy; DTG/3TC: Dolutegravir/lamivudine; ITT: Intention to treat; TND: Target not detected.

In addition, no difference in TND VL rates was observed between the treatment group at weeks 4, 24, and 48, irrespective of baseline HIV-1 VL and CD4⁺ T-cell count. Meanwhile, at week 12, statistically significant differences in TND VL rates were observed across subgroups. In the subgroup of participants with HIV-1 RNA ≥100,000 copies/mL, 21.8% (12/55) and 68.1% (47/69) in the DTG + 3TC and B/F/TAF groups, respectively, achieved TND VL values (χ² = 26.302, P <0.001); the corresponding proportions were 36.6% (30/82) and 87.8% (65/74), respectively (χ² = 42.909, P <0.001), for those with baseline HIV-1 RNA <100,000 copies/mL. In the CD4⁺ T-cell count <200 cells/μL subgroup, the B/F/TAF group was associated with significantly higher CVS than the DTG + 3TC group at week 12 (DTG + 3TC, 20.5% [8/39] vs. B/F/TAF, 78.7% [48/61], χ² = 32.677, P <0.001). Consistent results were observed in those with baseline CD4⁺ T-cell count ≥200 cells/μL (DTG + 3TC, 34.7% [34/98] vs. B/F/TAF, 78.0% [64/82], χ² = 33.834, P <0.001; Figure 2).

Figure 2.

Proportion of participants with plasma HIV-1 RNA TND status at each visit according to baseline (ITT analysis). (A) HIV-1 RNA <100,000 copies/mL; (B) HIV-1 RNA ≥100,000 copies/mL; (C) CD4⁺ cell count <200 cell/µL; (D) CD4⁺ cell count ≥200 cell/µL. B/F/TAF: Bictegravir/emtricitabine/tenofovir alafenamide; DTG/3TC: Dolutegravir/lamivudine; HIV: human immunodeficiency virus; ITT: Intention to treat; TND: Target not detected.

We also identified a significant difference in time to TND VL values between regimens. The median time to TND VL values for participants who initiated ART on the B/F/TAF regimen was 12.0 (95% confidence interval [CI] 11.3–12.7) weeks relative to 24.0 (95% CI 23.5–24.5) weeks for those who initiated on the DTG + 3TC regimen (Z = 63.869, P <0.001; Figure 3).

Figure 3.

Kaplan–Meier survival analysis of B/F/TAF and DTG/3TC groups from baseline to TND; P-value was obtained using the log-rank test. ART: Antiretroviral therapy; B/F/TAF: Bictegravir/emtricitabine/tenofovir alafenamide; cART: Combination ART; DTG/3TC: Dolutegravir/lamivudine; TND: Target not detected.

Comparison of immunological outcomes of DTG + 3TC vs. B/F/TAF

Overall, the CD4⁺ cell count and CD4/CD8 cell ratio increased significantly from baseline to week 48 in both groups. There were no statistical differences in terms of immunological response between the two groups at each visit time point. At week 48, the median changes from baseline in the CD4⁺ cell count and CD4/CD8 cell ratio were 143.5 (IQR: 65.0, 267.3) cell/µL and 0.35 (IQR: 0.16, 0.53) in the DTG/3TC group, respectively, compared to 156.0 (IQR: 84.0, 247.0) cell/µL and 0.29 (IQR: 0.18, 0.45) in the B/F/TAF group, respectively [Figure 4]. However, for the HIV-1 RNA ≥100,000 copies/mL stratum, a significant difference was observed between regimens with respect to CD4⁺ cell count but not CD4/CD8 cell ratio recovery at 48 weeks (DTG + 3TC, 127.0 [IQR: 62.0, 215.0] vs. B/F/TAF, 182.0 [IQR: 123.5, 266.0], Z = -2.599, P = 0.009). No between-regimen differences were found in elevation of the CD4⁺ cell count and CD4/CD8 cell ratio from baseline to week 48 in subgroups stratified by baseline CD4⁺ cell count (<200 vs. ≥200 cells/μL).

Figure 4.

CD4⁺ cell count and CD4/CD8 cell ratio changes from baseline to week 48. (A) Changes in CD4⁺ cell count; (B) Changes in CD4/CD8 cell ratio. Error bars represent IQR. ART: Antiretroviral therapy; B/F/TAF: Bictegravir/emtricitabine/tenofovir alafenamide; cART: Combination ART; DTG/3TC: Dolutegravir/lamivudine; IQR: Interquartile range.

Comparison of safety profile of DTG + 3TC vs. B/F/TAF

Participants treated with B/F/TAF reported lower rates of drug-related AEs (7/143, 4.9%) than those on DTG + 3TC (18/137, 13.1%; Table 2). All these recorded AEs were considered to be grade 1 or 2, most frequently occurring in the first 4 weeks and then waning over time. Among participants who initiated DTG + 3TC, the most commonly recorded AEs were fatigue (8.0%, 11/137), insomnia (6.6%, 9/137), backache (2.2%, 3/137), and somnolence (1.5%, 2/137), whereas gastrointestinal toxicities such as nausea (0.7%, 1/143), anorexia (1.4%, 2/143), and diarrhea (0.7%, 1/143) were frequently reported in the B/F/TAF-treated subjects [Table 2]. No serious AEs and deaths were documented in both groups. One patient in the DTG + 3TC group experienced a drug-related insomnia, leading to regimen modification. No regimen discontinuation occurred in the B/F/TAF group.

Table 2.

Safety results of B/F/TAF and DTG/3TC groups from baseline to week 48.

AEs	B/F/TAF (n = 143)	DTG/3TC (n = 137)	Statistic χ2	P-value
Fatigue	0	11 (8.0)	239.124	<0.001
Insomnia	2 (1.4)	9 (6.6)	4.957	0.026
Somnolence	0	2 (1.5)	–	0.239*
Backache	0	3 (2.2)	1.436	0.231
Nausea	1 (0.7)	0	–	1.000*
Anorexia	2 (1.4)	0	–	0.498*
Diarrhea	1 (0.7)	0	–	1.000*
Rash	1 (0.7)	0	–	1.000*
AEs that led to study drug discontinuation	0	1 (0.7)	–	0.489*
SAEs	0	0	–	–

Data were shown as n (%). ^*P for B/F/TAF vs. DTG + 3TC based on Fisher exact test. AEs: Adverse events; B/F/TAF: Bictegravir/emtricitabine/tenofovir alafenamide; DTG/3TC: Dolutegravir/lamivudine; SAEs: Severe adverse events; –: Not applicable.

All lipid parameters were documented increases from baseline to the final follow-up in both treatment groups, although without statistical significance. Changes in median triglycerides and HDL cholesterol at week 48 significantly favored DTG/3TC vs. B/F/TAF. No between-regimen differences in median change from baseline were observed for other lipid parameters (total cholesterol, low-density lipoprotein cholesterol; Table 3). The overall median weight change from baseline to week 48 was +1.5 (IQR: -1.0, 4.6) kg in the DTG + 3TC group and +2.5 (IQR: 0.0, 5.0) kg in the B/F/TAF group (Z = -2.133, P = 0.033).

Table 3.

Changes from baseline in renal and fasting lipid parameters at week 48.

Renal and fasting lipid parameters	B/F/TAF (n = 143)		DTG/3TC (n = 137)		Statistic Z	P-value*
	Baseline values	Median change from baseline	Baseline values	Median change from baseline
Cr (µmol/L)	77.0 (67.0, 83.0)	10.0 (0, 29.0)	75.0 (65.0, 84.0)	8.3 (-1.0, 20.0)	-0.148	0.882
eGFR (mL/min)	111.3 (101.5, 121.1)	-9.7 (-19.6, 0)	112.8 (96.6, 125.0)	-8.9 (-20.0, 1.0)	-0.543	0.587
TG (mmol/L)	1.32 (1.00, 1.80)	0.18 (-0.23, 0.86)	1.41 (0.91, 2.14)	0.01 (-0.51, 0.70)	-2.487	0.013
CHOL (mmol/L)	4.03 (3.51, 4.66)	0.35 (-0.25, 0.96)	4.00 (3.43, 4.57)	0.24 (-0.34, 1.01)	-0.048	0.961
HDL-C (mmol/L)	1.00 (0.87, 1.23)	0.16 (-0.07, 0.33)	1.07 (0.89, 1.40)	0.07 (-0.17, 0.30)	-2.205	0.027
LDL-C (mmol/L)	2.58 (2.15, 3.11)	0.42 (-0.23, 0.84)	2.49 (1.94, 3.04)	0.28 (-0.38, 0.89)	-0.788	0.409

Data were shown as median (interquartile ranges).^*P-values were from the two-sided Wilcoxon rank sum test used to compare the median change from baseline in the two treatment groups. B/F/TAF: Bictegravir/emtricitabine/tenofovir alafenamide; CHOL: Total cholesterol; Cr: Creatinine; DTG/3TC: Dolutegravir/lamivudine; eGFR: Estimated glomerular filtration rate; HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol; TG: Triglyceride.

At week 48, the median increases from baseline in creatine and the median decreases in eGFR did not differ significantly between the treatment groups [Table 3].

Discussion

The introduction of the second-generation INSTI-based regimens has offered optimal treatment options for PWH in China. Herein, most of the patients are treated with tenofovir + lamivudine + efavirenz (TLE), which is a preferred initial regimen provided by the National Free Antiretroviral Treatment Program. The clinical use of INSTIs in China today is largely dominated by DTG-based two-drug regimens (2DRs) or three-drug regimens (3DRs) and B/F/TAF, which are potentially beneficial for patients infected with HIV, especially for those faced with OIs, DDIs, or DRAEs. So far, several observational studies have compared the clinical outcomes between TLE and DTG + 3TC or TLE and B/F/TAF in treatment-naïve individuals, respectively, and all have consistently showed that either DTG + 3TC or B/F/TAF was superior to TLE in terms of rapid decline in VL or regimen persistence.^[14–16] However, we still lack comparative data in regard to efficacy and safety of DTG + 3TC and B/F/TAF in ART-naïve individuals in a real-life scenario. Therefore, we conducted this study to fill this knowledge gap.

In this study of ART-naïve PWH in China, both the regimens demonstrated excellent rates of TND, desirable CD4⁺ cell recovery, and favorable tolerability through 48 weeks. No differences in virological response rates were observed at weeks 24 and 48. The finding was in line with the results from two separate Spanish cohorts whose primary end points were at 24 and 48 weeks, respectively, thereby reporting a similar effectiveness profile of DTG + 3TC compared to B/F/TAF or to other first-line ART regimens in naïve individuals.^[17,18] These real-life data verified that the simplified dual regimen DTG + 3TC offered comparable high virological efficacy vs. B/F/TAF after 48 weeks in ART-naïve PWH.

Still, one question is discussed controversially on whether the dual regimen DTG/3TC is sufficiently potent to provide rapidity of initial VL decline indicated by a short time to viral suppression. The result of the GEMINI-1 and -2 studies has revealed a median of 57 days to reach TND status (HIV-1 RNA <40 copies/mL) in both DTG + 3TC and DTG + FTC/TDF groups in the overall study population.^[19] Likewise, an open-label, multicenter, randomized, and pilot clinical trial conducted in Spanish has confirmed a similar rapid initial VL decay in both B/F/TAF and DTG/3TC groups, and no between-regimen difference was observed in HIV-1 RNA decline from baseline or the percentages of participants with HIV-1 RNA <20 copies/mL at each time point (days 3,7,14,98, and weeks 12,24) in blood plasma, semen plasma, and rectal fluid.^[20] The data currently available from randomized controlled trials (RCTs) and pilot clinical trials thus supported the potency of DTG + 3TC in initial rapid viral suppression comparable to those obtainable with BIC- or DTG-based 3DR. Interestingly, contrary to the findings in those trials, we found that B/F/TAF was associated with a faster HIV-1 RNA decline than DTG/3TC, with most participants achieving TND VL within 12 weeks. Furthermore, the median time to TND VL values was observed to be 12 weeks in the B/F/TAF group, apparently shorter than that observed in the DTG + 3TC group (24 weeks). Indeed, the median time to reach virological suppression (VL <50 copies/mL) observed in other retrospective studies of Asian populations was similar to that in our findings, 184 days in the DTG + 3TC group and 90 days in the B/F/TAF group.^[21,22] Differences in baseline characteristics may contribute to mixed results of RCTs and real-life studies. Both the prevalence of comorbidities and higher baseline VL characteristic of the study populations may explain why HIV-1 RNA decay with DTG + 3TC during initial treatment was slow in the present real-life study than in the GEMINI study. Also, a small sample size with only 25 participants (16 in 3TC/DTG, 9 in B/F/TAF) in the Spanish pilot study may possibly limit the statistical power of the result to indicate differences between regimens. Altogether, our findings supported that B/F/TAF had a greater advantage over DTG + 3TC in early VL suppression. We speculated that the discrepancy between B/F/TAF and DTG/3TC regimens in the rapid rate of VL decline may be attributed to their different pharmacological properties. First, among currently available short-acting INSTIs, BIC has the lowest in vitro IC_{50, 90, 95} values and the longest half-life and presents a high inhibitory quotient (IQ) similar to that of DTG.^[12] Second, BIC is only available in a fixed-dose combination (FDC) tablet with FTC and TAF, when compared to the dual regimen DTG + 3TC, the inclusion of TAF may presumably enhance the overall potency of the triple combination, thereby improving the early antiviral efficacy of B/F/TAF.

High VL and/or a low CD4⁺ cell count at baseline is/are thought to be correlated with more treatment failures and slow VL reduction.^[23] However, there were no identified virological failures in our study. The rates of TND at weeks 24 and 48 were consistently high and without between-regimen difference within the stratified subgroups, regardless of baseline HIV-RNA levels or CD4⁺ cell counts. Corresponding to the overall analysis at week 12, B/F/TAF therapy was superior to DTG/3TC therapy in terms of response rates across various subgroups analyzed, and a relatively lower response rate was observed in participants with high VL (≥100,000 copies/mL) and those with low CD4⁺ count (<200 cells/mL) for both treatment groups. These results were generally in agreement with previous reports, where a high VL/low CD4⁺ cell count at baseline did not jeopardize the sustained and durable efficacy of either B/F/TAF or DTG + 3TC but did have some impacts on early VL decay.^[24,25]

The CD4⁺ cell count and CD4/CD8 ratio have been established as important markers of immune recovery and prognosis in people receiving ART. A growing body of evidence suggests that initiating ART with an INSTI results in greater immune recovery than initiating ART with other anchor agents.^[26,27] Data from the present study showed a great benefit in the recovery of CD4⁺ cell count and CD4/CD8 ratio 48 weeks after initiation of B/F/TAF and DTG + 3TC and with no observed inter-group difference. The finding was consistent with the results from several clinical trials, which reported no significant difference in CD4⁺ cell count changes between B/F/TAF- and DTG-based 3DR or between DTG + 3TC and B/F/TAF.^[28,29] Additionally, among participants with advanced HIV (CD4⁺ count <200 cells/mL), we found similar CD4⁺ cell count and CD4/CD8 ratio changes at 48 weeks between DTG + 3TC and B/F/TAF. Studies in this special population have yielded mixed results. While a multicenter European observational study reported no difference in the improvement of this marker between regimens,^[30] the OPERA cohort revealed a faster CD4⁺ cell count recovery linked to B/F/TAF initiation than other INSTIs-based or boosted darunavir-based regimens.^[11] In the subgroup of VL ≥100,000 copies/mL, B/F/TAF presented a desirable CD4 cell count elevation than DTG + 3TC at week 48, although changes in the CD4/CD8 ratio was similar in both groups.

There were a numerically lower proportion of participants with drug-related AEs with B/F/TAF vs. DTG + 3TC through week 48. The result was similar to other comparative RCTs comparing B/F/TAF and DTG-containing 3DR, where drug-related AEs (fatigue, nausea, and neuropsychiatric symptoms) during the first 48 weeks were more common in the DTG group.^[31–33] In our cohort, participants receiving a DTG + 3TC regimen were more prone to experience fatigue and neuropsychiatric events, whereas participants receiving B/F/TAF were more likely to develop gastroenterological disorders. All drug-related AEs were considered grade 1 or 2 and were quickly resolved. Only one patient in the DTG + 3TC group modified the regimen due to insomnia. Indeed, the overall rates of AEs and discontinuation were extremely low in our cohort when compared to other real-life cohort studies. One possible explanation is that the documentation of AEs can vary depending on the scales and heterogeneity of the study populations used and how they are reported.

ART regimens are thought to have a negative effect on the lipid profile. However, INSTIs have a relatively minor impact on lipid levels.^[34] Indeed, we observed small increases from baseline lipid parameters at week 48 in both treatment groups. Inter-group comparison was significant for triglycerides and HDL cholesterol, which favored the DTG+3TC group. The elevated levels of total, LDL cholesterols were comparable between the two groups over 48 weeks. Another concern regarding metabolic health was excess weight gain specifically associated with the use of DTG- or BIC-based ART. In our study, participants who started on B/F/TAF gained more weight than those who started on DTG + 3TC. The excess weight gain observed in both study groups raises the question of whether DTG + 3TC and B/F/TAF may adversely impact lipid levels via changes in body mass. Furthermore, as TAF has also been related to weight gain after ART initiation, it is not yet understood whether weight gain is partly driven by synergistic effects of BIC and TAF. Moreover, whether lipid changes and weight gain increase the risk of metabolic syndrome is still unclear, and further longitudinal research is needed. DTG and BIC are also found to be inhibitors of the renal transporters (organic cation transporter 2 and multidrug and toxin extrusion protein 1), which can cause a small increase in serum creatinine, but are not clinically relevant.^[35] Our study demonstrated a consistent trend wherein both groups exhibited small incremental changes in serum creatinine and decreases in eGFR at week 48 relative to baseline, but inter-group differences were not significant.

A strength of this study is that it is the first to compare initial VL decay kinetics in ART-naïve PWH under DTG + 3TC and B/F/TAF regimens and utilizes a more stringent but exploratory TND measure. Nonetheless, our study has some limitations. Firstly, as a retrospective observational study, the results were inevitably subject to residual confounders due to discrepancies between dual DTG + 3TC and triple B/F/TAF. Second, missing data in the DTG + 3TC group, with a relatively limited sample size and short length of follow-up, may have introduced bias that impaired the findings' statistical power. Finally, median time to TND VL values was estimated in weeks, rather than days, using the Kaplan–Meier method, which may possibly lead to some deviation in results. Hence, our findings need to be validated in more large-scale prospective randomized clinical studies.

In conclusion, our results showed a comparably excellent response rate based on TND status at week 48 between groups and a significant higher rate for B/F/TAF at week 12, with no apparent role for initial higher VL or lower CD4⁺ cell count. B/F/TAF is preferred over DTG + 3TC due to a rapid decline in VL. A faster CVS may be of clinical value as for averting onward secondary HIV transmission and helping clinicians to select more appropriate individualized DRs depending on the clinical scenario, thus reinforcing sustained viral suppression and adherence to lifelong ART. Although both regimens are safe and well-tolerated, INSTI- and TAF-associated long-term metabolic health consequences of weight gain and lipid parameter changes still warrant further study.

Conflicts of interest

None.