Indirect Treatment Comparison of Long-Acting Injectable Cabotegravir as Pre-exposure Prophylaxis When Compared with no Pre-exposure Prophylaxis for HIV Prevention

Abstract

Introduction

The efficacy of long-acting injectable cabotegravir (hereafter referred to as cabotegravir) versus daily oral tenofovir disoproxil fumarate/emtricitabine (TDF/FTC) for pre-exposure prophylaxis (PrEP) was demonstrated in two phase 3 randomized controlled trials (RCTs), HPTN 083 and HPTN 084. As these trials did not have a no-PrEP group, this analysis aimed to perform an indirect treatment comparison (ITC) of cabotegravir versus no PrEP via the common comparator of oral TDF/FTC.

Methods

A systematic literature review identifying trials of oral or both oral and injectable PrEP reporting HIV acquisition and oral PrEP adherence, measured by detectable TDF/FTC plasma levels, was conducted (November 1, 2023). Heterogeneity in oral TDF/FTC adherence level between trials was expected to confound ITC estimates; therefore, a meta-regression of adherence and resulting oral PrEP effectiveness, or reduction in HIV acquisition, was incorporated into an ITC using a joint Bayesian model framework.

Results

The analysis included ten RCTs. The meta-regression showed a strong relationship between oral TDF/FTC adherence and effectiveness. The predicted effectiveness of oral TDF/FTC versus no PrEP was greater for HPTN 083 (77%) compared with HPTN 084 (47%), reflecting the higher level of adherence observed in HTPN 083 (86%) compared with HPTN 084 (56%). Based on the ITC, the predicted effectiveness of cabotegravir versus no PrEP was similar for both populations investigated in HPTN 083 (92%) and HPTN 084 (93%).

Conclusions

The ITC of cabotegravir versus no PrEP predicted similar estimates of cabotegravir effectiveness in the HPTN 083 and 084 trials, suggesting a very high level of efficacy despite differences in population, setting, underlying rate of HIV acquisition, and oral TDF/FTC adherence. These estimates support the generalizability of the cabotegravir results from both HPTN trials to other populations and regions than those in which these trials were conducted.

Graphical abstract available for this article.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s40121-025-01172-9.

Key Summary Points

Why carry out this study?

The superiority of long-acting injectable cabotegravir for pre-exposure prophylaxis (PrEP) versus oral tenofovir disoproxil fumarate/emtricitabine (TDF/FTC) as PrEP has been demonstrated in two phase 3 randomized controlled trials, HPTN 083 (men who have sex with men and transgender women) and HPTN 084 (cisgender women).

A no-PrEP group was not included in the HPTN 083 and 084 trials; thus, indirect estimates are required to inform a cabotegravir versus no-PrEP comparison in reducing HIV acquisition.

Cabotegravir could be highly beneficial for populations without a suitable existing oral PrEP option, such as individuals with suboptimal adherence, taking medications with contraindications, who have difficulties swallowing, or who have cultural beliefs or personal circumstances that make the possession and use of oral PrEP problematic; thus, the primary objective of this analysis was to perform an indirect treatment comparison (ITC) of cabotegravir versus no PrEP via the common comparator of oral TDF/FTC as PrEP.

What was learned from this study?

Variation in adherence to oral TDF/FTC as PrEP was highly predictive of the effectiveness of TDF/FTC versus no PrEP.

The ITC revealed similar estimates of effectiveness for cabotegravir versus no PrEP in the HPTN 083 and 084 trials, suggesting a very high level of efficacy despite the differences in the population, setting, underlying rate of HIV acquisition, and adherence to oral TDF/FTC, which supports the generalizability of the HPTN 083 and 084 trial results of cabotegravir to other populations and regions.

Digital Features

This article is published with digital features, including a graphical abstract, to facilitate understanding of the article. To view digital features for this article, go to 10.6084/m9.figshare.29098205.

Introduction

Since 2015, the World Health Organization (WHO) has recommended tenofovir disoproxil fumarate (TDF)–based oral pre-exposure prophylaxis (PrEP) as part of a comprehensive prevention approach [1], with TDF/emtricitabine (FTC) as the most widely available usage form [2]. Oral PrEP effectiveness has been shown to have a strong correlation with adherence, indicating reduced adherence limits protection against HIV acquisition [3–5]. Notably, the Vaginal and Oral Interventions to Control the Epidemic (VOICE) trial found only 29% of women receiving oral PrEP (TDF/FTC) had detectable tenofovir plasma levels and no evidence of a treatment effect between oral PrEP and placebo in HIV incidence was observed [4]. Furthermore, a meta-analysis revealed that among 16 studies reporting adherence via plasma drug concentrations, the estimated pooled rate of suboptimal adherence was 43% in 2663 people receiving oral PrEP [6].

In 2022, long-acting injectable cabotegravir (hereafter referred to as cabotegravir), the first injectable PrEP regimen, was recommended by WHO [7]. The superior effectiveness of cabotegravir for PrEP versus oral TDF/FTC for PrEP has been demonstrated in two phase 3 randomized controlled trials, HPTN 083 (men who have sex with men and transgender women) and HPTN 084 (cisgender women), with few HIV acquisitions in people receiving cabotegravir in both trials [8, 9]. Additionally, when administered every 2 months, cabotegravir for PrEP has demonstrated improved adherence compared with once-daily oral TDF/FTC as PrEP. Cabotegravir could also be highly beneficial for populations without a suitable existing oral PrEP option, such as individuals with suboptimal adherence, taking medications with contraindications, or having difficulties swallowing [10]. Additionally, the every-2-month dosing schedule of cabotegravir for PrEP may be beneficial for populations who have cultural beliefs or personal circumstances that make the possession and use of oral PrEP problematic [11]. An additional long-acting injectable for PrEP, lenacapavir, has recently undergone evaluation in two phase 3 clinical trials (PURPOSE 1 and PURPOSE 2) [12, 13]. However, it is not yet approved and available as a comparator in clinical practice and is not the focus of the present study.

As a no-PrEP group was not included in the HPTN 083 and 084 trials [8, 9], it is essential to conduct an indirect treatment comparison (ITC) to evaluate the relative value of cabotegravir versus no PrEP. Additionally, evidence synthesis (combining information from multiple trials that investigated the same topic to comprehensively understand trial findings [14]) comparing cabotegravir for PrEP with no PrEP may be important as part of the standard health technology assessment process. By using evidence synthesis effectively, policymakers, healthcare institutions, clinicians, researchers, and the public can make more informed decisions about health and healthcare. It is also important to have a contemporary and transparent analysis based on a systematic literature review (SLR) to support ongoing health technology assessment activities and to meet regulatory agency requirements. As the effectiveness of oral TDF/FTC observed in clinical trials varies because of differing levels of adherence [15], it is important to account for differences in adherence when conducting this ITC of cabotegravir for PrEP with no PrEP.

The primary objective of this analysis was to perform an ITC of the effectiveness of cabotegravir for PrEP compared with a no-PrEP option via the common comparator of oral TDF/FTC. Secondary objectives were to identify and report levels of oral TDF/FTC adherence heterogeneity across clinical trials, model the relationship between effectiveness of TDF/FTC and adherence, and use this relationship in the meta-regression to predict the effectiveness of cabotegravir versus no PrEP and the background risk of HIV acquisition.

Methods

Study Design

This study was an ITC including meta-regression adjustment for study-level covariables. Bayesian methods were used to allow the uncertainties in the individual study estimates and meta-regression coefficients to be jointly represented in the final estimates of comparative effectiveness.

Ethical Approval

This ITC is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Data Sources

The clinical trials included in this analysis were identified in a recently conducted SLR, with searches run on November 1, 2023, using MEDLINE, EMBASE, and The Cochrane Central Register of Controlled Trials utilizing population, intervention, comparison, outcomes, and study criteria (Table 1; search strategies are available upon request). A supplementary search of key conferences was also conducted. A PRISMA diagram detailing the SLR search process is presented in Figure S1 [16], and trial identification was conducted by 2 systematic reviewers. Trials reporting adherence based on detectable tenofovir in plasma were eligible for inclusion in the ITC. The available trials form a connected network (Fig. 1), allowing an indirect comparison between cabotegravir for PrEP and no PrEP via the common comparator of oral TDF/FTC as PrEP, with direct comparisons available between cabotegravir and TDF/FTC and between TDF/FTC and no PrEP. Eligible trials investigated the effectiveness of oral TDF/FTC versus placebo (in this article, the terms 'placebo' and 'no PrEP' are used interchangeably). The HPTN 083 and 084 trial data from Landovitz et al. and Delany-Moretlwe et al. were utilized and all incident HIV acquisitions in the cabotegravir group were included, also comprising individuals who acquired HIV before enrollment, with no recent cabotegravir exposure, and before receiving cabotegravir injections [8, 9].

Table 1.

Systematic literature review eligibility criteria

PICOS	Inclusion criteria
Population	Cisgender women, men who have sex with men, and transgender women aged ≥ 18 years who have a greater likelihood of acquiring HIV Adolescents who have a greater likelihood of acquiring HIV
Interventions/Comparators	Long-acting injectable PrEP (e.g., cabotegravir for PrEP) Oral PrEP (e.g., TDF/FTC, TAF/FTC) Placebo or no PrEP
Outcomes	HIV incidence HIV acquisitions averted Adherence to PrEP Adverse events Incidence of other STIs Behavioral changes (e.g., condom use) Drug resistance
Study design	Randomized controlled trials
Other	English language only

Cabotegravir long-acting injectable cabotegravir, FTC emtricitabine, PICOS population, intervention, comparison, outcomes, and study criteria, PrEP pre-exposure prophylaxis, STI sexually transmitted infection, TAF tenofovir alafenamide, TDF tenofovir disoproxil fumarate

Fig. 1.

Trials included in the analysis. Cabotegravir long-acting injectable cabotegravir, FTC emtricitabine, TDF tenofovir disoproxil fumarate. ^aIncluding men

Outcomes

The primary outcome was the percent effectiveness ([1 – relative risk (RR)] × 100) of cabotegravir for PrEP compared with a no-PrEP option in reducing HIV acquisition. An indirect estimate of the comparative effectiveness of cabotegravir for PrEP compared with a no-PrEP option is presented as percent effectiveness. To reduce the risk of confounding, the indirect comparison incorporated a meta-regression of adherence to oral TDF/FTC as a study-level covariable describing modification of regimen effect for TDF/FTC as PrEP versus no PrEP.

The secondary outcome was the effectiveness of oral TDF/FTC expressed as a function of adherence. Adherence to oral TDF/FTC was estimated based on plasma samples. The measure of adherence used was the proportion of samples with detectable levels of TDF/FTC, which was chosen because it is an objective measure that was consistently reported across the identified clinical trials. Additionally, the background incidence of HIV acquisition for individuals not receiving PrEP in the HPTN 083 and 084 trials was estimated.

Data Analysis

Because the effectiveness of oral TDF/FTC is strongly dependent on the level of adherence to this regimen [17], heterogeneity in adherence levels may confound estimates from an ITC between cabotegravir for PrEP and no PrEP. Thus, a meta-regression was performed to characterize the relationship between oral TDF/FTC adherence and TDF/FTC effectiveness versus no PrEP. Adherence meta-regression specifications are presented in the Supplementary Methods. Different functional forms, including with or without assigned sex as a variable [18], were considered for the meta-regression, as well as whether study location and population should be included. The logarithmic functional form was ultimately preferred and was determined to be the best-fitting equation to employ moving forward. The levels of adherence to oral TDF/FTC seen in the HPTN 083 and 084 trials were then input into the meta-regression equation to generate estimates of TDF/FTC effectiveness versus no PrEP in these clinical trial populations.

The ITC was conducted on the log RR scale of HIV acquisition. A log RR of < 1 indicated a reduced risk of HIV acquisition with the PrEP regimen. The log RR of cabotegravir for PrEP versus no PrEP was estimated as: log RR_{cabotegravir vs. no PrEP} = log RR_{cabotegravir vs. oral TDF/FTC} + Log RR_{oral TDF/FTC vs. no PrEP}. The relationship between oral TDF/FTC adherence and TDF/FTC effectiveness was specified by the following equations:

1. Log RR = α + β.Adℎerence (0–1)

2. % effectiveness = (1 − RR) × 100

The ITC and meta-regression analyses were implemented jointly as a hierarchical Bayesian model. Parameters were estimated using Gibbs sampling as implemented in Just Another Gibbs Sampler, model burn-in was 50,000 samples, results were monitored for 50,000 samples, and three chains were run.

Results

Identified Clinical Trials and Regimen Effect Estimates

Ten clinical trials [4, 8, 9, 19–25] were identified from the SLR and included in the analysis. The risk of bias assessment for the SLR is available in Figure S2. Details for the ten trials included in the ITC, including the populations studied and the location of trial sites, are presented in Table 2. Location and population were heterogeneous between the identified trials. Among the included trials, the RR of HIV acquisition ranged from 0.12 to 0.34 for cabotegravir compared with 0.14 to 1.04 for oral TDF/FTC as PrEP (Table 3). Adherence to oral TDF/FTC, as detected in plasma, was highly variable and ranged from 0.29 to 0.88.

Table 2.

Baseline characteristics of included clinical trials

Trial	Trial location	Design	Regimena	N	Age, mean (SD)/median [IQR], years	Population, %
						Men who have sex with men	Transgender women	Male	Female	Prefer not to answer
HPTN 083 (2021) [8]	Global	Double-blind	Cabotegravir	2282	26 [22–32]	88	12	–	–	< 1
			TDF/FTC	2284	26 [22–32]	87	13	–	–	< 1
iPrEx (2010) [25]	Global	Double-blind	TDF/FTC	1251	–b	100	–	–	–	–
			Placebo	1248	–b	100	–	–	–	–
PROUD (2016) [24]	England	Open-label	TDF/FTC	273	35 [30–43]	100	–	–	–	–
			Deferred PrEP	267	35 [2942]	100	–	–	–	–
IPERGAY (2015) [21]	Europe	Double-blind	Event-driven TDF/FTC	199	35 [29–43]	100	–	–	–	–
			Placebo	201	34 [29–42]	100	–	–	–	–
HPTN 084 (2022) [9]	Africa	Double-blind	Cabotegravir	1614	25 [22–30]	–	–	–	100c	–
			TDF/FTC	1610	25 [22–30]	–	–	–	100c	–
VOICE (2015) [4]	Africa	–	TDF/FTC	1003	25.2 (5.2)	–	–	–	100	–
			Placebo	1009	25.3 (5.2)	–	–	–	100	–
FEM-PrEP (2012) [23]	Africa	Double-blind	TDF/FTC	1062	23 [18–35]	–	–	–	100	–
			Placebo	1058	23 [18–35]	–	–	–	100	–
TDF2 (2012) [22]	Africa	Double-blind	TDF/FTC	611	–b	–	–	54d	46d	–
			Placebo	608	–b	–	–	54d	46d	–
Partners PrEP Study continuation (2014) [19]	Africa	Double-blind	TDF	2215	33 [29–40]	–	–	62d	–	–
			TDF/FTC	2212	34 [28–40]	–	–	64d	–	–
Bangkok Tenofovir Study (2013) [20]	Asia	–	TDF	1204	–b	–	–	80e	–	–
			Placebo	1209	–b	–	–	80e	–	–

Cabotegravir long-acting injectable cabotegravir, FTC emtricitabine, PrEP pre-exposure prophylaxis, TDF tenofovir disoproxil fumarate

^aTDF/FTC was administered orally

^bAge reported in categories (no mean/median available)

^cHeterosexual

^dPeople who inject drugs

^eCisgender women

Table 3.

Trial regimen effect estimates

Trial	Regimena	Population	RR (95% CI) of HIV acquisition	TDF/FTC adherence, proportion of samples with detectable levels of TDF/FTC
Trials that have a placebo comparator
Partners PrEP Study continuation (2014) [19]	TDF/FTC	Heterosexual males	0.16 (0.06, 0.46)	0.81
		Heterosexual females	0.34 (0.16, 0.72)	0.81
Bangkok Tenofovir Study (2013) [20]	TDF	Males who use drugs	0.62 (0.32, 0.82)	0.66
		Females who use drugs	0.21 (0.03, 0.83)	0.66
iPrEx Trial (2010) [25]	TDF/FTC	Men who have sex with men	0.56 (0.37, 0.85)	0.50
VOICE (2015) [4]	TDF/FTC	Females	1.04 (0.73, 1.49)	0.29
IPERGAY (2015) [21]	Event-driven TDF/FTC	Men who have sex with men	0.14 (0.02, 0.60)	0.86
TDF2 (2012) [22]	TDF/FTC	Heterosexual males	0.51 (0.19, 1.22)	0.77
		Heterosexual females	0.20 (0.03, 0.75)	0.77
FEM-PrEP (2012) [23]	TDF/FTC	Females	0.94 (0.59, 1.52)	0.36
PROUD (2016) [24]	TDF/FTCb	Men who have sex with men	0.14 (0.04, 0.36)	0.88
Trials that have a comparator of TDF/FTC
HPTN 083 (2021) [8]	Cabotegravir	Men who have sex with men/transgender women	0.34 (0.18, 0.62)	0.86
HPTN 084 (2022) [9]	Cabotegravir	Cisgender females	0.12 (0.05, 0.31)	0.56

Cabotegravir long-acting injectable cabotegravir, FTC emtricitabine, PrEP pre-exposure prophylaxis, RR relative risk, TDF tenofovir disoproxil fumarate

^aTDF/FTC was administered orally

^bTDF/FTC deferred for 1 year

Base Case Relationship Between Adherence to and Effectiveness of oral TDF/FTC

The relationship between adherence to and effectiveness of oral TDF/FTC as PrEP versus no PrEP in reducing HIV acquisition is shown in Fig. 2. There appeared to be a strong relationship between adherence and effectiveness by trial, location, and population. There was no obvious deviation from the overall trend by location or population; however, there were insufficient trials to allow the estimation of location- or population-specific slopes.

Fig. 2.

The relationship between adherence to and effectiveness of oral TDF/FTC for PrEP versus no PrEP in reducing HIV acquisition by a trial, b location, and c population. FTC emtricitabine, PrEP pre-exposure prophylaxis, TDF tenofovir disoproxil fumarate

Model fit

Lower deviance information criteria (DIC) indicated a better fitting model; a difference of ≥ 2 was considered significant. The DIC was 230.79 for the linear relationship, 233.31 for the logarithmic relationship + sex, and 231.61 for the logarithmic relationship; thus, the model including assigned sex as a covariable had a higher DIC than the other 2 models (by ~ 2), suggesting that the inclusion of assigned sex as a covariable does not improve model fit (lower DIC indicates a better fitting model). Based on these observations, the logarithmic model without the assigned sex covariable was selected as the base case model (α = 0.9064 and β = − 2.7828) as there was no evidence that including the assigned sex covariable improved model fit. The use of a logarithmic functional form constrained the predicted RRs to be positive (estimated percent effectiveness to be < 100%).

Relationship Between Predicted Effectiveness of Oral TDF/FTC Versus no PrEP and Adherence

In the base case model, the mean (2.50–97.50% credible interval [CrI]) predicted effectiveness of oral TDF/FTC versus no PrEP was greater for HPTN 083 (77% [66.10–85.17]) compared with HPTN 084 (47% [35.32–57.96]; Fig. 3; Table 4), aligned with the higher level of adherence to TDF/FTC observed in HTPN 083 (86%) compared with HPTN 084 (56%; Table 3). The predicted relationship between adherence and effectiveness for the meta-regression models is presented in Figure S3.

Fig. 3.

The relationship between effectiveness and adherence in the base case (logarithmic relationship) meta-regression model. Cabotegravir long-acting injectable cabotegravir, FTC emtricitabine, PrEP pre-exposure prophylaxis, TDF tenofovir disoproxil fumarate

Table 4.

Base case predicted effectiveness and underlying risk of HIV acquisition^a

Model	Trial	Mean	2.50% Crl	97.50% Crl
Predicted % effectiveness of cabotegravir vs. no PrEP	HPTN 083	91.88	84.13	96.40
	HPTN 084	92.72	83.44	97.45
Predicted % effectiveness of TDF/FTC vs. no PrEP	HPTN 083	76.89	66.10	85.17
	HPTN 084	47.44	35.32	57.96
Predicted underlying risk of HIV acquisition (no PrEP event rate/100 PY)	HPTN 083	5.54	3.18	9.01
	HPTN 084	3.57	2.35	5.14

Cabotegravir long-acting injectable cabotegravir, CrI credible interval, FTC emtricitabine, PrEP pre-exposure prophylaxis, PY person-year, TDF tenofovir disoproxil fumarate

^aPredicted effectiveness was defined as prevention of HIV acquisition

Predicted Effectiveness of Cabotegravir Versus no PrEP and Underlying risk of HIV Acquisition

Based on the ITC (cabotegravir versus oral TDF/FTC), the mean (2.50–97.50% CrI) predicted effectiveness of cabotegravir versus no PrEP was similar for HPTN 083 (92% [84.13–96.40]) and HPTN 084 (93% [83.44–97.45]; Table 4). The underlying rates of HIV acquisition (logarithmic relationship) for people not receiving PrEP were estimated as 5.5 per 100 person-years and 3.6 per 100 person-years for the HPTN 083 and 084 trial populations, respectively.

Discussion

The ITC of cabotegravir versus no PrEP estimated the mean (CrI) effectiveness in reducing HIV acquisition to be 92% (84.13–96.40) in men who have sex with men and transgender women (HPTN 083 population) and 93% (83.44–97.45) in cisgender women (HPTN 084 population). Estimates of effectiveness were similar for cabotegravir versus no PrEP in the HPTN 083 and 084 trials, despite the differences in the population, setting, underlying rate of HIV acquisition, and oral TDF/FTC adherence [8, 9], supporting the generalizability of the HPTN trial results to other populations. Additionally, a conservative approach was used by including all HIV acquisitions reported in the HPTN trials regardless of when they happened (e.g., before enrollment).

The relationship between oral PrEP effectiveness and adherence is well established [15, 26, 27]. The present analysis supports this relationship, demonstrating variation in adherence to oral TDF/FTC as PrEP was highly predictive of TDF/FTC effectiveness versus no PrEP. The variation in adherence also appears to explain a large degree of the heterogeneity observed in trial results. A range of adherence to oral TDF/FTC was reported across trials, and a strong relationship between TDF/FTC adherence and effectiveness by trial, location, and population was observed. However, there was no obvious deviation from the overall trend by location or population. The predicted effectiveness of oral TDF/FTC versus no PrEP was greater for HPTN 083 compared with HPTN 084, reflecting the higher level of TDF/FTC adherence observed in HTPN 083 compared with HPTN 084.

Variable adherence to oral TDF/FTC has been observed in real-world practice [28–31] and is lower than observed in clinical trials; therefore, the effectiveness of TDF/FTC may also be lower in practice. Additionally, people may have difficulties with adherence to oral PrEP because of factors such as stigma, discrimination, and pill burden [32]. However, data from real-world studies showing high oral PrEP adherence as measured by detectable tenofovir in urine or blood are limited due to a lack of robustness as a result of typically being self-reported [33–35].

In contrast, high adherence to cabotegravir for PrEP has been observed in clinical practice. Over a median follow-up of 10 months, real-world adherence data for cabotegravir from the OPERA cohort (n = 764) found that among the 85% of people who completed the initiation sequence, 69% received all cabotegravir injections on time, with most continuation injection delays being short (median delay, 3 days) [36]. Overall, 19% of people who completed the initiation sequence had a missed injection, or ≥ 128 days without an injection. Over a 2-year follow-up, real-world use of cabotegravir for PrEP from the Trio Health Cohort (n = 526) found that among the 75% of people who received continuation injections (≥ 3 cabotegravir injections), 65% received all injections on time, with a median of 1 delayed injection and a median delay of 12 days [37]. Missed injections occurred in 3% of people, with a maximum of 1 missed injection in individuals with continuation injections. The adherence data from these cohorts support the high effectiveness of cabotegravir for PrEP (> 99%) observed in the real world [36, 37].

The results of this analysis may be used in future models to estimate the impact of cabotegravir on HIV incidence in different epidemiological contexts. As with any meta-analysis, although the results of the ITC appear to be robust to the specification of the meta-regression component, limitations exist. The effectiveness of cabotegravir for PrEP versus no PrEP was estimated using indirect comparison methods; thus, the validity of the current analysis relies on the validity of the consistency assumption (RR_{A vs. C} = RR_{A vs. B} × RR_{B vs. C}) and the validity of the meta-regression used to predict the effectiveness of oral TDF/FTC versus no PrEP as a function of adherence to TDF/FTC when used to predict the effectiveness of TDF/FTC in the HPTN 083 and 084 trial populations. A sensitivity analysis excluding PROUD, Bangkok study, and IPERGAY was performed, due to differences in the reporting of adherence and trial design, included population, or oral dosing, respectively; however, the exclusion of these trials had minimal impact on the results of the analysis compared with the base case analysis. The HTPN 083 and 084 trials did not include no PrEP groups, and there are no available trials directly comparing cabotegravir for PrEP versus no PrEP to validate the predictions of effectiveness used within this analysis. However, the results of the present ITC align with published estimates in previous modeling studies that estimated cabotegravir effectiveness using a counterfactual placebo comparator [38, 39]. There were insufficient trials available to include other covariables (such as geography or population) in the meta-regression model alongside adherence; however, there was no obvious deviation from the overall trend in adherence and effectiveness by population or location, suggesting adherence is the most important determinant of oral PrEP effectiveness. Lastly, the model assumes that adherence to cabotegravir will be similar to the adherence observed in the HPTN 083 and 084 trials.

Conclusions

In conclusion, variation in adherence to oral TDF/FTC as PrEP was highly predictive of the effectiveness of TDF/FTC versus no PrEP, and the ITC of cabotegravir versus no PrEP suggested very high effectiveness of cabotegravir in reducing HIV acquisitions, with predicted effectiveness rates of 92% in men who have sex with men and transgender women and 93% in cisgender women. Similar estimates of the effectiveness for cabotegravir for PrEP versus no PrEP were observed in the ITC for both HPTN 083 and 084 trial populations despite the differences in the population, setting, underlying rate of HIV acquisition, and oral TDF/FTC adherence, suggesting the generalizability of the HPTN trial results to other populations.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Conceptualization: Paul O'Brien, Sarah-Jane Anderson, Laure Dupont-Benjamin, and Melanie Schroeder; Data curation: Neil Hawkins and Juliette Thompson; Funding acquisition: Laure Dupont-Benjamin and Melanie Schroeder; Investigation: Neil Hawkins, Juliette Thompson, and Eric Manalastas; Methodology: Laure Dupont-Benjamin and Melanie Schroeder; Project administration: Paul O'Brien, Laure Dupont-Benjamin, and Melanie Schroeder; Supervision: Paul O'Brien, Laure Dupont-Benjamin, and Melanie Schroeder; Validation: Neil Hawkins and Eric Manalastas; Writing—original draft preparation: Paul O'Brien, Sarah-Jane Anderson, Laure Dupont-Benjamin, and Melanie Schroeder; Writing—review and editing: Neil Hawkins, Paul O'Brien, Juliette Thompson, Sarah-Jane Anderson, Eric Manalastas, Laure Dupont-Benjamin, and Melanie Schroeder.

Funding

This study, including the journal’s Rapid Service Fee, was funded by ViiV Healthcare.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflicts of Interest

Neil Hawkins, Juliette Thompson, and Eric Manalastas have received consulting fees from ViiV Healthcare. Paul O'Brien is an employee of ViiV Healthcare and may own stock in GSK. Sarah-Jane Anderson is an employee of GSK and owns stock in GSK and Haleon. Laure Dupont-Benjamin is an employee of ViiV Healthcare; owns stock in AbbVie, GSK, and Haleon; and has received consulting fees from the International Union Against Tuberculosis and Lung Disease. Melanie Schroeder is an employee of ViiV Healthcare; owns stock in GSK and Haleon.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.