Bringing doctors together

or

A free, private network for verified physicians

Leading voices

Dr. Lance Okeke

HIV/AIDS Medicine · Duke University

View

Dr. Aniruddha Hazra

HIV/AIDS Medicine · UCMC

View

Dr. Shital Patel

HIV/AIDS Medicine · BCM

View

Dr. Monica Gandhi

HIV/AIDS Medicine · UCSF

View

Dr. Ashwin Balagopal

HIV/AIDS Medicine · Johns Hopkins

View

Dr. Ofole Mgbako

HIV/AIDS Medicine · NYU Langone

View

Dr. Joshua Barocas

HIV/AIDS Medicine · CU Anschutz Medical Campus

View

Dr. Chris Beyrer

Global Health · Duke University

View

Dr. Keith Kaye

General ID · RU

View

Dr. Helen Boucher

Transplant/Immunocompromised States · Tufts

View
  • Discover your community

  • Stay current

  • A profile that represents you

  • Find your people

About RoonPrivacyTerms

    Overview

    Cabotegravir Maintains Protective Efficacy in the Setting of Bacterial Sexually Transmitted Infections: A Secondary Analysis of HPTN 083.

    Feb 25, 2026DOI: 10.1093/cid/ciae572Journal ArticleCID

    Written by: Meredith E ClementBrett HanscomDaniel HainesJose A BazanNuntisa ChotirosniramitRyan KofronSharon MannheimerKenneth H MayerMayara Secco Torres SilvaLydia Soto-TorresAlex R RinehartJames F RooneyAndrea JenningsKailazarid Gomez-FelicianoMarybeth McCauleyBeatriz GrinsztejnRaphael J Landovitzfor HPTN 083 Study Team

    Find these authors on Roon

    RL
    Dr. Raphael J. Landovitz
    HIV/AIDS Medicine · UCLA

    METHODS

    HPTN 083 was a double-blind, double-dummy randomized controlled trial with participants assigned 1:1 to either active CAB-LA with TDF/FTC placebo or active TDF/FTC with CAB-LA placebo. The trial included a 5-week oral-tablet lead-in phase, an injection phase with the initial 2 injections 4 weeks apart followed by injections every 8 weeks thereafter for approximately 3 years, and a “tail phase” (referring to decline in plasma levels of CAB-LA) in which participants received open-label TDF/FTC for 48 weeks starting 8 weeks after the last injection. Study visits occurred at weeks 2 and 4 of the oral lead-in phase, with a safety visit at week 5. During the injection phase, oral tablets were dispensed and injections administered at weeks 5, 9, and 17 and every 8 weeks thereafter until week 153. Safety assessment visits occurred at weeks 6 and 10 and then 2 weeks following each injection visit. The complete study design and details of HPTN 083 have been previously reported [13]. At the first preplanned interim end-point analysis, the randomized portion of the trial was stopped for efficacy by an independent data and safety monitoring board, and the trial was continued with an open-label design. This analysis includes data collected during the blinded phase, from December 2016 until May 2020. The HIV testing algorithm performed at all on-study visits included an HIV rapid test (cleared by the United States Food and Drug Administration) and a laboratory-based HIV antigen and antibody test. Additional retrospective analysis of stored specimens occurred at a central laboratory [14]. Serologic testing for syphilis and nucleic acid amplification testing for rectal and urethral gonorrhea and chlamydia were conducted every 6 months, and at investigator discretion for exposures or STI symptoms. Sites used validated nucleic acid amplification testing, although there was variability in type of test kit used by site. The date of first evidence of HIV infection and all reactive serologic tests for syphilis were centrally adjudicated for maximum assurance of treatment of prevalent syphilis upon study entry and appropriate treatment of incident or relapsed cases while on study. After syphilis diagnosis was made, sites followed titers every 3 months to ensure an appropriate response. If titers were increasing or did not decline four-fold, participants' ongoing sexual risk factors were evaluated and retreatment was conservatively pursued on a case-by-case basis if reinfection could not be reasonably exonerated.The number and proportion of participants experiencing any incident STI and multiple incident STIs were tabulated overall and by arm, region, and country. Incidence rates for gonorrhea, chlamydia, and syphilis were calculated as the number of infections per 100 person-years (PY), including repeat infections. Person-time included time from enrollment to the last visit with an STI test before 15 May 2020. STI incidence rates were calculated for each individual STI separately (syphilis, urethral and rectal chlamydia, and urethral and rectal gonorrhea) and for all STIs combined. Participants classified as having multiple STIs could have had 2 different STIs on the same date or repeat infection of the same STI on a different date; having the same STI diagnosed at different sites (eg, rectal and urogenital gonorrhea) on the same date was also considered as multiple STIs. The proportion of infections experienced by each percentile of the study population was also calculated.Combined STI incidence rates were assessed by demographic characteristics including age, race, ethnicity, gender cohort, education, treatment arm, recreational drug use (yes/no), injection drug use (yes/no), alcohol use (Alcohol Use Disorders Identification Test-Concise; scores of ≥4 represent potential active alcohol abuse, dependence, or heavy drinking [15]), country, region, reported condom use for anal sex in the past month, partner number in the past month, current marital status, and baseline STI status. Poisson regression was used to model the association between these factors and STI incidence. A multivariable model included only variables that were significant in univariate models at alpha = 0.10. Separate subset models were run for North America/United States and Latin America to more accurately evaluate race (US and Latin America) and ethnicity (US only), as these are the regions in which race/ethnicity data were collected separately. In the multivariable model, country was excluded to avoid high correlation with region. The incidence analysis included participants in the modified intent-to-treat population who had baseline STI testing results and at least 1 STI test result at a follow-up visit before 15 May 2020.Cox proportional hazards modeling with STI status as a time-varying covariate was used to evaluate potential interactions between STI status and the relative efficacy of CAB-LA versus TDF/FTC. Predictor terms included study arm (CAB-LA vs TDF/FTC), STI status, and an interaction term between study arm and STI status. Because the precise dates of STI infection and completion of STI treatment were unknown, STI status was determined for each time interval between STI tests, with all intervals classified as STI-positive or STI-negative. For the initial efficacy analysis, STI status was time-varying, with intervals before and after each positive STI test defined as STI-positive, and all others defined as STI-negative. When baseline STI testing results were positive, STI-positive status was carried forward until the first follow-up visit when STI testing was performed, or if no follow-up STI testing was performed, STI-positive status was carried forward until the participant's last available visit date (censor date).Two sensitivity analyses were conducted. The first sensitivity analysis used a non–time-varying approach that classified participants as either ever having an STI (including at baseline), or never having an incident STI. The second sensitivity analysis was a time-varying model that carried STI positive status backward in time to the last negative STI test, but not forward, reflecting an assumption of rapid STI treatment and cure following diagnosis. For these “maintenance of efficacy analyses,” participants from the modified intent-to-treat population were included who had at least 1 STI test performed, for any bacterial STI, at baseline or follow-up.

    RESULTS

    STIs were diagnosed in approximately 41% of the 3859 participants included in the analysis, and proportions of participants with any STI were similar across randomized study arms (39.1% for CAB-LA and 41.8% for TDF/FTC, Table 1). Across regions, Africa had a higher proportion of participants with any STI (55.5%) compared to North America (40.3%), Latin America (40.0%), and Asia (39.6%). For countries within the Asian region, Vietnam had a higher proportion of participants with any STI (50.0%) compared to Thailand (35.7%); for countries within the Latin American region, Brazil had a higher proportion with any STI (49.9%) relative to Argentina (36.7%) and Peru (31.4%); the North American and African Regions are composed of 1 country only (United States and South Africa, respectively).Proportion of STIs Diagnosed Among Those Tested, and STI Incidence Rate Among Participants at all Sites During the HPTN 083 Primary Blinded Period, N = 3859Abbreviations: CAB-LA, long-acting cabotegravir; TDF-FTC, tenofovir disoproxil fumarate/emtricitabine; STI, sexually transmitted infection.The overall STI incidence rate was 50.7 infections/100 PY and was similar between the 2 arms (49.9/100 PY in the CAB-LA arm and 51.6/100 PY in the TDF/FTC arm, Table 1). Incidence rates of individual bacterial STIs were high, and highest for syphilis (16.7/100 PY), rectal chlamydia (16.7/100 PY), and rectal gonorrhea (11.0/100 PY).STI incidence rates among subgroups are shown in Figure 1. Participants with an STI at baseline had higher rates of incident STIs during the study compared to those without a baseline STI (83.9/100 PY vs 43.1/100 PY). Additionally, STI incidence tended to be higher among younger participants (56.4/100 PY among participants in the 18–29 year-old age group vs 40.9/100 PY in participants aged 30 and older), participants with less education (65.8/100 PY among participants with primary-school education vs 48.1/100 PY among those with college education), and among participants living with a primary partner but not married (57.1/100 PY) compared to participants in a formal/legal partnership (39.9/100 PY). Racial differences within region were substantial, with Black/Mixed Race participants in Latin America having higher rates than Native or White participants (92.1/100 PY vs 49.2/100 PY and 66.7/100 PY, respectively). Black participants in the United States had a higher STI incidence rate relative to non-Black US participants (50.1/100 PY vs 34.4/100 PY). Additionally, Hispanic/Latinx participants in the United States had higher STI incidence rate relative to non-Hispanic/Latinx US participants (44.0/100 PY vs 41.0/100 PY).Sexually transmitted infection incidence rates by baseline demographic factors.Overall, multiple STIs were identified in approximately 18% of participants, and the proportions were similar by study treatment arm (17.7% for CAB-LA and 18.1% for TDF/FTC). Notably, 79% of STI diagnoses occurred in 25% of participants. Africa had a higher proportion experiencing multiple STIs (30.0%) relative to Asia (18.9%), North America (18.6%), or Latin America (16.1%).In univariate models, age, region, country, ethnicity, education level, baseline STI status, partner number, marital status, and reported condomless anal sex were significantly associated with STI incidence at the alpha = 0.10 level. In the final multivariable model, only age, region, education level, marital status, and baseline STI status were statistically significant at P < .05 (Table 2). In the US and Latin America subset models, race was also statistically significant at P < .05.Association Between Baseline Factors and STI Incidence Using Poisson Regression Among Participants at all Sites During the HPTN 083 Primary Blinded Period, N = 3859Bold text = statistically significant.Abbreviations: AUDIT-C, Alcohol Use Disorders Identification Test-Concise; CI, confidence interval; IRR, incidence rate ratio; MSM, men who have sex with men; STI, sexually transmitted infection; TGW, transgender women.The maintenance-of-efficacy analysis included 4558 participants. In both the CAB-LA arm and the TDF/FTC arm, HIV incidence was numerically lower when STIs were absent relative to when STIs were present, although this finding was not statistically significant. However, independent of absence or presence of STIs, HIV incidence was lower with CAB-LA versus TDF/FTC—with hazard ratios (HR) of 0.31 and 0.37, respectively; there was no significant interaction between STI status and CAB prevention efficacy (P = .75, Figure 2)Association between bacterial sexually transmitted infections and cabotegravir efficacy: HIV incidence by arm, hazard ratios comparing cabotegravir to tenofovir disoproxil fumarate/emtricitabine, and P value for interaction between STIs and CAB efficacy. Abbreviations: CAB, cabotegravir; CI, confidence interval; HIV, human immunodeficiency virus; HR, hazard ratio; STI, sexually transmitted infection; TDF-FTC, tenofovir disoproxil fumarate/emtricitabine.For the first sensitivity analysis model, CAB-LA maintained efficacy relative to TDF/FTC when STIs were present (HR = 0.34) and when STIs were absent (HR = 0.32); the interaction term was nonsignificant (P = .90). For the time-varying sensitivity model with backward-only STI imputation, the HR comparing CAB-LA to TDF/FTC efficacy was stronger when STIs were present (HR = 0.23) compared to when STIs were absent (HR = 0.39), although the difference was nonsignificant (P = .48).

    DISCUSSION

    In HPTN 083, a large, global PrEP trial comparing CAB-LA to daily oral TDF/FTC that enrolled over 4500 transgender women and cisgender MSM, we found a high incidence of STIs, particularly those most closely associated with HIV acquisition, including syphilis, rectal chlamydia, and rectal gonorrhea. The central finding in this analysis was that CAB-LA maintained robust protective efficacy relative to TDF/FTC in the setting of bacterial STIs. This is a critical and heartening finding given the intertwined HIV and STI epidemics and ongoing global efforts to dramatically reduce HIV incidence amidst surging STI rates and highlights the need for novel STI prevention interventions that are globally scalable and implementable.In the maintenance-of-efficacy analyses, there was no evidence for interaction between randomized arm and incident STIs on HIV incidence. Over the past 4 decades, extensive investigations have documented the association between bacterial STIs and increased HIV transmission and acquisition. Epidemiologically, STI diagnosis may be representative of sexual behaviors that increase the probability of HIV acquisition and/or sexual activity within networks enriched with HIV infection. Yet numerous studies demonstrate a link between STI history and subsequent HIV acquisition independent of sexual behavior or other confounding factors [16–18]. Biological mechanisms have plausibility to explain how STIs promote HIV transmission. Ulcerative STIs—infections such as syphilis that cause genital ulcers—can lead to breakdown in mucosal integrity and exposure of mucosal lymphocytes, the main targets for HIV infection; whereas inflammatory STIs—infections such as gonorrhea and chlamydia that cause mucosal inflammation—can lead to cellular infiltration at the site of infection, also increasing potential target cells for HIV transmission [4, 19]. In further support of this biological relationship, it has been demonstrated that men with STIs are more susceptible to HIV viral variants with reduced fitness relative to men without STIs [20]. As additional PrEP agents advance through clinical development, it will be important to interrogate each one's capacity to protect against HIV infection in the setting of increased STI-induced susceptibility.Similar to other PrEP trials and demonstration projects, overall STI incidence in HPTN 083 was elevated [21–23]. Although STI incidence was not as high in as in some studies, which have reported rates upward of 90 infections per 100 PY, the rate in HPTN 083 was notably high for syphilis, which has been commonly linked to subsequent HIV acquisition among MSM [24, 25]. An evaluation of STI incidence rates is vital to the interpretation of PrEP trials to provide a surrogate of behavior among participants that serves as a proxy for potential for HIV exposure [26].Given high STI rates in populations taking PrEP, many studies have attempted to ascertain whether there is truth in the concept of “risk compensation”—that is, an increase in condomless sex when acquisition of HIV is mitigated by PrEP use [27–29]. It must be noted that comparisons of STI rates in populations taking versus not taking PrEP are fraught with detection bias (ie, more frequent STI screening in PrEP populations) as well as selection bias (ie, recent STI diagnosis is an indication for PrEP). One recent meta-analysis that carefully assessed pre-PrEP and post-PrEP sexual behaviors and STI rates found a decrease in condom use and an increase in STI diagnoses associated with PrEP, demonstrating the importance of effective STI prevention strategies for those receiving PrEP [30, 31]. Another study, however, demonstrated the benefit of engaging in frequent STI screening per guidelines [31]; using a network-based mathematical model, investigators demonstrated a decline in gonorrhea and chlamydia rates with more PrEP uptake—even when reduced condom use was factored in—as a result of more frequent screening and treatment, and ultimately slower transmission [32]. The elevated and rising rates of STIs and the remarkable effectiveness of PrEP for HIV prevention have reenergized the search for biomedical STI prevention interventions [33–37].In prior PrEP studies, factors associated with incident STIs are similar to what we demonstrated here, including younger age, race, and baseline STI status [38, 39]. Similarly, reported condom use and partner number were associated with incident STIs in our analysis [40]. We found that one-quarter of participants accounted for three-quarters of the diagnosed STIs over the study period. This finding is similar to findings from the Pre-exposure Prophylaxis Expanded study, an open-label PrEP study in Australia, and makes the case for STI prevention efforts concentrated among a subset of persons taking PrEP [39]. Such intensive efforts may include more frequent testing [32], education on symptoms and self-examinations [41], services aimed at expedited partner notification and treatment [42, 43], and/or biomedical prevention with vaccines or antibiotic prophylaxis (eg, doxycycline postexposure prophylaxis) [33–37]. Although not statistically significant, HIV incidence was higher in the presence of STIs, regardless of study arm in HPTN 083, a finding that has been seen in other PrEP trials, further highlighting the need for strengthened STI prevention efforts.A limitation of this study was that STI status was not always available on the date of first evidence of HIV infection, which was determined using retrospective testing on stored specimens and often predated HIV detection at the study site. We therefore conducted 2 sensitivity analyses, 1 as a more conservative estimate of STI-positive status (characterizing more person-years as STI-positive) and the other as a less conservative estimate (characterizing fewer PY as STI-positive), and results were consistent. We also did not have information on frequency of symptomatic versus asymptomatic infection, although the relationship between symptomatic STIs and inflammation and associated HIV transmission risk is unclear [44]. Finally, oropharyngeal STI screening was not included in the study procedures, and although historically not directly associated with HIV acquisition, may represent an important site for asymptomatic infection and secondary transmission to nonoral anatomic sites.We found a high rate of STIs among participants in HPTN 083, with infections concentrated among one-quarter of participants. Furthermore, STIs were more frequent among certain subgroups, and these data may be helpful in guiding strategic implementation of new biomedical STI prevention strategies, including doxycycline postexposure prophylaxis [45]. Critically, we demonstrated that CAB-LA maintained protective efficacy in the presence of STIs, and future PrEP agents should be similarly evaluated. It is reassuring that even as efforts continue to identify effective STI mitigation strategies, currently approved systemic PrEP agents appear to maintain their protective efficacy in the presence of such bacterial STIs.
    RLDr. Raphael J. Landovitz
    Leading Voice

    Dr. Raphael J. Landovitz

    Infectious Disease · HIV & AIDS Medicine
    University of California, Los Angeles
    raphaell
    Los Angeles, CA, USA