CROI 2023 and HIV prevention: Drs. Connie Celum and Renee Heffron
Connie Celum, MD, MPH |
Professor, Global Health & Medicine, Adjunct Professor Epidemiology
University of Washington
Renee Heffron, PhD, MPH
Professor, Department of Medicine
Director, Center for AIDS Research
University of Alabama at Birmingham
• Oral PrEP
2023 marked the 30 year anniversary of CROI as well as the first in-person CROI in 4 years and was a memorable CROI. The opening session included noteworthy sessions from Alan Perelman (Modeling Dynamics of HIV for Treatment and Cure), Kevin De Cock's N'Galy-Mann lecture (HIV and Global Health in a Pandemic Era, Yvette Raphael (Community and Advocates as Equal Partners in Research and Development), and Tony Fauci (CROI at 30: Chronical of HIV/AIDS Research Progress). The plenary sessions were outstanding and one that included discussion of HIV prevention was Monday's plenary by Ambassador John Nkengasong (20 years of PEPFAR: Looking Back, Striding Forward). Oral abstract sessions included lively discussions about STI interventions (Monday). Three exciting symposia focused on long-acting PrEP (Symposium 3: Delivering on the Succss of Injectable PrEP), STI and HIV prevention (Symposium 6: Getting PrEPped, (and PEPped) for Sex: From HIV to STIs to MPOX), and a highly relevant session about science communication (Symposium 9: Science Communication in the Age of Misinformation). A special session on Tues highlighted the results from the MOSAICO HIV vaccine trial and future directions for HIV vaccine trials.
• Long-acting and novel PrEP strategies
• HIV testing (incidence assays + algorithms and self-testing)
• Prevention of Mother to Child transmission
• STDs prevention (doxycycline, Bexsero quadrivalent meningococcal vaccine)
Oral PrEP: As in past CROI meetings, exciting data about oral PrEP delivery were presented at CROI 2023.
PrEP adherence, persistence, and impact
In abstract 163, Marrazzo conducted a pooled analysis of the effectiveness of oral FTC/TDF in women in more real world contexts from 11 demonstration projects in 6 countries which were conducted from November 2012 to December 2020. They quantified PrEP adherence by tenofovir-diphosphate (TFV-DP) levels in dried blood spots or TFV in plasma, electronic pill cap-monitoring, pill-counts, self-report, and/or study-reported adherence scales. Of the 6296 cisgender women included in their pooled analysis, the median age at PrEP initiation was 23 years with 78% women from Africa and 21% from Asia. A total of 32 women acquired HIV for an HIV incidence of 0.72/100 per woman-year. Younger women, married women and those without children were more likely to acquire HIV. The adherence trajectory modeling indicated that 17% were highly adherent, 22% consistently moderate, 39% moderate and declining, and 21% consistently low adherence. Women >25 years were over 2 times more likely to remain highly adherent than those <18. No new HIV infections were observed in women with high, steady adherence. These multinational data indicate that oral FTC-TDF has high effectiveness in cisgender women with consistent high adherence, and that adherence support should be provided to help women use oral PrEP effectively in order to achieve high levels of protection.
In abstract 979: Mustanski presented about the predictive values of urine PrEP TFV test show strong clinical and research utility. A urine rapid point-of-care (POC) tenofovir (TFV) antibody-based test was compared well to LC-MS/MS of urine among a subset of 73 PrEP users in the RADAR community-cohort study of young men and transgender women who have sex with men were invited to this sub-study. At 3 monthly visits, participants completed a survey of daily PrEP use in the prior 7 days and provided DBS for TFV-diphosphate (TFV-DP) to estimate average dosing over the prior month and emtricitabine-triphosphate (FTC-TP) to assess recent dosing in the prior 2-3 days. A urine POC TFV test qualitatively assessed recent adherence (past 4-7 d) was performed.The two objective metrics of short-term adherence performed similarly well in predicting longer term adherence via TFV-DP (see Table for predictive values). Self-report overestimated longer term adherence, but was better for short term adherence metrics. In multivariable logistic regression analyses, the urine assay was a significant predictor of DBS TFV-DP (OR = 30.2, p < .0001); self-report did not add significantly to prediction. The urine POC TFV assay had excellent predictive values and self-report did not add significantly to prediction of adherence. The POC assay provides results in several minutes to enable same-visit counseling and adherence support interventions, requires no specialized training, and is projected to be low-cost.
In abstract, 973 Ngure presented about the acceptability and feasibility of urine based TFV testing adherence in Kenya. The investigators conducted semi-structured in-depth interviews with 20 women on PrEP who were enrolled in the POC assay arm of the PUMA trial and their seven providers in Thika, Kenya. Interviews were conducted after the 12-month study follow-up period. Most participants reported that the POC test improved their PrEP adherence since they wanted to receive positive results and that positive results led to less worry of acquiring HIV. Women liked that the POC test was clinic-based because subsequent counseling and interpretation of results was aided by clinic providers. The providers reported that having real-time adherence results enabled them to tailor counseling to individual needs, what some referred to as 'evidence-based counseling'. Concerns expressed by participants included perception of lack of trust among providers and embarrassment associated with providing a urine sample. Provider concerns included the POC test not measuring long term adherence and potentially affecting retention of women with adherence challenges. The initial interpretation of results was challenging for providers, although they reported improvements with more familiarity with the test. Additionally, providers reported that the POC test would be more feasible to use if the kits were widely available and marketed for clinical use.
In abstract 980, Zhu summarized oral and injectable PrEP use in the US from 2013-22, in which they utilized an algorithm to identify PrEP users in the US from the Longitudinal Prescription Database (IQVIA). By September 2022, more than half of PrEP users were prescribed generic FTC/TDF. Although few PrEP users (0.5%) were prescribed CAB-LA, a larger proportion of women were prescribed CAB-LA compared with oral PrEP.
In abstract 976, Shahmanesh reported on a pilot study of strategies to increase uptake of biomedical HIV prevention by South African youth through community-based sexual reproductive health services (SRH). In a 2x2 factorial RCT of representative sample of youth in South Africa SRH, the interventions (community STI testing with peer support and sexual reproductive health services) increased demand for differentiated biomedical HIV prevention (defined as a linkage to AYFS within 60 days) by 60% (aOR 1.61; 95% CI 1.32-1.95). Peer support, STI testing and mobile AFYS were acceptable and feasible and STI testing and SRH increased uptake of differentiated HIV prevention and care, although did not reduce the prevalence of HIV.
In abstract 983, Suprasert reported on the lack of improvement in PrEP use among persons who inject drugs (PWID) in San Francisco between 2018 and 2022. The PrEP indicator estimates were comparable to or significantly worse in 2022 than those of 2018, specifically while awareness of PrEP was reported by over half and was stable (54.9% vs. 54.1%; p=0.796), a small minority discussed PrEP with a health care provider (5.9% vs. 12.9%; p<0.001), and an even smaller proportion used PrEP (1.5% vs. 2.9%; p=0.147). Additional efforts are needed to increase PWID and health care providers' focus on PrEP for HIV prevention for PWID.
In abstract 984, Thomas reported the effect of intimate partner violence (IPV) on ART and PrEP adherence in 149 serodifferent couples recruited from 12 public ART clinics in Kampala, Uganda. The investigators found that low adherence was significantly associated with IPV; specifically, low adherence was 2.8-fold more common when PrEP users reported IPV and 6.9 fold more common when ART users reported IPV, compared to those who did not report IPV. Screening questions about IPV could be useful for ART and PrEP providers to be aware of IPV in order to provide more adherence support to their clients who have experienced IPV.
In vitro models with PBMCs and genital tissues can provide useful insights about PrEP agents and dosing strategies, including event-driven PrEP. In abstract 995, Herrera presented on PrEP with TDF-FTC and TAF-FTC for insertive sex using a foreskin in an in vitro model. Oral on demand PrEP with TDF-FTC or TAF-FTC provided ex vivo protection of foreskin and PBMCs. Compared to the control arm, all PrEP dosing arms significantly decreased p24 antigen levels after ex vivo HIV-1BaL challenge of foreskin tissue (all p< 0.0001) and PBMCs (p 0.002 to < 0.0001) with no differences between drugs (TDF-FTC or TAF-FTC) or timings observed. Ex vivo PEP dosing did not increase protection in tissue but further reduced p24 levels in PBMCs. There was no difference in time to, or duration of protection between PBMCs and explants using TAF-FTC and TDF-FTC, although higher levels of TFPdp were obtained with TAF-FTC than TDF-FTC in the foreskin and PBMCs (p≤0.0001). Foreskin CD4+ cell TFVdp levels were 42.7|10.9 fmol/106 cells (TAF|TDF: p=ns). PEP protection was lost with dosing after 48h exposure.
In abstract 996, Hiransuthikul reported on drug-drug interactions in the iFACT 3 study relevant to gender-affirming hormones, with a focus on TAF-based PrEP and hormone therapy in TGW. In 18 TGW, plasma estrogen, testosterone, FTC and TFV exposures trended lower when FTC/TAF was administered with feminizing hormone therapies. However, the AUC and Cmax GMRs of FTC and TFV were within the bioequivalence range, indicating no clinically significant DDI from FHT towards FTC/TAF-based PrEP.
PrEP in pregnancy and breastfeeding
PrEP is an important HIV prevention intervention during pregnancy and postpartum period in high HIV prevalence settings, given the higher risk of HIV acquisition during pregnancy and postpartum periods. In abstract 767 Mogaka presented about the high acceptability of STI testing and expedited partner therapy among pregnant women in western Kenya initiating PrEP. All participants were HIV-negative, ≥18 years, initiating PrEP that day within routine antenatal care, screened for syphilis, and had high HIV risk scores (corresponding to 8.9 HIV infections per 100 person-years). From February to September 2022, women at a subset of facilities were offered chlamydia and gonorrhea (CT/NG) testing using Xpert CT/NG® with same-day results, and those with diagnosed with CT or NG were offered immediate directly observed treatment and expedited partner therapy. Overall, 3% of women had syphilis, and 2% reported having a partner known to be living with HIV; 92% reported unknown partner(s) HIV status. Among women offered CT/NG testing, all accepted testing, and the CT and GC prevalence were 8% and 5%, respectively. Among women with any STI, only 14% reported STI symptoms (e.g., abnormal vaginal discharge and/or vulvar burning/itching). Frequency of having any STI was higher among women < 20 years compared to women ≥20 years (20% vs. 4%, p=0.002). All participants with STIs accepted directly-observed treatment for STIs while 88% of women with STIs accepted EPT for their partners. Encouraging findings were that one month after accepting EPT, all women had dispensed EPT to male partners and all partners accepted, and no social harms were reported.
In abstract 768, Nyema reported the feasibility of integrating PrEP into antenatal care for pregnant women without HIV in South Africa. Overall, 2448 pregnant women without HIV were screened at their first ANC visit, 2039 (61%) were interested in study participation, and 409 (12%) declined participation. Of those interested and screened (n=2039), 839 (41%) were ineligible and 1200 (59%) were eligible and enrolled. Of those enrolled, 1013 (84%) accepted a PrEP prescription at first ANC visit and 187 (16%) did not. Among women who accepted a PrEP prescription at enrollment, 829 (82%) returned at 1 month and confirmed they had initiated PrEP. Of 829 women who initiated PrEP, 76% returned and received a repeat prescription at 1 month, 58% at 3 months, 44% at 6 months, and 35% at 9 months. Of the 187 women who did not accept a PrEP prescription at enrolment, 104 (56%) returned and initiated PrEP sometime during their follow-up visits. Risk factors associated with PrEP discontinuation compared to PrEP persistence at 9 months included perceived low risk of HIV at baseline and earlier gestational age at baseline. The authors concluded that given the high PrEP initiation at first ANC visit, followed by rapid drop off in PrEP continuation, especially in the postpartum period, there is an urgent need for PrEP
integration into ANC and postpartum care as well as interventions to support PrEP continuation.
In abstract 769, Anand presented on PrEP continuation among peri-conception, pregnant and lactating women in Kenya from the Partners Scale-Up Project, a stepped- wedge cluster-randomized pragmatic trial of PrEP delivery in 25 public HIV clinics in Kenya. A total of 2640 women initiated PrEP, of whom 80% were in serodifferent relationships. In this large real-world program, intention to conceive was associated with better continuation on PrEP, and there was no difference in PrEP continuation during pregnancy and breastfeeding periods. PrEP continuation at 1, 3 and 6 months was 59%, 45%, and 36% respectively.
In abstract 771, Pintye presented data on longitudinal PrEP adherence among Kenyan women who initiated PrEP during pregnancy from the PrIMA cluster RCT in western Kenya. Among a random subset of participants, DBS quantifying TFV-DP concentrations were tested from all visits with any self-reported PrEP use in the last 30 days. TFV-DP benchmarks among women in the 2nd trimester of pregnancy and postpartum were defined by thresholds from directly observed studies (in IMPAACT 2009). PrEP adherence was higher during pregnancy than postpartum, with dosing of 2-6 tablets per week in 64% and 7 doses/week in 8%. Adherence was also associated with abuse, side-effects, and perceived risk. Interventions should prioritize sustaining adherence in the postpartum period and increasing knowledge of partner HIV status.
Long-acting and novel PrEP strategies
Numerous presentations focused on new PrEP agents, including those with known efficacy (injectable cabotegravir and dapivirine ring) and those under development (oral islatravir, rectal and vaginal inserts with eltegravir, cabotegravir implants, and injectable lenacapavir).
Dapivirine Rings: safety in pregnancy and breastfeeding
Two landmark studies from the Microbicides Trial Network presented data on the safety of dapivirine ring when used during pregnancy and breastfeeding. In Abstract 127, Bunge and colleagues from the DELIVER study followed 300 pregnant people in two cohorts across 4 countries in east and southern Africa who were enrolled either post 36 weeks of gestation or 30-35 weeks gestation and randomized to use either the DPV ring or daily oral FTC/TDF. Pregnancy outcomes were similar across arms and pregnancy complications were rare and did not differ by arm. In a separate analysis comparing pregnancy compliations against the background rate in each setting, there was also no difference. In Abstract 785, Maxensia Owor and colleagues from MTN presented the first data of dapivirine ring safety and drug detection during breastfeeding among 197 mother-infant pairs who were randomized to DPV ring and FTC/TDF and followed for 12 weeks. Drug concentrations of DPV in breastmilk were highest in the first week of use and fell to 596pg/ml at 12 weeks. In infant plasma, DPV concentrations at 12 weeks were 10.7pg/ml, demonstrating a low level of dapivirine transfer through breast milk to infants. During the trial, there were no adverse events related to use of either product. Together the results from these two trials continue to show safety and acceptability of the DPV ring during pregnancy and breastfeeding.
Injectable PrEP received considerable attention at the conference, including new data on injectable cabotegravir resistance with delayed injections, description of a novel syndrome, "LEVI," to describe delayed HIV seroconversion following CAB use, HIV prevention efficacy among Black men who have sex with men and transgender women, and safety and acceptability among adolescent females.
In Abstract 159, Dr. Marzinke and colleagues presented data on CAB pharmacokinetics among participants in HPTN 084 who received injections late (i.e., outside of the visit window schedule). Of 224 occurrences, including 205 that were 12-18 weeks after the most recent injection, injections given 4-6 weeks after the last injection resulted in 98% of people achieving >4x PA-IC90 and 87% achieving 8x PA-IC90, injections given 6-8 weeks after the last injection resulted in 95% achieving > 4x PA-IC90 and 84% achieving >8x PA-I IC90, and when injections were 8-10 weeks after the last injection, 88% achieved >4x PA-I IC90 and 62% achieved >8x PA- IC90. The team conculeded that CAB may offer 6 weeks of "dosing forgiveness" to persons born female at birth.
Dr. Eshelman described the LEVI syndrome - long-acting early viral inhibition syndrome - a scenario among persons who acquire HIV in the setting of using CAB-LA for HIV PrEP. Using retrospective testing among a small sample of people with delayed diagnoses of HIV, they describe suppressed HIV viremia, very low or undetected HIV DNA, and that INSTI resistance was common among people with delayed diagnosis due to continued use of CAB-LA during the period following infection. Pepole with this syndrome also experienced minimal or no clinical acute retroviral syndrome symptoms.
In Abstract 162, Dr. Sybil Hosek and colleagures from the HPTN 084-01 study presented results of CAB-LA safety among 55 female adolescents ages 12-17 years from 3 African countries. All participants received 5 injections of CAB-LA over a 33 week period before entering a phase with choice of injectable CAB or oral FTC/TDF (open label extension period). Adherence to injections was 100% and only 2 injections were given late. During the period with CAB-LA for everyone, there were no product related serious adverse events, no product discontinuations related to adverse events, no incident HIV infections, and no events of weight gain, hepatotoxicity, hypersensitivity, rash, seizures, or pancreatitis. Similar to data from studies in people >18 years, 26% of participants in 084-01 experienced an injection site reaction at least once. In surveys, acceptability was high and 92% chose to continue CAB-LA during the open label extension period.
In a pre-specified sub-analysis of the landmark HPTN 084 study of injectable CAB efficacy among men who have sex with men and transgender women, in Abstract 161 Dr. Scott presented results from the subset of Black/African American participants from the US sites. Results from 844 participants demonstrate a high level of efficacy of CAB relative to oral FTC/TDF (hazard ratio = 0.28, 95% CI 0.096-0.83) but HIV incidence was higher in both arms due (incidence of 2.11% per year in FTC/TDF arm and 0.58% per year in the CAB arm) to lower levels of adherence measured through drug levels of tenofovir and timeliness of CAB injections. Injection site reactions were slightly less common among Black particpiants relative non-Black participants.
In a poster presentation of Abstract 980, Zhu and colleagues presented data on CAB-LA prescriptions through September 2022. In September 2022 alone, there were 186,367 persons prescribed PrEP and 0.5% (N=1,353) were prescribed CAB.-LA Alongside these data were those from Abstract 981 by Dr. Hazra and colleagues describing the first known real world breakthrough HIV seroconversion case while using CAB-LA. With retrospective testing, the patient was known to have on-time CAB dosing and lab monitoring, yet was HIV viremic 91 days after their first injection (64 days after the second injection) and antibody positive on day 100. The patient's primary partner has known resistance to NRTI and INSTIs yet has had undetectable HIV viremia for 24 months. The patient reported using "illicit" IM testosterone. After initiating an ART regimen of DRV/C + DTG (based on the primary partner's regimen), the patient had undetectable HIV viremia (28 days after antibodies first detected).
Monthly oral islatravir for HIV Prevention
In Abstract 192, Squires presented data about the impact of oral Islatravir on lymphocyte counts, synthesizing data across 7 clinical trials that were testing Islatravir for efficacy in HIV prevention as well as combined with other antiretrovirals for treatment. In total, participants randomized to receive islatravir experienced declines in lymphocyte counts, including declines in CD4 counts among people with HIV who received islatravir in combination with other antiretrovirals and declines were greater in people receiving greater amounts of islatravir. Studies of Islatravir for HIV prevention have been discontinued but studies using smaller amounts of islatravir in combination with other antiretrovirals for HIV treatment (e.g. once weekly dosing) have resumed.
There were a number of studies presenting pre-clinical and phase 1 data on novel HIV prevention product. Tenofovir alafenamide fumarate and elvitegravir (TAF/EVG) is being formulated as a fast dissolving insert for rectal and vaginal use. Phase 1 safety data were presented on the insert in 21 participants and demonstrated overall safety with a regimen of 1 insert, a 7-day washout period, then use of 2 inserts (Abstract 164). PK data from rectal fluid and tissue show high levels of TAF, TFV and EVG for 24 hours after insertion. In addition to the safety data on TAF/EVG inserts, data from a study of macaques support development of TAF/EVG as a vaginal insert formulated for post-exposure protection for women. In a study to optimize the window of use (Abstract 990), macaques were challenged with SHIV once/week for 13 weeks followed by insertion of TAF/EVG or placebo 8 versus 24 hours after vaginal challenge (6 macaques/group). Macaques in the group receiving TAF/EVG 8 hours after challenge experienced the lowest frequency of seroconversion. Authors conclude that TAF/EVG inserts could advance into clinical testing for women as a possible agent for post exposure prophylaxis.
Abstract 989 presented data on a study of islatravir formulated as a biodegradable subdermal implant. Data from female pigtail macques with either a 86 mg implant or two 45 mg implants (inserted as 1 implant per arm) demonstrated frequent implant site reactions that persisted through 8 weeks in the group receiving the 86mg implant and no reactions in the other group. Steady state levels of ISL were much higher in the 86mg group. Macaques in the group with 2 implants were fully protected from vaginal challenge with SHIV. The authors conclude that while the 86mg implant may induce toxicity, the 2x 45mg implant option warrants further study in clinical safety studies.
Cabotegravir is also being studied as an in situ forming implant for HIV prevention. In Abstract 991, authors presented pharmacokinetic data as well as visualization data from X-rays done up to 180 days after injection into female mice. Data show detection of CAB for 180 days in numerous compartments (plasma, vaginal, cervical, and rectal tissues) and that CAB declines after the implant is removed but was not eliminated by 30 days post removal. The implant was able to be visualized for up to 86 days. Next steps for this product include testing in macaques.
Lenacapavir (LEN) is being developed as a long acting (6-monthly) injection for HIV prevention, and in numerous HIV treatment formultions. Abstract 992 presented data on LEN when used for rectal SHIV protection in macaques. Twenty macaques received one of 5 doses of LEN and those (n=11) that achieved LEN exposure above PA-EC95 were challenged with SHIV 7 weeks after LEN dosing. Three animals became infected and animals with greater exposure to LEN were all protected from HIV infection with the difference being significantly significant.
Innovations in HIV testing continue to be highlighted at CROI. This year, there were a sufficient number of studies on measuring HIV incidence via different avidity platforms and algorithms to warrant a poster discussion session (Abstracts 848-849) and additional posters. Surveillance data from Cambodia (Abstract 848) demonstrate use of the RITA algorithm, including viral load testing among those testing positive. Of >50,000 individuals tested for HIV during recent 2-year surveillance, 13% were newly diagnosed for HIV and 3.1% of those were determined to be recent infections (i.e., RITA-recent). Men who have sex with men, transgender people, and entertainment workers were more likely to be diagnosed as RITA-recent whereas people who use drugs were less likely to be RITA-recent. In Malawi, data from HIV recency testing revealed that 4.9% of people newly diagnosed with HIV were recent infections (Abstract 849) with the highest prevelance in breastfeeding women, people tested at STI clinics, and people with a prior test within the past 6 months, and people ages 17-18 years. Data from Rwanda (Abstract 199) demonstrate the utility of using recency testing to elicit contact information for sexual partners of people identify additional new and long term infections. In a study of 1,238 newly diagnosed people, 98 were classified as recently infected and 1738 sexual contacts were identified as part of the recency testing (among those with recent and long term infections). Recently-infected people identified a greater number of sexual partners. Among sexual partners contacted, HIV positivity was 20% among those with a recently infected partner (4% were recently infected) and 15% among those with a partner with long term infection (0.8% were recently infected).
HIV self testing continues to also be an important topic and 4 posters highlight progress operationalizing and delivering HIV self testing (abstracts 928, 929, 931, 932) as well as mathematical modeling of the promise and potential impact of self-testing (abstract 1051). Data from a large study in Cameroon among > 36,000 individuals receiving HIV self test kits through 5 types of locations (Antenatal, postnatal, and maternal and child clinics; partners of people with HIV; Workplace; Community; HIV testing services) highlighted the utility of the tests to reach many people and identify new HIV diagnoses (Abstract 928). The greatest yield for testing numbers was within sites for 1) antenatal, postnatal, and maternal and child clinics; 2) partners of people with HIV; and 3) HIV testing services.
In Abstract 929, authors compared the performance of blood-based HIV self-test kits administered in pharmacies in Kenya to standard of care rapid test kits. Among 1500 people enrolled, including 83 who had HIV, the blood-based HIV self-test performed with 98.8% sensitivity and 93.8% specificity based on client interpretation of test results, which were similar to pharmacist interpretations. Among the few results with discrepant result interpretations, about 1/3 were weak positive results.
In India, a web-based HIV self-test distribution service utilized by >5000 clients demonstrated high utilization of the service with 87% of people contacting the service receiving a kit (45% via courier), 82% reported their test result via virtual counseling, and 5% of those screened positive for HIV (Abstract 931). Use of the platform was found to be acceptable and >80% of the subset completing an additional survey responded that they would recommend the service.
In a study of people who inject drugs that was conducted along the US-Mexico border (Abstract 932), with >500 participants, 81% expressed willingness to use an HIV self-test again in the future and 81% expressed willingness to distribute a self-test kit within peer networks.
In a modeling study of scale-up of a community based program of PrEP delivery accompanied by HIV self testing (Abstract 1051), authors showed the primary impact to be a 50% reduction in new HIV infections averted using 20-year projections. Blood-based HIV self tests would result in <0.01% incorrect HIV-negative results and subsequent inappropriate PrEP initiations (relative to very similar amounts in the rapid diagnostic tests). Overall, this would yield no substantial difference in the amount of NRTI resistance.
Prevention of Mother to Child transmission (PMTCT)-
In an oral presentation of a cluster-randomized study conducted in Mozambique and Tanzania (abstract 129), authors demonstrated overwhelming effectiveness for a point-of-care viral load to identify mothers with viremia and infants at higher risk of HIV acquisition. In 6512 mothers, 22% were estimated retrospectively to have high HIV viral load at delivery. In the trial arms, 19% of women in the intervention arm and 4% in the control arm at high-risk of HIV transmission were identified (p<0.0001). Provision of enhanced prophylaxis to infants was more common in those in the intervention arm (59.8% versus 31.4%, OR 3.75, 95% CI 1.34-10.49).
In an oral presentation (Abstract 131), an intervention combining enhanced viral load counseling and peer support delivered to pregnant and breastfeeding women was demonstrated to increase the frequency of viral suppression at 12 months (70% at baseline, 95% after 12 months, with an intervention effect of 25% improvement, 95% CI 22-28%, p<0.001).
In a poster presentation, Abstract 777 identified a potential gap in the WHO threshold for recommending prophylaxis to HIV exposed and uninfected infants while they are breastfeeding. In a study of 679 mother-infant pairs from Zambia and Burkina Faso, they found that mothers with a viral load of 40-1000 copies/ml were 6 times as likely (95% CI 3.7-9.6) as mothers with viral load <40 copies/ml to have a viral load measurement >1000 copies/ml during breastfeeding. The authors suggest that changing the threshold to 40 copies/ml would reduce the risk of transmission during breastfeeding.
In a poster presentation related to the use of broadly neutralizing antibodies (bNAbs) to prevent maternal to child HIV transmission (Abstract 780), authors explored the potential for viral escape using samples archived from women who transmitted virus perinatally during studies conducted during pre-ART era. They found a high frequency in plasma of broadly bNAbs in the ART-untreated women at a rate similar to postnatal transmitting women which might signal the potential for viral escape and the need for multispecific bNAbs that are synergistic with ART to be used to curb perinatal transmission.
A number of oral presentations summarized important findings of clinical trials of doxycycline PEP for STI prevention. In abstract 119, Molina presented on the ANRS 174 DoxyVac study, which was an open-label randomized trial to prevent STIs in MSM on PrEP using a factorial design of doxycycline as post-exposure prophylaxis (doxyPEP) and the Bexsero quadrivalent meningococcal vaccine among MSM in France. At the time of the interim DSMB review in September 2022, 546 participants were randomized: 92 to receive Bexsero vaccine only: 181 doxyPEP only, 181 both interventions, and 92 were randomized to receive the standard of care. All participants were also taking HIV PrEP. Men given doxyPEP reported that they had used PEP after 83% of the times they had sex as specified during the study period. The average time taken between sex and taking PEP was 27 hours with 6% reporting side effects, all gastrointestinal (e.g., nausea or stomach pain). The annual incidence of STIs was 35.4% in men not receiving doxyPEP, and only 5.6% in men receiving PEP . An important finding was that doxyPEP reduced the rates of chlamydia and syphilis by 88% and 87%, and GC by 51%. With efficacy against urogenital but not pharyngeal infecitons. Importantly, a novel finding was that independent of the effect of doxyPEP, the Bexsero vaccine reduced the incidence of GC by 50%, which is an important finding in a prospective RCT and is supported by past observational studies.
The DoxyPEP study among MSM and TGW from San Francisco and Seattle was stopped early at an interim analysis in May 2022 due to meeting pre-specified efficacy thresholds. In the DoxyPEP study, doxycycline reduced the incidence of any of the STIs in any three-month period by 65% overall. Efficacy against chlamydia and syphilis was 88% and 87%, respectively, in HIV-negative MSM and TGW, and 77% and 74% in PWH. There has been interest in the impact of doxyPEP on antimicrobial resistance (AMR) in GC, given the lower efficacy (55%) of doxyPEP against GC. In abstract 120, Luetkemeyer presented on antimicrobial resistance in N. Gonorrhoeae, commensal Neisseria and S. Aureus in the Doxy PEP study. The 56 samples came from 17 people who were diagnosed with gonorrhoea in the year before study entry and 39 people diagnosed during study follow-up. Of the 39, 20 were taking doxycycline PEP. These relatively small numbers of resistance tests limited the statistical significance of the findings. Three of the baseline gonorrhoea samples had low-level resistance to doxycycline, meaning the MIC was twice what it is for non-resistance samples, and there was one sample with high-level resistance, meaning an MIC over 16 times higher compared with no resistance. Doxycycline concentrations should probably be enough to overcome low-level resistance but the drug would almost certainly fail against bacteria with high-level resistance. With Staph aureus, the percentage with nasopharyngeal colonization decreased significantly from 44% at baseline to 30% in those on doxyPEP, with a significant increase in the proportion of S. aureus samples that were resistant to doxycycline from 5% to 13% in those on doxyPEP.
The long-awaited results of doxy-PEP for STI prevention were presented by Stewart who reported the lack of efficacy of doxyPEP in Kenyan cis-gender women in abstract 121. The trial, conducted from 2020 through 2022, enrolled 449 non-pregnant cisgender women aged 18 to 30 who were taking HIV PrEP, who were randomly assigned to take doxycycline PEP or receive standard care (quarterly STI testing and treatment after diagnosis). Each week women received text messages asking about the frequency of sex and doxycycline use, and they were tested for STIs quarterly. The women had a median age of 24 years, had been on HIV PrEP (dialy oral FTC/TDF) for a median of 7 months, about 60% were using hormonal contraception, and 37% reported transactional sex. At baseline, 18% of women had a bacterial STI: 14% chlamydia, 4% gonorrhea and <1% syphilis. Almost all follow-up visits were completed, and the weekly survey response rate was 81%. Doxycycline was suspended during pregnancy and 80 women became pregnant during the study. STI rates were high during the study with an annual incidence of 27%, comparable to rates among men who have sex with men in high-income countries. The disappointing finding was that doxyPEP did not significantly reduce the risk of STIs: 50 cases of all STIs (GC, CT or syphilis were observed in the doxyPEP group and 59 cases in the standard care group (p = 0.51). For chlamydia, there were 35 new cases in the PEP group and 50 cases in the standard care group (p = 0.16). For gonorrhoea, there were 19 and 12 cases in the two groups, respectively (p = 0.19). The pattern was similar when accounting for time women did not take doxycycline during pregnancy. There was only one case of syphilis. The reasons for lack of efficacy in cisgender women are being explored in terms of plasma, vaginal secretions, cervical, and hair concentrations of doxycycline.
In abstract 118, Haaland presented data on the mucosal pharmacology of doxycycline for bacterial STI prevention in men and women who provided blood samples and mucosal swabs up to seven days after receiving a 200 mg dose of oral doxycycline. Rectal, vaginal and cervical biopsies and urethral swabs were collected 24 hours after dosing. Doxycycline concentrations peaked sooner in vaginal compared with rectal secretions (8 versus 48 hours), but there were no differences in blood levels between men and women. Doxycycline concentrations in rectal, vaginal and cervical tissue stayed well above the minimum levels need to inhibit chlamydia and syphilis for 3-4 days, and for 2 days for gonorrhea. Drug concentrations were higher in rectal compared with vaginal tissue. Future studies will evaluate the mucosal pharmacology of different doxyPEP dosing regimens.
In abstract 122, Traeger modeled the potential impact and efficiency of doxy-PEP among people with or at risk of HIV using the Fenway database. They evaluated different scenarios for prescribing doxyPEP and found that prescribing doxyPEP to MSM and TGW after a bacterial STI diagnosis instead of focusing on specific groups "at higher risk" would reduce STI spread while minimising the number of people needing to take doxyPEP. They estimated that prescribing doxyPEP for 12 months following an STI diagnosis could have averted 42% of all STIs.