A Dual Prevention Pill for HIV & Pregnancy Prevention: A Pilot Study Among Adolescent Girls and Young Women in Zimbabwe

Abstract

Adolescent girls and young women (AGYW) in sub-Saharan Africa are at risk for HIV/unintended pregnancy. Adherence and continuation to effective oral pre-exposure prophylaxis (PrEP) are limited in AGYW. We hypothesized that an over-encapsulated dual prevention pill (DPP) combining PrEP + oral contraceptive (OC) would be preferred, acceptable, and improve adherence compared to PrEP alone. We randomized HIV-uninfected, 16–24 year-olds 1:1 to use DPP or 2-pill regimen (2PR -PrEP and OCs separately) for three 28-day cycles each. We compared preference for DPP vs. 2PR (exact binomial test) and regimen effects on four acceptability domains: use attributes, product attributes, side effects, impact on sex (Wilcoxon signed-rank tests). Adherence was compared by regimen and sequence (mixed-effects logistic regression) using self-report, pill-count, tenofovir-diphosphate levels in dried blood spots (DBS), indicating ≥ 4 doses per week(≥ 500 fmol/punch at Month 1 and ≥ 700 fmol/punch at Months 2–6). 26/30 participants (mean age 19.4years) completed the study (Nov 2022–Sept 2023). 62% preferred DPP, 38% 2PR (p = 0.24). Most rated both regimens as acceptable, with no differences between them (all p > 0.05). Adherence was high by self-report (98%) and pill-count (97%), but low based on DBS results (DPP: mean 392 fmol/punch, 2PR: mean 384 fmol/punch); with only 10% consistently adherent. There was no difference in adherence by regimen, though odds were higher in period 1 vs. 2 (AOR 3.7; 95% CI 1.10–12.8). Both regimens were safe; 2 pregnancies occurred during the DPP regimen, with no HIV seroconversions. No significant differences in preference, acceptability, and adherence between DPP and 2PR were found in this study. Adherence was low and waned over time. A larger study with a smaller co-formulated DPP may better inform its impact on HIV and pregnancy prevention.

The study has been registered at ClinicalTrials.gov Identifier NCT04778514.

Introduction

Young African women bear a disproportionate and double burden of HIV and unmet need for contraception. In 2023, women and girls (15–24 years old) in eastern and southern Africa accounted for 27% of all new HIV infections and were three times as likely to acquire HIV than their male peers [1]. In sub-Saharan Africa (SSA) one adolescent girl or young woman (AGYW) aged 15–24 years old becomes infected with HIV every three minutes [2]. In Zimbabwe, which has an HIV prevalence of 10.5% in 15–49 year old adults, more young women (4.0%) are living with HIV than young men (2.8%) [3].

Oral pre-exposure prophylaxis (PrEP) with Truvada^® and its generic equivalents (tenofovir disoproxil fumarate [TDF]/emtricitabine [FTC]) is safe and highly effective in reducing HIV acquisition when taken as prescribed [4, 5] and is approved for use in more than 140 countries globally, including Zimbabwe [6]. Since its approval by the US Food and Drug Administrative Agency in 2012, global scale up has been slow though several countries have surpassed 100,000 PrEP initiations, accounting for 91% initiations in SSA [7]. However, low adherence and continuation rates among young African women means that those at highest HIV risk are still not using PrEP effectively [8, 9]. These young women face insurmountable product-related, individual, interpersonal, and structural challenges that limit access, uptake and persistent use of PrEP. Importantly, stigma is often cited as a reason for non-use of PrEP [10, 11]; women fear being regarded as HIV-infected or promiscuous if they take the same drugs used to treat HIV. AGYW, in particular, have voiced concerns about the consequences PrEP use will have on their sexual relationships, as PrEP use signals mistrust and infidelity, which can potentially result in relationship dissolution or violence [12]. Therefore, strategies are needed to increase PrEP use by AGYW at risk of HIV.

Many sexually active women at risk of HIV would be more likely to use protection if it also prevents pregnancy, highlighting the need for a contraceptive multipurpose prevention technology (MPT) [13–17]. Contraceptive MPTs may increase uptake and use of HIV prevention technologies, since preventing unintended pregnancy, may be more acceptable and often women’s primary concern compared to HIV prevention [18, 19]. As of August 2024, there were 21 MPT candidate products in the pipeline, most in the early stages of development [20].

A co-formulated dual prevention pill (DPP) comprised of oral PrEP and oral contraception is likely to be the fastest route to an approved MPT as it combines two marketed, approved drugs and would only require a bioequivalence study versus a large Phase 3 efficacy trial for regulatory approval [21]. The DPP in development contains the active pharmaceutical ingredients in a 28-day combined oral contraceptive (COC) regimen (150 mcg levonorgestrel [LNG], 30 mcg ethinyl estradiol [EE]) and a generic equivalent of Truvada^® (300 mg TDF, 200 mg of FTC). We hypothesized that a single DPP combining PrEP and a COC would be preferred, acceptable and increase PrEP adherence compared to PrEP alone. To test our hypothesis, we conducted a small pilot crossover study among AGYW in Zimbabwe where 28.2% of women aged 15–49 using modern contraception rely on COCs [22] with the prevalence is even higher among married women (41%) [23]. Because the actual co-formulated DPP is still in development, we over-encapsulated Truvada^® and a generic COC for acceptability testing (Fig. 1).

Fig. 1.

Over-encapsulated dual prevention pill (DPP)

Methods

Study Design

In a single-site, two-arm, randomized, open-label crossover trial among AGYW at risk of HIV and unintended pregnancy in Chitungwiza, Zimbabwe, we compared a single over-encapsulated DPP to a two-pill regimen (2PR) – Truvada^® and COC tablets taken separately – each used daily for three consecutive 28-day cycles by current COC users. Primary objectives were to assess and compare preference and acceptability of the DPP versus the 2PR; the secondary objective was to determine if adherence to the DPP was different from the 2PR. Detailed methods have been previously described [24].

Study Population

AGYW were recruited in Chitungwiza through community sensitization, snowballing, self-referral, and from sexual reproductive health clinics. Eligible participants were healthy, sexually active, HIV-uninfected, non-pregnant 16 to 24-year-old cisgender, heterosexual females who had been using COCs for at least 3 months prior to screening, and who were at risk of acquiring HIV based on self-assessment and clinical judgement. Exclusion criteria were primarily based on contraindications for PrEP and COCs, as described in detail previously [24]. Eligibility was evaluated and confirmed by experienced study clinicians.

Study Products and Procedures

At Enrollment/Visit 1, participants were randomly assigned to the sequence of study regimens: Sequence 1 = single DPP once daily for three 28-day cycles followed by the 2PR, two separate oral PrEP and COC tablets once daily for three 28-day cycles; Sequence 2 = 2PR followed by the DPP (Fig. 2). All participants were directly observed in the clinic when taking the first dose of each regimen. Truvada^® was supplied by Gilead (Foster City, CA, USA) and the COCs were purchased from Mylan (Hyderabad, India). The DPP was manufactured and packaged by PCI Pharma Services (Rockford, IL, USA).

Fig. 2.

Study schema

Monthly follow up visits included risk reduction counseling for HIV and other sexually transmitted infections (STIs); product dispensations/returns; blood collection for rapid HIV testing, dried blood spots (DBS) to determine tenofovir-diphosphate (TFV-DP) levels as a measure of PrEP adherence [25], and creatinine testing (at enrollment, 3 and 6 months), urine pregnancy testing, and behavioural questionnaires via computer assisted self-interviewing (CASI). Although there is no evidence of drug-drug interactions between PrEP and COCs [cites], adverse events (AEs) monitored to ensure participants’ safety in accordance with good clinical practice. AEs were recorded based on reported and/or observed signs and symptoms or laboratory findings and were graded according to the US Division of AIDS (DAIDS) Table for Grading the Severity of Adult and Pediatric Adverse Events Corrected Version 2.1 July 2017 [26].

Statistical Analysis

We summarized baseline demographic and other characteristics descriptively. We assessed preference using Fisher’s exact test to compare the proportion of women preferring the DPP to the 2PR after using each regimen for three 28-day cycles, as reported via CASI questionnaires. We evaluated acceptability by measuring and comparing the effect of each regimen on four acceptability domains (ease of use, product attributes, side effects, effect on sex) with three to six Likert response categories (such as 1, dislike very much to 5, like very much); differences between regimens were compared using Wilcoxon signed-rank tests.

Due to a survey administration error, nine participants were issued the questionnaire for the wrong regimen at the end of the first crossover period (DPP: n = 4; 2PR: n = 5). Invalid responses from these participants were not included in any analyses. In total, 26 participants received the correct questionnaire for the DPP, 25 for the 2PR, and 21 for both regimens. Participants who did not complete both treatment regimens, or who refused or otherwise did not respond to questions about one or both regimens were excluded from paired analyses. Missing data were not imputed.

We used three metrics to assess monthly medication adherence: self-report, pill count, and PrEP (TFV-DP) drug levels in DBS. On monthly CASI questionnaires, participants were asked to report how many doses they missed since their last visit; how good a job they did at taking their medication the way they were supposed to (ranging from 1-very poor to 6-excellent); and how often they took the medication the way they were supposed to (ranging from 1-never to 6-always). We compared the proportion of participants reporting any missed dose during either regimen using McNemar’s test. We measured adherence by pill count from pharmacy returns, calculating the proportion of doses taken versus expected based on the number of days between clinic visits. Pills not returned and not reported as lost were assumed to have been taken by the participant.

We defined “adherent” per DBS as TFV-DP levels indicative of ≥ 4 doses per week, corresponding to ≥ 500 fmol/punch at Month 1 and ≥ 700 fmol/punch at Months 2–6 (lower limit of quantification, 16.6 fmol/punch). We compared adherence per CASI responses by treatment regimen, treatment period and randomization sequence using mixed effects models with participant as a random effect and treated scaled responses (1-6) as continuous variables. We compared the percent of doses taken in each treatment period (adherence per returned pill count) by calculating the paired difference in proportion of days where a dose was taken and used the signed-rank test to determine if there was any difference in pill taking between the two regimens (H₀: p_DPP– p_2PR = 0). We compared adherence to the DPP versus the 2PR per DBS using a mixed-effects logistic regression model adjusting for randomization sequence (DPP or 2PR first) and treatment period (Period 1, Months 1–3 or Period 2, Months 4–6) as fixed effects and participant as a random effect.

All analyses were performed using SAS version 9.4 (Cary, NC, USA) or Stata version 14.2 (College Station, TX, USA).

Each participant provided written informed consent in Shona before undergoing any study procedures. For participants aged 16–17 years, the study team obtained informed consent directly from emancipated minors. For non-emancipated minors in the same age group, informed consent was obtained from a parent or legal guardian, along with assent from the minor. In accordance with the laws of Zimbabwe which criminalize sexual activity with individuals under the age of 18, the Community Advisory Board advised study team against seeking a waiver of parental or guardian consent for non-emancipated minors. This recommendation was made to ensure the protection and safeguarding of minors participating in the study.

Results

Participant Background Characteristics

As shown in Fig. 3, we screened 40 AGYW, 34 (85%) were eligible and 30 enrolled (November 2022-September 2023). Six women were excluded due to HIV (n = 1), pregnancy (n = 1), breastfeeding an infant under 6 months (n = 1), inability to meet study requirements (n = 1), and other health concerns (n = 2). Of the 30 participants, 26 completed the study; four withdrew early (two due to relocation/job conflicts, two due to pregnancy). One participant received the wrong product at enrollment, resulting in 16 participants in Sequence 1 and 14 participants in Sequence 2 (versus 15 in each).

Fig. 3.

Consort diagram

Baseline characteristics are summarized in Table 1. Mean age was 19.4 years (range 16–24) and did not differ significantly by randomization group. All participants but one had completed secondary education. Nearly all participants in both groups had given birth to one or more children. Marital status was the only characteristic that differed by group (p = 0.047); 31% of participants in Sequence 1 had never been married whereas 50% of participants in Sequence 2 were divorced and only one had never been married.

Table 1.

Baseline characteristics of study participants by randomization sequence (n = 30)

	SEQUENCE 1 DPP* then 2PR** (n = 16)		SEQUENCE 2 2PR then DPP (n = 14)		Total		Fisher's Exact Test
	n	%	n	%	n	%	P-value
Age (mean, SD)	19.4	2.4	19.5	2.5	19.4	2.4	0.882ª
Age category
16–17 years	5	31.2	3	21.4	8	26.7	0.896
18–19 years	4	25.0	5	35.7	9	30.0
20 + years	7	43.8	6	42.9	13	43.3
Marital status
Never married	5	31.2	1	7.1	6	20.0	0.047
Married	9	56.2	5	35.7	14	46.7
Divorced	2	12.5	7	50.0	9	30.0
Other	0	0.0	1	7.1	1	3.3
Educational completion
Primary	0	0.0	1	7.1	1	3.3	0.467
Secondary	16	100.0	13	92.9	29	96.7
In the past 4 weeks, how often did you go a whole day and night without eating?
Never	6	37.5	6	42.9	12	40.0	0.339
Rarely	5	31.2	2	14.3	7	23.3
Sometimes	1	6.2	4	28.6	5	16.7
Often	4	25.0	2	14.3	6	20.0
How often do you have a place to stay or store your things regularly?
Never	5	31.2	1	7.1	6	20.0	0.246
Rarely	4	25.0	2	14.3	6	20.0
Sometimes	1	6.2	1	7.1	2	6.7
Often	6	37.5	10	71.4	16	53.3
Do you currently earn and income of your own?
No	7	43.8	8	57.1	15	50.0	0.715
Yes (part-time)	9	56.2	6	42.9	15	50.0
Yes (full-time)	0	0.0	0	0.0	0	0.0
Number of sex partners, last 3 months†
0	0	0.0	1	7.1	1	3.4	0.355
1	9	60.0	8	57.1	17	58.6
2	4	26.7	1	7.1	5	17.2
3 +	2	13.3	4	28.6	6	20.7
Lifetime parity
0	1	6.2	0	0.0	1	3.3	0.485
1	13	81.2	14	100.0	27	90.0
2 +	2	12.5	0	0.0	2	6.7
Expected preference at baseline
Contraceptive pills and PrEP	3	18.8	2	14.3	5	16.7	0.729
DPP	9	56.2	7	50.0	16	53.3
Either equally	3	18.8	5	35.7	8	26.7
Neither	1	6.2	0	0.0	1	3.3
Any sexually transmitted infection***	6	37.5	4	28.6	10	33.3	0.709
Frequency of vaginal sex since last visit†
No vaginal sex since my last visit	3	20.0	4	28.6	7	24.1	0.506
Once or twice since my last visit	6	40.0	4	28.6	10	34.5
3–10 times since my last visit	2	13.3	5	35.7	7	24.1
Daily or almost daily	3	20.0	1	7.1	4	13.8
Refused	1	6.7	0	0.0	1	3.4
Used condom during last act of vaginal sex (of people who had vaginal sex)
Yes	3	27.3	4	40.0	7	33.3	0.659
No	8	66.7	6	60.0	14	66.7
Frequency of anal sex since last visit†
No anal sex since my last visit	14	93.3	13	92.9	27	93.1	1.000
Once or twice since my last visit	1	6.7	0	0.0	1	3.4
3–10 times since my last visit	0	0.0	1	7.1	1	3.4
Condom used during last act of anal sex (of those who had anal sex)
Yes	0	0.0	0	0.0	0	0.0	1.000
No	0	0.0	1	100.0	1	50.0
Refused	1	100.0	0	0.0	1	50.0
In the last 3 months, did you have sex with anyone to receive financial/material support?†
Yes	4	26.7	4	28.6	8	27.6	0.833
No	11	73.3	9	64.3	20	69.0
Refused	0	0.0	1	7.1	1	3.4
Ever used other contraceptive besides contraceptive pills‡	3	18.8	4	28.6	7	23.3	0.675
Male condoms	1	33.3	1	25.0	2	28.6
Female condoms	0	0.0	1	25.0	1	14.3
Injectable contraceptives	0	0.0	0	0.0	0	0.0
Implants	0	0.0	2	50.0	2	28.6
Morning after pill (emergency contraception)	0	0.0	1	25.0	1	14.3
Withdrawal	1	33.3	0	0.0	1	14.3
Has ever used anything to avoid getting HIV‡	5	31.2	4	28.6	9	30.0	1.000
Male condoms	5	100.0	2	50.0	7	77.8
Female condoms	0	0.0	1	25.0	1	11.1
Abstinence	1	20.0	0	0.0	1	11.1
PrEP	0	0.0	0	0.0	0	0.0
Traditional herbs/medicines	0	0.0	1	25.0	1	11.1
Other	0	0.0	1	25.0	1	11.1
Thinking about your life right now, how important is it to you to avoid becoming pregnant
Very important	14	87.5	13	92.9	27	90.0	1.000
Somewhat important	1	6.2	0	0.0	1	3.3
A little important	1	6.2	0	0.0	1	3.3
Don't know	0	0.0	1	7.1	1	3.3
Which is more important for you?†
Prevention of pregnancy	0	0.0	0	0.0	0	0.0
Prevention of HIV	1	6.7	1	7.1	2	6.9	1.000
They are both equally important	14	93.3	13	92.9	27	93.1
How worried are you that you may get HIV in the next 3 months?
Very worried	7	43.8	10	71.4	17	56.7	0.224
Somewhat worried	6	37.5	2	14.3	8	26.7
Not worried at all	3	18.8	1	7.1	4	13.3
Refused	0	0.0	1	7.1	1	3.3

*Dual prevention pill

**2-pill regimen

***STIs assessed include: Trichomoniasis, Chlamydia, Gonorrhea, Hepatitis B, Hepatitis C, Syphilis; some participants were found to have multiple concurrent STIs

^†Total sample size less than 30 due to missing responses

^‡Among participants who said they had ever used a contraceptive method besides contraceptive pills (n = 7), or had ever used anything to avoid getting HIV (n = 9) Multiple responses accepted; responses may not total to 100%

^aSignificance assessed by unpaired T-test used

Food insecurity and housing instability were common with 37% of participants reporting sometimes/often going an entire day and night without eating and 40% never/rarely having a place to stay or store their things in the past four weeks. Half of the participants said they earned their own income, all of whom reported working part-time. 90% of participants stated that preventing pregnancy was “very important” and 83% said they were somewhat or very worried about getting HIV in the next three months. Many participants had multiple risk factors for HIV acquisition. At baseline, 33% of participants had one or more STIs, including gonorrhea (3%), syphilis (10%), trichomoniasis (7%) and chlamydia (30%). Over one third (38%) reported having two or more sex partners in the three months before screening and 28% reported transactional sex during the same period. Of the 21 participants who had vaginal sex since screening, 33% reported using a condom the last time they had sex; 30% of all participants had ever used any HIV prevention method.

Preference

At baseline, 53% of participants expected to prefer the DPP, 17% expected to prefer the 2PR, 27% expected to like both regimens the same, and one participant expected to like neither (3%). Of the 26 participants who completed the study, 62% preferred the DPP and 38% preferred the 2PR (p = 0.24).

Acceptability

Nearly all participants reported that both regimens were somewhat/very easy to use every day (83% DPP, 91% 2PR) with no difference by regimen (Table 2). Mean scores on questions regarding most product-specific attributes (taste, smell, color, shape) ranged from 3.3 to 3.9 (3 = Neither like nor dislike, 4 = Like) with no significant differences between regimens. More participants said they had no trouble swallowing PrEP due to its size versus the DPP (38% versus 13%, p = 0.08), and none of the participants reported either PrEP or the DPP was so big that they could not swallow it at all. All participants said they felt very protected against HIV when using both regimens. Approximately one-quarter said they had experienced any side effects while using either regimen (26%, DPP; 29%, 2PR).

Table 2.

Acceptability of the dual prevention pill (DPP) and 2-pill regimen (2PR) after 3 months of use, by regimen (n = 26)

^aSignificance tested only among participants who answered this question about both regimens. Excludes participants who completed the incorrect form at the end of period 1, early terminations and refusals. Please see consort diagram for additional detail

Most participants reported that both regimens improved or had no impact on their sex lives with no significant differences by regimen (Table 2). At the end of the study, 93% of participants said they would be willing to use the DPP if it were available, 96% said they would recommend it to friends, and 100% said they would be more likely to use it if it also prevented STIs other than HIV. Most (72%) said they would be willing to pay a small fee for the DPP, although 12% said they would only use it if it were free and 16% said they would not use the DPP even if it was free. Participants had varying opinions as to where to get the DPP, including public HIV testing and counseling clinics (38%), public family planning clinics (25%) and pharmacies (16%). Participants also had different opinions about their preferred service providers, with nurses (22%) and counselors (22%) being most popular, followed by public facility doctors (17%) and community health workers (17%). Private facility doctors (13%) and pharmacists (9%) were less frequently selected as the preferred healthcare providers from whom to receive the DPP.

Adherence

For both regimens, participants rated themselves 4.5 (p = 0.64), on average (4 = good to 5 = very good) at taking their medication the way they were supposed to (Table 3). Similarly, on average, participants said they usually/almost always took their medication the way they were supposed with similar mean scores between regimens (DPP = 5.0, two pills = 4.7, p = 0.11; 4 = usually, 5 = almost always). A minority of participants reported that they ever missed any medication with no difference by regimen; 21% reported having ever missed taking the DPP and 29% had ever missed taking PrEP (McNemar’s test: p = 0.71). Reports of missed pills correspond to overall self-reported drug taking on more than 90% of study days (DPP = 95% and two pills = 98%; p = 0.12). Returned pill counts, which were consistent with self-reports, indicated that participants took over 95% of their expected doses (DPP = 95%; 2PR = 98%; p = 0.12).

Table 3.

Adherence to the dual prevention pill (DPP) and PrEP per self-report, pharmacy returns and dried blood spot (DBS) results, by visit (n = 30)

	DPP	PrEP	p-value
	Mean, SD (min–max)	Mean, SD (min–max)
Self-report: How good a job did you do at taking your medication in the way you were supposed to
Very poor (1) to Excellent (6)
Month 1	4.4, 1.13 (1–6)	4.3, 1.39 (1–6)
Month 2	4.6, 1.26 (2–6)	4.3, 1.35 (1–6)
Month 3	4.7, 0.96 (3–6)	5.0, 1.26 (1–6)
Overall	4.5, 1.13 (1–6)	4.5, 1.36 (1–6)
Self-report: How often did you take the medication in the way you were supposed to
Never (1) to Always (6)
Month 1	4.9, 1.58 (1–6)	4.8, 1.41 (1–6)	0.11*
Month 2	5.3, 0.96 (4–6)	4.7, 1.35 (1–6)
Month 3	4.9, 1.28 (1–6)	4.5, 1.83 (1–6)
Overall	5.0, 1.30 (1–6)	4.7, 1.51 (1–6)
Self-report: # of days pills were missed
Month 1	0.7, 2.88 (0–14)	0.5, 1.44 (0–6)
Month 2	0.3, 1.19 (0–6)	1.1, 4.99 (0–25)	0.88*
Month 3	0.05, 0.22 (0–1)	0.3, 1.11 (0–5)
Overall	0.4, 1.86 (0–14)	0.7, 3.14 (0–25)
Self-report: Proportion of days where pills were taken
Month 1	0.98, 0.10 (0.53–1)	0.98, 0.062 (0.73–1)	0.97*
Month 2	0.99, 0.044 (0.79–1)	1.00, 0.02 (0.93–1)
Month 3	1.00, 0.012 (0.96–1)	0.99, 0.042 (0.81–1)
OVERALL	0.98, 0.05 (0.83–1.0)	0.97, 0.09 (0.54–1)
Returned pill count: Proportion of days where pills were taken
3-month regimen period	0.95, 0.10 (0.49–1)	0.98, 0.03 (0.88–1)	0.12**
DBS: TFV-DP corresponds to ≥ 4 doses per week	n (%)	n (%)
Month 1	9/27 (33%)	7/26 (27%)
Month 2	7/26 (27%)	7/28 (25%)	0.67***
Month 3	4/26 (15%)	6/27 (22%)
OVERALL	5/28 (18%)	5/28 (18%)
TFV-DP in DBS (fmol/punch)
Month 1	408, 473.4 (0–1741)	400, 378.1 (0–1303)
Month 2	396, 514.2 (0–2146)	432, 603.3 (0–2613)	0.66****
Month 3	370, 558.0 (0–2009)	323, 490.3 (0–2185)
OVERALL	392, 509.4 (0–2146)	384, 497.8 (0–2613)

*For treatment regimen effect from mixed effects regression, adjusting for treatment period and randomization sequence

**Based on Wilcoxon Signed Rank Test testing paired difference in response (DPP-2 pills)

***From mixed effects logistic regression, adjusting for treatment period and randomization sequence

****From mixed effects regression of log-transformed TFV-DP value, adjusting for treatment period and randomization sequence

Based on DBS results, the same proportion (18%) of participants had adherent TFV-DP levels at all three visits during when using the DPP and when using the 2PR, however, only 10% of participants were consistently adherent to both regimens across all six visits. On average, TFV-DP concentrations were 392 fmol/punch (Standard Deviation [SD] = 509.4) for the DPP versus 384 fmol/punch (SD = 497.8) for the 2PR (p = 0.66). There was considerable variability between participant concentrations ranging from 0 to 2613 fmol/punch (Figs. 4a and b). Results from the logistic mixed-effects model showed no difference in adherence by treatment regimen overall (p = 0.67); however, participants were nearly four times more likely to be adherent during the first treatment period compared to the second treatment period (Odds Ratio [OR]: 3.7, 95% Confidence Interval [CI]: 1.097–12.810, p = 0.035). Participants randomized to Sequence 1 (DPP followed by 2PR) had nine times the odds of adherence compared to participants who were randomized Sequence 2 (2PR followed by DPP), although the difference was not significant (OR: 9.3 95% CI: 1.00-87.5, p = 0.050).

Fig. 4.

Individual levels of tenofovir diphosphate by sequence and visit. Color: Publish on web in color, Print in color

Safety

We collected data from a total of 165 28-day cycles: 82 DPP, 83 2PR (Table 4). There were no deaths, serious adverse events, study discontinuations due to AEs, HIV seroconversions, or incident STIs. There were two pregnancies, both detected during the DPP regimen. Twenty-three participants reported AEs when using the DPP and 13 during the 2PR, most of which were mild or moderate, and all resolving spontaneously or with minimal intervention. The most common AEs were gastrointestinal issues (dyspepsia, vomiting, nausea, gastritis); dizziness; infections; reduced creatinine clearance; and reproductive issues (menometrorrhagia, vaginal discharge).

Table 4.

Adverse events by severity, relatedness, and system organ class (n = 30)

	Participants on DPP* (N = 29)		Participants on 2PR** (N = 28)
	n	%	n	%
Total number of subjects with treatment emergent adverse events	23	79.0	13	46.0
Severity
Grade 1 (Mild)	3	10.0	1	4.0
Grade 2 (Moderate)	19	66.0	11	39.0
Grade 3 (Severe)	1	3.0	1	4.0
Relationship to DPP
Not related	11	38.0	–	–
Unlikely	0	0.0	–	–
Possibly related	3	10.0	–	–
Probable	8	28.0	–	–
Highly probable	8	28.0	–	–
Insufficient data to assess	0	0.0	–	–
Relationship to COC***—Zinnia-F
Not related	–	–	5	18.0
Unlikely	–	–	5	18.0
Possibly related	–	–	2	7.0
Probable	–	–	0	0.0
Highly probable	–	–	1	4.0
Insufficient data to assess	–	–	0	0.0
Relationship to PrEP—Truvada
Not related	–	–	6	21.0
Unlikely	–	–	4	14.0
Possibly related	–	–	1	4.0
Probable	–	–	0	0.0
Highly probable	–	–	2	7.0
Insufficient data to assess	–	–	0	0.0
Primary System Organ Class Preferred Term
Cardiac disorders	–	–	1	5.0
Palpitations	–	–	1	4.0
Ear and labyrinth disorders	1	3.0	1	5.0
Vertigo	1	3.0	1	4.0
Gastrointestinal disorders	7	21.0	2	9.0
Dyspepsia	1	3.0	1	4.0
Gastritis	1	3.0	–	–
Nausea	1	3.0	–	–
Vomiting	4	11.0	1	4.0
General disorders and administration site conditions	1	3.0	1	5.0
Malaise	1	3.0	1	4.0
Infections and infestations	8	24.0	5	23.0
Gastroenteritis	1	3.0	1	4.0
Nasopharyngitis	1	3.0	–	–
Oral candidiasis	1	3.0	–	–
Periodontitis	1	3.0	–	–
Respiratory tract infection	1	3.0	1	4.0
Tinea versicolour	–	–	1	4.0
Tonsillitis	1	3.0	2	8.0
Upper respiratory tract infection	3	8.0	1	4.0
Urinary tract infection	1	3.0	–	–
Investigations	8	24.0	6	27.0
Blood creatinine increased	1	3.0	2	8.0
Creatinine renal clearance decreased	8	21.0	5	21.0
Haemoglobin decreased	1	3.0	–	–
Musculoskeletal and connective tissue disorders	3	9.0	1	5.0
Arthralgia	1	3.0	–	–
Back pain	2	5.0	1	4.0
Nervous system disorders	4	12.0	4	18.0
Dizziness	2	5.0	1	4.0
Migraine	1	3.0	–	–
Neuropathy peripheral	–	–	1	4.0
Tension headache	1	3.0	2	8.0
Renal and urinary disorders	–	–	1	5.0
Proteinuria	–	–	1	4.0
Reproductive system and breast disorders	2	6.0	–	–
Menometrorrhagia	1	3.0	–	–
Vaginal discharge	1	3.0	–	–

*Dual prevention pill

**2-pill regimen

***Combined oral contraceptive

Discussion

In this first small pilot study evaluating an over-encapsulated DPP in Zimbabwe, there were no significant differences in preference, acceptability, adherence or safety between the DPP and the 2PR. Although 62% of participants preferred the DPP, a considerable proportion (38%) preferred two separate pills, reinforcing the importance of choice and not relying on a single solution. In our study, most women did not report experiencing side effects, there were no serious adverse events, and most AEs were mild/moderate and were within the expected AEs in the population among users of TDF/FTC and COCs [27]. As with separate oral PrEP and COCs, the most frequently reported AEs were mild to moderate gastrointestinal events [28], which were self-limiting or resolved with minimal intervention, a reassuring fact that should be incorporated into future counseling messages.

Adherence was low, overall, and waned over time as has been seen in previous studies in this population [29, 30]. Adherence was higher in the first treatment period, which is not unexpected. It is unclear, however, why adherence was higher among those randomly assigned to using the DPP first followed by the 2PR. This may reflect a novelty effect associated with receiving the new product first, as noted by Bhushan et al., [31] where prior experience with family planning methods influenced preferences for MPT delivery forms; for instance, women with a history of oral contraceptive use showed greater preference for tablets. However, in our study, this effect was not statistically significant and was largely driven by three consistently adherent participants in Sequence 1.

As previously noted in PrEP studies in African cisgender women, subjective adherence measures are inconsistent with objective biomarker findings [32, 33]. Self-report is subject to social desirability and recall bias and may be compounded by misinterpretation and/or misunderstanding of questions [34, 35]. Notwithstanding the limitations of subjective measures, the objective biomarkers that we are relying on may also be affected by environmental factors. Whereas FTC can be taken with or without food, the oral bioavailability of TDF is optimized when taken with food, particularly a high-fat meal [36]. A considerable proportion of participants indicated that they frequently faced food insecurity which could have resulted in fluctuations in the absorption of TDF, potentially leading to TDF blood levels far below the pre-defined threshold indicative of adherence, as has been seen in HIV treatment [37]. Given the high number of participants that had levels below the thresholds associated in prior studies with adherence (at least 4 pills/week) and HIV prevention, this phenomenon requires future investigation to determine its impact on HIV risk and medication adherence.

AGYW in this study and other parts of SSA often face housing instability, which negatively affects HIV PrEP uptake, adherence, and persistence. Crowded living conditions can limit access to private space for safe drug storage, making it difficult for some women to disclose PrEP use to partners or roommates [38]. Housing instability and disrupted routines make it harder for young women to start, adhere to, and continue oral PrEP. In addition, approximately one third of participants reported no sexual activity between the screening and enrollment visits, and research shows that changes in AGYW’s risk perceptions and behaviors (such as sexual activity, transactional sex, number of partners, partner’s HIV status) affect PrEP adherence [33, 35]. Participants might have been more motivated by the additional benefits provided by the study, such as reimbursement and health check-ups, rather than the products themselves, which could have influenced their adherence [39]. Comprehensive HIV prevention for AGYW requires a holistic approach that addresses social determinants of health, including economic instability such as food insecurity, malnutrition, and poor housing, as well as behavioral risk factors.

Our study had several limitations. First and foremost, the DPP study product in the trial was an over-encapsulated pill that is considerably larger (approximately 6 mm) with bulkier packaging than the co-formulated tablet currently under development [40]. Second, this small pilot study was not powered to detect small differences in acceptability and was further compromised as nine participants did not receive the correct questionnaire at the end of one product use period. Lastly, adherence per the DBS biomarker was low, which limits our ability to interpret acceptability and preference data. We were limited in our ability to compare contraceptive adherence between the two arms as there is no reliable objective marker for oral contraceptive adherence. Assessing the effect of the DPP on contraceptive effectiveness of a combined DPP will require a larger study where pregnancy incidence can be compared. These limitations provided fewer opportunities to observe events of interest: preference, acceptability and adherence. Our study also had several strengths. This well-controlled and stringently monitored study contributes initial evidence about acceptance, preference, and adherence to a DPP in Zimbabwe, and globally, where no other data exist. We successfully engaged AGYW, a population with a high HIV burden, who valued dual protection and were motivated to join the study due to their perceived risk of HIV and unintended pregnancy. Our findings highlight women’s interest in MPTs for simultaneous HIV and pregnancy prevention.

Conclusion

This pilot study was the first to assess acceptability of an over-encapsulated dual prevention product as a proxy for the DPP in development. We found high willingness and desire for a multipurpose product, but low adherence to the daily oral regimen. Future research with larger samples and the smaller co-formulated DPP may provide better insights into its potential impact on HIV and pregnancy prevention. Future research and market introduction plans for DPP should also explore social and structural support to encourage effective use, and easy access to the product.

Abbreviations

2PR

2-Pill Regimen

AE

Adverse Event

AGYW

Adolescent G irls and Young Women

AIDS 2024

The 25th International AIDS Conference

AOR

Adjusted Odds Ratio

CASI

Computer Assisted Self-Interviewing

CI

Confidence Interval

CIFF

The Children’s Investment Fund Foundation

COC

Combined Oral Contraceptive

CROI

Conference on Retroviruses and Opportunistic Infections

DBS

Dried Blood Spot

DPP

Dual Prevention Pill

EE

Ethinyl Estradiol

FTC

Emtricitabine

HIVR4P

HIV Research for Prevention Conference

INTEREST

International Conference on HIV Treatment, Pathogenesis, and Prevention Research

LNG

Levonorgestrel

MPT

Multipurpose Prevention Technology

OC

Oral Contraceptive

OR

Odds Ratio

PrEP

Pre-Exposure Prophylaxis

SD

Standard Deviation

SSA

Sub-Saharan Africa

STI

Sexually Transmitted Infection

TDF

Tenofovir Disoproxil Fumarate

TFV-DP

Tenofovir-Diphosphate

Author Contributions

NMM—Conceptualization; Supervision; Investigation; writing original draft, Editing. JBZ—Statistical analysis, Writing—Review & Editing. CM—Writing—Review & Editing. AD—Writing—Review & Editing. JJ—Writing—Review & Editing. TC—Writing—Review & Editing. PM—Writing—Review & Editing IVB—Project Administration, Writing review and editing. BAF—Conceptualization; Supervision; Writing—Review & Editing. LH—Oversaw methodology and implementation and Writing—Review & Editing. SM—Oversaw methodology and implementation and Writing—Review & Editing. MP—Statistical analysis, Study design, Writing—Review & Editing. All authors have read and approved the final manuscript.

Funding

Research reported in this publication was supported by the National Institute of Mental Health of the National Institutes of Health under Award Number R34MH119982. The Children’s Investment Fund Foundation (CIFF) Award Number 1903-03681 and 2106-06589 supported the manufacturing of the study product, as well as supplementing the US government funding. Gilead donated Truvada tablets for both studies. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or CIFF.

Data Availability

Data supporting these findings will be made available with a unique doi on the Population Council’s Dataverse site https://dataverse.harvard.edu/dataverse/popcouncil prior to publication.

Declarations

Competing Interests

All authors have no competing interests to declare that are relevant to the content of this article.

Ethical Approval

The study was sponsored by the Population Council (ClinicalTrials.gov Identifier NCT04778514). The protocol (PC #952) was reviewed and approved by the Population Council’s Institutional Review Board (New York, NY, USA), the Zimbabwe Ministry of Health and Child Care, Medical Research Council of Zimbabwe (REF MRCZ/A/2576), Medicines Control Authority of Zimbabwe, Research Council of Zimbabwe, Chitungwiza City Health Ethics Committee; and the Joint Research Ethics Committee for the University of Zimbabwe Faculty of Medicine and Health Sciences and Parirenyatwa Group of Hospitals. Each participant provided written informed consent in Shona before undergoing any study procedures.