Use of a Vaginal Ring Containing Dapivirine for HIV-1 Prevention
in Women
A full list of authors and affiliations appears at the end of the article.
Abstract
BACKGROUND—Antiretroviral medications that are used as prophylaxis can prevent
acquisition of human immunodeficiency virus type 1 (HIV-1) infection. However, in clinical trials
among African women, the incidence of HIV-1 infection was not reduced, probably because of
low adherence. Longer-acting methods of drug delivery, such as vaginal rings, may simplify use of
antiretroviral medications and provide HIV-1 protection.
METHODS—We conducted a phase 3, randomized, double-blind, placebo-controlled trial of a
monthly vaginal ring containing dapivirine, a non-nucleoside HIV-1 reverse-transcriptase inhibitor,
involving women between the ages of 18 and 45 years in Malawi, South Africa, Uganda, and
Zimbabwe.
RESULTS—Among the 2629 women who were enrolled, 168 HIV-1 infections occurred: 71 in
the dapivirine group and 97 in the placebo group (incidence, 3.3 and 4.5 per 100 person-years,
respectively). The incidence of HIV-1 infection in the dapivirine group was lower by 27% (95%
confidence interval [CI], 1 to 46; P = 0.05) than that in the placebo group. In an analysis that
excluded data from two sites that had reduced rates of retention and adherence, the incidence of
HIV-1 infection in the dapivirine group was lower by 37% (95% CI, 12 to 56; P = 0.007) than that
in the placebo group. In a post hoc analysis, higher rates of HIV-1 protection were observed
among women over the age of 21 years (56%; 95% CI, 31 to 71; P<0.001) but not among those 21
years of age or younger (-27%; 95% CI, −133 to 31; P = 0.45), a difference that was correlated
with reduced adherence. The rates of adverse medical events and antiretroviral resistance among
women who acquired HIV-1 infection were similar in the two groups.
CONCLUSIONS—A monthly vaginal ring containing dapivirine reduced the risk of HIV-1
infection among African women, with increased efficacy in subgroups with evidence of increased
adherence. (Funded by the National Institutes of Health; ClinicalTrials.gov number,
NCT01617096.)
More than half of the 35 million persons currently living with human immunodeficiency
virus type 1 (HIV-1) infection are women. A majority of these women reside in sub-Saharan
Africa,
1
a region that has some of the highest incidences of HIV-1 infection in any
population worldwide.
2–4
The use of antiretroviral medications as pre-exposure prophylaxis
Address reprint requests to Dr. Baeten at the International Clinical Research Center, Department of Global Health, University of
Washington, 325 Ninth Ave., Box 359927, Seattle, WA 98104, or at jbaeten@uw.edu.
The authors’ full names, academic degrees, and affiliations are listed in the Appendix.
*
A complete list of the members of the Microbicide Trials Network 020–A Study to Prevent Infection with a Ring for Extended Use
(MTN-020–ASPIRE) Study Team is provided in the Supplementary Appendix, available at
NEJM.org.
HHS Public Access
Author manuscript
N Engl J Med
. Author manuscript; available in PMC 2017 June 01.
Published in final edited form as:
N Engl J Med
. 2016 December ; 375(22): 2121–2132. doi:10.1056/NEJMoa1506110.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
is a promising approach to the prevention of HIV-1 acquisition.
5
Several clinical trials of the
antiretroviral tenofovir showed such protection against HIV-1.
2,6–8
However, in three trials
involving African women, adherence to tenofovir-containing pills and vaginal gels was low,
and HIV-1 protection was not shown.
3,4,9
Across trials of tenofovir-based prophylaxis, a
sizable proportion of participants were not adherent, a finding that emphasizes the need for
additional options, particularly ones that women can control and longer-acting approaches
that do not require daily or coitally dependent use.
5
Vaginal rings can provide sustained and controlled release of medications. For example,
rings containing exogenous hormones are licensed for contraception and estrogen
replacement.
10
For HIV-1 prevention, an antiretroviral-containing vaginal ring could provide
long-acting HIV-1 protection while reducing systemic exposure to the active pharmaceutical
ingredient and delivering the anti–HIV-1 agent at the site of viral transmission. Dapivirine is
a non-nucleoside HIV-1 reverse-transcriptase inhibitor that has activity against a broad range
of HIV-1 subtypes. In two phase 1 trials,
11,12
genital biopsy tissue samples obtained from
women using dapivirine vaginally in the form of a ring, films, and gels were substantially
less susceptible to HIV-1 when challenged ex vivo than were tissue samples obtained from
placebo-treated women. A monthly vaginal ring containing dapivirine was found to be safe
and acceptable in phase 1 and 2 studies, with typical plasma levels of the drug that were
lower by a factor of 1000 than levels in women receiving oral dapivirine.
11–1
5
We
conducted a phase 3, randomized, double-blind, placebo-controlled trial of the dapivirine
vaginal ring among African women.
METHODS
Study Population
From August 2012 through June 2015, we enrolled and followed healthy, sexually active,
non-pregnant, HIV-1–seronegative women between the ages of 18 and 45 years at 15
research sites in Malawi, South Africa, Uganda, and Zimbabwe (Tables S1 and S2 in the
Supplementary Appendix, available with the full text of this article at
NEJM.org).
16
The
primary objectives were to determine the efficacy and safety of the dapivirine vaginal ring as
compared with a placebo ring; after insertion, the ring is used for 4 weeks and then replaced
with another. Community members from each site provided input into trial design and
conduct. The trial protocol, which is available at NEJM.org, was approved by the ethics
review committee at each site (Table S3 in the Supplementary Appendix). All participants
provided written informed consent.
Study Procedures
At enrollment, women were assigned in a 1:1 ratio, with the use of fixed-size block
randomization, stratified according to site, to receive either a silicone elastomer vaginal
matrix ring containing 25 mg of dapivirine or a placebo vaginal ring. Both the dapivirine and
placebo rings were manufactured by QPharma under contract with the International
Partnership for Microbicides. The dapivirine and placebo rings were indistinguishable, and
with the exception of staff members at the central statistical and data management center,
investigators and participants were unaware of the randomization assignments until
Baeten et al. Page 2
N Engl J Med
. Author manuscript; available in PMC 2017 June 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
completion of the trial. Women were taught how to insert and remove the vaginal ring and
counseled to wear it for the entire month.
Women returned for monthly follow-up visits, which included HIV-1 serologic testing,
safety monitoring, and individualized adherence counseling (Table S4 in the Supplementary
Appendix). At each visit, a new ring was provided, and the ring that had been used during
the previous month was collected. Women were tested monthly for pregnancy, and the study
ring was withheld from women who became pregnant; they resumed use of the study ring
when no longer pregnant or lactating. All participants received a package of HIV-1
prevention services, including counseling with respect to HIV-1 risk reduction, partner
HIV-1 testing, treatment of sexually transmitted infections in participants and partners, and
free condoms. (Details regarding the trial design are provided in the Supplementary
Appendix.)
Objective Assessment of Adherence
Plasma samples that were collected quarterly were tested for the presence of dapivirine with
the use of a validated ultra-performance liquid chromatography–tandem mass spectrometry
assay (Clinical Pharmacology Analytical Laboratory), with a lower limit of quantification of
20 pg per milliliter.
17
To aid in distinguishing cases in which the ring was removed during
the month and then reinserted before a clinic visit, the detection of a plasma dapivirine level
of more than 95 pg per milliliter (a level nearly always achieved within 8 hours of
continuous use) was used to define adherence.
13,14
While the trial was ongoing, plasma
samples were assayed and results were reviewed by the trial leaders. To preserve blinding,
samples from both the dapivirine group and the placebo group were tested, and results were
summarized only as the percentage of samples with dapivirine detected, overall and for each
site. After the first year of the trial, testing for residual dapivirine in used rings was initiated
with the use of acetone extraction and high-pressure liquid chromatography (Parexel).
Women were defined as being adherent if the returned ring contained less than 23.5 mg of
dapivirine (i.e., with >1.5 mg released).
13,14
Primary End Points
The primary efficacy end point was HIV-1 infection, identified with the use of a standard
seroconversion algorithm (Fig. S1 in the Supplementary Appendix). The study ring was
temporarily withheld while confirmatory testing was pending and was permanently
discontinued if testing confirmed HIV-1 acquisition. Archived plasma samples from visits
before seroconversion were tested for HIV-1 RNA on polymerase-chain-reaction (PCR)
assay, and participants with detectable HIV-1 RNA at the time of enrollment were excluded
from the primary analysis. Participants completed a final study visit 4 weeks after the last
product-use visit to assess for delayed HIV-1 seroconversion, and women who tested
positive for HIV-1 at that visit and who had detectable HIV-1 RNA at the last product-use
visit were included in the primary analyses because HIV-1 infection had occurred during the
product-use period.
Baeten et al. Page 3
N Engl J Med
. Author manuscript; available in PMC 2017 June 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
The primary safety end point was a composite of any serious adverse event, any grade 3 or 4
adverse event, and any grade 2 adverse event that was assessed by the trial clinicians as
being related to dapivirine.
Study Oversight
The National Institutes of Health funded the trial. The authors designed the trial, gathered
and analyzed the data, prepared the manuscript, and were responsible for the decision to
submit the manuscript for publication. The International Partnership for Microbicides
supplied the vaginal rings, was the regulatory sponsor, and participated in the design of the
trial, the interpretation of the results, and the preparation of the manuscript. The ring
manufacturer, QPharma, had no role in the design or implementation of the trial. The authors
vouch for the accuracy and completeness of the data and analyses.
Statistical Analysis
The trial was designed with power of 90% to detect a risk of HIV-1 infection that was 60%
lower in the dapivirine group than in the placebo group, with a one-sided alpha level of
0.025. Like other trials of new HIV-1 prevention interventions,
6, 7,1
8
this trial was powered
so that the lower boundary of the 95% confidence interval would exclude a 25% reduction in
risk, with the primary analysis comparison for the trial planned against a standard null of 0%
(Table S5 in the Supplementary Appendix). Under these assumptions, a minimum of 120
HIV-1 acquisition events would be required to achieve the statistical power posited in the
design of the trial. An end-point–driven design was used, and the trial continued until the
target number of HIV-1 end points had been accrued and all participants had been followed
for a minimum of 12 months, in accordance with regulatory guidance regarding compilation
of safety information for new HIV-1 prevention strategies.
19
An annual incidence of HIV-1 infection of 3.9% in the placebo group was assumed, and a
sample size of 3476 women was planned. After the trial started, another HIV-1 prevention
trial that was conducted at several sites in our trial showed an HIV-1 incidence of more than
5% per year.
3
As a result, in October 2013, the sample size for this trial was recalculated to
approximately 2600 women, and the statistical analysis plan was modified accordingly. At
the same time, the analysis plan was further modified for a fully powered analysis that
would exclude all data from 2 of the 15 sites, since these 2 sites had shown lower-than-
anticipated participant retention and lower product adherence (with adherence levels of
<50%, according to measurement of plasma dapivirine levels) than at the other sites. Further
enrollment at the 2 sites was discontinued, but enrolled participants were permitted to
continue in follow-up. The sample-size recalculation and plan to exclude data from the 2
sites were approved by the independent data and safety monitoring board and reviewed by
regulatory agencies. Interim statistical monitoring, which was performed on the basis of data
from all 15 sites, used the Lan–DeMets spending approach to adjust the O’Brien–Fleming
sequential monitoring boundaries.
20,21
The primary analysis of HIV-1 protection was performed according to the intention-to-treat
principle with the use of Cox regression, stratified according to site, to estimate the relative
rates of time until HIV-1 acquisition. Two analyses were defined: one included data from all
Baeten et al. Page 4
N Engl J Med
. Author manuscript; available in PMC 2017 June 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
15 sites and a second excluded data from the 2 sites at which enrollment had been
discontinued early (i.e., 13 sites were included). Prespecified subgroup analyses were
planned. When it was determined that age was significantly related to the efficacy of HIV-1
protection, a post hoc analysis was designed to characterize more fully that relationship by
dividing the population into age-categorized thirds containing approximately equal numbers
of participants with HIV-1 infection, thus balancing the statistical power for the exploratory
subgroups. All analyses were conducted with the use of SAS software, version 9.4 (SAS
Institute), and R software, version 2.15.1 (R Project for Statistical Computing). A P value of
less than 0.05 was considered to indicate statistical significance, and all P values are two-
sided.
RESULTS
Study Participants
Of 5516 women who underwent screening, 2629 were enrolled: 1313 in the dapivirine group
and 1316 in the placebo group (Fig. 1). The median age was 26 years (interquartile range, 22
to 31). Less than half (41%) were married, and 85% had completed some secondary
schooling. Nearly all (99.5%) reported having had a primary sex partner during the 3 months
before trial enrollment, and 17% reported more than one partner; 57% reported the use of a
condom with the most recent sex act. Transactional sex in the previous year was reported by
6%, and anal sex during the previous 3 months by 2%. Nearly two thirds (64%) reported that
their primary partner was aware that they would be using a vaginal ring in a research trial.
Characteristics were similar in the two groups (Table 1).
Follow-up and Adherence
The rate of retention of participants for assessment of incident HIV-1 infection was 85% or
more during follow-up (Fig. 1), with 2614 participants (99.4%) completing at least one post-
randomization HIV-1 test and 4280 total person-years of follow-up accrued for assessment
of HIV-1 incidence. The median follow-up was 1.6 years (interquartile range, 1.1 to 2.3),
and the maximum follow-up was 2.6 years; 1024 women contributed more than 2 years of
follow-up. The most common reason for not dispensing the study ring was pregnancy, which
occurred at an incidence of 3.9 per 100 person-years in the dapivirine group and 4.0 per 100
person-years in the placebo group (P = 0.82).
In the dapivirine group, the drug was detected in 82% of plasma samples at levels of more
than 95 pg per milliliter (Fig. S2 in the Supplementary Appendix). Detection increased
during the first year of use and was relatively stable thereafter. In the subgroup of visits in
which returned rings were available, 84% contained less than 23.5 mg of dapivirine, and
dapivirine levels in plasma and in returned rings were correlated (Fig. S3 in the
Supplementary Appendix). In general, for visits at which plasma dapivirine levels were less
than 95 pg per milliliter, the residual dapivirine levels in used rings were similar to levels in
unused dapivirine rings, whereas residual dapivirine levels in used rings were lower for
visits at which plasma dapivirine levels were more than 95 pg per milliliter. However, a
range of residual dapivirine levels was observed, with low levels observed for some visits
Baeten et al. Page 5
N Engl J Med
. Author manuscript; available in PMC 2017 June 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript