From the Departments of *Medicine and †Pediatrics, Mount Sinai School of Medicine, New York, New York
Supported by Public Health Service grant HD43733 and by grant M01-RR-00071 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH).
The contents of this are solely the responsibility of the authors and do not necessarily represent the official views of NCRR or NIH.
Correspondence: Marla J. Keller, MD, Mount Sinai School of Medicine, One Gustave L. Levy Place, Box 1090, New York, NY 10029. E-mail: email@example.com.
Received on February 15, 2007, and accepted May 11, 2007.
Objectives: To determine whether applicator staining could be incorporated into a microbicide study as a marker of adherence.
Goal: To test whether observers could identify intravaginally inserted applicators and compare the observers’ ratings to subjects’ self reports.
Study Design: Subjects participating in a 14-day microbicide trial to assess the safety of 0.5% PRO 2000 or matched placebo gel returned used and unused applicators. Each applicator was individually dyed or batched and immersed in a 0.4% trypan blue waterbath. Masked observers rated the applicators as intravaginally inserted or not; results were compared to subjects’ self-reports. To determine if the dye would allow observers to differentiate whether applicators had been filled before intravaginal insertion, staff either filled applicators and discarded gel ex vivo or asked volunteers to intravaginally insert applicators without gel and mixed these with a random sampling of used and unused study samples.
Results: 358 of 360 applicators were returned; 306 were reported to have been vaginally inserted and 52 unused. Observers agreed with the participant’s self reports in 98–99% of cases. No differences were observed between applicators stained individually or in batches. The ability to distinguish between applicators that were presumed to have been used properly, intravaginally inserted without gel, filled with gel and emptied ex vivo, or unused ranged from 76 to 87%.
Conclusions: Staining of applicators is accurate, inexpensive, and correlated well with self-report in a study of short duration. This method could prove useful in assessing adherence to product dosing in future microbicide trials.
ACCORDING TO UNAIDS, an estimated 39.5 million people were living with human immunodeficiency virus (HIV) by the end of 2006, and an estimated 4.3 million people were newly infected with HIV in 2006.1 Condoms are effective at preventing transmission, but social barriers limit their use. Developing novel strategies to prevent HIV and other sexually transmitted infections is a critical global health priority. A major focus has been the development of drugs for vaginal application. Several topical microbicides have been advanced to large-scale Phase IIb/III clinical trials. Compliance with study procedures including product dosing, condom use, and frequency of sexual intercourse is difficult to assess in microbicide trials. The accuracy of self-report and diary review of sexual issues vary greatly and must be interpreted with caution. In a nonoxynol-9 clinical trial, researchers reported that participants completed monthly booklets intended to capture each act of coitus and/or gel use in the waiting room immediately before study visits.2 A rapid, inexpensive method of accurately assessing protocol adherence in clinical trials would greatly enhance interpretation of study results.
Potential methods for assessing adherence to product dosing include measurement of drug levels postapplication or measurement of functional drug activity after application.3 These strategies are limited by the need to obtain vaginal secretions (swab or cervicovaginal lavage) within a short time after application. Additionally, the number of cells isolated from swabs or cervicovaginal lavage may be insufficient to detect drugs that are rapidly internalized such as tenofovir. Although measuring intracellular drug levels using biopsy samples may be included in small pilot dosing studies, this would not be a realistic approach to monitor compliance in any large-scale clinical trial. Recent efforts have been directed at evaluating the vaginal applicators to determine whether they have been exposed to the genital tract. In a recently published study, intravaginally applied single-use Microlax applicators containing a placebo formulation could be correctly distinguished from unused applicators by staining with 1% trypan blue, which detected exposure to vaginal mucous.4 Building from this observation, we incorporated this assay with modifications in a randomized double-blind placebo-controlled 0.5% PRO 2000 study and compared results obtained to participant diaries and verbal reports.5
Materials and Methods
Participants were 24 healthy, HIV negative women enrolled in an IRB-approved study to evaluate the mucosal immune response to daily use of 0.5% PRO 2000 versus a matched placebo gel in a 14-day study.5 Details of the study design have been previously described.5 Sixteen individual tubes of PRO 2000 or placebo gel and sixteen single-use, unfilled vaginal applicators were dispensed to each study participant. Participants were instructed how to fill the applicator, intravaginally insert and dispense the gel. The first dose of gel was prepared and administered by the participant under the guidance of the study clinician. Study subjects were instructed to apply 1 dose of gel each subsequent night for a total of 14 days and to return the used, unrinsed applicators in separate plastic bags at the day 7 and day 14 study visits. They were also asked to return all unused applicators. Each subjects’ study diary was reviewed at each visit. Reminder telephone calls were placed to the participant 2–3 times per week during gel use.
Staining and Examination of Applicators
Two similar methods were adopted to stain applicators. In the method used for the first 4 participants, each applicator was held by the plunger tip, sprayed with a solution of 0.4% trypan blue for 5 seconds, allowed to air dry for 10 minutes, and rinsed gently with distilled water. The day 0 applicator (inserted under clinical observation by each enrolled participant) served as the positive control and an unused applicator served as the negative control (Figs. 1A,B). In an alternative method, which was employed for the remaining 20 participants, applicator staining was performed in batches. 200 mL of 0.4% trypan blue was poured into a shallow pan and 5–10 applicators at a time were submerged into the dye and left for ∼1 minute, removed, immediately rinsed in a shallow pan of distilled water, and air dried for 5–10 minutes. Gloves, masks, and eye protection were worn by laboratory personnel to avoid possible contamination. Two judges, who were masked to diary results or verbal reports of applicator and gel use, evaluated the applicators independently.
The color change in response to trypan blue staining presumably reflects exposure to vaginal mucous,4 but could be modified by the presence or absence of gel. To determine if the observers could differentiate between intravaginally applied applicators that had or had not been prefilled with gel, a substudy was designed. Four sample sets, each containing 18–20 stained applicators, were generated. Each set was comprised of a random assortment of the following: (a) stained applicators returned by study participants and reported and judged to have been unfilled and unused; (b) stained applicators returned by study participants and reported and judged to have been filled and applied intravaginally; (c) stained applicators that were unfilled and intravaginally inserted by volunteers; and (d) stained applicators that were filled and dispensed ex vivo by study staff. The location of each applicator type varied in each sample set. The observers were not given information regarding how many applicators of each type were included in the sample sets, but were shown a stained applicator representing each of the 4 conditions.
Evaluation of Subjects Samples
A total of 358 of 360 (99%) applicators were returned for evaluation, and participants reported that 306 had been filled with gel and intravaginally dispensed, and 52 unused (Table 1). Verbal report and diary review matched 100% about drug application. 100% of the participants reported using the gel the evening before the day 7 and day 14 visit and this was confirmed by clinician visualization of gel byproducts in the vaginal fornices in all of the subjects (Figs. 2A,B). The 2 independent observers’ evaluation of whether applicators had been exposed to the vagina based on trypan blue staining is summarized in Table 1. 98% and 99% of applicators that had been reported to have been properly used by the subject stained positively with trypan blue as determined by observers 1 and 2, respectively. Both observers reported that none of the reportedly unused returned applicators stained positively with trypan blue. For the first 4 participants, the returned applicators (n = 58, of which 48 were reported as “used” and 10 as “unused”) were stained individually using the spray technique. In an effort to reduce labor and time, the applicators returned by the next 20 subjects were batched for staining. No differences in sensitivity of trypan blue assay were noted between the 2 methods. No difference in the ability to detect intravaginal insertion of applicators was noted between subjects who had received PRO 2000 or placebo gel (Table 1).
To determine if gel exposure modifies applicator staining, thus permitting the observers to differentiate between unfilled or filled and inserted or uninserted applicators, a substudy was designed as described in Methods. Results indicate that both observers correctly identified all of the unfilled and uninserted applicators (Fig. 3). However, both observers were less able to determine whether applicators had been filled and inserted when they were asked to consider other options, e.g., had the applicator been filled with gel and emptied ex vivo or had the applicator been inserted into the vagina without gel? Accurate identification of filled and inserted applicators decreased for both observers (69 and 93%) (Fig. 3).
This study extends the initial observation that trypan blue staining could distinguish intravaginally inserted from uninserted applicators.4 We incorporated this assay into a 14-day microbicide safety study. We found that visual examination of trypan-blue stained applicators by 2 observers correlated with participant diary review and self-report in 98–99% of cases. Comparing trypan blue staining with self-reports is predicated on the assumption that the participants in this short study reliably reported adherence. We were able to confirm gel use on 3 of the 14 days. Administration of the first dose was directly observed, and gel byproducts were visualized in the vagina of all participants on day 7 and 14 (Fig. 2). However, we cannot be certain that the gel was used as reported on the other 11 days. In the absence of a gold standard such as the ability to directly measure drug levels, evaluating any new adherence tool is challenging. Although the observers accurately differentiated intravaginally inserted from uninserted applicators when only these 2 possibilities were considered, the substudy suggests that the observers were less accurate when also asked to consider applicators that had been unfilled and inserted or filled with gel and dispensed ex vivo. Although not tested, it is likely that this technique would not identify the possible scenario, in which an applicator was filled, inserted intravaginally and removed without dispensing the gel. In addition, the assay provides no information regarding the timing of gel applicator about intercourse.
We modified the originally described assay in 2 respects. First, we modified the assay to perform evaluations in batches, which would be important if one were to incorporate this assay into a larger clinical trial.
In addition, we used a lower concentration of trypan blue in this study (0.4%), which was as effective as the 1% concentration used in the previous study. This lower concentration should minimize potential risk to laboratory personnel from accidental inhalation or contact with trypan blue. Alternative dyes may also prove less expensive and safer. For example, a simple food dye may be just as effective as trypan blue in distinguishing intravaginally inserted from uninserted applicators.6
Advantages of this assay are that it is inexpensive, simple, and could easily be replicated in clinical settings by minimally trained staff in the developing world. However, the inability to differentiate filled from unfilled vaginally inserted applicators or to identify filled applicators in which the gel was dispensed ex vivo coupled with the inability to provide information about the timing of application about coitus limits the utility of this assay. Despite these limitations, this assay may better assess subject adherence than diaries or self report alone and may provide an important tool in evaluating the results obtained in clinical trials.
1. Joint United Nations Programme on HIV/AIDS and World Health Organization. AIDS Epidemic Update. 2006.
2. Van Damme L, Ramjee G, Alary M, et al. Effectiveness of COL-1492, a nonoxynol-9 vaginal gel, on HIV-1 transmission in female sex workers: A randomised controlled trial. Lancet 2002; 360:971–977.
3. Keller MJ, Zerhouni-Layachi B, Cheshenko N, et al. PRO 2000 gel inhibits HIV and herpes simplex virus infection following vaginal application: A double-blind placebo-controlled trial. J Infect Dis 2006; 193:27–35.
4. Wallace A, Thorn M, Maguire RA, et al. Assay for establishing whether microbicide applicators have been exposed to the vagina. Sex Transm Dis 2004; 31:465–468.
5. Keller MJ, Guzman E, Hazrati E, et al. PRO 2000 elicits a decline in genital tract immune mediators without compromising intrinsic antimicrobial activity. AIDS 2007; 21:467–476.
6. Govender S, Skoler S, Maguire R, et al. Evaluation of microbicide applicators to determine vaginal use in the Carraguard™ phase 3 clinical trial. Oral Presentation. Microbicides 2006, Cape Town, South Africa.