Duration of Polymerase Chain Reaction–Detectable DNA After Treatment of Chlamydia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis Infections in Women

Abstract

Background 

To avoid positive results attributable to residual DNA, the Centers for Disease Control and Prevention recommends avoiding repeat testing with nucleic-acid based tests within 3 weeks after treatment of chlamydial (Chlamydia trachomatis [CT]) or gonococcal (Neisseria gonorrhoeae [GC]) infection. We retrospectively analyzed the duration of detectable DNA from a longitudinal cohort of adolescent women after diagnosis and treatment of infection with CT, GC, or Trichomonas vaginalis (TV).

Methods 

Vaginal swabs were obtained weekly from young women for up to 12 weeks (observation period) after treatment of CT, GC and TV infections. Swabs were tested using a commercially available first generation nucleic acid amplification test (NAAT) for CT and GC, and a laboratory developed NAAT for TV. Kaplan-Meier statistics were used to estimate median time to the first negative DNA-based polymerase chain reaction (PCR) result.

Results 

Observation periods were available for analysis for 195, 82 and 102 treatments for CT, GC, and TV infection, respectively. Median time to a first negative PCR result for CT, GC, and TV was 9 (range 0–84), 6 (0–76), and 7 (0–84) days, and by day 21, 89%, 95%, and 85% were negative, respectively.

Conclusions 

Data from this retrospective analysis indicate that greater than 85% of these young women did not have detectable CT, GC, or TV DNA by day 21 post-treatment. This data may be useful to clinicians for patient management and post-treatment testing purposes.

In Brief

This study examined the length of time required to obtain a negative amplified DNA result after appropriate treatment of Chlamydia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis.

The Centers for Disease Control and Prevention provide guidelines for laboratory diagnosis of sexually transmitted infections (STIs).¹ These guidelines were updated in 2010, identifying nucleic acid amplification tests (NAATs) as the preferred diagnostic method for Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (GC) infections.² As diagnostic technologies have moved beyond culture of organisms, new issues have emerged as a result of the ability to detect DNA or RNA without knowing whether viable organisms capable of causing disease or transmission to sexual partners are present. One of the gaps in current knowledge is the duration of detectable DNA or RNA after successful antimicrobial therapy for these infections. This information will help clinicians to determine when to retest clients that continue to seek care after treatment of these infections. The current treatment guidelines recommend against retesting within 3 weeks after the completion of therapy because scant data regarding duration of detectable nucleic acids exist.³

Given the limited amount of data available for chlamydial (CT), gonococcal (GC), and trichomonal (no studies were found) shedding after treatment, we retrospectively analyzed samples collected as part of a longitudinal cohort of adolescent girls who provided weekly vaginal samples. Here we present an analysis of the duration of detectable DNA using weekly results obtained immediately after diagnosis and treatment.

MATERIALS AND METHODS

Population and Clinical Management

As part of a 10-year (1999–2009) longitudinal study of a cohort of adolescent women that was designed to increase our understanding of the behavioral and developmental epidemiology of STI, vaginal samples for NAAT were collected in clinic on a quarterly basis and weekly for multiple 12-week periods over the course of the project. This study and the population have been described in detail.⁴ All aspects of the Young Women’s Project were approved by the Indiana University-Clarian Institutional Review Board. All women provided written informed assent, and a parent or guardian provided permission for participation.

Women enrolled in the study were seen in a clinical setting every 3 months for diagnosis and management of STI and for collection of behavioral data. In addition to quarterly clinic visits, participants were asked to keep daily diaries and to self-collect weekly vaginal swabs for two 12-week periods each year. Samples collected during the periods of intense observation were batched for testing immediately before the subsequent clinic visit. Results were made available to study staff before each participant’s appointment to facilitate treatment of any infections identified. Participants interacted with study personnel on a weekly basis during the intensive observation periods and were routinely encouraged to seek health care through the study or through their primary providers should they experience any symptoms of STI. Polymerase chain reaction (PCR; Amplicor CT/NG; Roche Molecular Diagnostics, Indianapolis, IN) was used to detect chlamydia and gonorrhea DNA from these samples, and a modification of this assay was used to detect trichomonas DNA.^5,6

Prescriptions or directly observed therapy (DOT) were provided as part of the study protocol. Directly observed therapy was available for chlamydial infections from study initiation through May 2004 and for gonococcal infections from study inception through January 2002. Infections identified outside these periods and all trichomonal infections were managed by provision of a prescription. The most common prescriptions provided by study personnel included azithromycin (1 g, single dose) for chlamydia, ciprofloxacin (500 mg, single dose), or cefixime (400 mg, single dose) for gonococcal infections and metronidazole (2 g, single dose) for trichomoniasis.⁶

Samples and Diagnostic Testing

Cervical or vaginal swabs were collected by study personnel during clinic visits depending on the need for a full pelvic examination. Cervical swabs were placed in chlamydia transport medium (M4; Remel; Lenexa, Kansas, TX) at the clinic and vaginal swabs were transported dry. During the weekly follow-up visits, vaginal samples were self-obtained by participants using dry Dacron swabs and collected by field staff for transportation to the laboratory. Vaginal swabs were eluted using molecular grade water and tested using Amplicor PCR for detection of chlamydia or gonorrhea and a modification of this assay for detection of trichomonas DNA.^5,7

Analysis

Analyses were restricted to the set of weekly swab collection periods that began with a positive test result at the clinic visit and for which either DOT or provision of a prescription was documented. The date of the treatment was used as the start of the observation time, whereas the date of collection of the first negative swab sample was used as the end date of DNA shedding. For some individuals, follow-up time stopped accruing (i.e., time was right censored) because they (1) never had a negative result before their next scheduled clinic visit (at which point they were considered a treatment failure and retreated) or (2) sought retreatment during the weekly sampling period because of continued symptoms of infection. In weekly collection periods where 2 documented treatment events were within 2 weeks of each other, as a simplifying assumption, the date of the second event was used as the start of observation time. Potential explanations for repeat treatments prescribed within 2 weeks include the following: the initial prescription was lost, and therapy was not initiated until the second treatment event; continued symptoms after the initial treatment; or 2 different providers each issuing a prescription without knowledge of the other doing so. For periods with multiple treatment events not occurring within 2 weeks of each other, the date of the second treatment stopped the follow-up time for the prior treatment and began a new observation period. Median time to negative was estimated using Kaplan-Meier curves. A participant-level analysis of time to first negative result was also conducted using only each participant’s first observation period as started by their first treatment. Median time to a negative result was also estimated for the subset of periods for which we had DOT for chlamydial and gonococcal infections. Participant-level rates of a negative result by days 14 and 21 were estimated using each participant’s first period among these periods with DOT.

RESULTS

The study population was predominately African American (89.1%) with a mean age at enrollment of 15.3 years. For chlamydial infections, 195 observation periods from 123 participants were available for analysis (Fig. 1). One observation period was contributed by 77 (62.6%) participants, 2 observation periods by 29, and 3 or more by 17 participants, respectively. The median time to the first PCR negative sample was 9 days (range, 0–84; Fig. 2A). Using only the first observation period for each participant, 88 (71.5%) of women were negative for chlamydia PCR by day 14 and 109 (88.6%) were negative by day 21 (Table 1). In the subset analysis restricted to the 80 observation periods in which DOT was provided, the median days to negative was 7 (range, 0–84). In the DOT subset of 59 women, 74.6% were negative by 14 days and 91.5% were negative by 21 days.

TABLE 1: Total and Participant-Level Analysis of CT Time to Clearance

Figure 1: Flowchart of participant inclusion criteria for CT, GC, and TV analysis.

Figure 2: Time to first negative swab after appropriate treatment of CT or GC or TV.

For gonococcal infections, 82 periods of observation were contributed by 66 participants (Fig. 1). Of the participants, 54 (81.8%) had only 1 treatment period during participation in the overall study, whereas 8 participants contributed 2 treatment observations and 4 contributed 3 observation periods. For this infection, the median time to a negative PCR result was 6 days (range, 0–76 days; Fig. 2B). In the participant-level analysis, 51 (87.6%) women were negative by day 14 and 95.0% were negative by day 21 (Table 2). When analysis was restricted to the 19 observations with DOT, the median and range were unchanged. In the participant-level analysis of only DOT, 12 (79.8%) of women had end of shedding by day 14 and 14 (93.3%) women had end of shedding by day 21.

TABLE 2: Total and Participant-Level Analysis of GC Time to Clearance

Trichomonal infections were identified in 102 observation periods contributed by 64 participants (Fig. 1). Forty (62.5%) contributed only 1 observation period, whereas 15 and 9 women contributed 2 or more than 2 observations, respectively. For this organism, shedding was no longer detectable within a median of 7 days (range, 0–84 days; Fig. 2C). There was no period of DOT for this infection, and therefore, no subset analysis was possible. For trichomonal infections, 42 (75.0%) women had cleared residual DNA by day 14 and 46 (85.0%) had cleared by day 21 (Table 3).

TABLE 3: Total and Participant-Level Analysis of TV Time to Clearance

DISCUSSION

Few studies have assessed the potential duration of chlamydia, gonorrhea, or trichomonas nucleic acid shedding detectable by commercially available methods. Two seminal studies were performed when amplification methods for diagnosis of chlamydial infections were first introduced. The first of these followed 20 women weekly and performed both culture and a laboratory-developed PCR assay to detect both viable organisms and chlamydial DNA. The women were evaluated weekly after the completion of a 7-day course of doxycycline. Immediately after the week of therapy, all culture samples were negative and 10 of 20 PCR samples were negative. By the end of the first week after the completion of therapy, only 3 of 20 PCR samples remained positive and all were negative by the second week after the end of treatment. Although these data indicated that retesting at 3 weeks was unlikely to detect residual DNA after treatment, the results may not be generalizable because the sample size was small.⁸ The second study evaluated urine samples from 33 high school students who were followed up after treatment with either azithromycin (single dose) or doxycycline.⁹ Although up to 14.3% of samples continued to have detectable DNA up to 2 weeks after the completion of therapy, only 8 participants provided follow-up samples at that point.

For gonorrhea, the time to a negative test result was measured in 87 men and 43 women after a documented infection and single-dose therapy. Women (vaginal swabs) and men (urines) were sampled daily for up to 21 days, and samples were tested by the Ligase Chain Reaction assay (no longer available). The median time to a negative result by amplified testing was 1 day for men and 2 days for women, with all urine results being negative by day 6 and vaginal swabs by day 9.¹⁰

The data we present suggest that within 2 weeks of treatment, more than 72% to 88% of patients will no longer have detectable DNA. Within 3 weeks, 89%, 95%, and 85% of women no longer had detectable levels of chlamydial, gonococcal, and trichomonal DNA, respectively. In a recent study by Renault et al.,¹¹ the time to clearance of CT ribosomal RNA was examined using a commercially available assay (Gen-Probe APTIMA Combo 2, San Diego, CA). In this study, women with a diagnosed urogenital CT infection were treated with azithromycin and then asked to obtain self-administered vaginal swabs on days 0, 3, 7, 10, and 14. The vaginal specimens were tested for CT ribosomal RNA at each time point; 48 (79%) of 61 participants were negative at day 14. Linear regression analysis predicted a time to clearance of 17 days (95% confidence interval [CI], 16–18). This is similar to what was observed in our cohort where approximately 75% of the participants who received DOT were negative for CT by day 14.

Limitations to these analyses include weekly, versus daily, collection of specimens and perhaps the generalizability of these data to other patient populations. It is unclear how older women or men may clear DNA, and we present no data regarding clearance of RNA. There are currently no data that compare DNA to RNA shedding after treatment. The influence of differences in target quantity (higher for RNA) and nucleic acid stability (higher for DNA) is unknown but worthy of further study. In addition, in those who did not receive DOT, we cannot exclude a delay between the date of providing a prescription and the date the participant actually took the drug. In some cases, a participant may have failed to fill the prescription and, thus, never had treatment. These would tend to bias to longer durations in the overall analysis. Finally, it is important to contextualize these results as retrospective data analyses. The PCR assay used for chlamydia and gonorrhea was a first-generation assay, and the trichomonas assay was developed before the availability of commercial trichomonas NAATs that now exist.

There are several strengths of this study including a large number of observational periods for analysis, the collection of vaginal swabs instead of urine from females, and the use of a well-characterized NAAT for detection of CT, GC, and Trichomonas vaginalis (TV). The duration of detectable nucleic acid after treatment of these STIs is conservative and may, in fact, be shorter because vaginal specimens were only obtained at weekly intervals for testing. Positive results after this period are likely to be the result of continued exposure (i.e., to either viable or nonviable organisms) or possible treatment failure.⁶ These data should be used by clinicians to better manage those patients who request testing after treatment by guiding the discussion with the patient. For those with repeatedly positive results, testing and treatment of partners or treatment with alternate regimens may be appropriate based on the sexual history elicited during such conversations. This will be especially important in terms of managing gonococcal infections, which are becoming increasingly resistant to currently available antibiotics and may require new treatment follow-up paradigms for settings where culture is not available. Further studies with second-generation assays using both DNA and RNA targets would be useful.