Incidence and predictors of reinfection with trichomoniasis based on nucleic acid amplification testing results in HIV-infected patients

Abstract

Trichomonas vaginalis infection contributes to HIV transmission. The study objective was to determine the incidence and predictors of T. vaginalis reinfection among HIV-infected women in Birmingham, Alabama. A retrospective cohort study of women at an urban HIV clinic from August 2014 to March 2016 with T. vaginalis by nucleic acid amplification test (NAAT) was conducted. Time to first episode of reinfection was evaluated using Kaplan–Meier survival curves. The association of various predictors was evaluated by univariate and multivariable Cox proportional hazards analyses. Of 612 HIV-infected women at the UAB HIV clinic tested for T. vaginalis by the Aptima TV assay, 110 (18.0%) were identified with prevalent T. vaginalis infection. Overall, 25/110 (22.7%) had a first episode of T. vaginalis reinfection by NAAT with a rate of 3.7 reinfections per 100 person-months (95% confidence interval [CI]: 2.3, 5.2). In univariate analysis, only an HIV viral load (VL) ≥200 copies/ml approached statistical significance (hazard ratio = 2.26; 95% CI: 0.97, 5.29, p = 0.06). After adjusting for age and race, the association of HIV VL ≥200 copies/ml remained strong (adjusted hazard ratio = 2.49; 95% CI: 0.99, 6.27, p = 0.05). T. vaginalis reinfection was high among HIV-infected women in this sample, necessitating enhanced disease control efforts in this high-risk population.

Keywords

HIV sexually transmitted infection reinfection

Introduction

Trichomonas vaginalis, the most common nonviral sexually transmitted infection (STI), is associated with adverse pregnancy outcomes and increased HIV transmission.^1–3 It is the most prevalent STI in HIV-infected women (5–42%), with African-American race, and history of STIs being risk factors.^4–7 T. vaginalis is not reportable. With the exception of HIV-infected women who should be screened at entry to care and at least annually thereafter, there are no screening guidelines.⁸

Traditionally, women are tested for T. vaginalis using a wet mount of vaginal fluid (point-of-care test using microscopy with 50–60% sensitivity)⁹ with the gold standard (prior to the availability of the T. vaginalis nucleic acid amplification test [NAAT]) being culture of vaginal fluid.¹⁰ The T. vaginalis NAAT is a highly sensitive and specific test that can be used on multiple genitourinary specimen types in women (urine, vaginal [including self-collected], and endocervical specimens).^11–13 For HIV-infected women with T. vaginalis, the seven-day metronidazole regimen (500 mg BID) is more effective than a 2 g dose of metronidazole.⁸ Retesting is recommended within three months following initial treatment of HIV-infected women with T. vaginalis.

The University of Alabama at Birmingham (UAB) HIV clinic is an urban clinic servicing >4000 patients in the southern United States, an area with high HIV/STI rates.^14,15 The Aptima TV assay¹¹ became available for use in women at this clinic in August 2014. HIV-infected women at this clinic are screened for T. vaginalis by NAAT at entry to care and annually thereafter, in accordance with the 2015 CDC STD Treatment Guidelines,⁸ and more frequently if symptomatic. Availability of the highly sensitive Aptima TV assay has allowed recent investigation of T. vaginalis prevalence (17.2%) and predictors of infection (age ≤40, current cocaine use, and CD4 cell count < 200 cells/mm³) among HIV-infected women at this clinic.¹⁶ Data on T. vaginalis reinfection and predictors of reinfection among HIV-infected women based on NAAT test results are unavailable yet needed, given the adverse outcomes associated with T. vaginalis infection. Other studies on T. vaginalis reinfection in HIV-infected populations included women diagnosed by wet mount, culture, Papanicolaou (Pap) smear, or urinalysis^6,17,18 before NAAT was available.

With this in mind, the objective of this study was to determine the incidence and predictors of a first episode of T. vaginalis reinfection in HIV-infected women based on NAAT testing results.

Methods

This was a retrospective cohort study of HIV-infected women at the UAB HIV clinic between August 2014 and March 2016 with at least one diagnosis of T. vaginalis based on the Aptima TV assay (Hologic, San Diego, California). T. vaginalis NAAT testing was performed on genitourinary specimens (urine, vaginal, and/or endocervical). A first episode of T. vaginalis reinfection was defined as at least two positive T. vaginalis NAAT results during the study, including a positive result after metronidazole or tinidazole was prescribed for a first episode of infection and/or a positive result after an interval negative result following a first episode of infection. August 2014 was selected as the study start as this was the date of Aptima TV assay availability at UAB.

Data were obtained from the HIV Clinic Cohort Observational (electronic) Database Project (www.uab.edu/medicine/1917cliniccohort/), a prospective clinical cohort containing sociodemographic, psychosocial, and clinical data on HIV-infected patients.^19,20 Study criteria included HIV-infected women with an initial positive T. vaginalis NAAT during the study period. The following additional data (closest to the time of initial positive T. vaginalis NAAT) were obtained for women meeting inclusion criteria: sociodemographics (age, race, insurance); alcohol abuse (classified as ‘no risk,’ ‘low risk,’ and ‘at risk’ by the Alcohol Use Disorders Test-Consumption questionnaire²¹); and substance use (current, prior) due to marijuana, cocaine, and IV drug use (IVDU) (using the Alcohol, Smoking, and Substance Involvement Screening Test).²² Current sexual activity, condom use in the past six months, number of sexual partners in the past six months, and history of sex under the influence of alcohol and drugs in the past six months were obtained. In addition, data on Chlamydia and gonorrhea NAAT test results at the time of the first positive T. vaginalis NAAT or within four weeks of this test were obtained as were data on current use of antiretroviral therapy (ART), CD4 cell count, HIV viral load (VL), and subsequent T. vaginalis NAAT results. Finally, T. vaginalis treatment data (i.e. evidence of a metronidazole or tinidazole prescription at the first episode of T. vaginalis infection) were obtained by chart review. Treatment for T. vaginalis is not given as directly observed therapy; rather, medication is e-prescribed to the clinic pharmacy where patients pick it up.

For initial descriptive statistics, HIV-infected women with T. vaginalis by NAAT were dichotomized into those with and without reinfection. Categorical variables were reported as frequencies with percentages. Continuous variables were reported as mean with standard deviation (SD) or median with quartiles depending upon a normal/skewed distribution, respectively. Time (months) to first episode of T. vaginalis reinfection, the outcome of interest, was evaluated using Kaplan–Meier survival curves (time-to-event analysis) and compared across stratified variables using the log-rank test. The duration of survival started with the first positive T. vaginalis NAAT test. The end date for women with reinfection was the date of the first episode of reinfection during the study timeframe. Women without reinfection were censored at the end of the study (March 2016); two patients who died were censored on the day of death. Patients lost to follow-up (i.e. no primary care visit within six months of the study end date) were censored at six months from their last clinic visit/first positive test date or censored at the end of the study period, whichever came first.

The association of various predictors with T. vaginalis reinfection was evaluated by univariate and multivariable Cox proportional hazards analyses reporting unadjusted and adjusted hazard ratios (HRs and aHRs), respectively, with 95% confidence intervals (CIs). Due to the small sample size and to keep the model parsimonious, only the following variables were included in a multivariable model based on their clinical significance: age, race, and HIV VL. The proportional hazards assumption was tested by forming a variable×time interaction term for various predictors; the p-value was 0.36 indicating no violation of the assumption. Multicollinearity of these variables was examined with variance inflation factor (VIF) by adjusting the linear combinations by the weight matrix used in the maximum likelihood algorithm; the VIF for all the variables was <1.1 indicating no multicollinearity. Statistical significance was set at 0.05 (two-tailed).

As 58/110 women were retested for T. vaginalis by NAAT during the study period, sensitivity analyses were conducted. Two multivariable models were examined in sensitivity analyses¹: restricting analysis only to the 58 women who were retested for T. vaginalis by NAAT and² assuming 10% (n = 5) of the 52 women who were not retested were diagnosed with T. vaginalis reinfection had they been retested; the women were selected randomly with 1000 replications.

All analyses were performed using SAS, version 9.3 (SAS Institute, Cary, NC). This study was approved by the UAB institutional review board (protocol # X150123007).

Results

Of 612 HIV-infected women tested for T. vaginalis by NAAT between August 2014 and March 2016, 110 (18.0%) had prevalent T. vaginalis infection. Table 1 shows the characteristics of these 110 women, stratified by their T. vaginalis reinfection status. Mean age was 40.7 years (SD = 9.6), 88.2% were African-American, 90.9% were on ART, 82.6% had a CD4 cell count ≥200 cells/mm³, and 68% had an HIV VL ≥200 copies/ml. Data on IVDU were missing for 73% and not included in the analysis. Evidence of a metronidazole prescription at the time of first episode of T. vaginalis infection was available for 101/110 (91.8%) of women; 13/101 (12.9%) received the 2 g dose while 88/101 (87.1%) received the seven-day dose (data not shown). There were no tinidazole prescriptions. Treatment data were missing for nine women.

Table 1.

Characteristics of HIV-infected women with a diagnosis of trichomoniasis^a at the UAB 1917 HIV clinic, stratified by T. vaginalis reinfection status, August 2014–March 2016.

Characteristic	Reinfection (n = 25) n (%)	No reinfection (n = 85) n (%)	Total (N = 110) N (%)
Age, years
<25	0 (0)	4 (4.7)	4 (3.6)
25–40	15 (60.0)	39 (45.9)	54 (49.1)
>40	10 (40.0)	42 (49.4)	52 (47.3)
Race
African-American	22 (88.0)	75 (88.2)	97 (88.2)
Caucasian	3 (12.0)	10 (11.8)	13 (11.8)
Insurance type
Private	2 (8.0)	25 (29.4)	27 (24.5)
Public	8 (32.0)	30 (35.3)	38 (34.6)
Uninsured	15 (60.0)	28 (32.9)	43 (39.1)
Unknown	0 (0)	2 (2.4)	2 (1.8)
Alcohol abuse^b
At risk	2 (8.0)	16 (18.8)	18 (16.4)
No/low risk	5 (20.0)	23 (27.1)	28 (25.4)
Unknown	18 (72.0)	46 (54.1)	64 (58.2)
Cocaine use
Ever	5 (20.0)	15 (17.6)	20 (18.1)
Never	4 (16.0)	24 (28.2)	28 (25.5)
Unknown	16 (64.0)	46 (54.1)	62 (56.4)
Marijuana use
Ever	6 (24.0)	20 (23.5)	26 (23.6)
Never	2 (8.0)	16 (18.8)	18 (16.4)
Unknown	17 (68.0)	49 (57.7)	66 (60.0)
Chlamydia diagnosis^c
Yes	2 (11.8)	3 (5.3)	5 (6.8)
No	15 (88.2)	54 (94.7)	69 (93.2)
Gonorrhea diagnosis^c
Yes	0 (0)	1 (1.8)	1 (1.3)
No	18 (100.0)	56 (98.2)	74 (98.7)
On ART
Yes	24 (96.0)	76 (89.4)	100 (90.9)
No	1 (4.0)	9 (10.6)	10 (9.1)
CD4 cell count
≥200 cells/mm³	19 (76.0)	71 (84.5)	90 (82.6)
<200 cells/mm³	6 (24.0)	13 (15.5)	19 (17.4)
HIV viral load
<200 copies/ml	8 (32.0)	50 (59.5)	58 (53.2)
≥200 copies/ml	17 (68.0)	34 (40.5)	51 (46.8)
Currently sexually active
Yes	7 (28.0)	33 (38.8)	40 (36.4)
No	1 (4.0)	5 (5.9)	6 (5.5)
Unknown	17 (68.0)	47 (55.3)	64 (58.1)
Condom use, past six months
Safe sex	1 (4.0)	13 (15.3)	14 (12.7)
Unsafe sex	4 (16.0)	18 (21.2)	22 (20.0)
Unknown	19 (80.0)	54 (63.5)	73 (67.3)
No. of sexual partners, past six months
None	1 (4.0)	5 (5.9)	6 (5.4)
One partner	4 (16.0)	25 (29.4)	29 (26.4)
>1 partner	3 (12.0)	8 (9.4)	11 (10.0)
Unknown	17 (68.0)	47 (55.3)	64 (58.2)
Sex while using alcohol or drugs, past six months^d
Ever	2 (8.0)	10 (11.8)	12 (10.9)
Never	5 (20.0)	22 (25.9)	27 (24.6)

ART: antiretroviral therapy; CD4: CD4 T lymphocytes; HIV: human immunodeficiency virus; UAB: University of Alabama at Birmingham (Birmingham, AL).

Additional missing data: CD4 (no reinfection = 1); HIV viral load (no reinfection = 1); Chlamydia diagnosis (reinfection = 8, no reinfection = 28); gonorrhea diagnosis (reinfection = 7, no reinfection = 28).

^aDiagnosed by nucleic acid amplification test (NAAT).

^bDetermined by the AUDIT-C: Alcohol Use Disorders Identification Test-Consumption.

^cBy NAAT at the time of or within four weeks of initial T. vaginalis NAAT.

^dOnly among those who reported currently being sexually active (n = 40); missing data (no reinfection = 1).

Among the 110 HIV-infected women with a first episode of T. vaginalis infection by NAAT, 25/110 (22.7%) had a first episode of reinfection (58/110 [52.7%] received a second T. vaginalis NAAT during the study timeframe; reasons why the remainder of the women were not retested are unknown). Of these 25 women, 22 met the definition for reinfection based on a metronidazole prescription for treatment at their first infection and three met the definition based on an interval negative T. vaginalis NAAT. No significant differences regarding various characteristics (Table 1) were found between women retested (n = 58) versus those not retested (n = 52) except women who were retested were more likely to be on ART (96.6% versus 84.6%).

Overall, the median time to T. vaginalis reinfection (using a survival curve) was 16.1 months (95% CI: 11.1, 17.7). The total cumulative follow-up time for the 110 women was 674.9 months; thus, with 25 reinfections, the T. vaginalis incidence rate was 3.7 reinfections per 100 person-months (95% CI: 2.3, 5.2). Only five women were lost to follow-up.

Among the 25 women with a first episode of T. vaginalis reinfection (Table 1), the majority were African-American (88.0%) and between 25 and 40 years (60.0%). Over half (60.0%) were uninsured and 20.0% reported a history of cocaine use. The majority (96.0%) were on ART and 76.0% had a recent CD4 cell count ≥200 cells/mm³. In contrast, 68.0% had a recent HIV VL ≥200 copies/ml. The seven-day metronidazole dosing regimen had been prescribed for 20/25 (80.0%) of these women at their initial T. vaginalis episode while 2/25 (8.0%) had been prescribed the 2 g dose (data not shown). Treatment data were missing for three women. The median time to T. vaginalis reinfection was shorter in those with an HIV VL ≥200 copies/ml (12.9 months; 95% CI: 6.4, 17.7) compared to those with HIV VL <200 copies/ml (16.1 months; 95% CI: 16.1, undefined) (p = 0.05) (Figure 1).

Figure 1.

Survival curves for time to first episode of T. vaginalis reinfection stratified by HIV VL (<200 versus ≥200 copies/ml) among HIV-infected women at the UAB 1917 HIV clinic, August 2014–March 2016. UAB: University of Alabama at Birmingham (Birmingham, AL).

Table 2 shows the univariate and multivariable Cox proportional hazards analyses evaluating the association of various predictors with a first episode of T. vaginalis reinfection. In univariate analyses, only an HIV VL ≥200 copies/ml was found to approach statistical significance (HR = 2.26; 95% CI: 0.97, 5.29, p = 0.06). Adjusting for age and race, the association of HIV VL ≥200 copies/ml remained strong (aHR = 2.49; 95% CI: 0.99, 6.27, p = 0.05). These results remained robust in sensitivity analyses (Table 3).

Table 2.

Predictors of first episode of reinfection with trichomoniasis^a among HIV-infected women at the UAB 1917 HIV Clinic, August 2014–March 2016.

Predictor	N = 110n	Unadjusted hazard ratio^b (95% CI)	p-value	Adjusted hazard ratio^c (95% CI)	p-value
Age, years
≤40	58	1.39 (0.62, 3.14)	0.43	1.01 (0.42, 2.47)	0.98
>40*	52	1.00		1.00
Race
African-American	97	1.95 (0.45, 8.36)	0.37	2.41 (0.55, 10.61)	0.24
Caucasian*	13	1.00		1.00
Insurance type
Uninsured	43	1.39 (0.60, 3.19)	0.44	–	–
Insured^d,*	65	1.00
Alcohol abuse^e
At risk	18	0.71 (0.14, 3.64)	0.68	–	–
Unknown	64	1.58 (0.58, 4.30)	0.37	–	–
Low/no risk*	28	1.00
Cocaine use
Ever	20	1.92 (0.51, 7.17)	0.33	–	–
Unknown	62	1.58 (0.52, 4.79)	0.42	–
Never*	28	1.00
Marijuana use
Ever	26	2.82 (0.56, 14.20)	0.21	–	–
Unknown	66	2.23 (0.53, 10.10)	0.26	–
Never*	18	1.00
On ART
Yes	100	2.91 (0.39, 21.79)	0.30	–	–
No*	10	1.00
CD4 cell count (cells/mm³)
<200	19	0.89 (0.32, 2.41)	0.81	–	–
≥200*	90	1.00
HIV viral load (copies/ml)
≥200	48	2.26 (0.97, 5.29)	0.06	2.49 (0.99, 6.27)	0.05
<200*	61	1.00		1.00
Currently sexually active
Yes	40	1.70 (0.20, 14.28)	0.63	–	–
Unknown	64	1.95 (0.26, 14.88)	0.52	–	–
No*	6	1.00

ART: antiretroviral therapy; CD4: CD4 T lymphocytes; CI: confidence interval; HIV: human immunodeficiency virus; UAB: University of Alabama at Birmingham (Birmingham, AL).

Missing data: Insurance (no reinfection = 2); CD4 (no reinfection = 1); HIV viral load (no reinfection = 1).

*Reference category.

^aDiagnosed by nucleic acid amplification test (NAAT).

^bUnivariate Cox proportional hazards model.

^cMultivariable Cox proportional hazards model (N = 109).

^dInsured=private + public.

^eDetermined by the AUDIT-C: Alcohol Use Disorders Identification Test-Consumption.

Table 3.

Sensitivity analyses for the multivariable models predicting a first episode of T. vaginalis reinfection.

Multivariable model^a	Model 1^b		Model 2^c
Sample size (N)	58		109^d
Number of events (n)	25		30
Predictor	Adjusted hazard ratio^b (95% CI)	p-value	Adjusted hazard ratio^c (95% CI)	p-value
Age, years
≤40	0.73 (0.30–1.79)	0.49	1.02 (0.47–2.23)	0.73
>40*	1.00		1.00
Race
African-American	1.92 (0.43–8.50)	0.39	2.38 (0.60–9.55)	0.26
Caucasian*	1.00		1.00
HIV viral load (copies/ml)
≥200	2.24 (0.88–5.68)	0.09	2.03 (0.91–4.55)	0.12
<200*	1.00		1.00

CI: confidence interval; HIV: human immunodeficiency virus. * = Reference category.

^aCox proportional hazards model.

^bRestricting analysis to the 58 patients who were retested.

^cAssuming 10% (n = 5) of the 52 patients who were not retested could have been diagnosed with T. vaginalis reinfection had they had been retested. The 10% of the patients were randomly sampled with 1000 replications.

^dMissing data: HIV viral load (no reinfection = 1).

Discussion

Of 612 HIV-infected women at the UAB HIV clinic tested for T. vaginalis by NAAT, 110 (18.0%) had a T. vaginalis infection, and 25 (22.7%) of them had T. vaginalis reinfection by NAAT. Only an HIV VL > 200 was associated with T. vaginalis reinfection, both in univariate and multivariate analysis and this remained robust in sensitivity analyses.

The high proportion of T. vaginalis reinfection among HIV-infected women in this study could be due to several factors. Male (and female) partner(s) of the infected patients may not have been treated (the most common cause of reinfection), leading to reinfection. Concurrent treatment of T. vaginalis for all sexual partners is essential for cure and prevention of transmission and reinfection.⁸ Per the 2015 Centers for Disease Control STD Treatment guidelines,⁸ expedited partner therapy (EPT) may have a role in partner management for trichomoniasis. Although legal in the state of Alabama, EPT has not yet been routinely utilized by clinics, including the UAB HIV clinic, and did not impact this study. We were, therefore, unable to obtain data on partner treatment for T. vaginalis. Partner notification data for women with T. vaginalis were also not available.

Alternatively, it may be that the initial T. vaginalis infection diagnosed by NAAT in reinfected patients was inadequately treated due to noncompliance with therapy or persistent infection despite initial therapy. A randomized controlled trial of 270 HIV-infected women with T. vaginalis found that the one-time oral dose of 2 g metronidazole was less effective than metronidazole 500 mg orally twice daily for seven days.¹⁷ Of the 25 women in our study with T. vaginalis reinfection, 8.0% had received a 2 g dose of metronidazole at their first episode of T. vaginalis infection. This may have been suboptimal, leading to persistent T. vaginalis. ART (mostly nevirapine based) has also been associated with treatment failure among HIV-infected women with T. vaginalis, particularly those receiving the 2 g dose of metronidazole.¹⁸ In our study, only four women with an initial positive T. vaginalis NAAT during the study timeframe were on a nevirapine-containing ART regimen (data not shown). All four of these women were treated with the seven-day oral metronidazole regimen at the time of their initial infection and only one had T. vaginalis reinfection. Thus, being on a nevirapine-containing ART regimen did not substantially influence the results of our study. A subsequent study found that the ART effect on T. vaginalis treatment among HIV-infected women was not specific to regimens containing nevirapine.²³ In our study, 90.9% of HIV-infected women with a first episode of T. vaginalis infection and 96.0% of women with a first episode of T. vaginalis reinfection reported ART use, which may have influenced T. vaginalis treatment outcomes, particularly among women receiving the 2 g dose of metronidazole.

A prior study of 411 HIV-infected women in New Orleans, LA with trichomoniasis based on wet mount, Pap smear, or urinalysis found 36% of the women to be reinfected during a three-year follow-up with an incidence rate of 16.4 reinfections per 100 person-years.²⁴ A significant predictor of reinfection included history of another STI. Pregnancy during follow-up was negatively associated with reinfection. Given the relatively low sensitivity of the diagnostic methods used, the authors acknowledged that their reported incidence might be an underestimate of the true rate of new T. vaginalis infections in HIV-infected women. In addition, they hypothesized that women who became pregnant during follow-up may have been in more stable monogamous relationships or less likely to be sexually active due to pregnancy. Our proportion of T. vaginalis reinfections (43.1%) is perhaps larger than that seen in this study due to the better sensitivity of the T. vaginalis NAAT.¹¹ Pregnancy data on women in our study were not available. Thus, we were unable to determine if a negative association between pregnancy and T. vaginalis reinfection was observed in our sample.

Another study of HIV-infected women in St Louis, MO found that T. vaginalis reinfection (based on wet mount) was detected in 18.4% among 125 women attending at least one follow-up visit over a mean of 16 months.²⁵ Cumulative prevalence of all STIs at follow-up was high (38.4%) indicating participation in high-risk sexual behaviors despite HIV. Similarly, Capps et al.⁴ found that African-American race, history of sex with an IVDU, and history of STIs were associated with increased risk of reinfection with STIs (particularly trichomoniasis) among 323 HIV-infected women. In this study, 25% of the women had a new or recurrent STI over a median of 2.1 years. Interestingly, in our study we did not find that African-American race was associated with T. vaginalis reinfection, perhaps due to the small sample. We were not able to include IVDU in our analysis due to a large amount of missing data.

Of note, the median time to T. vaginalis reinfection was approximately three months shorter in those with an HIV VL ≥200 copies/ml compared to those with HIV VL <200 copies/ml in this study. Because this is a high-risk population, it could be that we should consider retesting HIV-infected women with T. vaginalis and an HIV VL >200 copies/ml more frequently than three months after initial therapy and annually thereafter. In addition, univariate and multivariable analyses found that an HIV VL of ≥200 copies/ml approached statistical significance for being a predictor of T. vaginalis reinfection. An elevated HIV VL could be a marker for other high-risk behaviors that may be associated with T. vaginalis reinfection, including nonadherence to ART therapy and T. vaginalis therapy. Interestingly, however, 96.0% of those patients reinfected with T. vaginalis reported current use of ART although only 32% had an HIV VL <200 copies/ml. T. vaginalis reinfection was not associated with lower absolute CD4 cell count, similar to prior studies,^5,26,27 and does not appear to be an opportunistic infection in HIV-infected patients.

This study has limitations. First, this was a convenience sample of HIV-infected women at our HIV clinic. It is not possible to know the true incidence of T. vaginalis reinfection among HIV-infected women at this clinic as not every patient was retested by NAAT during the study timeframe. Thus, the results should be interpreted cautiously. Reasons for the low rates of retesting are also unknown. Second, the retrospective nature of this study and the database used limited the amount of detailed sexual history and sexual behavior data obtained (including breakdown of specific genitourinary sites positive for T. vaginalis among infected patients, pregnancy data, etc.). Similarly, reasons why patients were reinfected are unknown (i.e. due to new infection from a new partner, treatment noncompliance, persistent infection due to suboptimal therapy, etc.). If due to treatment noncompliance or persistent infection due to suboptimal therapy, it is possible that some of the cases may not truly represent reinfection with T. vaginalis and instead represent prevalent infection. Finally, the results are from a convenience sample of HIV-infected women from one urban HIV clinic in the southern U.S. and may not be generalizable to other HIV-infected populations.

Nevertheless, this study has strengths including the use of the T. vaginalis NAAT in a high-risk population of HIV-infected women. The results demonstrate a high T. vaginalis reinfection rate after an initial episode of infection among HIV-infected women. This is likely contributing to further T. vaginalis and HIV transmission among this sample of women. Increased awareness of guidelines for T. vaginalis testing and treatment among HIV providers is needed. Standardization of T. vaginalis screening with the highly sensitive T. vaginalis NAAT and appropriate therapy for T. vaginalis-infected women with the seven-day oral metronidazole dosing regimen should be implemented. Consistent retesting practices by HIV providers for women with T. vaginalis are also needed, particularly for women with virological failure (HIV VL ≥200 copies/ml). Finally, enhanced disease control efforts for trichomoniasis are needed in HIV-infected populations.

Footnotes

Authors’ contribution

CAM, GAB, ART, and JRS were responsible for conception and design. GAB, ART, KH, and EFE were responsible for acquisition of data. CAM, GAB, ART, EFE, and JRS were responsible for analysis or interpretation of data. CAM and ART were responsible for drafting of the manuscript. CAM, GAB, ART, KH, EFE, and JRS were responsible for revising the manuscript for intellectual content. All authors were responsible for the final version of the manuscript to be published.

Acknowledgments

The authors would like to thank Anuj Kapil, Mohit Varshney, and Suneetha Thogaripally for their assistance with data management. This study was presented, in part, as poster presentation #34 at the 2016 Infectious Diseases Society of Gynecology Annual Meeting in Annapolis, MD on 11 August 2016.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Christina A Muzny has consulted for Lupin Pharmaceuticals. Greer A Burkholder has consulted for Definicare LLC and Medscape and received research support from Bristol-Myers Squibb and Amgen, Inc. Ellen F Eaton has received research support from the Bristol Myers Squibb Virology fellowship Grant and Merck & Co. Jane R Schwebke has been a consultant for and received research support from Lupin Pharmaceuticals, Hologic, BD Diagnostics, Cepheid, Quidel, Toltec, and StarPharma. The other authors have no potential conflicts of interest.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Christina A Muzny is supported by grant K23AI106957 from the National Institute of Allergy and Infectious Diseases and Greer A Burkholder is supported by grant K23HL126570 from the National Heart, Lung, and Blood Institute. Ellen F Eaton is supported by grant K12HS023009 from the Agency for Health Research and Quality. Biostatistical support for Ashutosh R Tamhane was made possible by an NIH award supporting the UAB Center for AIDS Research (CFAR) (P30 AI027767).

References

McClelland

Sangare

Hassan

et al . Infection with Trichomonas vaginalis increases the risk of HIV-1 acquisition. J Infect Dis 2007; 195: 698–702.

Cherpes

Wiesenfeld

Melan

et al . The associations between pelvic inflammatory disease, Trichomonas vaginalis infection, and positive herpes simplex virus type 2 serology. Sex Transm Dis 2006; 33: 747–752.

Cotch

Pastorek

2nd Nugent

et al . Trichomonas vaginalis associated with low birth weight and preterm delivery. The Vaginal Infections and Prematurity Study Group. Sex Transm Dis 1997; 24: 353–360.

Capps

Peng

Doyle

et al . Sexually transmitted infections in women infected with the human immunodeficiency virus. Terry Beirn community programs for clinical research on AIDS (CPCRA). Sex Transm Dis 1998; 25: 443–447.

Sorvillo

Kovacs

Kerndt

et al . Risk factors for trichomoniasis among women with human immunodeficiency virus (HIV) infection at a public clinic in Los Angeles County, California: implications for HIV prevention. Am J Trop Med Hyg 1998; 58: 495–500.

Monteiro

Bachmann

Desmond

et al . Incidence and risk factors for sexually transmitted infections among women in an Alabama HIV clinic. AIDS Res Hum Retroviruses 2004; 20: 577–583.

Alcaide

Feaster

Duan

et al . The incidence of Trichomonas vaginalis infection in women attending nine sexually transmitted diseases clinics in the USA. Sex Transm Infect 2016; 92: 58–62.

Workowski

Bolan

and Centers for disease C, prevention. Sexually transmitted diseases treatment guidelines, 2015. MMWR Recomm Rep 2015; 64: 1–137.

Krieger

Tam

Stevens

et al . Diagnosis of trichomoniasis. Comparison of conventional wet-mount examination with cytologic studies, cultures, and monoclonal antibody staining of direct specimens. JAMA 1988; 259: 1223–1227.

10.

Hobbs

Sena

Swygard

et al . Trichomonas vaginalis and trichomoniasis. In: Holmes

Sparling

Stamm

et al. (eds) Sexually transmitted diseases. 4th ed. New York: McGraw-Hill, 2008, pp.771–793.

11.

Schwebke

Hobbs

Taylor

et al . Molecular testing for Trichomonas vaginalis in women: results from a prospective U.S. clinical trial. J Clin Microbiol 2011; 49: 4106–4111.

12.

Van Der Pol

Williams

Taylor

et al . Detection of Trichomonas vaginalis DNA by use of self-obtained vaginal swabs with the BD ProbeTec Qx assay on the BD viper system. J Clin Microbiol 2014; 52: 885–889.

13.

Schwebke

Gaydos

Davis

et al . Clinical evaluation of the Cepheid Xpert(R) TV assay for detection of Trichomonas vaginalis with prospectively collected female and male specimens. J Clin Microbiol 2017; 56(2).

14.

2015 CDC STD surveillance, https://www.cdc.gov/std/stats15/std-surveillance-2015-print.pdf (2015, accessed 1 March 2018).

15.

HIV in the Southern United States, https://www.cdc.gov/hiv/pdf/policies/cdc-hiv-in-the-south-issue-brief.pdf (2016, accessed 1 March 2018).

16.

Muzny

Burkholder

Fry

et al . Trichomonas vaginalis nucleic acid amplification testing at an urban HIV clinic. Sex Transm Dis 2016; 43: 483–488.

17.

Kissinger

Mena

Levison

et al . A randomized treatment trial: single versus 7-day dose of metronidazole for the treatment of Trichomonas vaginalis among HIV-infected women. J Acquir Immune Defic Syndr 2010; 55: 565–571.

18.

Balkus

Richardson

Mochache

et al . A prospective cohort study comparing the effect of single-dose 2 g metronidazole on Trichomonas vaginalis infection in HIV-seropositive versus HIV-seronegative women. Sex Transm Dis 2013; 40: 499–505.

19.

Willig

Aban

Nevin

et al . Darunavir outcomes study: comparative effectiveness of virologic suppression, regimen durability, and discontinuation reasons for three-class experienced patients at 48 weeks. AIDS Res Hum Retroviruses 2010; 26: 1279–1285.

20.

Burkholder

Tamhane

Appell

et al . Short communication: viral suppression is associated with increased likelihood of colorectal cancer screening among persons living with HIV/AIDS. AIDS Res Hum Retroviruses 2015; 31: 519–524.

21.

Saunders

Aasland

Babor

et al . Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption–II. Addiction 1993; 88: 791–804.

22.

Newcombe

Humeniuk

Ali

Validation of the World Health Organization Alcohol, Smoking and Substance Involvement Screening Test (ASSIST): report of results from the Australian site. Drug Alcohol Rev 2005; 24: 217–226.

23.

Adamski

Clark

Mena

et al . The influence of ART on the treatment of Trichomonas vaginalis among HIV-infected women. Clin Infect Dis 2014; 59: 883–887.

24.

Niccolai

Kopicko

Kassie

et al . Incidence and predictors of reinfection with Trichomonas vaginalis in HIV-infected women. Sex Transm Dis 2000; 27: 284–288.

25.

Bersoff-Matcha

Horgan

Fraser

et al . Sexually transmitted disease acquisition among women infected with human immunodeficiency virus type 1. J Infect Dis 1998; 178: 1174–1177.

26.

Cu-Uvin

Jamieson

et al . Prevalence, incidence, and persistence or recurrence of trichomoniasis among human immunodeficiency virus (HIV)-positive women and among HIV-negative women at high risk for HIV infection. Clin Infect Dis 2002; 34: 1406–1411.

27.

Magnus

Clark

Myers

et al . Trichomonas vaginalis among HIV-Infected women: are immune status or protease inhibitor use associated with subsequent T. vaginalis positivity? Sex Transm Dis 2003; 30: 839–843.