1. RESULTS: After Discordant Analysis RESULTS: Discordant Testing
Board # 103
Andrew Hardick
The Johns Hopkins University
Division of Infectious Diseases
Department of Medicine
1159 Ross Bldg.
720 Rutland Ave.
Baltimore, MD 21205
(410) Fax (410)
Comparison of Cervical Samples to Vaginal Samples for the Detection of Trichomonas vaginalis Using a GEN-PROBE Research Assay for rRNA Andrew Hardick1, Billie Jo Wood1, Justin Hardick1, Rebecca Miller1, Mathilda Barnes1, Thomas C. Quinn1,2 Charlotte Gaydos1 1. The Johns Hopkins University, Baltimore, MD. 2. NIAID, NIH, Bethesda, MD.
RESULTS
Vaginal TMA TV
INTRODUCTION + -
RESULTS: After Discordant Analysis 15 2 5 71 +
Trichomonas vaginalis (TV), a protozoan parasite, is the most prevalent non-viral STI in the world
Estimated that in the U.S. alone, 3 million women contract the infection every year
Is a major cause of vaginitis and non-gonococcal urethritis
In women, infection ranges from asymptomatic to profound, acute, inflammatory disease
T. vaginalis usually infects the squamous epithelium of the genital tract
Can be found in the urethra, cervix, fallopian tubes, and pelvic organs
Vaginal swabs are traditionally used for diagnosis of T. vaginalis
Chlamydia and gonorrhea can both be detected in cervical swab specimens
Cervical TMA TV -
Infected Patients
Infected Patients
Overall Agreement: 92.5%
Kappa Statistic: 0.764
(strength of the agreement considered to be “good”) + - + - 17 3 1 72 16 1 2 74 + +
Vaginal TMA TV
Cervical TMA TV
Resolution of discordant TMA TV results to true positives by alternative testing
Of 5 TMA TV assay vaginal pos./ cervical neg. results:
1 sample was:
Wet preparation TV positive
LightCycler PCR TV positive
1 other sample was:
PCR alternative primer set TVK 3/ 4 positive for TV
Confirmed 2 out of 5 discordant samples as true positives - -
Vaginal TMA TV
Sensitivity % (17/18)
(95% CI: 70.6%;99.7%)
Specificity % (72/75)
(95% CI: 87.9%; 98.9%)
PPV % (17/20)
NPV % (72/73)
Cervical TMA TV
Sensitivity % (16/18)
(95% CI: 63.9%;98.0%)
Specificity % (74/75)
(95% CI: 91.7%; 99.9%)
PPV % (16/17)
NPV % (74/76)
DISCUSSION
Resolution of discordant TMA TV results to true positives by alternative testing
Of 2 TMA TV assay vaginal neg./ cervical pos. results:
1 sample was:
LightCycler PCR TV positive
PCR alternative primer set TVK 3/ 4 positive for TV
Confirmed 1 out of 2 discordant samples as true positives
Prevalence of T. vaginalis was high, 19.4% (18/93)
Overall agreement between cervical and vaginal samples to true T. vaginalis infected patients is high 94.1% (16/17)
This suggests that either vaginal or cervical samples would be sufficient for the detection of T. vaginalis
Objective
Determine if cervical swabs could serve as an adequate sample type for detection of Trichomonas vaginalis
METHODS
Specimens for this study were part of a larger study comparing a GEN-PROBE* APTIMA* format research assay (TMA* TV) to a research LightCycler* PCR assay for the detection of T. vaginalis
93 cervical and vaginal sample pairs were tested by the TMA TV assay
Cutoff for a positive result was ≥40,000 RLU (Relative Light Units)
A true positive was defined as 2 positive results by any 2 tests
Discordant samples were tested by:
Research LightCycler PCR TV assay
PCR for TV using alternative gene target primers TVK 3 and 4
Direct wet preparation for TV was performed at Clinics at time of sample collection
RESULTS: Discordant Testing
CONCLUSIONS
Vag +, CX -
TMA TV
Vag CX
Light-Cycler
Alt. PCR
TVK 3/ 4
Wet Prep.
1 True pos. + _
2 True pos. NA 3 4 5
Vag-, CX+ 2
T. vaginalis can be detected in both vaginal and cervical samples
Kappa statistic indicates the agreement between vaginal and cervical samples was “good”
No statistical difference in sensitivity and specificity estimates for vaginal and cervical specimens since 95% CI overlap
The TMA TV assay could be useful for detection of T. vaginalis in clinical settings
GEN-PROBE, APTIMA and TMA are trademarks of Gen-Probe Incorporated; LIGHTCYCLER is a trademark of Roche Diagnostics GmbH