Can mailed swab samples be dry-shipped for the detection of Chlamydia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis by nucleic acid amplification tests?

Abstract

Background

Dry-shipped and mailed vaginal swabs collected at home have been used in research studies for the detection of C. trachomatis (CT), N. gonorrhoeae (GC), and Trichomonas vaginalis (TV) by nucleic acid amplification tests (NAATs) in screening programs. A verification study was performed to compare the limit of detection of CT, GC, and TV on swabs that were dry-shipped to paired swabs that were wet-shipped in transport media through the U.S. mail.

Methods

The Centers for Disease Control and Prevention prepared inocula in sterile water to mock simulated urogenital swabs with high to low concentrations of CT and GC. Replicate swabs were inoculated with 100µl of dilutions, were dry transported or placed into commercial transport media (“wet”) for mailing for NAAT testing. The University of Alabama prepared replicate concentrations of TV, which were similarly shipped and tested by NAAT.

Results

All paired dry and wet swabs were detectable for CT. For GC, all paired dry and wet swabs were detectable for GC at concentrations ≥10³. At 10² and 10 CFU/ml, the 10 replicate GC results were variably positive. For TV, wet and dry shipped concentrations > 10² TV/ml tested positive, while results at 10 TV/ml were negative for dry swabs. Holding replicate dry swabs at 55°C 5 days before testing did not affect results.

Conclusion

NAATs were able to detect CT, GC, and TV on dry transported swabs. Using NAATs for testing home-collected, urogenital swabs mailed in a dry state to a laboratory may be useful for outreach screening programs.

Introduction

Chlamydia trachomatis (CT) is the most commonly reported communicable disease in the U.S., occurring most often among adolescent and young adult females. The Centers for Disease Control and Prevention (CDC) estimates there are approximately 3 million new cases each year (Centers for Disease Control and Prevention, 2010a). Acute chlamydia infections often have no symptoms, leaving many cases untreated. Serious sequelae such as pelvic inflammatory infections (PID) can be associated with untreated CT infections. The U.S. Preventive Services Task Force and major medical professional organizations recommend an annual screening test for chlamydia for all sexually active adolescents and young women 25 years of age and under, for all pregnant women, and for women at high risk (U.S.Preventive Services Task Force, 2007). Yet chlamydia screening remains an underutilized clinical preventive service with less that 50 percent of eligible women in commercial or Medicaid health plans being screened annually (National committee for Quality Assurance, 2010).

Likewise, Neisseria gonorrhoeae (GC) infections are highly prevalent with over 700,000 new infections every year in the U.S. and are also associated with the serious sequelae as CT (Centers for Disease Control and Prevention, 2010a). Trichomonas vaginalis (TV) infections are highly prevalent with estimates of 7–8 million infections annually in the United States and 180 million globally (WHO, 2010) representing the most common curable STI in sexually active women (Van Der Pol et al., 2005; Weinstock et al., 2004). Trichomonas infections have been associated with poor reproductive outcomes such as low birth weight and premature birth (Cotch et al., 1997; Schwebke and Burgess, 2004). Screening for all of these infections in women of reproductive age can impact these epidemics but many barriers exist, which could be overcome by offering home collection of urogenital samples to women.

Despite national screening recommendations and concerted efforts at control, rates of CT and GC among young women remain high, particularly amongst young minority women. Expanded outreach screening programs, beyond traditional clinics, are needed to reach young sexually active persons at risk. The advent of non-invasive specimens for CT and GC detection using NAAT technology has enabled the expansion of screening to non-clinical settings such as schools, mobile vans, and other settings outside the clinic. Because of the increasing prevalences of CT and GC, public health programs are needed that can implement novel screening programs to improve access to testing. Although there are no national surveillance data available for TV, because these infections are not reportable to CDC by law, research studies have demonstrated high prevalences in both young and older populations, as well as in minority groups (Van Der Pol et al., 2005; Weinstock et al., 2004).

Home collection of urogenital samples has been promoted by public health officials (Hobbs et al., 2008). Submission of such self-obtained vaginal samples (SOVS) collected at home could compliment screening programs, as well as remove barriers that women and men face in getting tested for STIs (Gaydos C.A. et al., 2009; Gaydos and Rompalo A.M., 2002). Since neither home-collected nor dry-shipped urogenital swab samples are cleared by the Federal Drug Administration (FDA) for use with commercial nucleic acid amplification tests (NAATs), we performed a verification study to compare the limit of detection of CT, GC and TV titered concentrations of organisms on swabs that were dry-shipped versus matching swabs that were transport media (wet)-shipped through the U.S. mail to a laboratory for testing.

Methods

Chlamydia and Gonorrhea Organisms

C. trachomatis genotype E was selected for this study based on genetic analysis of the outer membrane protein gene at the CDC and is a strain commonly found in genito-urinary tract infections among women (Millman et al., 2004). A single strain was tested since this was a pilot study and this strain is a very common serotype. The CT strain was grown in cultured buffalo green monkey kidney (BGMK) cells and harvested by differential centrifugation to remove cellular debris. The strains was serially diluted in sucrose-phosphate-glutamic acid (SPG) buffer and titrated for determination of the infectious end-point by calculation of inclusion-forming units (IFU). Inocula to spike simulated urogenital swabs were prepared in sterile water and represented dilutions ranging from 10⁷ to 1 IFU per ml, which were estimated by CDC to be representative of most clinical infections. (Michel CE, et al. 2007)

A clinical isolate of N. gonorrhoeae was selected on the basis of biochemical reference testing at the CDC. A single clinical strain was selected as being representative of current circulating strains since this organism is always genetically changing. The N. gonorrhoeae strain was grown on Modified Thayer-Martin selective medium incubated at 35°C for 24 hours under an atmosphere containing 5% CO₂. The strain was serially diluted in Mueller-Hinton broth and titrated for determination of the infectious end-point by calculation of colony forming units (CFU). Inocula were prepared in sterile water and represented dilutions ranging from 10⁷ to 10 CFU per ml.

Swabs

Becton Dickinson, BBL culture EZ, #220144 swabs were used. Presently, self sampling swabs collected at clinics are placed into manufacturers’ liquid (“wet”) transport media for shipment to the laboratory. We refer to these swabs for standard techniques as “wet”. There many research studies that have successfully used shipment of urogenital swabs in the dry state without any transport media. (Gaydos CA et al. 2002, 2006, 2009, Masek BJ et al 2009). We refer to these as “dry” swabs. For CT, there were replicate (2) “dry” swabs and matching “wet” swabs of each of concentration (10⁷, 10⁵, and 10³ IFU/ml, and 10 duplicates of 10 and 1 IFU/ml). For GC, there were replicate (2) “dry” and matching “wet” swabs each concentrations (10⁷, 10⁵, and 10³ IFU/ml, and 10 duplicates of 10 IFU/ml). Replicate swabs were inoculated with 100 µl of appropriate CT and GC concentrations. Thus, 100 ul of the 10⁷ /ml concentration yielded a swab representing 10⁶ organisms/swab for that test. One replicate swab was placed in the holding tube for “dry” transport, one replicate swab was placed into the holding tube for “heated transport”, and the matching swab made from the same concentration was placed into specimen transport media (STM) for wet transport, Aptima Combo 2 (Gen-Probe Inc., San Diego, CA). The replicate dry-shipped swabs were held at 55°C for 5 days before testing (see below). Also included were 8 swabs with only STM as negative controls.

Additionally, for GC, additional experiments with sets of the lower concentrations (10² and 10 CFU/ml) were repeated and transported two additional times.

Trichomonas organisms

Likewise, the University of Alabama at Birmingham (UAB) laboratory prepared concentrations of a clinical strain of TV, representing currently circulating strains, which was grown in Diamonds Media (Remel). TV organisms were counted in a hemocytometer counting chamber and dilutions were prepared to achieve various concentrations by serially diluting TV 10-fold in Diamonds Media and titrating for determination of the TV organisms/ml. Inocula were used to mock two simulated vaginal swabs from 10⁷ to 10² organisms/ml. There were 4 replicate swabs for concentrations of 10¹ and 10 replicates of 1 TV/ml. Amount placed on each swab was 100µl of each concentration. Negative control swabs containing STM transport media (Gen-Probe) were included for testing. The replicate dry-shipped swabs were held at 55°C for 5 days before testing, as for the CT and GC experiments.

Shipping Conditions

Matching swabs containing the various concentrations of organisms (two “dry” and one “wet” swab of each dilution) were placed in sealed biohazard bags and then into individual mailing envelopes at the CDC in Atlanta (CT and GC) or UAB (TV) and sent via United States Postal Service, first class, to The International STD Research Laboratory, Johns Hopkins University, Baltimore, MD.

Testing

Upon receipt, one of the “dry” swabs were placed into Aptima Combo 2 transport media for testing. These simulated urogenital swabs were (“wet” and one “dry” transported) were tested for CT and GC at the same time according to the routine Aptima Combo2 (Gen-Probe, Inc. San Diego, CA) protocol. Similarly, the same algorithm was followed for the TV swabs, which were tested by the ASR (Analyte Specific Reagents) Aptima TV assay (Gen-Probe).

High temperature environment

Upon receipt, for all three organisms, the replicate set of identical dry-shipped swabs with the various organism concentrations were heated at 55°C in an incubator for 5 days before being placed into transport media and tested, in order to simulate environmental transport conditions during higher ambient shipping temperatures (Papp et al., 2007).

Results

C. trachomatis

All concentrations of paired dry and wet swabs and heated dry swabs were detectable as positive for CT, except for one equivocal result for a 1 IFU/ml dry heated swab (Table 1).

Table 1.

Comparison of C. trachomatis (CT) concentrations from simulated swabs that were shipped in a dry or wet state and after heating the swab at 55°C for 5 days.

CT Concentration IFU/ml	CT wet	RLUa (×1000)	CT dry	RLUa (×1000)	CT dry heatedb	RLUa (×1000)
10 7	POS	801	POS	1306	POS	1373
10 7	POS	1328	POS	1272	POS	1338
10 5	POS	1303	POS	1244	POS	1329
10 5	POS	1307	POS	1273	POS	1325
10 3	POS	1248	POS	1148	POS	1320
10 3	POS	1190	POS	1168	POS	1328
10	POS	1161	POS	1128	POS	941
10	POS	1188	POS	566	POS	1318
10	POS	1178	POS	1154	POS	1049
10	POS	1271	POS	1229	POS	1173
10	POS	1081	POS	1145	POS	1005
10	POS	722	POS	756	POS	986
10	POS	1147	POS	1117	POS	1112
10	POS	1129	POS	1159	POS	1146
10	POS	1131	POS	1106	POS	1133
10	POS	1133	POS	1159	POS	921
1	POS	837	POS	507	POS	1169
1	POS	1026	POS	1023	POS	1093
1	POS	1111	POS	975	POS	1009
1	POS	1039	POS	938	POS	1171
1	POS	1007	POS	896	POS	1091
1	POS	1091	POS	1056	EQUIV	59
1	POS	1154	POS	1054	POS	1009
1	POS	1111	POS	981	POS	949
1	POS	1005	POS	935	POS	1056
1	POS	1058	POS	1000	POS	1133

^aRLU = Relative Light Unit measure for the Gen-Probe APTIMA test. Positive C. trachomatis specimens have RLU’s >100 thousands.

^bHeated test was conducted after the specimen was maintained at 55°C for 5 days.

Negative controls (16) tested were all negative.

N. gonorrhoeae

All paired dry and wet swabs were detectable for GC at all concentrations at and above 10 CFU/ml. For dry heated swabs, one 10³/ml was negative and at 10 CFU/ml results varied (5 positives, 3 equivocal, and 2 negatives) (Table 2).

Table 2.

Comparison of N. gonorrhoeae (GC) concentrations in Colony forming units (CFU)/ ml from simulated swabs that were shipped in a dry or wet state and after heating the swab at 55°C for 5 days.

GC Concentration CFU/ml	GC wet	RLUa (×1000)	GC dry	RLUa (×1000)	GC dry heatedb	RLUa (×1000)
10 7	POS	1302	POS	1205	POS	1313
10 7	POS	1288	POS	1324	POS	1328
10 5	POS	1325	POS	1282	POS	1244
10 5	POS	1313	POS	1306	POS	1176
10 3	POS	1275	POS	1912	NEG	56
10 3	POS	1285	POS	553	POS	173
10	POS	1228	POS	896	POS	214
10	POS	1202	POS	578	POS	276
10	POS	1167	POS	1008	POS	293
10	POS	1233	POS	862	NEG	46
10	POS	1273	POS	824	POS	223
10	POS	1257	POS	864	EQU	116
10	POS	1247	POS	399	EQU	112
10	POS	1234	POS	559	EQU	82
10	POS	1196	POS	707	POS	166
10	POS	1230	POS	854	NEG	54

^aRLU = Relative Light Unit measure for the Gen-Probe APTIMA test. Positive N. gonorrhoeae specimens have RLU’s >150 thousands.

^bHeated test was conducted after the specimen was maintained at 55°C for 5 days.

Repeat experiment data not in table. For 10² CFU/ml Dry heated, separate repeat experiments: Set A: 8 samples were negative, 2 were equivocal; Set B: 3 were positive, 2 were equivocal, and 5 were negative.

Repeat experiment data not in table. For 10¹ CFU/ml Dry heated, separate repeat experiments: Set A: 5 samples were positive, 3 were equivocal, and 2 were negative; Set B: 10/10 samples were negative.

Negative controls (16) tested were all negative.

Two different repeat experiments, which were conducted later for GC for dry heated swabs at 10² and 10 CFU/ml concentrations [concentration per swab corresponded to 10 and 1 CFU/swab each] demonstrated some variability. For the GC concentration for dry heated swabs at 10² CFU/ml [10 CFU/swab] for the first repeat experiment (Set A), there were 8 negative and 2 equivocal results. For the second repeat experiment at 10² CFU/ml [10 CFU/swab] (Set B), there were 3 positive results, 2 equivocal results, and 5 negative results.

For the GC concentration for dry heated swabs at 10¹ CFU/ml [1 CFU/swab] for the first repeat experiment (Set A), there were 5 positive, 3 equivocal results and 2 negative result. For the second repeat experiment at 10¹ CFU/ml [1 CFU/swab] (Set B) there were 10/10 negative results.

T. vaginalis

All concentrations from 10⁷ to 10³ TV/ml were positive in the wet, dry, and dry/heated swabs (Table 3). For 10² TV/ml, 2/2 were positive for wet transported swabs and 1 of 2 each were positive for both dry and dry/heated transported swabs. For the 4 replicates of TV at 10/ml, 4 were positive for the wet transported swabs and none were positive for dry or dry/heated transported swabs. This concentration corresponded to 1 TV per swab. For the 10 replicates at 1 TV/ml (corresponding to 0.1 TV per swab), 2/10 were positive in wet transported swabs and none were positive in dry or dry/heated swabs.

Table 3.

Comparison of T. vaginalis (TV) concentrations from simulated swabs that were shipped in a dry or wet state and after heating the swab at 55°C for 5 days.

TV Concentration/ml	TV wet	RLUa (×1000)	TV dry	RLUa (×1000)	TV dry heatedb	RLUa (×1000)
10 7	POS	1580	POS	1621	POS	4452
10 7	POS	1619	POS	1597	POS	4172
10 6	POS	1597	POS	1488	POS	3847
10 6	POS	1606	POS	1458	POS	2802
10 5	POS	1606	POS	1252	POS	2510
10 5	POS	1553	POS	1284	POS	2710
10 4	POS	1320	POS	908	POS	1436
10 4	POS	1370	POS	1002	POS	1427
10 3	POS	1076	POS	325	POS	349
10 3	POS	1099	POS	443	POS	520
10 2	POS	364	POS	71	POS	104
102	POS	718	NEG	11	NEG	36
10	POS	261	NEG	19	NEG	26
10	POS	257	NEG	25	NEG	2
10	POS	347	NEG	13	NEG	18
10	POS	410	NEG	32	NEG	9
1	POS	67	NEG	1	NEG	1
1	NEG	30	NEG	3	NEG	1
1	NEG	56	NEG	6	NEG	1
1	NEG	20	NEG	3	NEG	8
1	NEG	32	NEG	2	NEG	4
1	NEG	31	NEG	1	NEG	1
1	NEG	33	NEG	2	NEG	3
1	NEG	40	NEG	2	NEG	3
1	POS	115	NEG	1	NEG	1
1	NEG	23	NEG	2	NEG	1

^aRLU = Relative Light Unit measure for the Gen-Probe TRIC APTIMA test. Positive T. vaginalis specimens have RLUs >60 thousands.

^bHeated test was conducted after the specimen was maintained at 55°C for 5 days.

Negative controls (10) tested were all negative.

Discussion

The results of this series of experiments demonstrated that shipping simulated swabs with differing organism concentrations in a dry state provided identical results for chlamydia and similar results for gonorrhea and trichomonas. It was only at the very lowest concentrations which corresponded to 10 and 1 CFU/dry heated swabs for gonorrhea and 1 and 0.1 trichomonas/swab that variability was observed. In all likelihood, these levels were close to the lowest level of the limit of detection for the assays tested, since only 2/10 wet transport swabs at the 1 organism/ml [0.1 organism/swab] tested positive for trichomonas.

Vaginal swabs have been shown to perform as well as or better than endocervical specimens and better than urine specimens when screening for chlamydia and gonorrhea (Schachter et al., 2003; Shafer et al., 2003). Self-collected vaginal swabs can save significant clinician time, if the patient does not require a pelvic exam because of symptoms.

In July 2006, the NIH conducted a workshop to review published data and practices supporting the use of self-obtained vaginal swabs (SOVS) as specimens for the diagnosis of sexually transmitted infections (Hobbs et al., 2008). The expert panel concluded that such specimens were well accepted performed as well as or better than other specimen types for C. trachomatis and N. gonorrhoeae detection. In addition, citing successful studies of home collected, mailed SOVS (Cook et al., 2007; Gaydos et al., 2006), the panel stated that the validation of these specimens for home use was “urgently needed”. Home collection of urogenital samples sent by postal mail has been successfully used in many other prior research studies. (Gaydos C.A. et al., 2009; Grasek et al., 2010; Novak and Karlsson, 2006; Ostergaard et al., 2000; Ostergaard et al., 1998; van den Broek et al., 2010).

Dry shipped swabs for the detection of chlamydia and gonorrhea have been used previously in a STD clinic in a military population. There was no statistical difference in the sensitivity between wet or dry shipped swabs (Gaydos et al., 2002). Additionally, the strand displacement amplification assay (ProbeTec, Becton Dickinson, Sparks, MD) has achieved FDA clearance for dry-transported cervical swabs (Van Der Pol et al., 2001). Because of the earlier FDA clearance of dry-transported swabs, we have used dry-shipped vaginal swabs, collected at home after Internet recruitment, for home-collection kits in a research program (Gaydos C.A. et al., 2009; Gaydos C.A. et al., 2011; Masek et al., 2009). Pilot studies before the initiation of this program were performed in our laboratory to demonstrate the stability of chlamydia on dry swabs at room temperature (Hardick et al., 2005).

This method of shipping samples collected at home in a dry state offers an easy way to accomplish screening, while affording convenience, confidentiality, and privacy. For women for whom re-screening is desired, women clearly preferred home collection (75.5%) (Grasek et al., 2010). Since CDC recommends rescreening at 3 months after treatment for chlamydia or gonorrhea infection, the home collection with ry transport method has much appeal (Centers for Disease Control and Prevention, 2010b).

Despite national screening recommendations and concerted efforts at control, rates of CT among young women remain high, particularly amongst young minority women. Expanded outreach screening programs, beyond traditional clinics, are needed to reach young sexually active persons at risk. The advent of non-invasive specimens for CT and GC detection using NAAT technology has enabled the expansion of screening to non-clinical settings. Although there are no national surveillance data available for TV, because these infections are not reportable, research studies have demonstrated high prevalences in both young and older populations. (Gaydos C.A. et al., 2011; Weinstock et al., 2004). Because of the increasing prevalences of CT, GC, and TV, public health programs such as home collection sampling are needed that can complement novel screening programs to improve access to testing.

Since chlamydia screening has been identified as one of the top seven priorities in preventive health, based on health impact, cost effectiveness, and low current uptake, it seems reasonable to explore more home collection methods in order to facilitate asymptomatic screening methods for chlamydia, as well as for gonorrhea and trichomonas (Maciosek et al., 2006). Home collection of vaginal swabs with dry transport may well facilitate screening efforts for several sexually transmitted infections (STI) among women and help expand services to high-risk populations who have less access or inclination to utilize healthcare clinics. Also needed are more cost-effectiveness or comparative effectiveness research studies to determine the effectiveness of home-based STI screening compared to clinic screening (Huang et al., 2011; Smith et al., 2007). Consideration must be given to both the direct and indirect costs of screening, treatment, and potential STI complications, as well as partner notification programs.

This study was a pilot “proof of concept” phase one study and has limitations. We acknowledge that low numbers of strains and low numbers of replicates for the most part were tested. This precluded a statistical analysis of the comparability between wet and dry shipped swabs. A larger carefully designed phase 2 validation study is being performed. Other limitations of our study are that actual clinical samples were not used, but rather simulated swabs were employed. Clinical trials will be necessary to determine whether this slight difference in sensitivity has any clinical significance in screening programs, but one study using duplicate clinical samples did not detect significant difference between wet and dry transported samples for PCR for chlamydia and gonorrhea (Gaydos et al., 2002).

In summary, we demonstrated that dry-shipped simulated swabs appeared to perform as well as transport media wet-shipped swabs for all concentrations of chlamydia tested and for gonorrhea and trichomonas organisms at low, moderate, and high concentrations. The loss of sensitivity for gonorrhea (dry heated swabs) and trichomonas (dry and dry heated swabs) at the very low levels of organisms in the range of 10-0.1 organisms per swab may or may not be clinically significant. Dry shipped diagnostic urogenital swabs collected outside a clinic appear to be useful and may assist outreach STI screening programs.