OUP user menu

Comparing Two Methods of Detection for Chlamydia trachomatis in Liquid-Based Papanicolaou Tests

Angelique W. Levi MD, Danita Beckman CT (ASCP), Pei Hui MD, PhD, Kevin Schofield CT (ASCP), Malini Harigopal MD, David C. Chhieng MD, MBA, MSHI
DOI: http://dx.doi.org/10.1309/AJCP2B7XQTCNAMJP 236-240 First published online: 1 August 2012

Abstract

This study compared the performance of Chlamydia trachomatis testing using 2 methods: the BD ProbeTec Chlamydia trachomatis Qx Amplified DNA Assay (CTQ) on the BD Viper System with XTR technology (CTQ assay) and the Hybrid Capture (HC) 2 assay. A total of 1,054 Surepath and ThinPrep specimens were tested for C trachomatis nucleic acids using the CTQ assay and the HC2 assay. For positive and discrepant C trachomatis test results, confirmatory test for C trachomatis was performed using a reverse transcriptase–polymerase chain reaction. Of 1,054 liquid-based gynecologic cytology samples tested for C trachomatis using both assays, 1,041 tested negative on both. In 6 (0.57%) samples, findings were discordant. The CTQ assay and the HC2 assay had sensitivity rates of 100% and 66.7%, respectively, with comparable specificity (99.9%). The positive predictive values were 92.3% and 88.9% with the CTQ and HC2 assays, respectively. In this study, the CTQ assay was found to be more sensitive than the HC2 assay in detecting chlamydial infection; the CTQ assay also demonstrated a higher positive predictive value.

Key Words
  • Chlamydia trachomatis
  • Liquid-based Pap test
  • CTQ assay
  • Hybrid Capture assay

Chlamydia trachomatis infection is considered the most prevalent sexually transmitted disease in women in the United States. In 2010, 1.3 million chlamydial infections were reported to the Centers for Disease Control and Prevention (CDC) from 50 states and the District of Columbia.1 Underreporting is substantial because most people with Chlamydia are not aware of their infection and do not seek testing. An estimated 2.8 million infections occur annually in the United States. Several important complications can result from C trachomatis infection in women, the most serious of which include pelvic inflammatory disease, ectopic pregnancy, infertility, and adverse pregnancy outcomes.2 In addition, C trachomatis is also one of the most common causes of eye infections and pneumonia in neonates. Chlamydia is often known as a “silent” disease because the majority of infected patients are asymptomatic.2 When confined to the lower genital tract, the disease can be easily and effectively treated with antibiotics. However, treating pelvic inflammatory disease and infertility can be costly both psychologically and financially.

To detect chlamydial infections, health-care providers frequently rely on screening tests. The primary goal of Chlamydia screening is to reduce morbidity in individuals through early detection and treatment of uncomplicated lower genital tract infections. Annual screening of all sexually active women aged 25 years or less as well as older women with risk factors (ie, those who have a new sex partner or multiple sex partners) is recommended by various government and professional organizations including the CDC and the American Congress of Obstetricians and Gynecologists (ACOG).3,4 In addition, all pregnant women should be screened for chlamydial infection.

The Papanicolaou (Pap) test has become a routine procedure for women at their annual gynecologic visit because of its success in preventing cervical cancer and precursor lesions. In addition to its primary benefit as a cancer screening test, another benefit of the Pap test is the detection of cervicovaginal microorganisms. With the widespread use of liquid-based cytology, additional samples are readily available for other tests, such as screening for chlamydial infection.5,6 Assays that are used for diagnosing chlamydial infection are frequently based on the detection and identification of C trachomatis DNA. Most of the commercially available methods including the Hybrid Capture 2 (HC2) assay (Qiagen, Gaithersburg, MD) are approved by the US Food and Drug Administration (FDA) for ThinPrep Pap tests (Hologic, Boxborough, MA), but not for SurePath Pap tests (BD Diagnostic, Burlington, NC). Recently, the BD ProbeTec Chlamydia Trachomatis Qx Amplified DNA Assay (CTQ) on the BD Viper System with XTR technology (BD Diagnostic) was approved by the FDA for the detection of C trachomatis in both ThinPrep and SurePath liquid-based preparations. As part of a validation process, we compared the performance of the Viper System using XTR technology and the HC2 assay in the detection of C trachomatis using liquid-based preparations.

Materials and Methods

Specimen Collection and Storage

During a 1-month period, all liquid-based Pap tests with a concurrent request for C trachomatis testing were included in this study. The specimens consisted of both SurePath and ThinPrep samples. The specimens were collected routinely by gynecologists and obstetricians in PreservCyt solution (Hologic) and SurePath preservative solution (BD Diagnostic) for ThinPrep and SurePath specimens, respectively. The method of collection did not vary from the standard procedure used for obtaining samples for cytologic evaluation only. Although the specimens could be stored at room temperature for up to 21 days before testing, all samples were tested within 48 hours from the time of accession.

CTQ Amplified DNA Assay

An aliquot of 0.5 mL of SurePath or PreservCyt fluid was transferred to a liquid-based specimen dilution tube (BD Diagnostic) before cytology preparation. The tubes containing the 0.5-mL aliquots were loaded onto the Viper system, which then performed all the steps necessary for extraction and amplification of target DNA without further laboratory personnel assistance. DNA samples were extracted by transferring each sample to an extraction tube containing ferric oxide particles in a dissolvable film, together with fluorescent-labeled extraction control oligonucleotides; the same process was performed on the control specimens to confirm the validity of the extraction process for each individual specimen.

In the extraction tube, the bacterial cells and nuclei were disrupted, and the bacterial DNA released into solution, using a high-pH lysis buffer. An acidic buffer was then added to lower the pH and induce a positive charge on the ferric oxide particles, which in turn bound to the negatively charged bacterial and fluorescent-labeled control DNA in the solution. The ferric oxide particles with bound DNA were extracted using magnetic capture and a high pH buffer.

After the addition of a neutralization buffer to lower the pH for optimal amplification of the target, the DNA solutions were then transferred to priming microwells containing the specific primers and probes for the target DNA sequences, nucleotides, and other reaction components necessary for the assay. When present, C trachomatis DNA was detected by means of strand displacement amplification of the target sequence within open reading frame 3 in the presence of a fluorescent-labeled detector probe. The priming wells containing the mixture were then heated to 70°C for 10 minutes. To initiate the amplification and detection process, an aliquot of fluid would then be transferred from the priming microwells to the amplification microwells, which were preheated to 52.5°C. The fluorescent signals from each sample were then measured and positive or negative results were determined by comparing the maximum relative fluorescent signal obtained from a given specimen to a determined threshold.

HC2 Assay

The HC2 assay was performed as a combined test for C trachomatis and Neisseria gonorrhoeae. The assay detected DNA-RNA hybrids with signal amplification technology and therefore did not include a DNA amplification control. The initial denaturing step was performed manually by transferring an aliquot of 2 or 4 mL of SurePath or PreservCyt fluid, respectively, from the residual fluid after cytology preparation of a denaturing agent. The mixtures were then processed in batches of up to 88 samples (plus 8 controls) by incubating at 65°C for 45 minutes. The denatured samples were then loaded onto the Rapid Capture System (Qiagen), a programmable 96-well microplate processor for performing all steps before signal detection.

Signal detection was accomplished using a DML 2000 luminometer. Relative light units (RLUs) for the positive and negative controls were used to calculate the run-specific cutoff and results were reported as RLU/cut-off ratios. Samples with an RLU/cut-off ratio less than 1 and more than or equal to 2.5 were considered negative and positive, respectively. Samples with ratios more than or equal to 1 and less than 2.5 were considered equivocal and the samples were retested in duplicate. The samples were considered positive if the results of at least 2 of the 3 tests (the initial test and the 2 retests) were more than or equal to 1. An initial positive result meant that the specimen tested positive for either C trachomatis or N gonorrhoeae, or both. Reverse transcriptase–polymerase chain reaction (RT-PCR) was performed on specimens that yielded an initial positive result to identify whether the initial signal was caused by the presence of chlamydial or gonococcal DNA, or both.

RT-PCR

All samples that were positive for CTQ assay, HC2 assay, or both were tested with RT-PCR. Oligo-nucleotide sequences in RT-PCR testing for C trachomatis include the following: forward primer, 5′-CCA-CAGAATTCCGTCGATCA--CT-3′ ; reverse primer, 5′-TGCCGCTTTGAGTTCTGCTT--CT-3;' and the probe FAM-ATTCCCCACAGGCAGAGCTTGCCA. PCR reactions were performed on an ABI Prism 7000 Sequence Detection System (Life Technologies, Carlsbad, CA). For each PCR run, a master mixture was prepared on ice with 1× TaqMan buffer; 5 mmol/L of magnesium chloride; 200 mmol/L of deoxyadenosine triphosphate, deoxycytidine triphosphate, and deoxyguanosine triphosphate, and 400 mmol/L of deoxyuridine triphosphate; 300 nmol/L of each primer; 150 nmol/L probe; and 1.25 U of AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, CA). Ten microliters of each appropriately diluted DNA sample was added to 40 μL of the PCR master mixture. The thermal cycling conditions comprised an initial denaturation step at 95°C for 10 minutes, and 50 cycles at 95°C for 15 seconds and 65°C for 1 minute. Experiments were performed in duplicate for each data point for each run. Data were normalized to the quencher dye tetramethylrhodamine (TAMRA) and analyzed. The final results represented duplicate experiments of each specimen. Cases reaching the threshold level were further confirmed with a visual inspection of the output real-time PCR graphs for final reporting.

Statistical Analyses

The concordance between the CTQ and HC2 assays was analyzed using the κ statistic. The measure of agreement was generally interpreted as follows: less than 0.20, poor agreement; 0.21 to 0.40, fair agreement; 041 to 0.60, moderate agreement; 0.61 to 0.80, good agreement; and more than 0.80, excellent agreement.7 We also calculated the sensitivity and specificity, as well as positive and negative predictive values of both the CTQ and HC2 assays. A sample was treated as a true negative if results of both the CTQ assay and the HC2 assay were negative or if the result of the confirmatory RT-PCR test was negative. However, a sample was considered a true positive only if the result of the RT-PCR test confirmed a positive assay result.

Results

During a 1-month period, a total of 1,054 liquid-based gynecologic cytology samples were tested for C trachomatis using both the CTQ and HC2 assays; 99 (9.4%) of these were ThinPrep samples and 955 (90.6%) were SurePath samples. Table 1 summarizes the results of the CTQ assay and the HC2 assay. Among these samples, 1,041 were negative with both the CTQ assay and the HC2 assay. Eight were positive with both the CTQ and HC2 assays, as well as with RT-PCR. Five were positive with the CTQ assay only; 4 of these 5 cases were also positive with RT-PCR. One sample was positive with the HC2 assay, but negative with the CTQ assay and RT-PCR. As a result, 6 samples had discordant CTQ assay and HC2 assay results, accounting for 0.57% of all cases. The kappa statistic was 0.72. None of the samples that were positive with the HC2 assay were positive for Neisseria by RT-PCR or the BD ProbeTec Neisseria gonorrhoeae Qx Amplified DNA Assay.

The numbers of true negative and true positive cases were 1,042 and 12, respectively. Therefore, the prevalence of Chlamydia infection was 1.13% (95% confidence interval, 0.62% to 2.04%). Table 2 summarizes the performance of both methods in the detection of C trachomatis in liquid-based preparations. The CTQ assay demonstrated a significantly higher sensitivity than the HC2 assay, and a slightly higher positive predictive value (PPV) than the HC2 assay.

Discussion

Until recently, most commercially available C trachomatis assays, which were based on the detection of C trachomatis DNA, were approved by the FDA for ThinPrep liquid-based preparations only. In the winter of 2010, the BD Viper System with XTR Technology and the ProbeTec Qx Amplified DNA Assay was approved by the FDA for C trachomatis detection in both ThinPrep and SurePath specimens. In the spring of 2011, we decided to replace the HC2 assay with the CTQ assay to detect C trachomatis. When switching from one method to another, it is important to demonstrate that the new platform performs equally, if not better than, the existing one. Therefore, as part of the validation process, we compared the performance of the 2 assays for detecting C trachomatis with samples from a patient population with a relatively low prevalence of chlamydial infection (1.13%).

View this table:
Table 1
View this table:
Table 2

Both assays showed good agreement with a κ score of 0.72. Results were discordant in 6 cases. Using RT-PCR as the “gold standard,” the false-negative rates for the CTQ assay and the HC2 assay were 0.0% and 33.3%, respectively, whereas the false-positive rates for the CTQ assay and the HC2 assay were both 0.01%. The sensitivity and specificity for CTQ assay were 100% and 99.9%, respectively, which were comparable to the performance data published by the manufacturer, namely, 95.0% sensitivity and 95.3% specificity.8 Taylor et al9 also reported a sensitivity and specificity of 91.3% and 98.3% using endocervical swab specimens.

Based on our observations, the sensitivity of the CTQ assay in identifying C trachomatis was superior to that of the HC2 assay (100% vs 66.6%), while demonstrating similar specificity (100%). The probes and the primers used for the CTQ assay and PCR were complementary to sequences of C trachomatis cryptic plasmids, whereas the probes for the HC2 assay were complementary to both cryptic plasmid and genomic C trachomatis sequences. This may explain the “false-positive” case of HC2 assay. However, the differences in the probes would not explain the observed differences in sensitivities between the 2 assays. In general, it has been agreed that the sensitivities of nuclei acid amplification tests (NAAT) such as CTQ assay have consistently exceeded those of non-NAATs such as the HC2 assay.3 The limit of detection of the HC2 assay was 50 elementary bodies per milliliter10 with endocervical specimens whereas that of the CTQ assay was 14 elementary bodies per milliliter with urine specimens.11 In contrast to our observations, other studies have reported that the sensitivity for the HC2 assay with or without the use of the Rapid Capture System for detecting C trachomatis was above 90%.10,1215 One plausible explanation may be the types of samples being tested. Most of these studies use endocervical specimens collected with a cervical sampler and placed in a Digene sample transport medium (Qiagen), whereas the current study used aliquots from the residual fluid after cytology preparation.

Because RT-PCR was only used as a confirmatory test and to adjudicate discordant results in this study, it is possible that there were true-positive specimens in the test population that were not detected with either HC2 assay or CTQ assay. On the other hand, one specimen was considered a presumptive false positive with the CTQ assay because it was not detected with RT-PCR. It is possible that this specimen might actually contain C trachomatis DNA, but was not amplified in the PCR reaction because of the presence of an inhibitor. Because this “false-positive” case represented fewer than 0.1% of all specimens tested, the specificity of the CTQ assay for the detection of C trachomatis still approached 100%.

Although commercially available NAATs used for screening chlamydial infections are sensitive and specific for detecting C trachomatis in cytologic specimens, questions have been raised about their PPV, particularly in a screening population with a low prevalence of chlamydial infection.15 Because a false-positive test result for C trachomatis can have adverse social and psychological effects on a patient, the CDC has recommended performing a confirmatory NAAT when the PPV is less than 90%.3 The PPV of detection would increase unless the screening and confirmatory tests are falsely positive for the same reasons (ie, mislabeled specimens, cross-contamination, and cross-reactivity with a nonchlamydial organism). In the current study, the PPV was more than 90%; in theory, a confirmatory test would not be necessary. However, because our patient population has a very low prevalence of chlamydial infection and the 95% confidence intervals of PPV were between 62% and 100%, a confirmatory test with RT-PCR was required after an initial positive result with the CTQ assay. In a population with a low prevalence of chlamydial infection such as ours (~1%), our average daily volume of fewer than 1 of 50 samples would require confirmatory testing. In a population with a high prevalence of chlamydial infection (ie, ~18%), approximately 9 of 50 samples would require confirmatory testing, resulting in a substantial increase in workload.

The cost per test for the CTQ assay ($9.91) and the HC2 assay ($8.67) were comparable. For HC2 assay, each kit contained 96 wells. Because each kit required 8 control samples, each kit had the capacity to test up to 88 samples. If fewer than 88 samples were run, the remaining unused wells were discarded. Therefore, it might not be cost-effective in laboratories such as ours that performs fewer than 88 tests per kit. On the other hand, the CTQ assay had a very flexible format that allowed any number of specimens to be run without having to discard any unused test kits. The total cycle times from sample preparation to processing and reporting results for the HC2 assay and the CTQ assay were 16.5 hours and 4.5 hours, respectively. The actual hands-on time was 3.5 hours and 2 hours for the HC2 assay and the CTQ assay, respectively. After switching from the HC2 assay to the CTQ assay, our turnaround time for C trachomatis testing decreased by 50% from 4.4 days to 2.2 days (results not shown).

In conclusion, the CTQ assay is a cost-effective, quick, and reliable method for detecting C trachomatis in liquid-based cytology specimens. A confirmatory test is recommended in a population with a low prevalence of chlamydial infection to minimize the number of false positives. The CTQ assay can be easily adopted in a clinical laboratory setting because of its sensitivity, accuracy, high throughput, and short cycle time.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
View Abstract