Secondary Logo

Journal Logo

Articles

Multisite Pooling Study Using Ligase Chain Reaction in Screening for Genital Chlamydia trachomatis Infections

CLARK, AGNES M. BS,*; STEECE, RICHARD PhD,†; CROUSE, KAREN BA, MT(ASCP),‡; CAMPBELL, JOYCE BS; ZANTO, SUSANNE CLS (NCA),‖‖; KARTCHNER, DIANE BS; MOTTICE, SUSAN PhD; PETTIT, DENISE PhD#

Author Information
  • Free

Abstract

BECAUSE OF IMPROVED changes in technology, nucleic acid amplification tests have demonstrated a 20% to 30% increase in the number of infected patients identified. 1 A highly efficient method, DNA amplification exponentially amplifies copies of a specific DNA sequence. Plasmid DNA amplification, used in the ligase chain reaction (LCR) assay, is thought to provide greater sensitivity because up to 10 plasmids may be present in each chlamydial elementary body. 2,3 Ligase chain reaction testing has enabled sensitive and specific detection of Chlamydia trachomatis in urine as well as cervical and urethral swabs from individual patients. 1,4,5

The high cost of nucleic acid amplification tests (two to four times the reagent cost), as compared with their nonnucleic counterparts used to detect C trachomatis, necessitated the development of cost-saving measures in some laboratories to ensure the affordability and availability of these tests for health department and community clinics. Pooling of specimens for testing has been investigated as a way to reduce the cost associated with amplified test procedures such as LCR. The concept of pooling involves screening specimens in groups. If any pool is positive, the testing algorithm returns to retesting the individual specimens in that pool to determine the true positive specimen.

Most previous chlamydia pooling studies have been performed with urine specimens. Therefore, our objective was to evaluate the performance of LCR when single cervical specimens and pools of 5 and 10 cervical specimens were used to detect the presence of C trachomatis. Outcomes of interest were the accuracy of the results, the possible effects from dilution of individual samples resulting in reduced sensitivity, the effect of inhibitors, 6 and the effect of varying infection prevalence on cost efficiency.

Methods

Each of the six participating laboratories processed at least 500 consecutive cervical swabs, which were received for routine C trachomatis screening. Specimen collection, transfer, and preparation were performed according to the manufacturer’s instructions for the LCR assay (Abbott Labs, Abbott Park, IL). Samples were processed on arrival at the laboratory, or refrigerated and processed within the next 24 to 48 hours. Swabs were heated to 97 ± 2 °C for 15 ± 1 minutes to facilitate the release of DNA from the cells. Tubes were cooled at room temperature for 15 ± 5 minutes. As each tube was sampled, the swab was expressed and the sample aliquots pipetted. Then clean caps replaced those discarded.

For each group of 10 specimens, the individual processed samples (100 μl) were inoculated into individual amplification vials 1 to 10 in the order they were received in the laboratory. Next, 20 μl of each prepared sample from tubes 1 to 5 was inoculated into amplification vial 11. This procedure was repeated with the second five specimens (tubes 6 to 10) and added to amplification vial 12. For the pools of 10 specimens, 10 μl from each prepared sample was placed into amplification vial 13. The final test volume for each amplification vial was 100 μl.

All the controls were prepared according to the product insert. In addition, a positive processing control was included in each run. The processed samples then were amplified and the amplified products detected by LCR immunoassay technique, as previously described. 7 Each amplification-specimen mixture was placed in the Abbott LCx Thermal Cycler (Abbott Labs) for 40 cycles and pulse centrifuged for 10 to 15 seconds in a microcentrifuge. The tubes containing the amplified nucleic acid copies then were placed in an automated analyzer (Abbott LCx), which detects DNA present in any specimen.

Pools were not split between two runs. To account for any internal variation between machines and between runs, each set of individual specimens and pools of 5 and 10 specimens (13 amplification vials) were run at the same time in the same thermal cycler and LCx analyzer. Only amplification vials and reagent packs from identical lots, provided by Abbott Labs, were used in all participating laboratories. For individually tested samples, a sample-to-cutoff ratio (S/CO) of more than 1 was considered positive. Any individual reading with a S/CO between 0.80 and 1 was considered equivocal. Such a reading was repeated in duplicate. If equivocal on retest, the sample was classified as negative.

In the pools, a S/CO of 0.2 or more was defined as presumptive positive. 7 If a pool was positive, with corresponding individual specimens negative, the individual specimens were repeated in duplicate. If the individual specimen was positive, with corresponding pools negative, the pools were repeated in duplicate. Because technician time, salary, and indirect costs varied between laboratories, we calculated the estimated percentage of savings as the diagnostic reagent cost for each site on the basis of site-specific prevalence (positivity).

Results

Chlamydia prevalence (positivity) and number of positive samples varied by laboratory (Table 1). Altogether, 186 samples (mean S/CO, 3.30) were identified as positive for C trachomatis using individual specimens for testing. Analysis identified 164 pools of 5 specimens as C trachomatis positive and 137 pools of 10 specimens as positive. Of 951 pools, 10 pools (6 pools of 5 and 4 pools of 10 specimens) with a S/CO less than 1 contained individual samples that were positive (mean S/CO, 1.69). Therefore, to enhance the sensitivity of pooled specimen analysis, and to compensate for the reduced volume of each individual specimen in the pooling procedure or C trachomatis detection, the S/CO used for positive pools was 0.2 or more. 4 This allowed 99.5% of the positive specimens in pools of 5 and 98.9% of the positive specimens in pools of 10 to be detected.

Table 1
Table 1:
Individual Site Results

Overall, two individual samples had high negative readings below a cutoff (S/CO) of less than 1 (0.91, 0.97). The pool of 5 specimens (1.62) and the pool of 10 specimens (3.01) were positive for the first sample (0.91). The first sample (0.91) was positive on repeat testing (1.29). The pool of 5 specimens (0.37) was positive for the second sample (0.97), and the pool of 10 specimens was just below the cutoff (0.17). The second sample (0.97) was negative on repeat testing (0.02).

At site 1, one individual positive sample had an S/CO of 3.88. Its 5- and 10-specimen pools both were negative. On repeat testing, the individual specimen and pools were positive. For all the laboratories, the pooling of 5-specimen samples resulted in reagent cost savings of 41% to 62% (mean, 54%), and the pooling of 10-specimen samples resulted in reagent cost savings of 31% to 63% (mean, 47%) 8,9 (Table 2).

Table 2
Table 2:
Calculated Percentage of Reagent Cost Savings for Each Test Site*

Discussion

In this multicenter pooling study, we evaluated a testing strategy of pooling prepared cervical swab specimens for detection of C trachomatis using the LCR assay. The sensitivity and specificity of the pooling results were high, as defined by the same samples tested individually (Table 3).

Table 3
Table 3:
Accuracy of Pooling Endocervical Samples for Detection of Chlamydia trachomatis Infection in Women by Ligase Chain Reaction

The testing of pooled specimens showed results similar to those of individually run LCR samples. Sensitivity and specificity were not affected significantly by changing the number of samples per pool (5 versus 10 samples), or by lowering the sample-to-cutoff ratio to 0.2, although some pools (10 in 301) yielded S/CO results between 0.2 and 1.

At site 1, one specimen was positive by individual LCR, with a S/CO of 3.88, but both related pools were negative. When repeated, the individual sample and the two related pools were positive. This is consistent with a sampling error, in which the individual specimen most likely was not inoculated into the pooled amplification vial. Because pooling of specimens adds an additional processing step, extra care must be exercised by the laboratory to prevent sampling or pipetting errors.

In this study, if a pool was positive, with the corresponding individual specimens negative, the individual specimens were repeated in duplicate. The retests of the individual specimens were considered the correct answer because this was the Food and Drug Administration (FDA)-cleared procedure. Other studies have reported false negative results because of nonspecific inhibitors in the specimens. One possible alternate strategy that could be useful in testing pooled specimens would involve conducting dilution studies when individual specimens are negative, with a corresponding positive pool. According to the current study, this situation appears to be rare. However, several test sites that routinely test pooled specimens have incorporated testing algorithms to deal with the problem of occasional inhibitory specimens by rerunning the individual specimens from positive pools, both undiluted (e.g., 1:4) and diluted (e.g., 1:10), to check for possible inhibitors. The results of this retest then are reported. The sensitivity for individual testing was 98.5%: 100% for the 5-specimen pools and 99.5% for the 10-specimen pools (Table 3).

All six sites showed significant reagent cost savings by pooling of specimens into 5- and 10-specimen pools, as compared with individual screening. The potential reagent cost savings afforded by pooling was influenced by prevalence: the higher the prevalence, the lower the reagent cost savings. In general, pooling by 5 specimens detected the most positive specimens and resulted in the greatest overall reagent cost savings (Table 2). Although all the sites reported a reagent cost savings, they did not experience any additional personnel savings because all positive pools needed to be repeated.

The current study showed that the pooling of endocervical swabs for LCR by laboratory personnel should considerably decrease overall costs from those incurred by individual testing, allowing more specimens to be tested with the same allotted funds. This approach may be an especially important means of maintaining the ability to screen large numbers of persons with C trachomatis, even in populations with declining chlamydia prevalence.

However, in an area of high chlamydia prevalence, the high number of retested individual specimens from presumptive positive pools would result in greater cost and labor than individual screening. In addition, if pools are screened on day 1, and individual specimens are run on day 2, there will be a slight increase in the turnaround time. This increase in turnaround time may be a concern for clinicians. Therefore, each site considering the use of pooled specimens for C trachomatis detection should examine the prevalence of their population to determine optimum pooling size. In addition, because the FDA has not cleared the Abbott LCR for use with pooled specimens, each site that incorporates this procedure would be required to perform a verification study to satisfy federal Clinical Laboratories Improvement Act regulations.

References

1. Black C. Current methods of laboratory diagnosis of chlamydia trachomatis infections. Clin Microbiol 1997; 10: 160–184.
2. Girdner J, Cullen A, Salama G, He L, Lorincz A, Quinn T. Evaluation of the digene hybrid capture II CT-ID test for detection of Chlamydia trachomatis in endocervical specimens. J Clin Microbiol 1999; 37: 1579–1581.
3. Lee H, Chernesky M, Schachter J, et al. Diagnosis of Chlamydia trachomatis genitourinary infection in women by ligase chain reaction assay of urine. Lancet 1995; 345: 213–216.
4. Gaydos C, Howell R, Quinn T, Gaydos J, McKee K. Use of ligase chain reaction of urine compared to cervical culture for the detection of Chlamydia trachomatis infections in an asymptomatic military population of pregnant and nonpregnant females attending PAP smear clinics. J Clin Microbiol 1998; 36: 1300–1304.
5. Stary A, Schuh E, Kerschbaumer M, Gotz B, Lee H. Performance of transcription-mediated amplification and ligase chain reaction assays for detection of chlamydial infection in urogenital samples obtained by invasive and noninvasive methods. J Clin Microbiol 1998; 36: 2666–2670.
6. Kapala J, Copes D, Sproston A, et al. Pooling cervical swabs and testing by ligase chain reaction are accurate and cost-saving strategies for diagnosis of Chlamydia trachomatis. J Clin Microbiol 2000; 38: 2480–2483.
7. Kacena K, Quinn S, Howell R, Madico G, Quinn T, Gaydos C. Pooling urine samples for ligase chain reaction screening for genital Chlamydia trachomatis infection in asymptomatic women. J Clin Microbiol 1998; 36: 481–485.
8. Peeling R, Toye B, Jessamine P, Gemmill I. Pooling of urine specimens for PCR testing: a cost-saving strategy for Chlamydia trachomatis control program. Sex Trans Inf 1998; 74: 66–70.
9. Krepel J, Patel J, Sproston A, et al. The impact on accuracy and cost of ligase chain reaction testing by pooling urine specimens for the diagnosis of Chlamydia trachomatis infections. Sex Transm Dis 1999; 26: 504–507.
© Copyright 2001 American Sexually Transmitted Diseases Association