In vitro assessment of the APTIMA Combo 2 assay for the detection of Chlamydia trachomatis using highly purified elementary bodies

Chlamydia trachomatis infections are one of the most common sexually transmitted diseases worldwide. Unfortunately, in chlamydial infections the symptoms of infected persons are often mild or absent, and therefore asymptomatic and persistent infection by this pathogen has led to its widespread distribution and makes the organism difficult to control. Ascending infection by C. trachomatis of the female upper genital tract, known as chronic disease, causes salpingitis, which in turn leads to the eventual complications of ectopic pregnancy and tubal infertility (Cates & Wasserheit, 1991). Persistent infection is characteristic of C. trachomatis infection in vitro (Beatty et al., 1994, 1995; Hogan et al., 2004), and it was recently reported that persistent infection induced reactive arthritis (Rahaman et al., 1992; Gaston, 1995; Hanada et al., 2003). As a result, early diagnosis and treatment of C. trachomatis are important in symptomatic, asymptomatic and persistent infection to prevent complications and control the spread of this disease.

There are many diagnostic kits available for C. trachomatis infection, and the IDEIA PCE Chlamydia (DAKO Diagnostics) (Okadome et al., 1999) and the AMPLICOR C. trachomatis (Roche Diagnostic Systems) (Bass et al., 1993) are both used commercially for routine tests. Diagnosis of C. trachomatis infection has made remarkable progress due to advances in nucleic acid amplification (NAA) techniques. However, the first-generation NAA detection assays, such as AMPLICOR C. trachomatis or Chlamydia LCx (Abbot Laboratories), have had some technological problems regarding their performance. These include cumbersome specimen processing and the occurrence of false-negative results (Jensen et al., 1997; Goessens et al., 1997; Peterson et al., 1997; Notomi et al., 1998; Mahony et al., 1998; Toye et al., 1998). False-negative results might be due to the presence of inhibitory substances in clinical specimens, e.g. phosphate in urine specimens or Fe²⁺ ion from haemoglobin in erythrocytes, and therefore attention should be paid to inhibitor control of these commercial assays.

APTIMA Combo 2 (Gen-Probe Inc.) is a second-generation NAA test that utilizes target capture and transcription-mediated amplification. This assay amplifies target rRNA, which is present at a high copy number in all cells and bacteria, in order to increase the sensitivity of the assay and to allow the simultaneous detection of both C. trachomatis and Neisseria gonorrhoeae in endocervical swab and urine specimens (Gaydos et al., 2003). The method of target capture is selectively used to isolate target C. trachomatis and N. gonorrhoeae nucleic acids from potentially inhibitory biological substances in specimens. In the present study, we examined the sensitivity of the APTIMA Combo 2 assay and the inhibitory effects of phosphate and Fe²⁺ ion using highly purified elementary bodies (EBs) of C. trachomatis.

Preparation of purified EB suspension.
C. trachomatis serovar D/UW-3/Cx was prepared in McCoy cells and propagated according to a previously reported method (Nagayama et al., 1988). Purification of EBs was performed by the method of Miyashita & Matsumoto (1992), except that the Renografin gradient centrifugation was repeated three times to ensure the purity of the EB preparations. Purified EBs were suspended in sucrose/phosphate/glutamate (SPG) medium (75.0 g sucrose, 0.52 g KH₂PO₄, 3.07 g Na₂HPO₄.12H₂O, 0.72 g glutamic acid; distilled water to 1000 ml) and frozen at 80 °C until use.

Serial 10-fold dilutions of the suspension of purified EBs were inoculated on to McCoy cells and stained at 48 h after inoculation with a fluorescein-labelled species-specific monoclonal antibody (Syva Microtrak) to determine the number of infectious EBs present, defined as inclusion-forming units (i.f.u.) ml¹. The number of EBs in suspension was counted under a Hitachi H-7100 transmission electron microscope at a magnification of x8000. Concomitantly, the number of polystyrene beads (0.203 µm diameter; Seradyn) was counted using the same method under an electron microscope. Based on these counts, the mean number of EBs per field was determined and then converted based on the ratio of the number of polystyrene beads per field to the polystyrene bead concentration. As a result, the titre of the stock suspension was found to be 3.9 x 10⁷ i.f.u. ml¹, which was 1.56 x 10⁸ EB ml¹; i.e. 1 i.f.u. was equivalent to four EBs.

End-point determination.
Aliquots of purified EB suspension were diluted up to 3.9 x 10⁶ i.f.u. ml¹ in SPG medium and then serially diluted in each manufacturer's transport medium with a lysis agent provided with the APTIMA Combo 2 or AMPLICOR C. trachomatis kits, respectively. Serial 10-fold dilutions of an EB preparation with a concentration range of 500.000005 i.f.u. ml¹ were tested by both the APTIMA Combo 2 and AMPLICOR C. trachomatis assays to determine the concentrations closest to the cut-off values of the respective tests.

Effect of inhibitory substances.
In order to study the inhibitory effect of phosphate and Fe²⁺ ion, chlamydial dilutions of 50, 5 and 0.5 i.f.u. ml¹ were tested by adding three concentrations of phosphate or Fe²⁺ ion solution before specimen processing by both kits. In the case of phosphate, a stock solution of 60 mM phosphate (2.5 g KH₂PO₄, 14.85 g Na₂HPO₄.12H₂O, distilled water to 1000 ml) was used to prepare final concentrations of 6.0, 1.2 and 0.6 mM. In the case of Fe²⁺ ion, a stock solution of 100 µg FeCl₂ ml¹ (226.9 mg FeCl₂, distilled water to 1000 ml) was used to prepare final concentrations of 1.0, 0.5 and 0.1 µg ml¹. No inhibitory substances were added to the 0.5 i.f.u. ml¹ dilution for the AMPLICOR C. trachomatis assay.

The detection limit obtained in the APTIMA Combo 2 assay using highly purified EBs was 0.005 i.f.u. ml¹ (Table 1), which corresponded to 0.002 i.f.u. per assay, since 0.4 ml of the suspension was used in one assay. This corresponded to 0.008 EB per assay when converted to the number of EBs. The AMPLICOR C. trachomatis assay using an EB suspension of 5 i.f.u. ml¹ was the lowest concentration to yield a positive result (Table 1), which was equivalent to 0.5 EB per assay. Miyashita et al. (1996) and Notomi et al. (1998) reported that the sensitivity of the Chlamydia LCx assay was 2 EB per assay or 0.6 i.f.u. per assay, respectively. We obtained similar results, although we found that the AMPLICOR C. trachomatis assay was a little more sensitive. The APTIMA Combo 2 assay was 1000-fold more sensitive than the AMPLICOR C. trachomatis for the detection of serial dilutions of C. trachomatis D (Table 1). One of the reasons for the excellent sensitivity of this assay could be that the target of the APTIMA Combo 2 system is rRNA, as described above. Another reason for the different sensitivities of these assays could be that the diluents of the APTIMA Combo 2 and the AMPLICOR kit are different, containing the appropriate constituents for each kit. However, as the diluent is also used in the clinical materials, the difference in sensitivity between the APTIMA Combo 2 and the AMPLICOR kit might thus be reflected in the clinical cases.

Table 1. End-point determination of the APTIMA Combo 2 and AMPLICOR C. trachomatis assays for detection of C. trachomatis +, Positive result; , negative result; NT, not tested.

To examine the effect of inhibitory substances, phosphate was added to the samples of purified EB suspension before processing. The addition of 0.6, 1.2 or 6.0 mM phosphate had no effect on the results of the APTIMA Combo 2 assay, even when the concentration of the EB suspension was 0.5 i.f.u. ml¹ (Table 2). On the other hand, the AMPLICOR C. trachomatis assay of the sample containing 5 i.f.u. ml¹ (0.5 EB per assay) of EB suspension resulted in a false-negative result when 1.2 or 6.0 mM phosphate was added (Table 2). In addition, the sample containing EB suspension at 50 i.f.u. ml¹ gave a false-negative result with 6.0 mM phosphate (Table 2). The inhibitory effects of Fe²⁺ ion were also tested in both assays. No effect was observed for 0.1, 0.5 or 1.0 µg Fe²⁺ ion ml¹ in APTIMA Combo 2 assay (Table 3). In contrast, the inhibitory effect of Fe²⁺ ion on the AMPLICOR C. trachomatis assay was remarkable, with the addition of 0.1 µg Fe²⁺ ion ml¹ to the EB suspension samples of 5 i.f.u. ml¹ inhibiting the efficacy of the assay (Table 3).

Table 2. Inhibitory effect of phosphate on the APTIMA Combo 2 and AMPLICOR C. trachomatis assays

Table 3. Inhibitory effect of Fe2+ ion on the APTIMA Combo 2 and AMPLICOR C. trachomatis assays ±, Of three tests, one or two were positive.

With increased sensitivity, the occurrence of false-negative results as a result of inhibitors becomes important. The principle of NAA makes it necessary to consider phosphate as an inhibitor, since it most likely prevents amplification of the target nucleic acids. The concentration of phosphate in SPG medium is 12.4 mM and usually 6070 mM in the urine of a healthy adult. C. trachomatis L2 preparations with first-void urine have been reported to produce false-negative rates of 0.48 % using the APTIMA Combo 2 assay (Chong et al., 2003), while their prevalence has been reported to be 7.0 % in urine specimens using the AMPLICOR C. trachomatis assay (Mahony et al., 1998). In addition, clinical specimens frequently contain blood elements or tissue debris. We have observed that whole blood inhibits the LCx tests (data not shown). Whole blood contains erythrocytes, leukocytes, platelets and plasma, and haemoglobin in erythrocytes contains Fe²⁺ or Fe³⁺. Thus, we also examined the inhibitory effect of the Fe²⁺ ion on the APTIMA Combo 2 assay. Our results demonstrated that neither phosphate nor Fe²⁺ ion had an inhibitory effect on the APTIMA Combo 2 assay (Tables 2 and 3), even when the inoculum concentration was 0.5 i.f.u. ml¹. In contrast, in the AMPLICOR C. trachomatis assay, samples containing inocula at a concentration near to the cut-off value (5 i.f.u. ml¹) often resulted in false-negative results in the presence of phosphate and Fe²⁺ ion (Tables 2 and 3). Unknown inhibitors other than phosphate and Fe²⁺ ion may be present in clinical specimens, for example, hormones and/or enzymes, since inhibitory activity has also been found in cervical specimens using the AMPLICOR C. trachomatis assay (Verkooyen et al., 1996). The reason for false-negative results using first-void urine in the APTIMA Combo 2 assay could not be clearly elucidated; however, the prevalence was very low.

The unreliable results observed in the AMPLICOR tests occurred when the number of targets in the sample was low, in other words, near the cut-off value. When phosphate or Fe²⁺ ion was added to samples containing inoculum concentrations near to the cut-off value (5 i.f.u. ml¹), the results became more variable (Tables 2 and 3). In addition, several investigators have reported that results obtained using the AMPLICOR C. trachomatis assay may not be reproducible (Bauwens et al., 1993; de Barbeyrac et al., 1995; Peterson et al., 1997). It is possible that the reproducibility problems of the AMPLICOR test may have caused the variations in our results.

The APTIMA Combo 2 assay is thus considered to be a useful tool for reducing the prevalence and incidence of C. trachomatis infection, since this assay had a greater sensitivity than the AMPLICOR C. trachomatis assay. In addition, neither phosphate nor Fe²⁺ ion inhibited the APTIMA Combo 2 assay, and therefore this assay is also recommended for use with urine or blood-containing specimens.

We appreciate the financial and material support from Dr Katsuyasu Kudou, Fujirebio Inc., Tokyo, Japan.