< > Home : FAQ : Professional : Chlamydiales : Immunobiology : Infections : Diagnosis & Treatment : Links : Contact Us

Roche AMPLICOR^® CT/NG and Roche COBAS^® AMPLICOR^® CT/NG

PCR, the polymerase chain reaction , was conceived by Kary Mullis in 1983 [Saiki et al., 1985]. It was developed by a team of scientists from Cetus and became hailed as one of the most important discoveries in molecular biology. Mullis and others were awarded the Nobel Prize for chemistry in 1993. Subsequently the PCR patents for diagnostic applications were acquired by the Hoffman-La Roche group. The Roche AMPLICOR^® PCR for C. trachomatis was the first commercial nucleic acid amplification test to be introduced for the diagnosis of chlamydial infection. Subsequently a multiplex assay became available that permits simultaneous PCR amplification of C. trachomatis (CT) and N. gonorrhoeae (NG) target DNA as well as internal control DNA. Currently the test has two main formulations; the Roche AMPLICOR CT/NG MWP (multi well plate) PCR and the fully automated Roche COBAS^® AMPLICOR PCR. 

The PCR reaction is extremely well known and will only be dealt with in outline here in order to clarify the nature of the kits. 

The standard CT/NG MWP kit consists of specimen preparation kit, a common amplifier kit, and a choice of any combination of C. trachomatis, N. gonorrhoeae or an internal control as tests, making up a classic 96 well tray. Specific primers for CT and NG are biotin-labelled oligonucleotides of 20 - 30 base pairs. The optional internal control (IC) [Rosenstrauss et al., 1998] consists of non infectious DNA containing the same primer binding region as target DNA, but with a different probe binding region that enables the IC amplicon to be distinguished.  

Urine, swabs etc are treated with detergent to release the target chlamydial DNA, which is directed against the multi-copy cryptic chlamydial plasmid. The test is not isothermal like the TMA or SDA tests, but involves rapid thermal cycling in a specially programmed thermal cycler. Theoretically, this is a disadvantage of PCR compared to isothermal processes because of the time required for thermal cycling and equilibration. 

The PCR 'mastermix' consists of the standard mix of deoxynucleotide triphosphates, cofactors, and the amplification enzyme, taq polymerase, which is a DNA-dependent DNA polymerase. However deoxyuridine triphosphate (dUTP) is substituted for the more usual deoxythymidine triphosphate (dTTP) in the mastermix. Thus the amplified DNA product (amplicon) can be distinguished from native DNA because it contains dUTP. This is important because, in a busy clinical laboratory, as with other nucleic acid amplification tests, it is easy to contaminate a new test with amplicons from previous tests unless special measures are taken. Here, the PCR protocol incorporates the enzyme uracil-N-glycosylase (AmpErase^® ), which at the start of the test at temperatures below 55C degrades any prior contaminating amplicon at C1 of the deoxyuridine residues, but does not degrade native (target) DNA which lacks uridine. Statistical algorithms may also be helpful in determining those runs which have been contaminated. 

A single PCR cycle consists first of thermal denaturation of double strand to single strand DNA, followed by the annealing of specific primer followed by the taq polymerase dependent synthesis of new DNA. These steps occur at different temperature optima, successive cycles producing multiple copies of the target sequence. At the temperatures required for amplification and extension, the uracil-N-glycosylase is inactive so that it does not prevent the formation of new, deoxyuridine containing amplicons. Subsequently, any residual enzyme is chemically denatured by the addition of detergent. 

After amplification and alkaline denaturation to single strands, the biotinylated amplicons bind to capture probes, which, in the MWP format are covalently linked to bovine serum albumen and  bound to the solid surfaces of each microwell. The amount of biotin captured is proportional to the amount of specifically amplified amplicon that has been produced. Following a wash step to remove unbound biotin labelled DNA, the biotinylated amplicon is measured by the addition of a conjugate of avidin coupled to the enzyme horse radish peroxidase. Avidin, a protein in egg white, has enormously high affinity for biotin, a B vitamin. This latter stage is a conventional and well known enzyme immunoassay procedure, in which horse radish peroxidase in conjunction with hydrogen peroxide produces a blue and then yellow product from a colourless enzyme substrate, tetramethylbenzidine. This product can be measured in a photometer.

The automated COBAS AMPLICOR has a somewhat different format. It consists of separate A (amplification) tubes and (D) detection cups; liquid ready to use reagents and bar-coded reagent racks, automatic monitoring and automatic quality control. Double stranded DNA is chemically denatured by automatic addition of alkali. A portion of each sample is automatically transferred to the detection cup, which contains the appropriate test oligonucleotide probe on magnetic microparticles to which free amplicon hybridises. This hybridisation is critical for test specificity. The particles plus hybrids are automatically washed four times while immobilised with a magnet. The detection process is similar to before, but the final colour is read in a spectrophotometer equipped with a gallium aluminium arsenide light emitting diode. The assay results are calculated automatically, taking into account various controls, and may be output to printer or to a computer data network. To start the process the operator simply loads prepared samples, bar-coded reagents and pre-packaged detection cups into the analyser and presses start. However the analyser must be carefully maintained.

The sensitivity, specificity and predictive values for the Roche AMPLICOR for various prevalence levels, from the package insert are shown below:

Table 1. Chlamydia trachomatis

Table 2. Neisseria gonorrhoeae

The usual test samples are, from men: urine (or urethral swabs if necessary) and from women: cervical swabs or urine.  Like other nucleic acid amplification assays, PCR is susceptible to false negative results, particularly for female urine. In one study, the prevalence of inhibitors causing complete inhibition of amplification was 4.9% for PCR, 2.6% for LCR, and 7.5% for TMA. Logistic regression analysis revealed that inhibition in female urine was associated with: beta-human chorionic gonadotropin (odds ratio [OR], 3.3) and crystals (OR, 3.3) for PCR. Most (84%) inhibition was removed by storage overnight at +4 or -70 degrees C and dilution of the specimen 1:10 [Mahoney et al., 1998].  

Clinical studies

Here we are concerned only with studies from 1997 onwards. Some studies involving comparisons with other tests are dealt with elsewhere on the site.

In 1997 a major multicentre trial of both the automated COBAS AMPLICOR and semi-automated AMPLICOR CT/NG was performed. Test performance compared to culture was evaluated for 2,236 matched endocervical swab and urine specimens obtained from women and for 1,940 matched urethral swab and urine specimens obtained from men. Culture-negative, PCR-positive specimens that tested positive in a direct immuno-fluorescence test or by an alternative PCR were considered true positives [see: discrepant analysis]. The overall prevalences of chlamydial infection were 2.4% in women and 7.2% in men. The two AMPLICOR formats were concordant for 98.1% of the specimens. With the infected patient as the reference standard, the resolved sensitivities of COBAS AMPLICOR^® were 89.7% for endocervical swabs; 89.2% for female urines; 88.6% for male urethral swabs and 90.3% for male urines. The specificities were 99.4% for endocervical swabs; 99.0% for female urines; 98.7% for male urethral swabs  and 98.4% for male urines. The internal control revealed that 2.4% of the specimens were inhibitory when initially tested but valid results were obtained for 98.6% of the specimens because a majority of the inhibitory specimens were not inhibitory when a second aliquot was tested. It was concluded that both AMPLICOR formats for C. trachomatis testing exhibited equally high sensitivity and specificity for urogenital swabs and urine specimens and that these assays are well suited for screening for C. trachomatis infection [van der Pol et al., 2000].

Another major multicentre study of the fully automated COBAS AMPLICOR^® was performed on 2,014 endocervical swabs and 1,278 urine specimens obtained from women and for 373 urethral swabs and 254 urine specimens obtained from men, again with culture as the reference test and with an alternative PCR directed against the chlamydial major outer membrane protein (MOMP) gene as the arbitrator of discrepant results. Overall prevalence of chlamydial infection was 4.3% in cervical swabs and 11.0% in urethral swabs from men. For the subset of patients from whom both swab and urine specimens were available, the combined results for both specimen types were used to identify all infected patients. On this basis, the resolved sensitivities for the COBAS AMPLICOR^® PCR  test were 82.6% for endocervical swabs; 84.4% for female urines; 84.2% for urethral swabs from men and 89.5% for male urines. However the sensitivity of culture was only 56.5% for endocervical specimens and 63.2%  for male urethral specimens. Again the conclusion was that the COBAS AMPLICOR^® CT/NG test used for C. trachomatis showed equally high sensitivities and specificities with both urogenital swab and urine specimens and was well-suited for use in screening [Vincelette et al., 1999].  

In Europe, Basiri, Mardh and Domeika, 1997 used the Roche AMPLICOR^® CT/NG PCR to screen chlamydial or gonococcal infections in urines from 3,340 asymptomatic women attending European health care units for contraceptive advice or pregnancy termination. All samples were kept frozen (-20 degrees C) prior to testing, a procedure which also reduces the effects of urine PCR inhibitors. Chlamydia-positive samples were retested once by the plasmid-based PCR kit and also by a MOMP-based PCR. Discrepancies were resolved by using a direct immunofluorescence (DIF) test on urinary sediment. Samples positive for N. gonorrhoeae were retested by chromosomal primer-based PCR and verified by a 16S rRNA PCR. Overall 74 of 81 samples were positive for C. trachomatis by both plasmid- and MOMP-based PCRs; 6 samples were positive by plasmid-based PCR and DIF, and one sample was positive by both MOMP-based PCR and DIF. It was noted that a stong point of the test is the unique ability to identify urine or tissue inhibitors of amplification using the optional internal control [Basiri, Mardh and Domeika, 1997]. 

A new approach for screening is the use of intravaginal swabs transported through the mail in a dry rather than wet state. Gaydos et al., 2002 found that the sensitivity of dry swabs versus true positives was 91.3% among female military recruits and the specificity was 99.3%. This is attractive in non clinic settings, particularly where there are obstacles to accessing medical care [Gaydos et al., 2002 ]. It is also valuable [Stary, Personal communication] where the diagnostic laboratory, either for work scheduling purposes or reagent economy, wishes to retain flexibility as to which test will be used on receipt of the specimen. A second approach to screening is to combine a screen for cervical C. trachomatis infection with the routine, periodic, cervical cytology screen. The COBAS AMPLICOR^® CT/NG test was found to be satisfactory when used for testing cervical cytology samples collected in PreservCyt transport medium (ThinPrep Pap test: Cytyc Corporation, Boxborough, Mass., USA) [Bianchi et al., 2002; see: screening at cervical cytology exam]. 

In performance, the Roche COBAS AMPLICOR and Abbott LCx ^® ligase chain reaction were found to be broadly similar, with kappa coefficients of consistency of 0.93 and 0.95 for swab and urine specimens respectively [Puolakkainan et al., 1998]. A problem has been reported with the gonococcal component of the COBAS AMPLICOR^® CT/NG PCR cross reacting with certain strains of non-pathogenic Neisseria. In some of these strains, the target sequence is reportedly identical to N. gonorrhoeae whereas other strains have a small number of sequence mismatches within the regions recognised by the test. An equivocal zone-retesting algorithm has been suggested to cope with this problem [van der Pol et al., 2001]. 

Rectal specimens

No chlamydial diagnostic test is licensed for use for the detection of chlamydiae in rectal material. Nevertheless where chlamydial proctitis is suspected, some indication of the presence of chlamydial infection could be useful. Golden et al., 2003 tested rectal specimens from men who have sex with men for C. trachomatis by using the COBAS PCR (Roche Diagnostics) and ligase chain reaction LCR (Abbott laboratories). They also compared three different PCR specimen-processing procedures for these specimens. Chlamydiae were detected by one or more procedures in 22 of 186 specimens. All three PCR tests were positive for 17 specimens, all of which also tested positive by LCR.

NEXT: ligase chain reaction; the Abbott LCx^®

Golden, M. R., Astete, S. G., Galvan, R., Lucchetti, A., Sanchez, J., Celum, C. L., Whittington, W. L., Stamm, W. E., Holmes, K. K. & Totten, P. A. (2003). Pilot Study of COBAS PCR and Ligase Chain Reaction for detection of rectal Infections due to Chlamydia trachomatis. Journal of Clinical Microbiology 41, 2174 - 2175.