JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 1990, p. 1770-1773

Vol. 28, No. 8

0095-1137/90/081770-04$02.00/0 Copyright C 1990, American Society for Microbiology

Confirmatory Assay Increases Specificity of the Chlamydiazyme Test for Chlamydia trachomatis Infection of the Cervix JEANNE MONCADA,' JULIUS SCHACHTER,1t* GAIL BOLAN,2 JOSEPH ENGELMAN,2 LAWRENCE HOWARD,3 ISA MUSHAHWAR,3 GEOFF RIDGWAY,4 GILANFAR MUMTAZ,4 WALTER STAMM,5 AND AGNES CLARK5 Department of Laboratory Medicine, University of California, San Francisco, California 94143'; San Francisco City

Clinic, San Francisco, California 941032; Abbott Laboratories, North Chicago, Illinois 600643; Department of Clinical Microbiology, University College Hospital, London, England WCIE 6AU4; and Department of Medicine, University of Washington, Seattle, Washington 981045 Received 29 January 1990/Accepted 4 May 1990

Enzyme immunoassays for the detection of chlamydial antigens are commonly used to diagnose Chlamydia trachomatis infection. As is true for ail nonculture methods, the specificities of these tests are a concern. A confirmatory blocking assay (Abbott Laboratories, North Chicago, Ill.) was evaluated at four sexually transmitted disease test sites. This assay is designed to confirm true-positive Chlamydiazyme (CZ) specimens and to identify false-positive CZ reactions caused by cross-reacting bacteria. Cervical specimens were collected from 2,891 women. Chlamydia prevalence by tissue culture (TC) was 9.2% (266 of 2,891 specimens). Compared with TC, the sensitivity and specificity of CZ were 78.9% (210 of 266 specimens) and 98.2% (2,577 of 2,625 specimens), respectively. There were 48 CZ false-positive reactions. The direct fluorescent-antibody test (DFA) was positive for 31 of 48 false-positive reactions, indicating culture misses. Thus, when the standard was both TC and DFA, CZ sensitivity was 81.1% and CZ specificity was 99.3%. Of the 17 CZ-positive patients who were negative by both TC and DFA, 3 were negative on repeat CZ and Il of 14 were identified as false positive by the confirmatory assay. The confirmatory test was positive for CZ-positive women who were positive by TC or DFA. Use of the confirmatory test, which increased the specificity to 99.9%, would increase confidence in positive CZ results and make the test more useful for screening populations with a low prevalence of C. trachomatis infection.

The "gold standard" for identification of Chlamydia trachomatis is tissue culture (TC) isolation of the agent (11). However, this is a time-consuming, labor-intensive, and relatively costly technique. Direct antigen detection methods provide a rapid and less expensive alternative, and many laboratories now test for chlamydia by enzyme immunosorbent assay (EIA). Of the commercially available EIA kits, Chlamydiazyme (CZ; Abbott Laboratories, North Chicago, 111.) is the most widely used. A major limitation of the CZ assay in previous studies of urogenital samples has been its reported specificity of 88.6 to 97.5% when it is compared with the culture method (1, 3, 6, 15). These data suggest that EIA should be used mainly in populations with a high prevalence of C. trachomatis infections and not in low-risk populations, in whom the predictive value of positive tests is lower. However, a large reservoir of chlamydial infections is found in these low-risk populations. Clinically inapparent infections are common in both men and women (14, 16). If the problem of nonspecificity and low predictive value of a positive test could be solved, then EIA could be used as a screening test in low-risk populations. False-positive reactions can occur because TC is 2.0 >2.0 >2.0 >2.0

1.139 0.563 0.248 0.146

1.924 0.408 0.228 0.038

0.015 0.006 0.002 0.001

trachomatis antibody of a CZ kit to give a final concentration of 5 pug/mi. The blocking assay was performed exactly as described above for the CZ assay, except that the blocked antibody preparation was used in place of the unmodified anti-C. trachomatis antibody solution. The CZ retest and the confirmatory assay were run on the majority of TC- and CZ-positive samples and on all TC-negative and CZ-positive samples. Samples which had an initial absorbance of >2.0 were diluted 1/10 in the specimen dilution buffer before retesting. If the sample still yielded an absorbance of >2.0, the assays were repeated with further dilution until an absorbance reading of 50% was measured with up to 2 x 105 elementary bodies. Inhibition of 50% or more was hot observed with larger numbers of elementary .bodies. Blocking reaction with various serovars. Table 2 shows the results for all C. trachomatis serovars, various Chlamydia psittaci serovars, and Chlamydia pneumoniae in the regular CZ and confirmatory assays. The chlamydial LPS is a genus-specific antigen. Thus, as expected, all serovars within the three chlamydial species that were tested were detected in the CZ assay. AUl ODs were reduced by at least 95% compared with the ODs in the unblocked assay. Clinical study. The overall prevalence of C. trachomatis among the women tested was 9.2% (266 of 2,891 specimens) by culture (Table 3). The CZ assay was positive for 210 of 266 culture-positive specimens (sensitivity, 78.9%) and 48 of 2,625 culture-negative specimens (specificity, 98.2%). HowTABLE 3. CZ assay performance profiles by standard EIA and the blocking assay' No. of specimens

CZ test result

~+

+

+

-

210 56

TPc

Tpb

TC

48

241

17

2,577

56

2,577

+

_

241 56

2,591

3

a The sensitivities, specificities, positive predictive values, and negative predictive values for CZ and TC, CZ and TP, and CZ confirmed by the blocking assay and TP were 78.9, 98.2, 81.4, and 97.9S%; 81.1, 99.3, 93.4, and and 81.1, 99.9, 98.95, and 97.8%, respectively. 97.9%; b TP is TC and DFA resolution of discrepant results. c CZ results were confirmed by blocking assay.

ever, DFA was positive for 31 of these 48 specimens, indicating that they were missed by the culture method. Thus, when the standard was either TC or DFA positivity of discrepant results, the specificity of the CZ assay was 99.3% (2,577 of 2,594 specimens), the sensitivity was 81.1% (241 of 297 specimens), and the prevalence was 10.3% (297 of 2,891 specimens). TC sensitivity was 89.6% (266 of 297 specimens). Of the 17 CZ-positive patients who were negative by both TC and DFA, three were negative in the repeat CZ assay and 11 of 14 were negative by the confirmatory assay. All 17 specimens were also cytospin-FA stain negative. Reisolation attempts on three samples from San Francisco were negative. The confirmatory test was positive for all 147 specimens tested from CZ-positive women who were positive by TC or DFA. Use of the dual criteria of reproducibility of the positive CZ result and .50% reduction of OD in the confirmatory assay increased the specificity of the CZ assay to 99.9%. Table 3 shows the new performance profiles with the use of the confirmatory test.

DISCUSSION We found that 64.6% (31 of 48 specimens) of the initial false-positive CZ specimens were true positives, as indicated by a third test (DFA) and the confirmatory assay. These findings are not surprising, since TC is

Confirmatory assay increases specificity of the chlamydiazyme test for Chlamydia trachomatis infection of the cervix.

Enzyme immunoassays for the detection of chlamydial antigens are commonly used to diagnose Chlamydia trachomatis infection. As is true for all noncult...
834KB Sizes 0 Downloads 0 Views