Rapid detection of Chlamydia antigens 
Clinical and laborato,y study 
COMPARISON OF ANTIGEN DETECTION 
ASSAYS AND SERUM ANTIBODY TESTS 
FOR DIAGNOSIS OF INFECTION 
WITH CHLAMYDIA TRACHOMATIS 
P. Auer-Grumbach, D. Stuenzner, E.M. Haller, H. Kessler, K. Pierer and E. Marth 
ABSTRACT 
Objectives. Rapid antigen detection in patients with lower urinary tract infection caused by Chlamydia 
trachomatis by direct immunofluorescence and an enzyme-linked immunosorbent assay (ELISA) was compared 
with cell culture as well as polymerase chain reaction (PCR). Furthermore, the performance of two serum 
antibody ELISAs was compared with an immunoperoxidase test. 
Methods. A total of 143 urethral, conjunctival, and pharyngeal swabs were investigated. McCoy cell culture 
and a molecular assay including PCR served as standard methods. An elementary body direct fluorescent-
antibody assay (DFA) was performed on urethral and conjunctival specimens, and an ELISA antigen test was 
performed on urethral specimens. Serum samples of 49 patients were analyzed for lgA and IgG antibodies. 
Results. Antigen ELISA testing of urethral specimens was comparable to culture and PCR concerning 
specificity and sensitivity. The antibody ELISAs were significantly better in the detection of lgA in comparison 
with the immunoperoxidase test. 
Conclusions. The antigen ELISA was found to be the best rapid diagnostic assay and has sufficient 
specificity as well as sensitivity. PCR can serve as gold standard equivalent to cell culture . Serum antibody 
assays are only of value in clinical situations when antigen detection is not possible. 
KEY WORDS 
Chlamydia trachomatis, PCR, cell culture, antigen detection 
INTRODUCTION 
Clinical features of chlamydial infection include 
acute and chronic disorders as urethritis, cervicitis, 
endometritis, salpingitis, perihepatitis, and pelvic 
inflammatory disease (1-4 ). Sensitive, specific and 
rapid techniques for the diagnosis of infections 
produced by C. trachomatis are thus of great 
importance. Ascending endocervical infection is an 
important cause of infertility and ectopic pregnancy 
acta dermatovenerologica A.P A. Vol 4, 95, No 4 
and may promote the transmission of the human 
immunodeficiency virus (5). Furthermore, C. tracho-
matis is a frequent cause of newborn and adult 
keratoconjunctivitis (6,7), and even newborn respiratory 
tract infection has been reported (8). Its possible 
role as a triggering agent in rheumatic diseases 
remains in discussion (9). 
Culture on McCoy cell monolayers is considered 
the standard for detection of infectious elementary 
179 
Rapid detection of Chlarnydia antigem· 
bodies. Results of chlamydial isolation in McCoy 
celi culture are available after 48 - 72 hours. Molecular 
assays including PCR have been found to be a 
good alternative for the detection of C. trachomatis 
(10-12). However, development of specimens still 
needs more than four hours. More rapid techniques 
of chlamydial detection include use of a direct 
fluorescent-conjugated antibody staining (DPA), and 
an enzyme-linked immunosorbent assay (ELISA) (13-
15). Serum antibody tests have been introduced to 
facilitate diagnosis in clinical situations without antigen 
detection, but only elevated titers are indicative of 
active chlamydial infection (16). 
We previously observed that patients with urethritis -
caused by C. trachomatis have occasionally conjunc-
tivitis at the same tirne ( data not published). In the 
present study, standard detection of C. trachomatis 
was done with cell culture and PCR in urethral, 
pharyngeal and conjunctival specimens to evaluate 
concomitant infection. DFA was additionally performed 
on urethral and conjunctival specimens, antigen ELISA 
was done on urethral specimens. Serum IgA and 
IgG antibodies were measured with three different 
test systems to evaluate their performance in com-
parison with McCoy cell culture as well as PCR. 
MATERIALS AND METHODS 
Specimens 
Seventy-three urethral specimens were collected 
from 50 women and 23 men with symptoms of 
urethritis attending the outpatients clinic of the 
Department of Dermatology of the Graz University 
Hospital. The mean age of the patients was 26 
years (age range, 17 to 72 years). Three dacron-
tipped swabs were obtained from the urethra and 
used in randomized order for DP A, for ELISA and 
for cell culture and the molecular assay, respectively. 
The swab for molecular assay and culture was 
immediately transferred to 1.5 ml of 2-SP medium 
and stored frozen at -70°C until assayed. Swabs 
were also obtained from conjunctivae (28 patients) 
and pharynx (42 patients). Venous blood was collected 
in 10 ml vials (Sarstedt, Nuembrecht, Germany) 
from 71 patients with lower urinary tract infection. 
The sera were centrifuged for 10 minutes at 3000 
rpm and stored at -20°C until examination. 
Celi culture assay and PCR 
Cell culture and PCR were done as recently 
reported by Kessler et al (12). Shortly, shell vials 
180 
containing McCoy cell monolayers (ATCC, Rockville, 
MD, USA) were used for culture. Staining was 
performed with a direct fluorescent monoclonal 
antibody stain (Syva, Palo Alto, CA, USA). The 
presence of one or more typical inclusions was 
considered a positive result. 
Por the molecular detection, the Amplicor PCR 
assay (Roche Molecular Systems, Branchburg, NJ, 
USA) was used. If the OD 
450 
value was greater than 
0,25 OD units, the sample was considered positive. 
DFA and ELlSA-test 
Por DPA, the Chlamydia trachomatis Direct Spe-
cimen Test (Syva, Palo Alto, CA) was used. Slides 
were prepared immediately after specimen collection. 
Within 1 h of collection, staining as well as mounting 
and reading were done in accordance with the 
manufacturer's advice. The presence of 5 or more 
chlamydial elementary bodies was needed to be 
considered positive. 
The Chlamydia EIA (Pharmacia, Uppsala, Sweden) 
was done immediately after specimen collection. 
Specimens and controls were added to the wells of 
a inicrotitration strip within 15 minutes and processed 
according to the manufacturer's recommendations. 
Absorbance was measured at 405 nm. Values greater 
as the cut off value (mean absorbance value for 
three negative controls + 0.05) were considered 
positive. 
Serum antibodies 
The presence of serum antibodies of the IgA and 
IgG class was tested in 71 patients with indirect 
IPAzyme chlamydia immunoperoxidase test (Savyon, 
Beer-Sheva, Israel) with serum dilutions of 1:16 for 
IgA as well as 1:64 and 1:128 for IgG. In 49 
patients, IgA and IgG antibodies were determined 
additionally with the SeroELISA test (Savyon, Beer-
Sheva, Israel) and the rELISA test (medac, Hamburg, 
Germany). Both tests were performed according to 
the manufacturer's advice. 
The SeroELISA tests employs the L2 serovar 
broadly reacting antigen of C. trachomatis. In a first 
step, the antigenic material was bound in microtiter 
wells. Specific antibodies, if present in patient serum, 
bind to the attached antigen forming an antigen-
antibody complex. Then, horseradish peroxidase 
conjugated anti-human IgG (gamma- or alp,ha-chain 
specific) was added. The peroxidase conjugated 
antibody binds to the antibody moiety of the antigen-
antibody complex. In a third step, substrate was 
acta derrnatovenerologica A_P A. Vol 4, 95, No 4 
Rapid detection of Chlamydia antigem· 
added. It reacts with the peroxidase and changes its 
color to blue. After the enzymatic reaction was 
stopped, the absorbance was determined at 450 nm. 
Values greater than the cut off value (0,145 x 
[ absorbance of positive control minus absorbance of 
negative control] plus absorbance of negative control) 
are considered positive . 
The rELISA test is based on a fragment of the 
genus specific lipopolysaccharide in the outer 
membrane of elementary and reticulate bodies. The 
test principle is similar to the SeroELISA. Instead 
of horseradish, ABTS substrate is used as chromogen. 
The absorbance is determined at 405 nm. Each test 
is run with instrument blank, two negative controls, 
two positive controls and two borderline controls 
(human serum containing anti-chlamydial antibodies). 
Values greater than the cut off (mean of borderline 
controls) are considered positive. 
Statistics 
Sensitivities, specificities, prevalences (i.e. pretest 
odds of disease ), odds ( = likelihood) ratios for the 
diagnostic tests and post-test odds for disease were 
calculated using a 2x2 table and Bayes' theorem for 
ali tests (17) . Culture and PCR were taken as gold 
standard. Significance testing was done by using the 
chi-square analysis with Yates' correction for small 
numbers (18). 
RESULTS 
The positivity rate for lower urina! tract infection 
caused by Chlamydia trachomatis was 16,4% (12 
from 73) defined as positive culture and positive 
PCR assay. Culture and PCR revealed identical 
results. The differences in the test performance 
between culture as well as PCR and ELISA were 
not significant (p > 0.05), whereas DFA differed 
significantly (p < 0.05). DFA had a sensitivity of 
25%, and ELISA of 33%. Specifity was 98% in 
both tests (Table 1). Of 50 female patients 44 were 
negative for infections with Chlamydia trachomatis. 
In this group, Ureaplasma urealyticum was found in 
10 patients, Candida albicans in 6, Gardnerella 
~aginalis in 3, double infections in 4, and no germs 
m 17 patients. Of 23 male patients, 17 were negative 
for Chlamydia trachomatis. In 5 of these, Ureaplasma 
urealyticum was detected, 2 were infected by 
Mycoplasma hominis, and 2 had double infections. 
In 7 men, no germ could be isolated. Thus, 24 
patients (17 women and 7 men) could not be 
acta dermatovenerologica A .P A. Vol 4, 95, No 4 
diagnostically cleared. 
In conjunctival swabs, one female patient tested 
positive in both eyes with PCR and negative with 
culture and DFA. She was considered to have 110 
eye infection, and no urethral germ could be identified. 
She had chlamydial antibodies of the IgG class, but 
no IgA antibodies. No pharyngeal infection could 
be detected with PCR or McCoy culture. 
Three different antibody detection systems were 
compared. Seventy-one sera were tested with the 
IP Azyme, and 49 of them additionally with the 
other two ELISAs. Test performances were calculated 
with cell culture and PCR as reference methods. 
The results are shown in table 2 for IgA and IgG 
detection. Concerning IgA antibodies, IP Azyme and 
SeroELISA as well as IP Azyme and rELISA differed 
significantly (p < 0.05). However, SeroELISA and 
rELISA gave no significant difference. 
In the detection of IgG antibodies, a significant 
difference between IPAzyme titer of 1:64 and 
SeroELISA was found (p < 0.01), but no statistical 
difference between IPAzyme titer of 1:64 and rELISA 
was found. The difference between SeroELISA and 
rELISA was also significant (p < 0.02). There was 
no statistical difference in the comparison of IP Azyme 
titre 1:128 with SeroELISA or rELISA, respectively. 
Taken together, both ELISAs performed better than 
the immunoperoxidase test. In the detection of IgG 
antibodies, the SeroELISA proved to be superior to 
the rELISA. There was no difference between the 
two ELISAs in the detection of IgA antibodies. 
DISCUSSION 
Detection of C. trachomatis by cultivation on cell 
monolayers is still considered the "gold standard". 
This method bas the disadvantage of being a ]abor-
intensive and time -consuming procedure . The 
difficulties in transportation of live organisms and 
low copy number in the specimens may additionally 
complicate detection (19). Therefore, PCR has recently 
been introduced for detection of C. trachomatis and 
found to be a good alternative to culture (10-12). 
For outpatient clinics, cell culture as well as PCR 
are not feasible. Culture needs at minimum 50 
hours, and PCR can be performed within 5 hours. 
Stand-by tests like DF A and ELISA are easy to 
perform, but less sensitive and less specific than cell 
culture and PCR. Using DFA, results can be obtain.ed 
within one hour after sampling. The antigen ELISA 
needs three hours. A common standard for antibody 
181 
Rapid detection of Chlamydia antigen.s· 
Table 1: Comparison oj PCR, DFA and ELISA with cell culture in 73 urethral specimens. McCoy cell culture 
results are taken as gold standard. 
PCR + PCR - DFA + DFA - ELISA + ELISA -
culture + 12 o 3 9 4 8 
culture - o 61 1 60 1 60 
sensitivity 100% 25% 33% 
specificity 100% 98% 98% 
PTD 00 o 2.50 0.13 3.34 0.08 
PCR = polymerase chain reaction; DF A = elementa,y body direct fluorescent-antibody assay; PTD = post test 
likelihood far disease 
detection is not yet established. In this study, 
performance of commercially available kits was tested 
to evaluate the value of these tests in the spectrum 
of diagnostic tools. 
Sensitivity and specificity describe the performance 
of a test. A clinician is interested to reassure or 
reject a diagnosis. To facilitate this decision, we 
calculated prevalence (pretest likelihood of disease ), 
Table 2: Comparison oj serum antibody detection by indirect IPAzyme immunoperoxidase test, SeroELISA test and 
rELISA test with antigen detection by cell culture and PCR in urethral specimens. 
SeroELISA 
Immunoglobulin 
+ + 
antigen + 1 11 3 4 
antigen- 2 57 5 37 
sensitivity 8% 42% 
specificity 96% 88% 
PTD 0.41 0.19 0.51 o.os 
Immunoglobulin 
1:64 + 1:64 - + 
antigen + 6 6 2 5 
antigen - 26 33 8 34 
sensitivity 50% 28 % 
specificity 55% 80% 
PTD 0.19 0.13 0.21 0.13 
PCR = polymerase chain reaction; 
DFA= elementary body direct fluorescent - antibody assay; 
PTD= post test likelihoods for disease. 
A 
+ 
2 5 
7 35 
28% 
83% 
0.24 0.11 
G 
+ 
3 4 
19 23 
42% 
54% 
0.13 0.18 
182 acta dermatovenerologica A.P A. Vol 4, 95, No 4 
Rapid detection of Chlamydia antigens 
like liho od ( odds) ratios far positive and negative 
test fesults and the post-test likelihood of disease 
(PTD). Prevalence multiplied with likelihood ratios 
gives post-test likelihoods of disease. PTD is related 
to a clinical question and is independent from 
sensitivity and specificity. On the other hand, 
sensitivity and specificity describe the quality of a 
test in a laboratory situation. With this approach, 
the clinical infarmation content of the diagnostic 
tests is considerably increased. PTD help the clinician 
to decide whether an infection is caused by C. 
trachomatis or not (17). PCR and cell culture gave 
identical results in the investigation of urethral 
specirnens. PCR is therefare a good alternative to 
culture and should be readily accepted as an equivalent 
"gold standard". 
In comparison with culture and PCR, only ELISA 
antigen testing proved to be of sufficient sensitivity 
and specificity. ELISA sensitivity of 33% and DFA 
sensitivity of 25% were not very high in this study. 
This might be due to the fact that some patients 
came to the clinic after various antibiotic treatments 
initiated by their doctors and the lower sensitivity of 
these tests compared to the gold standard. The 
difference between positive PTD and negative PTD 
was greater in ELISA than in DF A. This may 
partially be due to test characteristics of the DF A 
test. It is a subjective method requiring trained 
staff. There is no possibility of automatization. 
Subjective influences as well as technical equipment 
of the tluorescent microscope influence the outcome. 
ELISA tests are ideal far routine work with larger 
sample sizes, because automatization is possible, but 
these tests need more tirne. Today, a "stand-by" 
diagnostic service can be offered with DFA, but at 
the cost of quality. We faund ELISA to be more 
sensitive than DFA. As ELISA can also be performed 
as a stand-by service, it is the best alternative to 
PCR in smaller laboratories. 
As there were no pharyngeal and conjunctival 
infections, an evaluation was not possible. Coincidental 
pharyngeal and conjunctival infections with Chlamydia 
trachomatis are obviously rare (16). 
If antigen detection is not possible, antibody 
detection may give useful infarmation to the clinician. 
Recently, new antibody tests have been developed 
to irnprove the diagnostic value either by measuring 
different immunoglobulin classes (20) or by using 
different antigens. The SeroELISA test is based on 
the L
2 
serovar and employs both the major outer 
membrane protein (MOMP) and a lipopolysaccharide 
(LPS) as antigens to detect early group specific 
antibodies against Chlamydia trachomatis (MOMP) 
and late genus specific antibodies against Chlamydia 
psittaci, pneumoniae and trachomatis (LPS) . The 
rELISA uses a recombinant genus specific LPS 
fragment as antigen and is designed to detect 
antibodies that arise earlier than anti-MOMP 
antibodies. These tests can not compete with antigen 
detection as shown by comparison of corresponding 
PTD values. Pattern and tirne course of antibody 
development do still not allow a diagnosis of acute 
chlamydial infection with sufficient confidence, although 
the performance of these newer tests has been 
irnproved compared to the IP Azyme test. Thus, 
despite better discrimination of the antibody ELISAs 
as shown by PTD values, they are stili only of value 
in chronic disease without the possibility of antigen 
detection to support clinical management. 
In summary, PCR is a good alternative to cell 
culture. Antigen detection with ELISA remains an 
alternative for stand-by service in outpatient clinics. 
Antibody testing maybe useful in clinical situations 
where the detection of the antigen is not possible, 
but can not compete with any antigen detection 
system. 
ACKNOWLEDGMENTS 
We want to thank Mss. E. Probst, H Schaupperl, E. 
Frohlich and M Joch for their technical assistance. 
REFERENCES 
l. Schachter J. Chlamydiae. In: Balows A, Hausler 
WJ Jr, Herrmann KL, Isenberg HD, Shadomy HJ, 
eds. Manual of Clinical Microbiology, 5th ed. 
Washington DC: American Society far Microbiology, 
1991; 1045-1053. 
2. Mueller BA, Luz-Jirnenez M, Daling JR, Moore 
DE, McKnight B, Weiss NS . Risk factors far tuba! 
infertility. Influence of history of prior pelvic infla-
acta dennatovenerologica A.P A. Vol 4, 95, No 4 
mmatory disease. Sex Transm Dis 1992; 19:28-34. 
3. Safrin S, Schachter J, Dahrouge D, Sweet RL. 
Long-term sequelae of acute pelvic inflammatory 
disease. A retrospective cohort study. Am J Obstet 
Gynecol 1992; 166: 1300-1305. 
4. Paavonen J, Wolner-Hanssen P. Chlamydia ti-a-
chomatis: a major threat to reproduction. Hum 
Reprod 1989; 4: 111-124. 
183 
Rapid detection of Chlamydia antigeni' 
5. Plummer F, Simonsen J, Cameron DJ et al. 
Cofactors in male-temale sexual transmission of human 
immunodeficiency virus type l. lnfect Dis 1991; 163: 
233-239. 
6. Sandstrom KI, Bell TA, Chandler JW, et al. 
Microbial causes of neonatal conjunctivitis. J Pediatr 
1984; 5: 706-711. 
7. Shafer MA, Schachter J, Moscicki AB, et al. 
Urinary leukocyte esterase screening test tor asymp-
tomatic chlamydial and gonococcal infections in males. 
JAMA 1989; 262: 2562-2566 
8. Schachter J. Chlamydiaceae: The Chlamydiae. In: 
Balows A, Hausler WJ Jr, Lenette EH, eds. Laboratory 
Diagnosis of lnfectious Diseases. Principles and 
Practice. Vol II: Viral, Rickettsial, and Chlamydial 
Diseases. 1st ed., New York (NY): Springer, 1988: 
847-863 
9. Scott Pollock P, Hunter Handsfield H. Arthritis. 
In: Holmes KK, Mardh PA, Sparling PF, Wiesner 
PJ eds. Sexually Transmitted Diseases. 1 st ed., New 
York (NY): McGraw-Hill, 1984: 745-758. 
10. Bass CA, Jungkind DL, Silverman NS, Bondi 
JM. Oinical evaluation of a new polymerase chain 
reaction assay tor detection of Chlamydia trachomatis 
in endocervical specimens. J Oin Microbiol 1993; 
31: 2648-2653. 
11. Talley AR, Garcia-Ferrer F, Laycock KA, 
Loeffelholz M, Pepose JS. The use of polymerase 
chain reaction tor the detection of chlamydial 
keratoconjunctivitis. Am J Ophthalmol 1992; 114: 
685-692. 
12. Kessler, HH, Pierer K, Stunzner D, Auer-
Grumbach P, Haller EM, Marth E. Rapid detection 
of Chlamydia trachomatis in conjunctival, pharyngeal, 
and urethral specimens with a new polymerase chain 
reaction assay. Sex Transm Dis 1994; 21: 191-195. 
13. Ridgway GL, Taylor-Robinson. Current problems 
in microbiology: chlamydial infections: which laboratory 
test? J Oin Pathol 1991; 44: 1-5. 
14. Siena MF, Clarke LM, Boyle JF. The laboratory 
diagnosis of chlamydial infections. Lab Med 1988; 
19: 311-314. 
15. Moncada J, Schachter J, Bolan G, et al. Evaluation 
of Syva's enzyme immunoassay tor the detection of 
Chlamydia trachomatis in urogenital specimens. Diagn 
Microbiol Infect Dis 1992; 15: 663-668. 
16. Csango PA, Sarov B, Schiotz H, Sarov I. 
Comparison between cell culture and serology tor 
detecting Chlamydia trachomatis in women seeking 
abortion. J Oin Pathol 1988; 41: 89-92 
17. Sackett DL, Haynes RB, Tugwell B. Oinical 
epidemiology. A basic science tor clinical medicine. 
1st ed. Boston: Little, Brown & Comp, 1985; 100-126. 
18. Altmann DG. Practical statistics for medical 
research. 1 st ed., London: Chapman and Hall, 
1991; 244-258. 
19. Schachter J . Biology of Chlamydia trachomatis. 
In: Holmes KK, Mardh PA, Sparling PF, Wiesner 
PJ, eds. Sexually transmitted diseases. 1st ed., New 
York: McGraw-Hill, 1984; 243-257. 
20. Theunissen JJH, van Heist BYM, Chien-A-Lien 
RAM, Wagenvoort JHT, Stolz E, Michel MF. 
Detection of IgG, IgM and IgA antibodies in patients 
with uncomplicated Chlamydia trachomatis infection: 
a comparison between enzyme linked immunofluo-
rescent assay and isolation in cell culture. Int J 
STD & AIDS 1993; 4: 43-48. 
AUTHORS' ADDRESSES 
184 
Piet Auer-Grumbach MD, Department of Dermatology and Venerology, University of Graz, 
Auenbruggerplatz 8. A-8036 Graz, Austria 
Doris Stuenzner MD, GhD, Department of Hygiene Laboratory Unit, University of Graz 
Harald Kessler MD, same address 
Karin Pierer MD, same address 
Egon Marth MD, PhD, Professor and Head of the Department of Hygiene, same address 
Eva-Maria Haller MD, Department of Ophthalmology Laboratory Unit, University of Graz, Austria 
acta dermatovenerologica A.P A. Vol 4, 95, No 4