International Journal of srD & AIDS 1992: 3: 355-359
ORIGINAL ARTICLE
Chlamydia trachomatis antigen detection by Chlamydiazyme combined with Chlamydia Blocking Reagent verification Barbre Zeeberg
I,
Ingrid Thelin 2, and Claes Schalen MOl
Departments of lMedical Microbiology and2Dermatovenereology, LundUniversity Hospital, 5-22185 Lund, Sweden Summary: Several options exist for the detection of chlamydial infection in a routine laboratory setting. Enzyme immuno assay (EIA) technology offers rapid turn around of results and is less technically demanding than chlamydial cell culture. In addition, recently introduced EIA confirmatory reagents have the potential to improve the accuracy of EIA detection. We have evaluated one such confirmatory reagent (Chlamydia Blocking Reagent" , Abbott Laboratories) to determine the accuracy of the Chlamydiazymev EIA with special regard to interpretation of low absorbance values. An initial series of 192 male urethral specimens showed that use of a lowered cut off level (absorbance value 0.05) compared with that recommended by the manufacturer increased sensitivity of the EIA from 0.73 to 0.83, thus motivating studies on this interpretative modification. Of 1101 EIA reactive specimens, 65% were determined to be chlamydia positive by the Chlamydia Blocking Reagent. The proportion of female cervical specimens that did not confirm positive was elevated compared with male urethral specimens, 43% vs. 5.7% respectively. In samples yielding absorbance from the recommended cut off level to 0.05 (approximately 50% below), the corresponding figures were 78% and 14% respectively. In 85 selected EIA reactive samples, examination by a direct immunofluorescence staining assay (OFA) (Micro'Trak'" , Syva Inc.) revealed elementary bodies in 85% of 67 blocking test positive and in 24% of 18 blocking test negative samples. The possibility that Gram-negative bacteria were responsible for unconfirmed EIA reactive specimens was investigated using bacterial suspensions. While EIA reactivity was noted with several strains for Gram-negative bacteria, both the blocking reagent and OFA correctly verified the absence of chlamydial antigen. We conclude that confirming Chlamydiazyme initial reactive samples with the Chlamydia Blocking Reagent improved the accuracy of chlamydial detection by identifying false positive specimens. Furthermore, by including confirmatory testing, a lower cut off level than recommended by the manufacturer seems applicable, at least for male urethral samples, thus increasing sensitivity of the Chlamydiazyme EIA. Keywords: Chlamydia trachomatis, antigen detedion, Chlamydiazyme
INTRODUCTION Due to a clinical requirement for rapid analyses, antigen detection tests are increasingly used as alternatives to cell culture in the diagnosis of ~hlamydia trachomatis. On a smaller scale, partiCUlarly in laboratories without facilities for cell culture, direct immunofluorescence assay (DFA), though time-consuming, is widely used; this method reportedly gives a high specificity but a low sensitivity compared with culturing under optimal conditions that may limit its use for low-prevalence Correspondence to: Claes Schalen, Department of Medical Microbiology, Lund University Hospital, S-22185 Lund, Sweden
populationsv". In laboratories processing high numbers of specimens, enzyme-linked immunosorbent assay (EIA) appears to be more feasible than DFA. Although a low sensitivity of available EIA kits has been reported, particularly for urethral specimens, their specificity seems surprisingly high1,5- 10. In contrast to DFA, the EIA method does not permit morphological confirmation, leading to erroneous interpretations, e.g. immunological cross-reactions between chlamydiae and various Gram-negative species'! as well as Fe-mediated bindin§ of immunoglobulins to Staphylococcus aureus' or is-haemolytic streptococci-', The basis of immunological cross-reactions between chlamydiae and Enterobacteriaceae or
356
International Journal of SID & AIDS Volume 3 September/October 1992
Acinetobacter is a common core region of the lipopolysaccharide (LPS)lO. It is claimed by some manufacturers that, in clinical specimens, such bacterial species seldom occur in amounts high enough to complicate the detection of C. trachomatis. The present study was undertaken in order to see to what extent cross-reacting bacteria might account for mis-interpretation of one EIA for C. irachomatis, based on an LPS-specific antiserum. In addition, a newly introduced monoclonal reagent, intended for EIA blocking, was evaluated with special reference to heterologous Gram-negative species and to the possibility of enhancing the performance of the EIA on clinical specimens. METHODS Patient samples From 175male patients with symptoms of urethritis, examined at the Department of Dermatovenereology, two urethral samples for cell culture and EIA, respectively, were obtained. In addition, 1101 consecutivemale urethral or female cervical samples, all positive by EIA, were examined by a recently introduced blockingstep (see below). These specimens were obtained from various STD and primary care centres in the district. Specimens to be examined by EIA were obtained with the swabs and transported in the tube supplied by the manufacturer. Specimens for chlamydial culture were transported in a medium previously describedw,
specimens diagnosed at the Clinical Microbiology Department. Identification was according to standard procedures. Bacterial isolates of staphylococci, streptococci, Enterobacteriaceae, Acinetobacter and Pseudomonas aeruginosa were cultured overnight at 37°C on horseblood agar. The following specieswere cultured on appropriate solid media and in a 6% CO2 atmosphere: Haemophilus influenzae, Moraxella caiarrhalis, Neisseria gonorrhoeae, Mycoplasma hominis, Candida albicans, Lactobacillus sp. Anaerobic species were cultured on blood agar for 2 days anaerobically. The cells were harvested and suspended in 1 ml Chlamydiazyme Specimen Dilution buffer to a cell density of approximately lOS cells/ml. as judged by viable counting. The bacterial cells were tested in the Chlamydiazyme EIA by adding 200ILl of undiluted suspension for each assay. For DFA a l000-fold dilution was applied to the microscope slide. TIle tests were then performed as recommended. RESULTS Examination of various bacterial strains in two detection tests for Chlamydia trachoma tis
Bacterialstrains representing 11 Gram-positive and 11 Gram-negative species (2 strains of each) did not reveal any reactivity by DFA. By EIA, all the Grampositive species (5 strains each of group A, B, C and G streptococci, Staphylococcus aureus, S. epidermidis and S. saprophyticus, 2 strains each of Lactobacillus sp., Peptococcus magna and Clostridium perfringens), Chlamydia! detection methods as well as 2 strains each of Candida albicans and Chlamydial cell culture was performed on cyclo- Mycoplasma hominis were negative. However, all the heximide-treated McCoy cells as described by Ripa Gram-negative species tested, except Pseudomonas et al.15 The cells were stained by iodine after 72 h aeruginosa, were found to be reactive to various incubation. degrees (Table 1) and, in particular, high absorbance The procedure for the EIA Chlamydiazyme'" values were noted for E. coli and Acinetobacter (Abbott) was as suggested by the manufacturer. strains. By titration, these two species required Whereas recommended cut off level was 0.11-0.115, 1()5-1()6 cells to reveal EIA reactivity (Figure 1). samples yielding absorbance> 0.05were considered reactive. The background level was below 0.02. Reactive samples were retested for confirmation in Table 1. EIA (Chlamydiazyme) reactivity of selected bacterial the presence or absence of a monoclonal antibody isolates" (Chlamydia Blocking Reagent'") according to the instructions. Absorbance Tested Positive (range) Further testing of 85 randomly selected EIA Species reactive samples by DFA was done following Acinetobacter sp. 10 8 >2.0-0.162 centrifugation of the remaining suspension at 2100 g Escherichia coli 20 >2.0-0.062 16 for 1/2 h. The pellet after decantation was applied Klebsiella sp. 4 0.766-0.065 1 to microscope slides and stained according to Enterobacter sp. 2 0.538-'0.389 2 the manufacturer with a monoclonal, fluorescein Citrobacter sp. 2 0.605-0.411 2 labelled IgM antibody directed against Chlamydia Proteus sp. 5 0.301 1 10 0.990-0.121 2 trachomatis (MicroTrak, Syva Inc., USA). A minimum Neisseria gonorrhoeae Pseudomonas aeruginosa 3 o of 10EB was required for positive judgement. Testing of selected bacterial species
Haemophilus inf!uenZJle Moraxella catarrhalis Bacteroides fragilis
2 2 2
Various Gram-positive and Gram-negative species were obtained from urine, genital or respiratory tract
* Approx. 107 CFU per test. All were negative after confirmation testing by the blocking reagent
2 2 2
0.081-0.056 >2.0-1.040 0.107-0.098
Zeeberg et al. C. trachomatis antigen detection 2.4
"l
1.6
c:n ~
"c0 e0
1.2
/
o
III
;2
0.8 0.4 0.0
Table 3. Negative outcome of blocking test among 1101 patient samples as related to degree of reactivity in EIA (Chlamydiazyme)
0
2.0
E
e
/ /~
Absorbance
0
--~-_.
7
6 5 1010g CFU/test
4
Male urethra
__ .--
Figure 1. Titration of two Acinetobacter and one E. coli isolates in (Chlamydiazyme). Overnight cultures were suspended in Chlamydia specimen dilution buffer and lO-fold dilutions in saline tested by EIA EIA
Selected EIA reactive strains (Table 1) were examined by the Chlamydiazyme confirmation step. In all cases unchanged absorbance values were noted following addition of the blocking reagent; thus the test was said to be negative for those strains. Need for confirmation step in clinical samples tested by EIA A total of 192 male urethral samples were tested by both EIA (Chlamydiazyme) and cell culture, yielding a prevalence of C. trachomatis of 21%. The sensitivity of EIA was 0.73 (29 out of 40) compared with a specificity of 0.99 (151 out of 152). Using a lowered cut off value of 0.05, however, the corresponding figures were 0.83 (33 out of 40) and 0.98 (149 out of 152) (Table 2). In an independent series, consecutive clinical samples displaying reactivity in EIA were tested by the confirmation step; specimens yielding absorbance values exceeding 0.050 were included. Out of 230 specimens from male urethra, 5.7% Were unaffected by the blocking reagent and the ~eactivity was thus considered to be caused by lrrelevant bacteria. The majority of the latter cases Were specimens with low absorbance, though 7 of them exceeded the cut off level. Out of 871 cervix specimens, 372 (43%) were negative after conftrmation with the blocking reagent. This proportion Table 2. Performance of ElA (Chlamydiazyme*) using lowered cut offt compared with cell culture in 192 male urethral specimens
Cervix ._--~----~._
0.050-0.079 0.080-CoV* CoV*-0.2 0.2-0.5 0.5-1.0 1.0-1.5 1.5-2.0 >2.0 Total
~ =-====~--. Acin.2
i
357
2/17 (12%)
80/97 (82%)
4/25 (16%)
1091144 (76%) 871166 (52%)
5/38 (13%) 1/41 (2%) 1/35 (1%)
0/23 0/24 0/27 13/230 (6%)
.._-_.-
59/161 (37%)
20/88 (23%) 10/40 (25%) 10/40 (25%) 3/133(2%) 372/871 (43%)
'Cut-off value, determined according to the manufacturer, was between 0.11-0.115
was much higher among specimens with an absorbance value below the cut off level (78%) than among those above (29%). However, among specimens with an absorbance value exceeding 1.0, as many as 17/198 (7.9%) were negative in the confirmation step compared with none of 74male urethra specimens (Table3). In 4 selected cases which turned negative after confirmation, bacterial culture from the sample, though the transport buffer contained a bacteriostatic agent, yielded growth of Gram-negative species. In 3 of these cases, the patient had been recently treated with an antibiotic for verified infection with C. trachomatis (Table 4). Comparison of EIA blocking and DFA for confirmation
Various EIA reactive cervicalor urethral samples were tested by either a repeat EIA with the Chlamydia Blocking Reagent or by DFA (Table 5). Out of 67 samples which turned out positive by the Chlamydia Blocking Reagent, DFAwas positive in 85%; agreement was complete for samples with high absorbance values, whereas, probably due to the high cut off of 10 EB selected, several samples with low absorbance were recorded as negative by DFA. Of 18 samples with a negative outcome by the blocking reagent 76% were also negative in DFA. All five samples which were negative by the blocking reagent but were positive by DFA had absorbance values below 0.6. Table 4. Four cases of unconfirmed reactivity in EIA (Chlamydiazyme) caused by Enterobacteriaceae species
Performance using cut off:!:
Absorbance
Bacterial culture
_.,._-----_._.~._-
Culture result +
Recommended + 29 1
11 151
Lowered +
Age/sex Sample ... ----
33 3
7
149
*Tested without confirmation step t Absorbance 0.05; recommended cut off level was 0.11-0.115 :t:Using recommended and lowered cut off levels, respectively, sensitivity was 0.73 and 0.83; specificity was 0.99 and 0.98; PVpos Was 0.97 and 0.91; PVneg was 0.93 and 0.96. The prevalence of Chlamydia trachomatis was 21%
2/m 22/f*
44/f* 42/f*
---~------
throat cervix cervix cervix
--.
Without With mAb mAbt* ~------
-----------------
0.919 0.370 0.118 0.149
0.886 0.364 0.123 0.145
Species -~--
------_.----
E. coli E. coli Klebsiella E. coli
Absorbance:t: n.d. 0.056 0.065 0.082
*Patient treated recently for verified Chlamydia trachomatis infection tchlamydia Blocking Reagent. +The isolated strain was tested for EIA reactivity as described in Materials & Methods
358
International Journal of STD & AIDS Volume 3 September/October 1992
Table 5. Relation between outcome of DFA examination and EIA confinnation test in 85 genital samples showing EIA (Chlamydiazyme) reactivity Absorbance"
No.
Positive blocking test abs. >2.0 abs.l.D-2-0 abs.0.5-1.0 abs. 2.0 abs.1.0-2.0 abs.0.5-1.0 abs.
ORIGINAL ARTICLE
Chlamydia trachomatis antigen detection by Chlamydiazyme combined with Chlamydia Blocking Reagent verification Barbre Zeeberg
I,
Ingrid Thelin 2, and Claes Schalen MOl
Departments of lMedical Microbiology and2Dermatovenereology, LundUniversity Hospital, 5-22185 Lund, Sweden Summary: Several options exist for the detection of chlamydial infection in a routine laboratory setting. Enzyme immuno assay (EIA) technology offers rapid turn around of results and is less technically demanding than chlamydial cell culture. In addition, recently introduced EIA confirmatory reagents have the potential to improve the accuracy of EIA detection. We have evaluated one such confirmatory reagent (Chlamydia Blocking Reagent" , Abbott Laboratories) to determine the accuracy of the Chlamydiazymev EIA with special regard to interpretation of low absorbance values. An initial series of 192 male urethral specimens showed that use of a lowered cut off level (absorbance value 0.05) compared with that recommended by the manufacturer increased sensitivity of the EIA from 0.73 to 0.83, thus motivating studies on this interpretative modification. Of 1101 EIA reactive specimens, 65% were determined to be chlamydia positive by the Chlamydia Blocking Reagent. The proportion of female cervical specimens that did not confirm positive was elevated compared with male urethral specimens, 43% vs. 5.7% respectively. In samples yielding absorbance from the recommended cut off level to 0.05 (approximately 50% below), the corresponding figures were 78% and 14% respectively. In 85 selected EIA reactive samples, examination by a direct immunofluorescence staining assay (OFA) (Micro'Trak'" , Syva Inc.) revealed elementary bodies in 85% of 67 blocking test positive and in 24% of 18 blocking test negative samples. The possibility that Gram-negative bacteria were responsible for unconfirmed EIA reactive specimens was investigated using bacterial suspensions. While EIA reactivity was noted with several strains for Gram-negative bacteria, both the blocking reagent and OFA correctly verified the absence of chlamydial antigen. We conclude that confirming Chlamydiazyme initial reactive samples with the Chlamydia Blocking Reagent improved the accuracy of chlamydial detection by identifying false positive specimens. Furthermore, by including confirmatory testing, a lower cut off level than recommended by the manufacturer seems applicable, at least for male urethral samples, thus increasing sensitivity of the Chlamydiazyme EIA. Keywords: Chlamydia trachomatis, antigen detedion, Chlamydiazyme
INTRODUCTION Due to a clinical requirement for rapid analyses, antigen detection tests are increasingly used as alternatives to cell culture in the diagnosis of ~hlamydia trachomatis. On a smaller scale, partiCUlarly in laboratories without facilities for cell culture, direct immunofluorescence assay (DFA), though time-consuming, is widely used; this method reportedly gives a high specificity but a low sensitivity compared with culturing under optimal conditions that may limit its use for low-prevalence Correspondence to: Claes Schalen, Department of Medical Microbiology, Lund University Hospital, S-22185 Lund, Sweden
populationsv". In laboratories processing high numbers of specimens, enzyme-linked immunosorbent assay (EIA) appears to be more feasible than DFA. Although a low sensitivity of available EIA kits has been reported, particularly for urethral specimens, their specificity seems surprisingly high1,5- 10. In contrast to DFA, the EIA method does not permit morphological confirmation, leading to erroneous interpretations, e.g. immunological cross-reactions between chlamydiae and various Gram-negative species'! as well as Fe-mediated bindin§ of immunoglobulins to Staphylococcus aureus' or is-haemolytic streptococci-', The basis of immunological cross-reactions between chlamydiae and Enterobacteriaceae or
356
International Journal of SID & AIDS Volume 3 September/October 1992
Acinetobacter is a common core region of the lipopolysaccharide (LPS)lO. It is claimed by some manufacturers that, in clinical specimens, such bacterial species seldom occur in amounts high enough to complicate the detection of C. trachomatis. The present study was undertaken in order to see to what extent cross-reacting bacteria might account for mis-interpretation of one EIA for C. irachomatis, based on an LPS-specific antiserum. In addition, a newly introduced monoclonal reagent, intended for EIA blocking, was evaluated with special reference to heterologous Gram-negative species and to the possibility of enhancing the performance of the EIA on clinical specimens. METHODS Patient samples From 175male patients with symptoms of urethritis, examined at the Department of Dermatovenereology, two urethral samples for cell culture and EIA, respectively, were obtained. In addition, 1101 consecutivemale urethral or female cervical samples, all positive by EIA, were examined by a recently introduced blockingstep (see below). These specimens were obtained from various STD and primary care centres in the district. Specimens to be examined by EIA were obtained with the swabs and transported in the tube supplied by the manufacturer. Specimens for chlamydial culture were transported in a medium previously describedw,
specimens diagnosed at the Clinical Microbiology Department. Identification was according to standard procedures. Bacterial isolates of staphylococci, streptococci, Enterobacteriaceae, Acinetobacter and Pseudomonas aeruginosa were cultured overnight at 37°C on horseblood agar. The following specieswere cultured on appropriate solid media and in a 6% CO2 atmosphere: Haemophilus influenzae, Moraxella caiarrhalis, Neisseria gonorrhoeae, Mycoplasma hominis, Candida albicans, Lactobacillus sp. Anaerobic species were cultured on blood agar for 2 days anaerobically. The cells were harvested and suspended in 1 ml Chlamydiazyme Specimen Dilution buffer to a cell density of approximately lOS cells/ml. as judged by viable counting. The bacterial cells were tested in the Chlamydiazyme EIA by adding 200ILl of undiluted suspension for each assay. For DFA a l000-fold dilution was applied to the microscope slide. TIle tests were then performed as recommended. RESULTS Examination of various bacterial strains in two detection tests for Chlamydia trachoma tis
Bacterialstrains representing 11 Gram-positive and 11 Gram-negative species (2 strains of each) did not reveal any reactivity by DFA. By EIA, all the Grampositive species (5 strains each of group A, B, C and G streptococci, Staphylococcus aureus, S. epidermidis and S. saprophyticus, 2 strains each of Lactobacillus sp., Peptococcus magna and Clostridium perfringens), Chlamydia! detection methods as well as 2 strains each of Candida albicans and Chlamydial cell culture was performed on cyclo- Mycoplasma hominis were negative. However, all the heximide-treated McCoy cells as described by Ripa Gram-negative species tested, except Pseudomonas et al.15 The cells were stained by iodine after 72 h aeruginosa, were found to be reactive to various incubation. degrees (Table 1) and, in particular, high absorbance The procedure for the EIA Chlamydiazyme'" values were noted for E. coli and Acinetobacter (Abbott) was as suggested by the manufacturer. strains. By titration, these two species required Whereas recommended cut off level was 0.11-0.115, 1()5-1()6 cells to reveal EIA reactivity (Figure 1). samples yielding absorbance> 0.05were considered reactive. The background level was below 0.02. Reactive samples were retested for confirmation in Table 1. EIA (Chlamydiazyme) reactivity of selected bacterial the presence or absence of a monoclonal antibody isolates" (Chlamydia Blocking Reagent'") according to the instructions. Absorbance Tested Positive (range) Further testing of 85 randomly selected EIA Species reactive samples by DFA was done following Acinetobacter sp. 10 8 >2.0-0.162 centrifugation of the remaining suspension at 2100 g Escherichia coli 20 >2.0-0.062 16 for 1/2 h. The pellet after decantation was applied Klebsiella sp. 4 0.766-0.065 1 to microscope slides and stained according to Enterobacter sp. 2 0.538-'0.389 2 the manufacturer with a monoclonal, fluorescein Citrobacter sp. 2 0.605-0.411 2 labelled IgM antibody directed against Chlamydia Proteus sp. 5 0.301 1 10 0.990-0.121 2 trachomatis (MicroTrak, Syva Inc., USA). A minimum Neisseria gonorrhoeae Pseudomonas aeruginosa 3 o of 10EB was required for positive judgement. Testing of selected bacterial species
Haemophilus inf!uenZJle Moraxella catarrhalis Bacteroides fragilis
2 2 2
Various Gram-positive and Gram-negative species were obtained from urine, genital or respiratory tract
* Approx. 107 CFU per test. All were negative after confirmation testing by the blocking reagent
2 2 2
0.081-0.056 >2.0-1.040 0.107-0.098
Zeeberg et al. C. trachomatis antigen detection 2.4
"l
1.6
c:n ~
"c0 e0
1.2
/
o
III
;2
0.8 0.4 0.0
Table 3. Negative outcome of blocking test among 1101 patient samples as related to degree of reactivity in EIA (Chlamydiazyme)
0
2.0
E
e
/ /~
Absorbance
0
--~-_.
7
6 5 1010g CFU/test
4
Male urethra
__ .--
Figure 1. Titration of two Acinetobacter and one E. coli isolates in (Chlamydiazyme). Overnight cultures were suspended in Chlamydia specimen dilution buffer and lO-fold dilutions in saline tested by EIA EIA
Selected EIA reactive strains (Table 1) were examined by the Chlamydiazyme confirmation step. In all cases unchanged absorbance values were noted following addition of the blocking reagent; thus the test was said to be negative for those strains. Need for confirmation step in clinical samples tested by EIA A total of 192 male urethral samples were tested by both EIA (Chlamydiazyme) and cell culture, yielding a prevalence of C. trachomatis of 21%. The sensitivity of EIA was 0.73 (29 out of 40) compared with a specificity of 0.99 (151 out of 152). Using a lowered cut off value of 0.05, however, the corresponding figures were 0.83 (33 out of 40) and 0.98 (149 out of 152) (Table 2). In an independent series, consecutive clinical samples displaying reactivity in EIA were tested by the confirmation step; specimens yielding absorbance values exceeding 0.050 were included. Out of 230 specimens from male urethra, 5.7% Were unaffected by the blocking reagent and the ~eactivity was thus considered to be caused by lrrelevant bacteria. The majority of the latter cases Were specimens with low absorbance, though 7 of them exceeded the cut off level. Out of 871 cervix specimens, 372 (43%) were negative after conftrmation with the blocking reagent. This proportion Table 2. Performance of ElA (Chlamydiazyme*) using lowered cut offt compared with cell culture in 192 male urethral specimens
Cervix ._--~----~._
0.050-0.079 0.080-CoV* CoV*-0.2 0.2-0.5 0.5-1.0 1.0-1.5 1.5-2.0 >2.0 Total
~ =-====~--. Acin.2
i
357
2/17 (12%)
80/97 (82%)
4/25 (16%)
1091144 (76%) 871166 (52%)
5/38 (13%) 1/41 (2%) 1/35 (1%)
0/23 0/24 0/27 13/230 (6%)
.._-_.-
59/161 (37%)
20/88 (23%) 10/40 (25%) 10/40 (25%) 3/133(2%) 372/871 (43%)
'Cut-off value, determined according to the manufacturer, was between 0.11-0.115
was much higher among specimens with an absorbance value below the cut off level (78%) than among those above (29%). However, among specimens with an absorbance value exceeding 1.0, as many as 17/198 (7.9%) were negative in the confirmation step compared with none of 74male urethra specimens (Table3). In 4 selected cases which turned negative after confirmation, bacterial culture from the sample, though the transport buffer contained a bacteriostatic agent, yielded growth of Gram-negative species. In 3 of these cases, the patient had been recently treated with an antibiotic for verified infection with C. trachomatis (Table 4). Comparison of EIA blocking and DFA for confirmation
Various EIA reactive cervicalor urethral samples were tested by either a repeat EIA with the Chlamydia Blocking Reagent or by DFA (Table 5). Out of 67 samples which turned out positive by the Chlamydia Blocking Reagent, DFAwas positive in 85%; agreement was complete for samples with high absorbance values, whereas, probably due to the high cut off of 10 EB selected, several samples with low absorbance were recorded as negative by DFA. Of 18 samples with a negative outcome by the blocking reagent 76% were also negative in DFA. All five samples which were negative by the blocking reagent but were positive by DFA had absorbance values below 0.6. Table 4. Four cases of unconfirmed reactivity in EIA (Chlamydiazyme) caused by Enterobacteriaceae species
Performance using cut off:!:
Absorbance
Bacterial culture
_.,._-----_._.~._-
Culture result +
Recommended + 29 1
11 151
Lowered +
Age/sex Sample ... ----
33 3
7
149
*Tested without confirmation step t Absorbance 0.05; recommended cut off level was 0.11-0.115 :t:Using recommended and lowered cut off levels, respectively, sensitivity was 0.73 and 0.83; specificity was 0.99 and 0.98; PVpos Was 0.97 and 0.91; PVneg was 0.93 and 0.96. The prevalence of Chlamydia trachomatis was 21%
2/m 22/f*
44/f* 42/f*
---~------
throat cervix cervix cervix
--.
Without With mAb mAbt* ~------
-----------------
0.919 0.370 0.118 0.149
0.886 0.364 0.123 0.145
Species -~--
------_.----
E. coli E. coli Klebsiella E. coli
Absorbance:t: n.d. 0.056 0.065 0.082
*Patient treated recently for verified Chlamydia trachomatis infection tchlamydia Blocking Reagent. +The isolated strain was tested for EIA reactivity as described in Materials & Methods
358
International Journal of STD & AIDS Volume 3 September/October 1992
Table 5. Relation between outcome of DFA examination and EIA confinnation test in 85 genital samples showing EIA (Chlamydiazyme) reactivity Absorbance"
No.
Positive blocking test abs. >2.0 abs.l.D-2-0 abs.0.5-1.0 abs. 2.0 abs.1.0-2.0 abs.0.5-1.0 abs.
