J Clin Pathol 1991;44:1027-1029 1027 Demonstration of Chlamydia trachomatis in colposcopic cervical biopsy specimens by an J Clin Pathol: first published as 10.1136/jcp.44.12.1027 on 1 December 1991. Downloaded from immunoperoxidase method J M Edwards, A R Campbell, A Tait, M Lusher Abstract Methods A total of 31 cervical biopsy specimens Endocervical swabs were taken from were taken from 29 women attending a nulliparous, adult women attending a genito- genitourinary medicine clinic, nine urinary medicine clinic. The swabs were women (11 biopsy specimens) were immediately transferred to the laboratory in known to have Chlamydia trachomatis Chlamydia culture medium and an enzyme cervicitis and 20 women were known to linked immunosorbent assay (ELISA) test be free of chlamydial infection. The was performed to detect the group specific specimens were routinely processed to lipopolysaccharide antigen (LPS) (Antigenz, paraffin wax and stained by an anti- Northumbria Biologicals Ltd, Newcastle). All Chlamydia immunoperoxidase tech- positive ELISA cases were confirmed with nique to localise the organisms. Of the 11 direct fluorescence (Microtrak, Syva). Eleven positive biopsy specimens three showed known positive and 20 known negative cases positive staining of elementary/reti- were brought back to the clinic one week culate bodies. In one case the surface later and a colposcopically directed cervical endocervical cells showed large in- biopsy specimen was taken from the squamo- clusions which were packed with columnar junction. chlamydial bodies. The specimens were fixed in formalin and The diagnosis of chlamydial infection processed to paraffin wax as normal. A routine is difficult to make clinically and in haematoxylin and eosin section was examined routine cytological and histological to check that all biopsy specimens included specimens but immunoperoxidase stain- the squamocolumnar junction. ing can clearly identify C trachomatis Two monoclonal antibodies were used on inclusions in cervical biopsy specimens the 31 cervical biopsy specimens plus several provided infection is severe. infected and non-infected McCoy cell mono- layer cultures. One method used the genus specific monoclonal antibody LPS (Novo- Infections caused by Chlamydia trachomatis Biolabs) which is the same antigen as is detec- http://jcp.bmj.com/ are now recognised as the most prevalent and ted in the ELISA technique performed in our among the most damaging of all sexually laboratory. The other used the species specific transmitted diseases.' Infection is often low monoclonal antibody MOMP (Syva Com- grade and chronic but is an important cause of pany, USA) which recognises the major outer pelvic inflammatory disease and its sequelae. membrane protein of C trachomatis elemen- Treatment is and simple effective but signs tary bodies.7 Both the commercially available on September 28, 2021 by guest. Protected copyright. and symptoms are often subtle. LPS and MOMP monoclonal mouse anti- Papanicolou stained cervical smear cell bodies are labelled with fluorocien isothio- changes are now regarded as too non-specific cyanate (FITC). We used two separate to allow Chlamydia to be differentiated from immunocytochemistry systems to convert to other cervical infections.2 Histological features an immunoperoxidase end point, a 1 in 100 associated with chlamydial infection in rabbit anti-mouse secondary antibody and a Department of cervical biopsy specimens have recently been 1 in 100 mouse anti-FITC monoclonal secon- Pathology, Royal described3 and their specificity awaits further dary. An avidin-biotin-complex (ABC) Preston Hospital confirmation. Use of electron microscopy to system was then used to bypass the FITC J M Edwards search for inclusions has a detection rate of signal and visualised using peroxidase and A R Campbell only 4%.4 diaminobenzidine (DAB). Both these systems Department of Genitourinary Fluorescent antibody staining of cervical gave identical results. Medicine, University smears is nearly as sensitive as culture All immunocytochemical experiments in- of Liverpool provided the smear is taken from the endo- cluded positive and negative control slides. To A Tait cervix,' but fluorescent antibody staining of begin with the positive control consisted of Department of cervical is less sensitive.3 A Virology, University of biopsy specimens infected McCoy cell monolayers. After Manchester peroxidase-antiperoxidase method has been positive results were achieved on paraffin wax M Lusher used on cervical smears with some success in sections these cases were used as positive Correspondence to: localising inclusions.6 We attempted to stain controls. The negative control was performed Mr A R Campbell, C trachomatis inclusions in wax Department of Pathology, for paraffin by omitting the primary antibody from the Royal Preston Hospital, embedded cervical biopsy specimens using an protocol. Preston, Lancashire of the fluores- PR2 4HG. immunoperoxidase adaptation cence to a more Accepted for publication technique give convenient and Results 29 May 1991 more permanent result. Both monoclonal antibodies LPS and MOMP 1028 Edwards, Campbell, Tait, Lusher J Clin Pathol: first published as 10.1136/jcp.44.12.1027 on 1 December 1991. Downloaded from p Figure 1 Cervical biopsy specimen stained by an Figure 2 Chlamydial inclusions are stained brown immunoperoxidase stain to illustrate chlamydial bodies on within endocervical cells and one cell (arrow) is in the the surface and within endocervical cells. process of releasing its chlamydial bodies onto the surface. showed infected cells in monolayer cultures. appointing (table). The MOMP antibody was The LPS antibody failed to identify any also positive on all infected monolayer cul- evidence of infection in the 11 positive biopsy tures, our pick-up rate on biopsy specimens of specimens. The MOMP technique detected three out of 11 corresponds closely to earlier three positive specimens: in two, chlamydial work using a fluorescein labelled MOMP bodies were present free on the endocervical antibody in which two cases out of 28 showed cell surface; in the third case (fig 1) a large chlamydial inclusions and one showed number of chlamydial inclusions were identi- elementary bodies attached to the surface of fied on the surface and within the endocervical the glandular epithelium.3 Direct fluorescence cells lining the canal (fig 2). In one cell antibody testing on smears has shown that the (arrowed) the inclusion appeared to be about species specific MOMP antibody produces to release its bodies on to the surface. The more intense fluorescence than the genus poor sensitivity was disappointing, so we specific anti-LPS monoclonal antibody9 and compared the positivity of the ELISA tech- we have also shown that MOMP is superior. nique performed at the initial visit with our C trachomatis activity waxes and wanes so results and we noted that our stains were only perhaps in some of our positive cases very few picking up the patients who were reported as organisms were present at the time of biopsy. having a strongly positive ELISA result. The We attempted to minimise this problem by case was scored as ++++, two cases as performing the index endocervical swab to +++; the six remaining Chlamydia cases identify the infected group and the biopsy a http://jcp.bmj.com/ had less strong ELISA results, with three week apart; a shorter period was logistically cases + + and three as +. impossible. We acknowledge the fact that our case and control numbers are small and that this curtails serious assessment of sensitivity Discussion and specificity, but the close correlation with We were unable to find a readily available the degree of positivity of the ELISA tech- immunoperoxidase labelled antibody and nique suggests that our immunoperoxidase on September 28, 2021 by guest. Protected copyright. therefore attempted to localise C trachomatis stain gives good identification of chlamydial by an adaptation of an immunofluoresence bodies provided infection is severe. The technique, as described. We confirm the gen- advantages are that tissue morphology is erally accepted view that the endocervical cells preserved, interpretation of the exact infection provide a desirable environment for the site is simple, and the stain is less transient organism to multipy whereas the squamous and more convenient to assess than a fluores- cells and connective tissue cells were spared. cence technique. We found only the columnar cells on the endocervical surface were infected whereas Our thanks to Dr S J Richmond, Department of Virology, University of Manchester, for her help and encouragement, the columnar cells within the crypts did not and also to Dr B Pratt, Medical Microbiology Department, contain inclusions, and this finding differs from University of Liverpool, and Dr I iAcDicken, Department of that of other workers.3 Histopathology, Women's Hospital, Liverpool. The LPS antibody stained the infected cells in the monolayers but failed to stain the Correlation between ELISA result and cervical biopsy positive biopsy specimens. The cells were immunoperoxidase staining in 29 patients fixed for about 10 minutes whereas fixation time for the biopsy specimens varied from one Immunoperoxidase Case No ELISA (LPS) (MOMP) to six weeks. We reduced the fixation time to 24 hours but the positive rate failed to 1 + + + + Inclusions 2 + + + Freebodies improve. Lundermose et al8 stained C psittaci 3 + + + Freebodies using an LPS antibody on formalin fixed 4, 5,.6 + + Negative 7, 8,9