Journal of Infection (I992) z5, Supplement z, 77-86

The differentiation o f Chlamydia species by antigen detection in s p u t u m specimens f r o m patients with c o m m u n i t y acquired acute respiratory infections Margaret Sillis,* Philippa White,* E. O. C a u l , t I. D. P a u l t and J. D. Treharne~ * PHLS/Microbiology Department, West Norwich Hospital, Bowthorpe Road, Norwich NR2 3 Tx, t Virology Department, Regional Public Health Laboratory, Myrtle Road, Kingsdown, Bristol and :~Section of Virology, Department of Pathology, Institute of Ophthalmology, Judd Street, London, U.K. Summary An amplified enzyme immunoassay (IDEIA III: Dako Diagnostics Ltd) for detecting genus-specific chlamydia antigen was evaluated prospectively on 286 respiratory specimens from 275 patients presenting with community-acquired pneumonia or persistent chest infection. Nineteen patients had evidence of recent chlamydial infection, having two or more positive sputum or serological markers. Sputa from two other patients were ELISA,positive in the absence of other positive criteria and were regarded as false-positive results. When compared with a direct immunofluorescence test for chlamydial elementary bodies (EBs) using a genus-specific monoclonal antibody, the ELISA gave a positive predictive value of 91% and a negative predictive value of 99 %- Non-specific problems with a wide variety of other micro-organisms isolated from the sputa were not encountered. Attempts to differentiate between Chlamydia psittaci, Chlamydia pneumoniae and Chlamydia trachomatis using genusspecific lipopolysaccharide reactive---and species-specific major outer membrane protein--monoclonal antibodies were encouraging and results were substantiated, in most patients, by the species-specific serological assays of the whole-cell-inclusion immunofluorescence or micro-immunofluorescence assays. The study demonstrated that antigen detection techniques offer scope for routine laboratories to diagnose chlamydial respiratory infections rapidly and reliably and may enable differentiation to species level. Although immunofluorescence offers marginally greater sensitivity and specificity when compared with ELISA, the latter is less subjective and less demanding. Sixty-eight per cent of these infections would have remained undiagnosed despite the general availability of ELISA tests.

Introduction T h e three chlamydia species--Chlamydia trachomatis, Chlamydia psittaci and Chlamydia pneumoniae--are i m p o r t a n t respiratory pathogens of humans. Chlamydia trachomatis causes a typically afebrile p n e u m o n i a in perinatallyinfected infants between the e n d week and 4th m o n t h of life, 1'2 chronic respiratory disease d u r i n g childhood in a proportion of these infants, ~'~ and has been isolated f r o m i m m u n o c o m p r o m i s e d patients with lower respiratory tract infection. 5 Chlamydia psittaci is associated with zoonotically t r a n s m i t t e d respiratory infection particularly from psittacine birds 6 and poultry. 7'8 A causal relationship between the newly described C. pneumoniae species 9 and pharyngitis 1° and atypically p n e u m o n i a 11 has been established, with seroconversion occurring m a i n l y in children between 5 and 9 years old and thereafter to 40 % seroprevalence in adult populations worldwide. 1~,13 oi63-4453/92/SlOo77 + IO $03.00/0

© I992 The British Society for the Study of Infection

78

M. SILLIS ET AL.

Culture and identification of these organisms from respiratory material 1°' 14 is not within the scope of a routine laboratory, and serological diagnosis by complement fixation test (CF) remains the prime diagnostic tool. However, early antimicrobial therapy may diminish or delay the production of CF antibody and thereby decrease the sensitivity of this test. 15 T h e genus-specific nature of this assay is an important limitation when species identification is required. More specific serological assays such as the whole-cell-inclusion immunofluorescence test (WHIF) 16 and the microimmunofluorescence test (micro-IF) 17,18 are not suited to routine laboratories and are prone to interpretative errors in inexperienced hands. Newer, rapid techniques of E L I S A and direct immunofluorescence (DIF), using monoclonal antibodies to detect chlamydial antigens, have been well-documented in the diagnosis of ocular 19'2° and genital 21-23 chlamydial infections but have not been widely using in the diagnosis of respiratory chlamydial infections with the exception of neonatal pneumonia. 2°, ~4,25 Reliable detection of chlamydial antigens in clinical material should result in more rapid instigation of therapy and better patient care. T o date, a major limitation of the use of these direct assays on sputum material has been the non-specific reactivity of many commercial assays with other microbial agents, e.g. Acinetobacter sp., 26 Escherichia coli 27 and particularly Staphylococcus aureus, ~s all of which may be found, not uncommonly, in the respiratory tract. More recently some commercial kits (e.g. I D E I A * III Dako) have incorporated purified Fab fragments to circumvent the Fc receptor-mediated non-specific reactivity. This has increased their potential use for chlamydial diagnosis using sputum material. Differentiation to species level of the infecting chlamydia has important clinical and epidemiological consequences. 11'29 Genus-specific and speciesspecific monoclonal antibodies are commercially available for the differentiation of C. psittaci from C. trachomatis. Species-specific C. pneumoniae MOMP-reactive monoclonal antibodies are commercially available, although not thoroughly evaluated on clinical material. We report a prospective study of the diagnostic potential of direct chlamydial antigen detection using sputum specimens from patients with communityacquired acute chest infections. Our preliminary findings have been reported elsewhere.3°' 31

Materials and methods Study population Patients under 65 years of age presenting with community-acquired pneumonia or acute persisting chest infection, with no known underlying or predisposing condition, were investigated. Sputum was obtained from all cases and serum specimens obtained whenever possible. Sputum samples All sputa had been submitted to the Norwich Public Health Laboratory for routine microbiological investigation of acute chest infection. T h e sputum was mucolysed using an equal volume of freshly prepared 2 % N-acetyl-Lcysteine32 for Io rain by digestion using a mechanical shaker. All samples were cultured routinely to establish a bacteriological profile. On receipt into the Virology Department it was often necessary to continue the digestion process

Chlamydia antigen detection in sputum

79

for a further 20 min depending on the tenacity of the specimen. This process of digestion is critical to the sensitivity and specificity of the E L I S A test. Incomplete digestion can result in a false-positive result if tenacious clumps of sputum adhere non-specifically to the solid phase trapping conjugated antibody. Failure to release chlamydial antigen from the specimen may also lead to a false-negative result. T h e digested specimen was transferred to a plastic V-bottomed universal container and washed with a minimum of twice its volume of phosphate buffered saline p H 7"2 (PBS). After centrifugation at 2ooo g for IO min the supernatant was discarded and the pellet resuspended in o'5 ml PBS (suspension A). A o'I ml-volume of suspension A was added to 2 ml o'2 M sucrose phosphate solution (2SP). a3 Many of the specimens took more than 24 h to arrive at the laboratory. This is suboptimal for detection of chlamydia by culture. ELISA

Approximately 0.2 ml of suspension A was added to 0-8 ml of the Dako I D E I A I I I transport medium and processed further according to the manufacturers' instructions. 21 Immunofluorescence

T h e cell density of o'I ml of suspension A was adjusted with PBS to achieve a single layer of cells for each smear. Three smears were prepared on P T F E coated slides (8 m m diameter wells), air-dried and fixed in acetone for io min. Initially one smear was stained using a genus-specific monoclonal antibody ( I M A G E N , Dako). If antigen was detected, further differential staining was carried out using a C. trachomatis species-specific M O M P antibody (Syva Microtrak) and a C. pneumoniae species-specific M O M P antibody ( I O L 2o7) (Institute of Ophthalmology). T h e whole area of the well was systematically scanned at a magnification of x 6oo using epi-illuminescence fluorescent microscopy. T h e presence of more than two chlamydial elementary bodies (EBs) was taken to indicate a positive result. Culture

Routine culture 2°, 14 was carried out by all three centres on selected sputa only. Duplicate M c C o y cell monolayer cultures were treated with cycloheximide, 2° emetine 1~ and mitomycin. ~4 After 72-96 h incubation coverslip monolayers were fixed in methanol. One of each coverslip monolayer was stained using either genus-specific monoclonal antibody or C. trachomatis species-specific monoclonal antibody. After 5 days one emetine-treated cell monolayer was stained with genus-specific monoclonal antibody and the other harvested and passaged. Passaged monolayers were harvested after a further 5 days and stained initially with the Syva M O M P antibody and, if found negative, then counterstained with the genus-specific reagent. Further subculture was discontinued after five negative passages. Cell culture fluids were tested by E L I S A at each passage level. Serology All sera were tested for genus-specific antibody by conventional C F test a5 and for presence of chlamydia-specific IgM. Preliminary examination of selected

80

M. SILLIS ET AL.

sera was carried out using W H I F 16 and micro-IF 17 assays without prior knowledge of E L I S A / D I F sputum results. CF A modification of the method of Bradstreet and Taylor, using microtitre plates, was employed in a conventional overnight fixation test that used 3 H D s o of Richardson's preserved complement (Tissue Culture Services Ltd) and haemolytic serum (Sera Lab. Ltd) Genus-specific (EAE) antigen was supplied by the Division of Microbiological Reagents (DMR), Central Public Health Laboratory, London NW9 5HT. Further serological assays ( W H I F and micro-IF) were carried out on sera with a CF titre > 32 or four-fold rise/fall. Chlamydia IgM ELISA All sera were examined for genus-specific IgM using the modified #-capture assay previously described) ° Positive criteria Sera from all patients with any of the following criteria were selected for "~THIF and micro-IF testing: antigen detected in sputum by E L I S A or D I F , chlamydia CF antibody titre > 32 or four-fold rise/fall or chlamydia genusspecific IgM detected. In the W H I F test indication of C. ~rachomatis infection was confirmed by a rising titre or an eight-fold, or greater, difference in titre between genus (CF) and L2 species antibody. Chlamydia pneumoniae infection was suggested by a rising titre or an eight-fold, or greater, difference in titre between genus ( C F / L 2 ) and C. pneumoniae species antibody. Chlamyclia psittaci infection was suggested when differences in titre between the genus (CF antibody) and species antibody titres to L2 and C. pneumoniae were less than four-fold and provided at least I month had elapsed from the patient's onset of illness. In the m i c r o - I F test an IgG antibody titre of > 256, a four-fold rise or fall in IgG antibody or detection of species-specific IgM were used as the criteria suggesting recent infection. Results Microbiology results No clinically significant pathogens were detected in 233 of the 286 sputa examined. In the remaining 53 a wide variety of organisms was detected including Haemophilus influenzae (n = IO), Streptococcus pneumoniae (n = 3), S. aureus (n = 6), Pseudomonas spp. (n = 3), Haemophilus parainfluenzae (n = 2), Klebsiella spp. (n = I), Aspergillus spp. (n = I), yeasts (n : I), coliforms (n = I), respiratory syncytial virus (n = 4), and influenza virus type A (n = 3). Chlamydia antigen detection results In total 19 patients had evidence of chlamydial infection on the basis of two or more of the positive criteria (see above). T w e n t y of 29. specimens examined

81

Chlamydia antigen detection in sputum Table I Results of chlamydia antigen detection on sputum specimens ELISA (IDEIA III) +ve DIF (IMAGEN)

+ ve -- ve

2o 2

-ve 2 262

from these 19 patients were ELISA-positive/DIF-positive and confirmed by at least one serological criterion in most cases (9o %). Two specimens were positive by D I F only (see Table I). Two specimens from patients with no other criteria supporting chlamydial infection were E L I S A positive only. T h e E L I S A test ( I D E I A III, Dako) compared to D I F (Imagen, Dako) gave a positive predictive value of 91%, and a negative predictive value of 99 %. All other sputum specimens were negative by E L I S A and D I F including eight specimens with heavy S. aureus growth which were tested to confirm the non-reactivity of this organism in the assay. One sputum specimen from a patent with serological evidence of chlamydial infection was negative by E L I S A and D I F . Sputum specimens from I I patients were not available to be tested with the C. pneumoniae monoclonal antibody. However, species differentiation was achieved using D I F on sputum from nine patients (see Table II). Two specimens gave a positive result using the C. trachomatis M O M P antibody and in one other specimen EBs were visualised using this antibody but the fluorescence was m u c h reduced. It was felt the latter was due to crossreactivity. Four specimens were reactive with the C. pneumoniae M O M P antibody although, in one specimen (Case IO), the number of brightly fluorescing EBs was m u c h less than observed when using the genus-specific antibody. In three cases C. psittaci was defined as the infecting species as shown by the non-reactivity of the C. pneumoniae and C. trachomatis M O M P antibodies on the sputum deposit. All remaining DIF-positive sputa were classified as C. pneumoniae/ C. psittaci cases as determined by a negative result using the C. trachomatis antibody. Culture of Chlamydia spp. was positive from the sputa of five patients--two confirmed as C. trachomatis and one as C. psittaci. T h e two other isolates were lost to contamination before species differentiation was confirmed (see Table II). Serology results With the exception of paired sera from one patient, none of the 54 sera obtained from 47 of the sputum-negative patients showed positive chlamydia serological criteria. T h e patient with paired sera had a four-fold rise in genusspecific antibody titre by the CF and W H I F tests but no genus-specific IgM or significant antibody by the M i c r o - I F test. Serum was unavailable from two of the sputum-positive cases.

C. psittaci C. trachomatis*

C. pneumoniae'~ C. pneumoniae

C. psittaci*

C. pneumoniae

C. psittaci

C. pneumoniae

C. trachomatis*

Case

I 3

IO

I3

14

16

I7

19

+2I +2 + II +2 +6 + I +2 +4

+ io

+2

+4 + 17 +4 +3 +6 384 I92 > 512 192 32 I92 < 16 192 384

< 8

32 32 1024 48 48

Genus (CF)

* Chlamydia isolated and identified. t Chlamydia isolated b u t n o t identified. ~: I g M positive. N.T., N o t tested.

II

A n t i g e n in sputum (DIF)

Time interval (weeks after onset)

512 512 1024 N.T. 64 32 512 512 512 512 64 I28 < 64 IO24 8OOO

L2

512 512 512 16000 64 512 64 2048 8OOO

96

N.T. N.T. IO24 N.T. 4096

Chlamydia pneumoniae

WHIF

o O O O 0 O O O O S e r u m unavailable

o

256 64 O O O O O O O

o

o O O 32 32

D-K

C. trachomatis

S e r u m unavailable o O O 32 32

A-C

Chlamydia trachomatis

64 ~6 32 64 O O O I6 O

o

o O O 32 32

LGV 1- 3

M i c r o - I F (IgG)

Table II Serological profile in cases where species identification was achieved

16 256 64 I28 O O O 65 4O96

16

o O 0 16 32

Chlamydia psittaci

64 64 64~ 128 O 0 0 O 64

o

64 32 512 256 256

C. pneumoniae

t~

O0

Chlamydia antigen detection in sputum

83

All of the remaining 16 sputum-positive cases had genus-specific IgM, eight had a CF antibody titre > i28 and three cases showed a four-fold rise/fall in titre. T h e remaining five patients had static CF titres between I6 and 96. W H I F and m i c r o - I F results are discussed in further detail only where chlamydial species identification was achieved on sputum deposits. Using criteria cited earlier one case (patient 3) had evidence of C. trachomatis infection, four had evidence of C. psittaci infection (patients Io, I3, I6 and I7) and two had evidence of C. pneumoniae infection (patients 11 and I4). Chlamydia pneumoniae infection was indicated in two cases (patients io and I4). In one patient (case I3) the initial response, in serum taken 2 months after acute illness, suggested C. trachomatis infection but after 5 months the m i c r o - I F serology indicated C. psittaci infection. Chlamydia psittaci infection was also indicated in patient I7. Significant Micro-IF antibody responses were not detected in the remaining patients. M i c r o - I F serology from the two ELISA-positive/DIF-negative patients demonstrated a low level of antibody to C. pneumoniae only. T h e W H I F serology showed a low level of chlamydial antibody in the serum from one patient but a raised titre to C. pneumoniae in the other patient. Discussion

Our findings suggest that antigen detection using E L I S A and D I F on sputum material is a rapid and reliable method for the diagnosis of respiratory chlamydial infections. D I F is the more sensitive and specific antigen detection method and has the advantage of direct visualisation of the characteristic morphology of chlamydial EBs and also enables assessment of specimen quality. However, considerable experience is necessary to gauge correctly the density of the specimen to obtain a single layer of cells on the smear. Prolonged washing after the staining may be necessary to eliminate a mucoidal haze which interferes with visualisation and in many cases, there may be low numbers of EBs in the smear. Thorough examination of smears by a skilled microscopist is necessary. T h e genus-specific E L I S A test is widely used in routine laboratories for the diagnosis of genital chlamydial infections and less skill in performance is required compared to the D I F . Of the 23 sputum specimens examined from I9 patients with evidence of chlamydial infection, 20 were antigen-positive by D I F and ELISA. With the exception of the two patients from whom serum was unavailable, serological investigations provided substantive evidence of chlamydial infection in all but two patients with chlamydial antigen-positive sputum. T h e two E L I S A negative/DIF-positive sputum specimens were weakly positive with only three EBs visualised in each smear. One of these specimens was salivary and probably inadequate and the other was obtained at intubation from a patient. Presumed false-positive reactions detected by E L I S A in the sputa from two patients were not confirmed by D I F , culture or by serological criteria. No other microbial agent could be identified as either the aetiological agent in these two patients or as the cause of the non-specific reactivity in the ELISA. A wide variety of organisms, including S. aureus detected in a number of sputa did not cause non-specific reactivity. A possible explanation of non-reactivity, on review, could be that transfer of minute amounts of agar to the sputum

84

M. S I L L I S

ET AL.

during routine bacteriological investigations occurred. Such contamination is well-known to interfere in ELISAs. Chlamydia was cultured from five patients. Two isolates were c o n f r m e d as C. trachomatis, one as C. psittaci and two were lost to contamination before differential staining using C. pneumoniae M O M P was carried out. T h e isolation of C. trachomatis from the sputum of one patient (case 3) was particularly surprising since she was an immunocompetent 63-year-old female with no relevant exposure history. W H I F serology in part substantiated the sputum findings but m i c r o - I F serology revealed static low level antibodies to C. pneumoniae only. T h e second C. trachomatis case was a 2-month-old baby who had respiratory symptoms since 2 weeks of age. Isolation of C. psittaci from one patient (case 13) was consistent with the W H I F serology results. However, m i c r o - I F testing of serum taken 2 months after onset of illness in this patient showed reproducible antibody with a specificity, greater against C. trachomatis strains than C. psittaci. A further serum taken at 5 months after onset showed a more characteristic profile of C. psittaci specificity. Clinical and exposure history failed to implicate concurrent C. trachomatis genital infection. One of the remaining patients, from whom species identification was apparently achieved by differential staining of the sputum deposit, failed to submit a serum specimen. Identification of C. psittaci in this patient (case I) was supported by strong evidence of bird exposure. Chlamydia psittaci infection in patient I6 was supported by W H I F serological criteria but no antibody response was detected by m i c r o - I F testing 6 weeks after onset of illness. It is possible that the C. psittaci strains used in the antigen pool in the M i c r o - I F test were too distantly related to detect homotypic antibody responses. Species antigen differentiation on sputum indicated that four patients had C. pneumoniae infections. In two cases (patients I I and 14) W H I F and microIF results were concordant and confirmed the sputum findings. In patient IO the M i c r o - I F results clearly supported evidence of C. pneumoniae infection but W H I F serology suggested C. psittaci infection. This patient's illness was strongly associated with bird exposure occurring within 2 weeks of acquisition of a new budgerigar. Differential staining of the sputum deposit revealed significantly fewer EBs stained by the C. pneumoniae antibody compared with genus-specific antibody. Dual infection with C. psittaci and C. pneumoniae is unlikely but cannot be excluded until further tests have been carried out. Chlamydia pneumoniae seemed to be convincingly identified in the sputum deposit of patient I7 using the C. pneumoniae (IOL2o7) antibody. However, the serological profile by m i c r o - I F and W H I F testing on sera taken over a 4week period after onset of illness showed a greater specificity to C. psittaci than C. pneumoniae. In addition, this patient had evidence of prolonged exposure to a sick bird. On balance, the weight of evidence obtained from the results of patient 17 suggests that the C. pneumoniae species may have been misidentified in the sputum deposits. Further examination of these sputum specimens by culture and polymerase chain reaction should provide definitive evidence of the accuracy and reliability of species differentiation on respiratory material and may clarify the clinical usefulness of the W H I F and m i c r o - I F assays. Serological profiles of many of the patients in this study suggest that genus-

Chlamydia antigen detection in sputum

85

specific a n t i b o d y appears first and species ( M O M P ) a n t i b o d y appears at least 4 - 6 weeks after onset o f illness. Species differentiation by serology is likely to be o f r e t r o s p e c t i v e clinical value only. I n conclusion, antigen det ect i on t e c h n i q u e s using s p u t u m specimens offer scope to r o u t i n e laboratories to diagnose chlamydial r e s p i r a t o r y infections rapidly and reliably. A l t h o u g h D I F offers marginally greater sensitivity and specificity w h e n c o m p a r e d with E L I S A the latter is less subjective and less demanding. O u r c u r r e n t r e c o m m e n d a t i o n , for det ect i on of Chlamydia spp. in s p u t u m , is to test r o u t i n e l y by E L I S A and confirm, b y visualisation o f EBs using genusa n d species-specific m o n o c l o n a l reagents. In o u r experience s e r u m specimens are received f r o m less t h a n IO ~/o o f patients p r e s e n t i n g to general practitioners with chest infections, and f r o m less t ha n 30 % of patients a d m i t t e d into hospital with r e s p i r a t o r y illness. C o n s e q u e n t l y , diagnosis o f chlamydial r e s p i r a t o r y infections is almost certainly u n d e r - d i a g n o s e d . In this study I3 (68 %) c h l a m y d i a - p o s i t i v e patients w oul d not have been diagnosed despite the general availability o f E L I S A tests. W e c o n c u r with o t h e r workers 36 that antigen det ect i on t e c h n i q u e s should be applied r o u t i n e l y to the investigation o f severe or persisting c o m m u n i t y a c q u i r e d chest infections to ensure a p p r o p r i a t e t h e r a p y is instigated as early as possible. (We are grateful to Miss Claire Stuart for typing this manuscript and Dako Diagnostics Ltd for supplying the I D E I A ~ Chlamydia kits.) References i. Harrison JR, English MG, Lee CK, Alexander ER. Chlamydia trachomatis infant pneumonitis. Comparison with matched controls and other infant pneumonitis. N Engl,7 Med I978 ; 298 : 7o2-7o8. 2. Beem M, Saxon E. Respiratory tract colonisation and a distinctive pneumonic syndrome in infants infected with Chlamydia trachomatis. N Engl J A4ed 1977; 296: 3o6-3IO. 3. Harrison H, Taussig LH, Fulginiti VA. Chlamydia trachomatis and chronic respiratory disease in childhood. Paed Infect Dis I982 ; I : 29-32. 4. Weiss SG, Newcomb RW, Beem MO. Pulmonary assessment of children after Chlamydia pneumonia of infancy. J Paediatr 1986; 5: 659-664. 5. Tack KJ, Rasp FL, Hanto D et al. Isolation of Chlamydia trachomatis from the lower respiratory tract of adults. Lancet I98O; i: II6-I2O. 6. Wreghitt TG, Taylor CED. Incidence of respiratory tract chlamydial infections and the importance of psittacine birds. Lancet I988; i: 582. 7. Andrews BE, Major R, Palmer SR. Ornithosis in poultry workers. Lancet 1981 ; i (822I): 632-634. 8. Strauss J. Microbiologic and epidemiologic aspects of duck omithosis in Czechoslovakia. Am J Ophthalmol 1967; 63 : 1246-1259. 9- Grayston JT, Kuo CC, Campbell LA, Wang SP. Chlamydia pneumoniae sp. nov. for Chlamydia sp. strain TWAR. Int J Syst Bacteriol 1989; 39: 88-90. IO. Grayston JT, Kuo CC, Wang SP, Altman J. A new Chlamydia psittaci strain called TWAR from acute respiratory tract infections. N Engl J Med 1986; 315 : 16I-I 68. I I. Grayston JT, Wang SP, Kuo CC, Campbell LA. Current knowledge on Chlamydia pneumoniae strain TWAR, an important cause of pneumonia and other acute respiratory disease. Eur J Clin Microbiol Infect Dis 1989; 8: 191-2o2. I2. Wang SP, Grayston JT. Microimmunofluorescence serological studies with the TWAR organism. In: Oriel JD, Ridgway G, Schachter J, Taylor-Robinson D, Ward M, Eds. Chlamydial infections. Cambridge: Cambridge University Press, I986: 337-340.

86

M. S I L L I S E T AL.

13. Forsey T, Darougar S, Treharne JD. Prevalence in human beings of antibodies to chlamydia IOL-2o7: an atypical strain of chlamydia. J Infect 1986; I2: 145-152. I4. Paul ID. The growth of Chlamydia in McCoy cells treated with emetine. Med Lab Sci I982; 39:I5-32. 15. Meyer KF, Eddie B. The evidence of tetracycline compounds on the development of antibodies in psittacosis. Am Rev Tuberculosis Pulmonary Dis 1956; 74: 566-571. 16. Richmond S, Caul EO. Fluorescent antibody studies in chlamydial infections. J Clin Microbiol 1975 ; I: 345-352. 17. Trehame JD, Darougar S, Jones B. Modification of the microimmunofluorescence test to provide a routine serodiagnostic test for chlamydial infection. J Clin Pathol 1977; 30: 51o-517. 18. Wang SP, Grayston JT. Human serology in Chlamydia trachomatis infection with microimmunofluorescence. J Infect Dis 1974; 13o: 388-397. 19. Mabey DCW, Robertson JN, Ward ME. Detection of Chlamydia trachomatis by enzyme immunoassay in patients with trachoma. Lancet I967; ii: 1491-1493. 2o. Potts MJ, Paul ID, Roome APCH, Caul EO. Rapid diagnosis of Chlamydia trachomatis infection in patients attending an ophthalmic casualty department. J Ophthalmol I986; 70 : 677-68o. 2I. Pugh SF, Slack RCB, Caul EO, Paul ID, Appleton PN, Gatley S. Enzyme amplified immunoassay: a novel technique applied to direct detection of Chlamydia trachomatis in clinical specimens..7 Clin Pathol 1985 ; 38 : 1139-1 I4I. 22. Kellogg J. Clinical and laboratory considerations of culture vs antigen assays for detection of Chlamydia trachomatis from genital specimens. Arch Pathol Lab Med 1989; 113: 453-459. 23. Paul ID, Caul EO. Evaluation of three Chlamydia trachomatis immunoassays with an unbiassed, noninvasive clinical sample. J Clin Microbiol I99O; 28: 22o-222. 24. Friis B, Kuo CC, Wang SP, Mordhorst CH, Grayston JT. Rapid diagnosis of Chlamydia trachomatis pneumonia in infants. Acta Path Microbiol Immunol Scand 1984; Sect. B. 92: I39-I43. 25. Paisley JW, Laver B, Melinkovich P, Glitterman B, Feiten D, Berman S. Rapid diagnosis of Chlamydia trachomatis pneumonia in infants by direct immunofluorescence microscopy of nasopharyngeal secretions. J Paediatr I986; IO9: 653-655. 26. Saikku P, Puolakkainen M, Leinonen M, Nurminen M, Nissinen A. Cross reactivity between Chlamydiazyme and acinetobacter strains (Letter). N Engl J Med 1986; 314: 922-923. 27. Goudswaard J, Sabbe L, van Belzen C. Interference by Gram-negative bacteria in the enzyme immunoassay for detecting Chlamydia trachomatis. J Infect 1989; 18: 94-95. 28. Krech T, Gerhard-Fsadni D, Hofmann N, Miller S. Interference of Staphylococcus aureus in the detection of Chlamydia trachomatis by monoclonal antibodies (Letter). Lancet 1985 ; i: iI6I. 29. Kuo CC, Grayston JT. In vitro drug susceptibility of Chlamydia TWAR. Antimicrob Agents Chemother I988; 32: 257-258. 3o. Sillis M, White P. Rapid identification of Chlamydia psittaci and TWAR (C. pneumoniae) in sputum samples using an amplified enzyme immunoassay. J Clin Pathol I99O; 43 : 26o. 3I. Sillis M, White P. Rapid diagnosis of psittacosis. Lancet I99O; i: 726. 32. Woodhams A, Mead G. A comparison between pancreatin and N-acetyl-L-cysteine as sputum liquefying agent for culture of organisms. Tubercle (Lond) I965 ; 46: 224-226. 33- Kuo CC, Grayston JT. Factors affecting viability and growth in HeLa 229 cells of Chlamydia sp. strain TWAR. jt Clin Microbiol 1988 ; 26: 812--8I 5. 34. Woodland RM, Kirton RP, Darougar S. Sensitivity of mitomysin C-treated McCoy cells for the isolation of Chlamydia trachomatis from genital specimens. Eur J Clin Microbiol 1987; 6: 653-656. 35- Bradstreet CMP, Taylor CED. Technique of complement fixation test applicable to the diagnosis of virus diseases. Monthly Bull Min Health P H L S I962; 2I: 96-IO4. 36. Riordan T, Lewin I, Oliver MH. Rapid diagnosis of psittacosis. Lancet 199o; i: 47I.