Previous Article | Next Article ![]()
Clinical Microbiology Reviews, October 2004, p. 982-1011, Vol. 17, No. 4
0893-8512/04/$08.00+0 DOI: 10.1128/CMR.17.4.982-1011.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Clinical Research Unit, London School of Hygiene & Tropical Medicine, London, United Kingdom,1 Special Programme for Research and Training in Tropical Diseases, World Health Organization, Geneva, Switzerland,2 National Microbiology Laboratory, Health Canada, Winnipeg, Canada3
SUMMARY INTRODUCTION EPIDEMIOLOGY Patterns of Distribution CAUSATIVE ORGANISM Historical Perspective Developmental Cycle Structure Cellular components important for diagnostic assays: MOMP and lipopolysaccharide. Taxonomy Classification and Tropism of C. trachomatis Strains Genome of C. trachomatis Chromosome. Plasmid. NATURAL HISTORY AND CLINICAL FEATURES CLINICAL DIAGNOSIS History Examination Differential Diagnosis Grading Systems Dawson, Jones, and Tarizzo 1981 (modified WHO system or FPC system). Thylefors, Dawson, Jones, West, and Taylor, 1987 (WHO simplified system). Comparability of Grading Schemes LABORATORY DIAGNOSIS Microscopy Cell Culture Direct Fluorescent Antibody Enzyme Immunoassay Serology Direct Hybridization Probe Tests PCR Ligase Chain Reaction Strand Displacement Assay Transcription-Mediated Amplification Quantitative PCR Sensitivity and Specificity of Laboratory Tests Correlation of Laboratory Tests with Clinical Signs of Trachoma Factors influencing the accuracy of the test. Factors influencing the accuracy of clinical diagnosis. Factors relating to the natural history of infection. Which Laboratory Tests Are Useful? COMMUNITY ASSESSMENT WHO Simplified Grading Scheme at the Community Level Trachoma Rapid Assessment Required Indicators from Community Assessment Sequential Sampling Place of Laboratory Tests in Community Assessment ACKNOWLEDGMENTS REFERENCES
| SUMMARY |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
| EPIDEMIOLOGY |
|---|
|
|
|---|
| CAUSATIVE ORGANISM |
|---|
|
|
|---|
The causative agent of trachoma was not visualized until 1907, when Halberstaedter and von Prowazek described the presence of inclusion bodies (Halberstaedter-Prowazek bodies) inside infected cells. They believed the organism to be a protozoon (90). The transmissibility of trachoma was by then already firmly established in the minds of the public. Hundreds of Russian and Austro-Hungarian First World War conscripts, for example, evaded military service by infecting their own eyes with discharges wiped from the eyes of trachoma patients (62). Meanwhile, unconvinced by the findings of Halberstaedter and von Prowazek, researchers nominated a variety of bacteria, fungi, and viruses as the underlying pathogen (62). It was not until 1957 in Peking that T'ang et al. completed the first successful isolation, using chicken embryos whose yolk sacs had been inoculated with material from infected human eyes (212). They were able to serially passage the organism in eggs and to use this material to infect the eyes of rhesus monkeys, producing characteristic clinical signs of trachoma and, on one occasion, inclusion bodies. Based on filtration experiments, they believed the trachoma agent to be a virus (212). T'ang et al.'s methods were successfully replicated by Collier and Sowa in the Gambia in 1958 (45). The isolates obtained were noted both to have the same antigen as and to physically resemble the agents of psittacosis and lymphogranuloma venereum (45) and the agent of some cases of cervicitis and mucopurulent conjunctivitis of the newborn (110).
As knowledge of the nature of these organisms accumulated, there was considerable debate over whether they represented a transitional remnant on the degenerate evolutionary pathway that Green had hypothesized (82) for the descent of viruses from bacteria, or whether they should be placed wholly within one or another of these classes. In 1966, Moulder published a comprehensive review of the growth, division, structure, chemical composition, and metabolism of the group, taking into account the definitions of viruses and bacteria that had recently been proposed by Lwoff and Stanier, respectively. He concluded fairly unequivocally that chlamydiae were intracellular bacteria, with a distinctive developmental cycle and unique structure (151). The Taxonomy Committee of the American Society for Microbiology unified these organisms in the genus Chlamydia and supported their status as bacteria (166). Today, some 15 major bacterial groupings are recognized, and the chlamydiae are the only ones whose members are all exclusively intracellular parasites of eukaryotes (66).
Elementary bodies of chlamydiae are spherical (or, rarely, pear-shaped) and 0.2 to 0.3 µm in diameter (42). They appear to bind to susceptible host cells via heparin bridges. There is considerable interest in identifying the chlamydial ligands involved in heparin binding: candidates include the major outer membrane protein and the cysteine-rich protein OmcB, both of which are found in the chlamydial outer membrane complex (203, 209). Successful attachment of the elementary body is followed by its entry into the host cell. Although hosts are nonprofessional phagocytes, their oxidative and glycolytic pathways must be intact for entry to occur, suggesting that they participate actively in elementary body ingestion (150). Tyrosine phosphorylation of host cell proteins and actin cytoskeletal rearrangement may be involved (69, 179). Although elementary bodies are taken up 10 to 100 times more efficiently than latex particles of the same size or Escherichia coli (34), the active participation of the elementary body appears to be minimal; elementary body envelopes are internalized just as efficiently as whole elementary bodies (65). The processes involved in attachment and uptake may differ between species of chlamydiae and even between variants of the same species (179, 196).
Once inside a host cell, the elementary body reorganizes into a 0.5- to 1.0-µm-diameter reticulate body within a membrane-bound vacuole known as an inclusion. The reticulate body grows and replicates by binary fission, remaining within the inclusion membrane (derived from the cytoplasmic membrane of the host) for the entire duration of the organism's intracellular phase. After a period of exponential growth, progeny differentiate back into elementary bodies. A number of late-phase proteins are synthesized during the reticulate body-to-elementary body transformation, including chlamydial outer membrane complex proteins OmcB and OmcA and two histone H1-like proteins, Hc1 and Hc2, which are involved in compaction of the chlamydial chromosome (42, 85). Elementary bodies are released into the extracellular environment by fusion of the membrane of the inclusion with that of the host cell or upon host cell lysis (30, 150). The specific processes involved in cellular exit are poorly defined. In tissue culture, the entire developmental cycle from attachment to exit takes between 48 and 72 h.
|
Reticulate bodies are larger than elementary bodies and contain diffuse, fibrillar DNA plus a high concentration of ribosomes. The cell envelope appears less complex than that of the elementary body, lacking the hexagonally packed structures of the elementary body periplasmic space. The surface of the reticulate body outer membrane, however, contains projections and rosettes at even higher densities than are seen on elementary bodies (139). The outer end of the projections appear to contact the inclusion membrane, leading to the hypothesis that projection-rosette complexes have a secretory function analogous to the type III secretory system found in other bacterial species (42, 104).
Most bacteria have peptidoglycan, a complex cross-linked polymer, in their cell envelope. In gram-negative organisms, peptidoglycan is found in the periplasmic space, while in gram-positive organisms it lies immediately outside the cytoplasmic membrane and may constitute up to 50% of total cell wall material (30). Its function is to help maintain cell shape and integrity despite the relatively high internal osmotic pressure of the bacterium. Penicillin and other ß-lactam antibiotics inhibit the growth of susceptible organisms by preventing the formation of peptide cross-links in peptidoglycan. This effect is mediated through bacterial penicillin-binding proteins. Chlamydiae produce penicillin-binding proteins, and attempts to grow them in the presence of penicillin result in the formation of aberrant inclusions, but, surprisingly, they do not appear to contain appreciable amounts of peptidoglycan (17, 35). This paradox raises interesting questions about the biology of the organism (41, 76) and demands an alternative explanation for the rigidity and osmotic stability of the elementary body. These properties are presently thought to be conferred by the chlamydial outer membrane complex.
The chlamydial outer membrane complex was first defined in 1981, when Caldwell et al. (35) reported the outcome of their experiments with the detergent Sarkosyl (sodium N-lauroyl sarcosine), which had been shown to selectively solubilize the cytoplasm and cytoplasmic membranes of gram-negative bacteria (71). Transmission electron microscopy of Sarkosyl-treated elementary bodies showed empty elementary body particles with an apparently intact outer membrane (35). Caldwell et al.'s method has since been used as the standard method for purification of the chlamydial outer membrane complex. Using sodium dodecyl sulfate-polyacrylamide gel electrophoresis of elementary body lysates, the same group identified a protein of about 40 kDa that was found in a number of chlamydial strains. They went on to show that this protein, which soon became known as the major outer membrane protein (MOMP), was one of the elements of the Sarkosyl-insoluble chlamydial outer membrane complex and that it constituted about 60% of the total protein mass of the elementary body cell wall. 125I labeling of the protein indicated that it was surface exposed (35, 95, 183).
Hatch et al., trying to extract MOMP from other components of the cell wall, noted that it could not be dissolved in sodium dodecyl sulfate or mercaptoethanol alone but was soluble in a solution containing both of these agents. This implied that disulfide bonds were important in binding MOMP to the chlamydial outer membrane complex and perhaps in maintaining the overall structural stability of the cell wall (95).
It was quickly recognized that the importance of MOMP was not confined to its structural role. Salari and Ward were able to extract MOMP from 14 of the 15 then-known serovars (see below) of the chlamydial species Chlamydia trachomatis and noted minor serovar-specific variations in its molecular weight (183). The existence of species- and subspecies-specific epitopes within the protein was noted (37, 206), before it became clear that MOMP also contained serovar-specific epitopes (138, 162). Even more important, polyclonal (36) and then monoclonal (168, 251) antibodies to MOMP were shown to neutralize infectivity of the live organism. These discoveries raised hopes that the protein would be useful in the development of a protective subunit vaccine. Unfortunately, such hopes have so far remained unfulfilled (129).
Although multiple strands of evidence suggest that MOMP is surface exposed, other data have localized parts of the molecule to the periplasmic space (10). These two conclusions are consistent with MOMP's being an integral membrane protein. It now seems likely that MOMP's physiological function is as a membrane channel or porin permeable to ATP. Prevention of uptake of host cell ATP could potentially be a mechanism by which antibodies to MOMP block cellular infection (248). Although the protein is thought to form trimeric aggregates within the outer membrane (23), its actual conformation within intact chlamydiae is unknown (94).
A model of the cell envelope structure has been suggested by Everett and Hatch (67) (Fig. 1). It has been hypothesized that the structural stability of elementary bodies is maintained by disulfide cross-linking between cysteine residues of MOMP and other membrane proteins (67, 160). A similar supramolecular structure is absent from reticulate bodies, which are osmotically fragile. Protein cross-linking appears to occur during the last 24 h of the intracellular phase of the life cycle (161).
All species of chlamydiae identified to date have a common lipopolysaccharide that differs from the lipopolysaccharide of other bacteria. The molecule is present in the outer membrane of the cell envelope throughout the life cycle (24, 145) and contains polysaccharide epitopes recognized by the human humoral immune system (30, 59).
In 1999, a paper by Everett et al. attempted to reclassify the family into two genera, Chlamydia and Chlamydophila, which together contain a total of nine species (66). This new nomenclature has proven controversial (J. Schachter, R. S. Stephens, P. Timms, C. Kuo, P. M. Bavoil, S. Birkelund, J. Boman, H. Caldwell, L. A. Campbell, M. Chernesky, G. Christiansen, I. N. Clarke, C. Gaydos, J. T. Grayston, T. Hackstadt, R. Hsia, B. Kaltenboeck, M. Leinonnen, D. Ocjius, G. McClarty, J. Orfila, R. Peeling, M. Puolakkainen, T. C. Quinn, R. G. Rank, J. Raulston, G. L. Ridgeway, P. Saikku, W. E. Stamm, D. Taylor-Robinson, S. P. Wang, and P. B. Wyrick, letter, Int. J. Syst. Evol. Microbiol. 51:249, 2001; K. Everett and A. Andersen, authors' reply to letter, Int. J. Syst. Evol. Microbiol. 51:251-253, 2001). Whether or not the proposed changes enter general use, the designation of the pathogen responsible for human trachoma will remain Chlamydia trachomatis.
The mouse pneumonitis biovar (which gains species status in the Everett et al. classification) includes two strains: MoPn, which is found in the respiratory tract of mice, and SFPD, which has been isolated from the intestines of hamsters (66). Neither strain is known to infect humans.
The other two C. trachomatis biovars preferentially infect humans. They are closely related. Four serotypes or serovars are currently included in the lymphogranuloma venereum biovar, and 15 are currently included in the trachoma biovar. Each of the 19 can be distinguished from the others on the basis of binding affinity for monoclonal antibodies. They can also be differentiated by polymorphisms in the sequence of MOMP or in the sequence of the gene omp1, which codes for MOMP. There is very limited variation in these sequences between isolates of any given serovar (66). Separation of the lymphogranuloma venereum and trachoma strains into two biovars is based on tissue tropism: lymphogranuloma venereum strains can invade lymphatic tissue, while trachoma strains are restricted to mucosal epithelial cells.
The lymphogranuloma venereum serovars (L1, L2, L2a, and L3) are rare. All are sexually transmitted, although the eye may also act as the portal of entry. Infection is associated with a suppurative adenitis, usually of the inguinal or perirectal nodes, as well as systemic symptoms. The disease is most commonly seen in tropical and subtropical areas (30).
The trachoma serovars of C. trachomatis are designated by the letters A through K, plus Ba, Da, Ia, and Ja (66, 207). Different serovars have different tissue preferences. Serovars A, B, Ba, and C are the usual ocular isolates from patients with clinical trachoma in regions where trachoma is endemic, while D to K, Da, Ia, and Ja are typically associated with genital tract disease. The latter are the commonest causes of urethritis and mucopurulent cervicitis in females and nongonococcal urethritis in males. They have also been linked to female pelvic inflammatory disease, infertility, ectopic pregnancy, and chronic pelvic pain; male epididymitis, prostatitis, and infertility; neonatal conjunctivitis and pneumonia; and various arthritides.
Strains are tissue selective rather than specific. Even where trachoma is endemic, genital serovars are occasionally found in the eye. Serovars D (15, 31, 127), E (31), F (31, 97, 127), J (93), K (127), and L2 (with serovar A coinfection) (31) have all been isolated from conjunctival swabs taken from individuals with typical clinical signs of active trachoma. Similarly, ocular C. trachomatis strains are sometimes isolated from the genital tract. Frost et al. determined the serovars of 435 isolates taken from male and female attendees at sexually transmitted disease, perinatal, and family planning clinics in Canada and found that 5% were serovar Ba and 2% were serovar C strains (73).
1,043,000-bp chromosome (202). The first gene to be analyzed was that coding for MOMP, which was designated omp1. In 1986, Stephens et al. sequenced omp1 from a C. trachomatis L2 strain after cloning and expressing the gene in an E. coli
bacteriophage (205). Comparison of this gene with that from C. trachomatis serovars that were subsequently sequenced revealed extensive omp1 sequence variation. Most of the polymorphisms were localized to four 40- to 90-bp-long variable domains (VDs), designated VD1, VD2, VD3, and VD4, regularly distributed among the relatively conserved constant domains (CDs). Examination of the accessibility of MOMP segments to digestion by proteolytic enzymes suggests surface exposure of variable domain-encoded peptide sequences, with localization of the protein's amino and carboxy termini inside the periplasmic space (10). Serovar specificity of C. trachomatis appears to be determined by particular residues within VD1, VD2, and VD4 (10, 19). Yuan et al. found that for each serovar, the variable domain coding for the most hydrophilic and charged amino acid sequence contained the serovar-specific epitope (250). Later studies, however, indicated that omp1 of a given serovar can incorporate multiple distinct serovar-specific epitopes, each of which may be found in a different VD (19). Collectively, these findings (plus the demonstration that anti-MOMP antibodies neutralize the organism, as discussed above) indicate that the omp1 gene product, MOMP, spans the outer membrane of the cell envelope and presents immunologically important epitopes, coded for by one or more VDs, at the cell surface. Heterogeneity in omp1 constant domains between urogenital and trachoma isolates of the same Ba and C serovars has been identified. The altered nucleotide sequences produce changes in the amino acid sequences of MOMP and could potentially play a role in determining the tissue tropism or virulence of the organism (74). More extensive analysis by Stothard et al. of 69 strains representing 17 serovars has not, however, supported an association between omp1 sequence and tissue tropism, disease presentation, or epidemiologic success (207).
The first complete C. trachomatis genome sequence (a serovar D isolate) was published by Stephens et al. in 1998 (202; R. S. Stephens, S. Kalman, C. Fenner, and R. David 1998, Chlamydia genome project [http://chlamydia-www.berkeley.edu:4231],accessed 15 January 2002). Notable findings included the localization of an entire set of genes required for peptidoglycan synthesis (despite the lack of demonstrable peptidoglycan in the organism, as discussed above) and genes encoding ATP biosynthetic pathways (despite C. trachomatis's apparent inability to make its own ATP). The genome sequence of a C. trachomatis mouse pneumonitis strain was subsequently also analyzed, which highlighted the presence of a plasticity zone near the chlamydial chromosome's origin of replication (176). This zone includes genes coding for enzymes involved in tryptophan synthesis. Ocular but not genital serovars of the C. trachomatis trachoma biovar have recently been found to carry a deletion or frameshift mutation at this locus. Ocular strains are therefore unable to synthesize tryptophan from exogenous indole (38, 70). This finding is the first known point of difference in the biosynthetic abilities of ocular and genital strains (70).
| NATURAL HISTORY AND CLINICAL FEATURES |
|---|
|
|
|---|
The degree of distress caused by ocular infection with C. trachomatis ranges from minimal to severe. Many infections are asymptomatic. In other cases, following an incubation period of 5 to 10 days, conjunctival infection produces an irritated, red eye and scanty mucopurulent discharge. Involvement of the cornea in the acute inflammatory process can cause pain and photophobia (184). In general, symptoms are milder than would be expected from the appearance of the eye (53).
The first sign of infection is a nonspecific vasodilation of conjunctival blood vessels (184). Specific changes may be noted after infection of several weeks duration (44), with the development of follicles subjacent to the conjunctivae of the fornices, the tarsal plates, and the limbus. Follicles are lymphoid germinal centers and are found immediately below the epithelial cell layer. They are grey or creamy masses 0.2 to 3.0 mm or more in diameter (47). It is not uncommon to find one or two follicles in normal healthy eyes, usually towards the lateral or medial canthi. Because the superficial layer of the conjunctival stroma lacks lymphoid tissue until about 3 months after birth (111), newborns are unable to mount a follicular response to ocular chlamydial infection (223). Papillae may also be noted at this stage: in mild cases, a few isolated, small red dots can be seen with the naked eye. With the aid of a slit lamp, papillae appear as small swellings of the conjunctiva, each with a central vascular core. When inflammation is severe, an intense papillary reaction on the tarsal conjunctiva is associated with a diffuse thickening of the conjunctiva, obscuration of the deep tarsal vessels, and, sometimes, eyelid edema. If the cornea is involved in the inflammatory process, a superficial punctate keratitis may be noted upon instillation of fluorescein into the conjunctival sac. Superficial infiltrates or pannus (subepithelial infiltration of fibrovascular tissue into the peripheral cornea, once thought to be found to at least some degree in every case of trachoma) (62) also indicate corneal inflammation. Follicles, papillae, and these corneal signs are features of active disease; the signs discussed below are all manifestations of late-stage trachoma. Note, however, that pannus may persist long after active disease has subsided.
Resolution of follicles may be accompanied by scarring of the subepithelial conjunctiva. Scar deposition is most prominent in the upper tarsal plate, although the conjunctival fornices, the bulbar conjunctiva, and the upper part of the cornea may also be involved. In areas where trachoma is endemic, upper tarsal plate scars derived from repeated episodes of infection can eventually accumulate to such an extent that they becomes visible macroscopically after eversion of the upper lid, appearing as white bands against the erythematous background of the conjunctiva. At the limbus, replacement of follicles by scar results in the formation of translucent depressions in the corneoscleral junction called Herbert's pits.
If sufficient tarsoconjunctival scarring accumulates, contraction of it over the years will cause the upper eyelid to turn inward so that the lashes rub against the globe. This is known as trichiasis. When the whole lid margin is turned in, the condition is known as entropion. Scars around the bases of hair follicles can pull individual eyelashes into contact with the cornea, even without entropion (184). Trichiasis is intensely irritating. Sufferers may use homemade forceps to remove their own lashes or attempt to keep their lids elevated with strips of cloth tied around their heads.
Besides being painful, trichiasis injures the cornea. In addition to the direct abrasive effect of the in-turned lashes, secondary bacterial and fungal infections of the cornea and corneal drying due to scarring of forniceal-mucous, lacrimal, and Meibomian glands accelerate epithelial damage. Collagenous scar is laid down as part of the repair process. Because scars are opaque, vision can be affected by scarring that involves the central part of the cornea.
| CLINICAL DIAGNOSIS |
|---|
|
|
|---|
Papillae are poorly specific for trachoma, particularly if there are no accompanying follicles. They form part of the conjunctival tissue's response to many acute and chronic inflammatory disorders and are therefore seen in bacterial, viral, and allergic conjunctivitides.
In areas where trachoma is endemic, pannus, conjunctival scarring, and trichiasis are nearly always attributable to trachoma. Herbert's pits are pathognomonic of past trachomatous inflammation. Corneal opacity, however, has many possible etiologies, including previous traumatic injury, viral, bacterial, and fungal infections, and a variety of other conditions. The prevalence of corneal scar is therefore not necessarily a good estimate of the contribution of trachoma to the overall burden of blindness and visual impairment.
The several variants of the MacCallan trachoma classification (130, 131) were probably derived from a description of the four stages of trachoma by Aetus of Amida in the sixth century (245). The MacCallan systems were durably popular in the ophthalmological literature of the first half of the twentieth century and still occasionally appear in papers in peer-reviewed journals (147) and consulting room reference texts (177), despite (i) implying that the clinical course is linear, ignoring trachoma's multicyclic nature; (ii) not mentioning corneal opacity or visual impairment and therefore having little prognostic value; and (iii) lacking clear definitions that would allow one stage to be reliably differentiated from the next (47, 52, 53, 109). Some of the later schemes are exceptionally complex: the 1966 proposal by the Fourth WHO Scientific Group on Trachoma Research, for example, took more than four pages to outline, recommending the grading of up to 19 signs, each of which had its own scale (4). Two classifications are in current general use and will be discussed here. The others have largely been superseded.
In the modified WHO system (53), the upper tarsal follicles (F) are graded F0 for no follicles, F1 for follicles present but no more than five in zones 2 and 3 together (Fig. 2), F2 for more than five follicles in zones 2 and 3 together but less than five in zone 3, and F3 for five or more follicles in each of the three zones. Upper tarsal papillary hypertrophy and diffuse infiltration (P) are graded P0 for absent, normal appearance; P1 for minimal, individual vascular tufts (papillae) prominent but deep subconjunctival vessels on the tarsus not obscured; P2 for moderate, more prominent papillae and normal vessels appear hazy even when seen by the naked eye; and P3 for pronounced, conjunctiva thickened and opaque, normal vessels on the tarsus are hidden over more than half of the surface. Conjunctival scarring (C) is graded C0 for no scarring on the conjunctiva; C1 for mild, fine, scattered scars on the upper tarsal conjunctiva or scars on the other parts of the conjunctiva; C2 for moderate, more severe scarring but without shortening or distortion of the upper tarsus; and C3 for severe scarring with distortion of the upper tarsus. Trichiasis/entropion (T/E) is scored as T/E0 for no trichiasis or entropion, T/E1 for lashes deviated towards the eye but not touching the globe, T/E2 for lashes touching the globe but not rubbing on the cornea, and T/E3 for lashes constantly rubbing on the cornea. Corneal scarring (CC) is scored CC0 for absent, CC1 for minimal scarring or opacity not involving the visual axis (determining whether or not corneal opacity "involves the visual axis" [the line between the fovea and the target] is impossible to determine by inspection [231, 236]: the term "over the pupil" might have been more practical, but the definitions here are reproduced as they appeared in the original publication) and with clear central cornea, CC2 for moderate scarring or opacity involving the visual axis, with the pupillary margin visible through the opacity, and CC3 for severe central scarring or opacity, with the pupillary margin not visible through the opacity.
|
|
|
Preliminary testing of this system by its developers (after testing and modifying a prototype) involved four observers each examining 179 cases (226). The interobserver agreement measurements found in this study are presented in Table 2. The degree of agreement is indicated with the kappa statistic, a measure of observer reliability for categorical data that estimates the extent of agreement not due to chance between two sets of observations of the same variable. Kappa has possible values between 1 and +1, with 1 indicating complete disagreement, and +1 complete agreement. Landis and Koch (118) set arbitrary divisions for describing the relative strength of agreement associated with this measurement, as follows: <0.00, poor; 0.00 to 0.20, slight; 0.21 to 0.40, fair; 0.41 to 0.60, moderate; 0.61 to 0.80, substantial; and 0.81 to 1.00, almost perfect.
|
|
|
|
| LABORATORY DIAGNOSIS |
|---|
|
|
|---|
However, as will be discussed in detail later, clinical signs of active disease do not necessarily mean that the conjunctiva is currently infected with C. trachomatis. Detection of the presence or absence of C. trachomatis is often desirable for research purposes. (Later in this paper, we will present an argument that there may be an additional role for laboratory assays in community assessment, particularly for certifying the elimination of trachoma as a public health problem.) The available assays include microscopy of conjunctival scrapings, isolation in cell culture, direct fluorescent antibody, enzyme immunoassay, serology, nucleic acid hybridization probes, and nucleic acid amplification tests. The characteristics of all of these assays are compared in Table 6. The sections below present the basic principles of the assays and their advantages and disadvantages for detecting ocular C. trachomatis. Laboratory diagnosis of urogenital C. trachomatis infections has been comprehensively reviewed by Black (26).
|
Despite the high specificity of isolation, a number of problems are associated with use of it as a diagnostic test. Inhibition of chlamydial growth in culture can, in theory, be caused by cytokines or antibodies produced by infected tissues and introduced into the culture medium with the clinical sample (180, 243). Even when using purified elementary body stock, which (by definition) should be free of inhibitors, the data suggest that only about one elementary body in several hundred (or more) is capable of successfully infecting tissue culture cells (105); the reasons for this are unclear. Because of the organism's stringent requirements for special transport medium, the need for strict maintenance of the cold chain, and the complicated nature of the culture protocol, there are multiple opportunities for variation in factors that impact isolation efficiency and, therefore, sensitivity (186). This makes it difficult to compare results between laboratories or even to compare one run to the next in the same laboratory. Additionally, chlamydial culture is expensive and time-consuming and requires special expertise.
Although cell culture is considered the gold standard for laboratory diagnosis, it is now accepted that isolation of C. trachomatis in cell culture is less than 100% sensitive (14, 43, 186, 189, 193).
For the clinician, specimen collection for the direct fluorescent antibody test is straightforward: conjunctival cells and exudates from a swab are smeared onto a slide in the field or clinic, fixed with methanol (189), and air dried. The slides are easy to transport and store (126). In the laboratory, the fixed sample is stained with fluorescein isothiocyanate-conjugated monoclonal antibody and examined under a fluorescence microscope. Performing an epithelial cell count provides a straightforward method for determining the adequacy of the sample (216). Samples taken from the upper tarsal conjunctiva yield a greater concentration of infected cells than those from the lower fornix (91).
The MicroTrak performed extremely well in initial sensitivity and specificity trials (232), and the fact that specimens could be transported to the laboratory at ambient temperature gave it a considerable advantage over tissue culture. As a result, the MicroTrak has been widely used in trachoma research studies (28, 31, 55, 57, 100, 135, 189, 210, 217, 240, 242, 244). Comparison between studies is made difficult by the fact that the threshold for defining a positive slide varies from one group of investigators to the next. The major practical disadvantages with the direct fluorescent antibody test are that the average time required to screen each slide is more than 20 min (216), and, because the criteria for positivity are so subjective, specificity is highly dependent on the competence of the microscopist (108).
Chlamydial antigens are eluted from collected swabs in specimen dilution buffer. An aliquot of sample eluate is placed onto a solid-phase support (such as a microtiter plate well or a polystyrene bead) to which antibodies that bind chlamydial antigens have been attached. Bound chlamydial antigen is then detected by the addition of a second antibody conjugated to a developer, such as horseradish peroxidase (103). Following a washing step to remove unbound components, a colorless substrate that is transformed by peroxidase to a colored product is added to the well. The presence of chlamydial antigen-antibody complexes is demonstrated by detecting the color change with a spectrophotometer.
Enzyme immunoassays do not require immediate refrigeration of clinical specimens following collection (128), and specimen processing can be completed in 4 h. High throughput can be achieved by batching. If processing is delayed, however, prolonged sample storage can reduce sensitivity (216). Specificity is also a problem. Chlamydial lipopolysaccharide shares epitopes with a number of other bacterial species (148). As a result, Staphylococcus aureus, Haemophilus aegyptius, Klebsiella pneumoniae, Gardnerella vaginalis, Neisseria gonorrhoeae, Escherichia coli, Streptococcus agalactiae, Moraxella lacunata, Chlamydia psittaci, the Salmonella enterica serovar Minnesota Re mutant, Acinetobacter lwoffii and Acinetobacter calcoaceticus var. anitratus can all react in this type of assay (182, 189, 220; M. M. Rothburn, H. Mallinson, and K. J. Mutton, Letter, Lancet ii:982-983, 1986). Conjunctival infection or sample contamination with any of these organisms could therefore produce a false-positive result. A confirmatory assay that selectively blocks binding of the chlamydia-specific epitope can be used to separate true positives from false positives and thereby increase the specificity of the test (148).
The Boots CellTech (later Dako) IDEIA (Boots CellTech, Slough, England) (12, 125, 128, 164, 238) incorporated a detection system with the potential for improved specificity, with murine monoclonal antibody to chlamydial lipopolysaccharide in place of the polyclonal antibody used in most other tests (133). The detection principles were later further altered by attaching multiple copies of an antilipopolysaccharide monoclonal antibody-alkaline phosphatase complex to a dextran backbone. In this format, designated polymer conjugate enhancement, each copy of lipopolysaccharide in the sample is able to capture multiple copies of the enzyme, resulting in dual amplification of the signal (39). There are insufficient data, however, to conclude that this test has better specificity or sensitivity than other enzyme immunoassays (113).
A number of rapid point-of-care tests with the enzyme immunoassay format are also available. Results are available 30 min after sample collection, but the consensus on these tests seems to be that they sacrifice sensitivity for speed (174, 246).
The microimmunofluorescence technique developed by Wang and Grayston (237) was the first method used to classify strains of C. trachomatis into serovars. The serovar-specific antigens delineated by this test can be used in an indirect fluorescent antibody assay to detect antichlamydial antibodies in serum or tears, with greater sensitivity than achieved with complement fixation (230). Serial dilutions of the sample are placed on glass slides to which antigens of different C. trachomatis serovars have been fixed (Individual Antigen Serovar Kit; Washington Research Foundation, Seattle, Wash.). Following incubation, the slides are probed with fluorescein-labeled anti-human immunoglobulin. Testing for the presence of immunoglobulin A, immunoglobulin G, and immunoglobulin M can be performed separately (132, 216).
Detecting antichlamydial antibody in serum is difficult, subjective, and tedious and has poor specificity and poor reproducibility. The potential advantage of distinguishing between acute, subacute, and chronic infection is not borne out, even with the use of paired acute- and convalescent-phase sera, because production of immunoglobulin M antibody is not stimulated by ocular reinfection with a previously encountered C. trachomatis serotype (230). Its uses are limited. Tear microimmunofluorescence has better correlation with clinical trachoma but suffers from low sensitivity and the same practical disadvantages (216).
The Gen-Probe PACE 2 (Gen-Probe, San Diego, Calif.) is a nucleic acid probe or hybridization probe test. The probe is a synthetic single-stranded DNA molecule complementary to a region of chlamydial rRNA. The sample is heated so that cells are lysed and rRNA is released. The probe, labeled with an acridinium ester, is added; it forms a stable DNA-RNA hybrid with its target sequence. Detection is performed with a hybridization protection assay: following binding, selection reagent is added, which hydrolyses the acridinium ester on unhybridized probes and thereby deactivates it; the acridinium ester on hybridized probes is protected within the double helix of the DNA-RNA complex. With subsequent addition of hydrogen peroxide, bound acridinium ester releases a pulse of light that can be detected with a luminometer. Because there are about 2,000 copies of rRNA per chlamydial cell, sensitivity is slightly better than that achieved by enzyme immunoassay (119, 247), without the need for amplification of nucleic acid.
A second hybridization probe test, the Hybrid Capture II (HCII; Digene, Gaithersburg, Md.), uses an enzyme immunoassay to achieve signal amplification and is therefore classed as a nucleic acid probe-signal amplification assay. Alkali is added to the clinical sample to lyse cells and denature double-stranded DNA (separate it into its two component strands). An RNA hybridization probe with nucleotide sequence complementary to chlamydial DNA is introduced; because the DNA has been denatured, the probe can anneal to its target region. The RNA-DNA complex is then captured by antibody bound to microwell plates. Antibodies labeled with alkaline phosphatase are added to bind to the bound complex. After washing, the presence or absence of bound DNA can be determined by adding a substrate for alkaline phosphatase: cleavage of the substrate results in emission of light (77; M. E. Ward, 2002, Molecular diagnostics: direct hybridization probe tests [http://www.chlamydiae.com/chlamydiae/restricted/docs/labtests/diag_hybridization.htm], accessed 1 July 2002). The test is rapid and reproducible (77), but, for the diagnosis of ocular chlamydial infection, direct hybridization assays have been largely overlooked in favor of the nucleic acid amplification tests discussed below.
There are three phases of the reaction. All take place in the same vessel but at different temperatures. The reaction mixture contains an aliquot of the sample, an excess of deoxyribonucleotide triphosphates, DNA polymerase, and two primers (short, synthetic oligonucleotides that flank the region to be amplified); one primer is complementary to the sense strand, and one to the antisense strand at the opposite end of the target sequence. First the mixture is heated to between 90 and 95°C to break apart the two strands of target DNA, allowing the primers access to complementary sequences on the target. At a reduced temperature, the primers anneal to these binding regions. The mixture is then heated again to enhance the activity of the DNA polymerase, which extends each chain from the 3' end of its annealed primer to produce two double-stranded copies of the target sequence. Both of the two newly synthesized primer extension products contain the appropriate primer-binding regions and, after heat-induced separation from the target, can themselves function as templates alongside the original templates in the next round of duplication. Multiple repetitions of the denaturation-annealing-extension process therefore result in exponential accumulation of the target. PCR is now usually automated in a thermal cycler, which rapidly and reliably changes the temperature of the reaction vessel to provide appropriate conditions for each stage of the amplification process.
PCR is ideally suited to the detection of DNA of fastidious and noncultivatable infectious agents because it does not rely on the presence of viable organisms in the sample. The first bacterium for which a PCR-based detection method was published was C. trachomatis (63).
A number of different nucleic acid sequences have been used as targets in PCRs for the detection of C. trachomatis. These include the chlamydial cryptic plasmid (pCT) (43, 97, 134, 163, 164), omp1, coding for MOMP (27, 63, 96, 134, 163), the gene coding for 16S rRNA (2, 43, 134), and omp2, coding for OmcB (239). With the exception of pCT, all of these targets are sequences found on the C. trachomatis chromosome, which includes two complete rRNA operons and single copies of omp1 and omp2.
PCR directed at plasmid genes (164) or omp1 (27) is thought to be both sensitive and specific for the diagnosis of C. trachomatis infection. Primers designed for the chlamydial rRNA gene amplify this DNA sequence in C. trachomatis, C. psittaci, and C. pneumoniae (43), which reduces the specificity of the assay. The omp2-based PCR (detecting the gene coding for OMC-B) has also been found to return positive results on samples containing any of these three species; subsequent restriction endonuclease digestion and gel electrophoresis permit species and strain identification of isolates (239), but this assay has not been used extensively in published studies to date.
Mahoney et al. (134) estimated that plasmid-based PCRs are between 10 and 10,000 times more sensitive than PCRs directed against C. trachomatis chromosomal genes. This is probably at least partly attributable to the presence of multiple copies of the plasmid per chlamydial cell (134, 163, 164). Bailey et al. (13) also suggested that using a plasmid target gives greater sensitivity. Using serial dilutions of DNA standards, they calculated the lower detection limit of their plasmid-based PCR to be 1 to 10 elementary bodies, compared with 10 to 100 elementary bodies for a PCR agai