Abstract
Making a prompt and accurate diagnosis of genital tract infections is the key to instituting appropriate treatment and the linchpin of sexually transmitted infection control. We present a brief history, not covering syphilis, of diagnostic events for each of six bacteria and one protozoan from the time of discovery up to the molecular revolution. The latter is touched upon but its impact will form the substance of a further presentation. Here, hindsight is helpful in understanding the way in which progress was made over 135 years, often when microbiology, not even seen as a distinct discipline, had a difficult time in providing what was required in terms of dependable diagnostic techniques. Gram-staining, growth on artificial media, growth in cultured cells, enzyme immunoassays, metabolic and immunofluorescence tests have all had their place and some still do despite the avalanche of the molecular era. Serology to determine the existence of organism-specific antibodies has been important in managing syphilis, but has only sometimes been helpful in supporting a diagnosis for other infections and has rarely been the primary deciding factor.
Introduction
Making a prompt and accurate diagnosis of genital tract infections is the key to instituting appropriate treatment and the linchpin of sexually transmitted infection (STI) control. During their evolution, there were times when diagnostic tests were non-specific, insensitive and slow but, nevertheless, the best available, as discussed here. The following is concerned, not with treponemes, but with the diagnosis of six common genito-urinary bacteria and one protozoan and the problems experienced with tests up to and just encroaching on the molecular era. Events in this period are to be covered in a separate article.
Neisseria gonorrhoeae
Gonorrhoea, a disease mentioned in ancient biblical times, defied aetiological understanding, being wrapped in confusion and mystery well into the 1800s. As an example of this, legend has it that John Hunter, a renowned surgeon, inoculated himself, although probably others, with gonococcal exudate and developed gonorrhoea and syphilis. 1 This fortified his ardent belief that the two diseases had the same cause. The confusion was not clarified until the microscopy observations of Albert Neisser were published in 1879. 2 He detected bacteria in urethral smears from subjects with gonorrhoea and the name given to them, Neisseria gonorrhoeae, was a tribute to his discovery. They were not observed in patients with syphilis and they caused gonorrhoea in subjects inoculated experimentally. The stain described by Hans Christian Gram in 1884, 3 used originally to study lobar pneumonia, was soon seized upon because it enabled gonococci to be seen more readily. There cannot be a better example of a technique standing the test of time since it is still in use today, 135 years after its inception. Gonorrhoea, in the late 1880s and thereafter, was treated in a variety of ways, often urethral irrigation in the male and success or failure, usually the latter, was sometimes assessed by Gram-staining, gonococci being sufficiently distinctive for this to be worthwhile. The culture of gonococci in vitro in 1882 4 heralded in a ‘golden’ period when about this time agar was discovered and then Petri dishes invented (1887), so enabling gonococcal colonies to be identified. However, it was more than 50 years before much improved treatment came with the advent of sulphonamide in 1937 and then penicillin in 1944. The gradual development of superior culture medium, 5 particularly Thayer–Martin, 6 by 1966 meant easier diagnosis of infection and better determination of the susceptibility of the bacteria to antibiotics. Use of an immunoassay, Gonozyme (Abbott Labs) from the early 1980s, 7 resulted in at least 30 subsequent publications of its use extending into the mid-1990s. In general, the assay was equivalent in sensitivity to Gram-staining for men and more sensitive in women, but less specific and rarely used routinely in women because of the worry about false-positive results. Perusal of the reports reveals that 70% had various adverse comments and a direct immunofluorescence test (Syva Labs) did not improve matters. 8 Other non-culture tests, for example Gen-Probe Pace-2 which was based on a single-stranded DNA probe designed to hybridize with gonococcal rRNA, 9 were comparable to culture in sensitivity, but no better. Fortunately, circumstances changed for the better around the turn of the century with the introduction of molecular tests with immeasurably superior sensitivity. Serological tests to measure antibodies in patients’ sera were first developed almost 50 years ago and have comprised at least seven methods. 10 The magnitude of the response in terms of antibody titre and duration tends to be greatest in the more severe infections, 11 but is otherwise quite brief. In comparison with organism detection, serological sensitivity and specificity, whatever the test, are poor and no test has gained clinical acceptance for screening, case finding or diagnosis.
Chlamydia trachomatis
Trachoma is recognizable in descriptions of blindness in ancient Chinese and Egyptian writings. However, it was not until 1907 that Ludwig Halbersteadter and Stanislaus von Prowazek, associates of Neisser, first described intracytoplasmic inclusions in conjunctival scrapings from orangutans inoculated with scrapings from patients with trachoma and recognized the involvement of an infectious agent. 12 In 1930, the first isolation of a chlamydial agent (Chlamydia psittaci) was accomplished. Twenty-seven years later, the genomically and biologically different agent, C. trachomatis, was isolated in embryonated hens’ eggs by T’ang et al. 13 in China. Soon after this breakthrough, Barrie Jones and Eric Dunlop and colleagues in London made isolations from patients with ocular and genital disease in hens’ eggs, understandably using the term ‘trachoma inclusion conjunctivitis’ (TRIC) agent. 14 Unfortunately, this technique was somewhat prone to causing laboratory cross-contamination and occasionally false-positive results. It was timely, therefore, that in 1965 Gordan and Quan reported 15 that they had achieved propagation in cell culture (irradiated McCoy cells), a procedure that was not afflicted by the unwanted drawback. Indeed, it became the ‘gold standard’ for detection for over two decades or more, being specific, although rather slow and relatively insensitive, 16 so that some chlamydia-positive patients went untreated, unless treated empirically. The mid-1980s saw the introduction of enzyme immunoassays, such as Chlamydiazyme (Abbott) and IDEIA (Cell Tech Diagnostics), 16 with more than 60 publications, almost up to the turn of the century, 17 attesting to the continued use of these tests, despite a lack of sensitivity. Some (TestPack: Abbott; Clearview: Unipath; Surecell: Kodak), although very rapid were, without question, unacceptably insensitive,17,18 so that a scenario could be envisaged of some women carrying a small organism load being regarded as chlamydia-negative and, as with insensitive cell culture, going untreated with possible devastating consequences. That is about a 17% risk of pelvic inflammatory disease (PID), a 7% risk of salpingitis, a 0.5% risk of tubal factor infertility and a 0.2% risk of ectopic pregnancy. 19 The ‘MicroTrak’(Syva) immunofluorescence technique, available a few years later, proved very sensitive for those who had the ability to detect <10 fluorescing chlamydial particles.16,18 However, for others who failed to detect such small numbers, the test was insensitive and, yet again, many subjects were erroneously regarded as chlamydia-negative with possible misfortune facing them. In the early 1990s, commercial PCR and the ligase chain reaction (LCR) test, both specific and very sensitive, 18 were introduced successfully and affirmed the insensitivity of any test that had been used before. However, irrespective of achievements in the molecular field, efforts to produce a rapid, sensitive and specific non-molecular point-of-care (POC) test continued. Although a two-step enzyme detection system (HandiLab-C) was eventually found to fail in these respects, 20 in 2003, after many years of perseverance, a research group in Cambridge, UK, announced 21 the development of a signal amplification system (SAS) which provided results rapidly in tests for C. trachomatis. While not particularly sensitive in test results published in 2007, 22 some physicians had become interested in its possible potential, 23 their notion being that a test with a sensitivity less than desired, but which provided a result allowing almost immediate treatment, was preferable to a slower molecular test of greater sensitivity. The argument that it would be helpful in settings without advanced diagnostic facilities, for example in some developing countries, 24 took a severe blow by the finding of extreme insensitivity by manufacture-independent groups testing in resource poor settings.25,26 Insensitive non-molecular tests have been a continuing blight, as seen by the poor performance of the BioChekSwab rapid test, 27 apparently an updated version of the HandiLab-C test. This raises the point that from the 1990s to the present time, some failed (insensitive) tests have been presented by manufacturers under another guise, providing fresh hope to users and patients. Of no benefit to either, the intricacies and machinations of this deceitful practice have come under withering and justified rebuke, 28 which should ensure that history is not repeated. Serological tests, mainly microfluorescence assays, 29 enabling assessment of IgM, IgG and IgA antibodies have been helpful in supporting clinical and microbial findings. This has been seen in the case of epididymitis 30 and upper genital tract disease in women if severe or chronic, 31 or in tubal infertility, 32 or in lymphogranuloma venereum, when higher than usual antibody titres might be expected. However, serology is not the ‘first port of call’ in making most diagnoses.
Mycoplasma hominis
This mycoplasma, isolated from an abscess of Bartholin’s gland in 1937 by Dienes and Edsall,
33
was the first of human origin to be discovered. Subsequently, isolation in culture and identification has been common practice and generally not regarded as difficult using a fluid or solidified agar medium of beef-heart infusion broth supplemented with yeast extract and animal serum, often horse. Commercial kits, for example IST (BioMerieux),
34
combining detection and antibiotic testing, may be useful for those who seldom undertake such work. However, whether seeking M. hominis is worthwhile is highly questionable as there is no evidence that M. hominis causes non-gonococcal urethritis (NGU) or other genital tract disease in men. In women, this mycoplasma seems an occasional cause of PID,35–37 but determining its exact role has proved very difficult.
37
Certainly, its detection in the vagina or urine of an individual with symptoms and/signs of PID does not mean that it is a cause. M. hominis organisms have also been found in a small number in the vagina of some healthy women and sometimes in a large number (≥105 organisms
Gardnerella vaginalis
This bacterium was isolated by Leopold in 1953 from the genital tract of women with cervicitis. 41 Two years later, Gardner and Dukes 42 found the bacterium in the vagina of women with BV and termed it Haemophilus vaginalis, a name subsequently changed to Gardnerella vaginalis. With the development of sensitive culture media, 43 G. vaginalis was found in women without signs of a vaginal infection, so that its mere presence was not a sound reason for treatment. Moreover, finding that many other bacterial species, for example M. hominis (mentioned above), Mobiluncus and Atopobium vaginae, were also strongly associated with BV, 44 became a complicating factor in diagnosing this condition. A further problem with G. vaginalis is that it does not comprise a homogeneous group of bacteria, being made up of four distinct molecular subgroups or clades, possibly amounting to separate bacterial species, which may have distinct roles in BV pathogenesis. 45 However, until this issue is finally settled, the diagnosis of BV may be made, not on clinical grounds, but accurately by observing changes in the cellular and microbial flora in Gram-stained vaginal smears.46,47 BV-positivity is more likely to imply the need for treatment than is G. vaginalis-positivity.
Ureaplasma spp
Ureaplasmas were isolated by Maurice Shepard in 1954 from the urethra of men with NGU. 48 They produced very small colonies on agar media, posing a difficulty in detection. This was overcome by finding that the organisms were unique in metabolizing urea, 49 hence the name. In liquid medium containing urea and a pH indicator, such metabolism resulted in a colour change, indicative of organism multiplication, and allowed a quantitative assessment to be made. 50 Twenty-three years after the discovery of ureaplasmas, Koch’s postulates were fulfilled 51 ; human experimentation with these organisms, aided by sound microbiological backing, had a much greater chance of success than did the gonococcal exploits of Hunter who, inevitably, lacked any microbiological support. Success was a boost to further work on the aetiology of disease, mentioned elsewhere, 52 and the initial suspicion 53 and final recognition 54 that the human ureaplasmas comprised two species, Ureaplasma urealyticum and U. parvum, prompted a further flurry of activity designed to define clinical associations. Serological tests based on enzyme immunoassays, metabolism inhibition and micro-immunofluorescence have supported but not supplanted ureaplasma organism detection. This has been seen, for example, in studies of NGU 55 and epididymitis. 56
Mycoplasma genitalium
During studies of acute NGU in the 1970s, some men were seen to respond to tetracycline therapy despite the failure to detect pathogenic bacteria in the urethra. 57 In addition, dark field microscopy of urethral smears sometimes revealed motile spiral forms resembling spiroplasmas which infect plants and insects and prompted the thought that there might be a human counterpart. In view of all this, urethral specimens were taken in 1980 by one of us (DT-R) to Joe Tully’s Laboratory at the National Institutes of Health, USA where they were inoculated into SP4 medium conducive to the growth of spiroplasmas and mycoplasmas. Spiroplasmas were not isolated but two strains of a glucose-fermenting mycoplasma were, as was a third after the specimens were taken back to the UK. The strains were found to be closely related but different serologically from all known mycoplasmas. The discovery of a new mycoplasma, later termed Mycoplasma genitalium, was announced at a meeting in Jerusalem in 1981 5 7 and published soon after in the Lancet. 58 Further isolation attempts in the same way, but with different batches of medium, failed. Indeed, no one was successful anywhere in the world and it was not until the use of PCR technology in 1991 that detection was achieved again. 59 During this ‘fallow’ decade, however, Koch’s postulates were fulfilled by demonstrating parametritis in baboons, salpingitis in marmosets and grivet monkeys and, most impressively, urethritis in male chimpanzees. 60 This was the forerunner of successful studies in the human field. 59 Serological tests have been used in different ways. A micro-immunofluorecence test was developed to measure antibody to M. genitalium and was used to provide the first indication that this mycoplasma might be involved in PID 61 and probably in NGU, 62 both studies prior to the molecular period. Later, use of an enzyme-linked immunosorbent assay provided evidence for the involvement of this mycoplasma in female infertility. 63
Trichomonas vaginalis
In 1836, in a book describing his microscopic examination of genital discharges, Alexandre Donné 64 reported the existence of often round, flagellated, motile organisms. They were thought to be harmless commensals and, remarkably, it was more than 100 years later before Koch’s postulates were fulfilled, 65 establishing T. vaginalis in the aetiology of vaginitis, and the concept of trichomoniasis as a STI becoming firmly established. For years, and even now, microscopic examination of a ‘wet mount’ prepared from a genital tract specimen has been a convenient rapid POC diagnostic test. A sensitivity of less than 45–60% is, however, much less than that of other methods, resulting in many missed infections. 66 Culture is more sensitive, at 60–80%,66,67 for detecting T. vaginalis, but requires complex media (for example, Diamond medium) and incubation for up to a week. Medium contained in a plastic envelope (‘pouch’), obtained commercially, 68 may be used for convenience. An immuno-based test (OSOM Trichomonas Rapid Test), 69 in view of the comments made about chlamydial immunoassays, has a surprisingly, but seemingly genuine, high sensitivity which out-performs microscopy. However, how such a single POC test fits into the diagnostic pathway of a modern sexual health clinic remains to be seen. Serological test: although specific antibody to T. vaginalis may be detected in sera and vaginal secretions by, for example, an enzyme-linked immunoabsorbent assay, 70 serology has never formed part of the diagnostic arsenal.
The current diagnostic situation
This turns largely on molecular tests which will be discussed in some detail in a future article, together with home-care ‘on-line’ testing and other aspects in the modern era.
Footnotes
Authors’ note
Some of the work referred to in this review was supported by the Medical Research Council and by the NIHR Health Protection Research Unit in Evaluation of Interventions at the University of Bristol. The views expressed are those of the authors and not necessarily those of the MRC, NHS, the NIHR, the Department of Health or Public Health England.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.