Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Marine microorganism wikipedia , lookup
Traveler's diarrhea wikipedia , lookup
Triclocarban wikipedia , lookup
Anaerobic infection wikipedia , lookup
Bacterial cell structure wikipedia , lookup
Human microbiota wikipedia , lookup
Infection control wikipedia , lookup
Urinary tract infection wikipedia , lookup
Bacterial morphological plasticity wikipedia , lookup
E European Society of Human Reproduction and Embryology Human Reproduction Update 1999, Vol. 5, No.4 pp. 373–385 Hypothesis on the role of sub-clinical bacteria of the endometrium (bacteria endometrialis) in gynaecological and obstetric enigmas D.A.Viniker Department of Obstetrics and Gynaecology, Whipps Cross Hospital, London, UK Unexplained infertility, recurrent abortion, dysfunctional uterine bleeding, pelvic pain, premenstrual syndrome, premature labour, placental insufficiency and pre-eclampsia are examples of common obstetric and gynaecological problems that frequently defy adequate explanation. Bacterial vaginosis, a non-inflammatory condition, is associated with premature labour, but antibiotics administered topically provide less effective prophylaxis than those administered orally. This would indicate that bacterial vaginosis might be a marker for significant genital tract bacteria, but some pathology is dependent on micro-organisms ascending out of reach of topical antibiotics. The author was led to consider the hypothesis that micro-organisms, possibly those associated with bacterial vaginosis, surreptitiously inhabit the uterine cavity (bacteria endometrialis) where they are culprits of some common gynaecological and obstetric enigmas. The objective of this review is to provide an initial theoretical examination of this hypothesis. Bacteria in the endometrium have been associated with infertility. Antiphospholipids have been linked to recurrent miscarriage and pre-eclampsia and with infections including Mycoplasma. Pre-eclampsia might be explained by an exaggerated host response to intrauterine micro-organisms or bacterial toxins. The hypothesis that one common factor, bacteria endometrialis, could provide a plausible explanation for a variety of obstetric and gynaecological mysteries is particularly intriguing. There is sufficient evidence to justify further investigation. Key words: endometrial bacteria/pre-eclampsia/recurrent miscarriage/subclinical bacterial infections/unexplained infertility TABLE OF CONTENTS Introduction Bacterial vaginosis A comparison of bacterial vaginosis and bacteria endometrialis Dysfunctional uterine bleeding and pelvic pain Bacteria endometrialis and unexplained infertility Pregnancy complications and sub-clinical bacteria of the endometrium Conclusions Acknowledgements References 373 373 374 375 376 377 381 382 382 certainly, each of these clinical syndromes has more than one initiating pathological process, which confound scientific analysis. At a recent conference, ‘The Problem with Prematurity II’, held at St Thomas’ Hospital, London, 7–9th September 1998, a variety of thought-provoking papers was presented. On reflection, several of these papers led the author to the above hypothesis. It is becoming apparent that micro-organisms associated with bacterial vaginosis (BV) are capable of initiating local symptoms without clinical evidence of inflammation, as well as a pathological process as potentially devastating as prematurity. Expanding knowledge about BV may enhance our ability to consider the possible implications of BV-type organisms higher up the genital tract. Introduction Numerous hypotheses have been expounded to explain a number of obstetric and gynaecological syndromes: invariably, they may meet some—but not all—of the criteria. Almost Bacterial vaginosis More than a century ago, Alfred Doderlein (1892) discovered that lactobacilli are inhabitants of the vagina, where they have a Address for correspondence: Department of Obstetrics and Gynaecology, Whipps Cross Hospital, Whipps Cross Road, Leytonstone, London E11 1NR, UK 374 D.A.Viniker symbiotic relationship with their host. The vaginal flora is normally dominated by lactobacilli, which are responsible for reducing the pH by metabolizing glycogen from squamous cells to lactic acid. The resultant acidic milieu provides protection against infection. BV is recognized as the most common cause of vaginal discharge and it was once believed to be no more than a nuisance that could be ignored in asymptomatic women. The discharge tends to be malodorous, particularly after intercourse. A remarkable feature of BV is the absence of a host reaction; thus, the suffix ‘osis’ rather than ‘itis’ is preferred because signs of inflammation are usually absent (Hay, 1994; Donders, 1998). A variety of organisms can be cultured from the vagina in asymptomatic, as well as symptomatic, women. Bacteriological observations on vaginal samples are therefore more difficult to interpret than those obtained from sites where sterility would be expected. In the 1950s, the investigation and management of vaginal discharge was revolutionized (Gardner and Dukes, 1955) when the organism now called Gardnerella vaginalis was first described. BV has become the adopted term to describe a clinical condition characterized by an overgrowth of predominantly anaerobic bacteria within the vagina and a concomitant reduction or absence of lactobacilli. A change in pH, possibly following menstruation or sexual intercourse, may be the trigger for BV to develop. Women with BV are found to have a group of micro-organisms present in high concentrations. This group includes the anaerobic organisms G.vaginalis, Mobiluncus, Bacteroides, Peptostreptococcus species and Mycoplasma hominis and Ureaplasma urealyticum. There are a number of strains of some of these micro-organisms (Grattard et al., 1995). Mycoplasmas are the smallest free-living micro-organisms, with a high incidence of antigenic variation and variable reactivity to monoclonal antibodies (Christiansen et al., 1997). Some BV organisms are described as being opportunistic. The clinical manifestations of BV may depend on a synergistic interaction of a variety of these organisms (Levett, 1989). It is unknown whether there is one particular microbe that is responsible for BV, or even whether the principal organisms are among those previously cited or are still to be identified (Hill, 1993). The diagnosis of BV can be made by the presence of three out of four of the following: a homogeneous discharge; pH >4.7; positive amine test; and ‘clue’ cell identification (Amsel et al., 1983; McGregor and French, 1998). The laboratory investigation for BV is dependent on microscopy (Donders, 1998) rather than culture. A Gram stain of vaginal fluid may assist diagnosis (Spiegel et al., 1983). The incidence of BV is increased in underprivileged communities. G.vaginalis can be recovered from the urethrae of male sexual partners (Gardner and Dukes, 1955). BV is mostly related to sexual activity, although it has been demonstrated in virgins (Bump and Buesching, 1988). Organisms associated with BV can be transmitted from mother to neonate (Dinsmoor et al., 1989; Gannon, 1993). The mode of transmission of BV organisms remains controversial. Recent studies (Holst, 1990) suggested that BV organisms may colonize the vagina from the woman’s intestinal tract. It is now considered that although BV is usually sexually transmitted, this is not exclusively the case (MacDermott, 1995). The Gram-negative organisms, including Bacteroides, are sensitive to the antibiotic metronidazole, whereas Mycoplasma, Ureaplasma and Mobiluncus are sensitive to macrolide antibiotics such as erythromycin, and to the tetracyclines. A comparison of bacterial vaginosis and bacteria endometrialis The bacteriological investigation of the vagina and endometrium are compared in Table I. This shows how the possible significance of bacteria endometrialis (BE) could have been missed or overlooked. Table I. A comparison of the bacteriology of the vagina and endometrium, showing how the possible significance of bacteria endometrialis could have been missed or overlooked Vagina Endometrium Rich in micro-organisms Relatively sterile. High vaginal swabs: routine clinical investigation Samples confined to research centres. Bacterial vaginosis: diagnosis by microscopy of wet preparation or Gram stain Wet preparations and Gram stain: not studied Culture unhelpful for bacterial vaginosis Culture only. Specialist centres required for Mycoplasma hominis and Ureaplasma urealyticum Possible marker for bacteria endometrialis Micro-organisms can be present even with negative cervical cultures Bacterial vaginosis: no clinical inflammation (‘-osis’, not an ‘-itis’) Micro-organisms can occur with negative clinical examination High vaginal swabs are frequently obtained in routine clinical practice, whereas the bacteriology of endometrial samples has been studied only by a few research groups. Many micro-organisms can colonize the vagina without being pathogenic. The bacteriological diagnosis of BV is dependent on microscopic assessment of a wet preparation of the discharge or a Gram stain rather than culture, whereas investigation of the endometrium has depended on culture alone. Despite extensive investigation, the significance of the varied patterns of bacteria found in the vagina remains controversial (Donders, 1998). Our knowledge about endometrial organisms is comparatively spartan and more difficult to interpret. There is an association between cervical colonization with U.urealyticum and chronic endometritis (Khatamee and Sommers, 1989). Micro-organisms can be present in the endometrium when cervical cultures are negative (Lucisano et al., 1992). Subsequently, endometrial microbiology and histopathology Sub-clinical bacteria of the endometrium were evaluated in women with symptomatic bacterial vaginosis but no clinical evidence of pelvic inflammatory disease (Korn et al., 1995). Plasma cell endometritis was found in 10 of the 22 women studied compared with one of the 19 controls (P < 0.01). Positive endometrial cultures for BV-associated organisms were significantly more common in association with plasma cell endometritis (P = 0.002), but eight of 30 women without histological changes also had positive cultures. It may be that BV is a marker for BE, but there is probably a poor correlation (Koren and Spigland, 1978). Bacteria have been identified in the endometrium in asymptomatic as well as symptomatic women (Table II). Dysfunctional uterine bleeding and pelvic pain Menorrhagia is a frequent gynaecological problem that can be debilitating. The term dysfunctional uterine bleeding is adopted when no clinically obvious causation can be found (Viniker, 1997a). Pelvic inflammatory disease is recognized as a cause of menstrual disturbance, although in clinical practice the diagnosis is usually dependent on evidence of salpingitis. The significance of chronic endometritis remains an area of debate. Some claim that even a few plasma cells in the endometrium 375 are diagnostic, but these cells may be found in healthy women (Fox, 1973a). It was concluded (Tindall, 1987) that it is illogical to apply the term to those with menorrhagia or pelvic pain. Pathogenic micro-organisms have been demonstrated in the endometrium without evidence of pelvic infection being visible at laparoscopy and with negative cervical cultures (Lucisano et al., 1992). An interesting hypothesis has been put forward (Profet, 1993) that menstruation is designed to flush out intrauterine infection: increased menstrual flow is a sign that the uterus has detected infection. In a double-blind study, metronidazole tablets (500 mg) or a placebo were administered daily to 42 gynaecological outpatients with irregular bleeding or discharge and BV (Larsson et al., 1990). Mobiluncus sp. was identified in 81% of cases before treatment. Metronidazole gave a cure rate of 76% compared with 5% in the placebo group; repeated therapy, when required, achieved a cure rate of 100%. It was concluded that it is important to treat BV in patients with symptoms other than malodorous discharge. The bacteriology of the endometrial cavity was evaluated immediately after hysterectomy in 99 women (Moller et al., 1995), with nearly a quarter of the endometria being shown to be colonized by one or more microorganism. Table II. Clinical studies of the bacteriology of the cervix/vagina and endometrial cavity Reference Clinical situation Organism Cervix/vagina N1 Idriss et al. (1978) Koren and Spigland (1978) Stray-Pedersen et al. (1982) Kundsin et al. (1986) Naessens (1987) Cararach et al. (1998) Normal Infertility -Unexplained -Explained Normal Infertile Normal Infertile Infertile Normal pregnancy Induced abortion Normal fertile Spontaneous abortion Recurrent abortion Uncontrollable Preterm labour Normal U.urealyticum Mycoplasma U.urealyticum Mycoplasma/ Ureaplasma U.urealyticum Culture (all relevant organisms) Preterm labour Normal Preterm labour Normal Preterm labour Normal Preterm labour Mycoplasma Gram-negative Bacteria Gram stain 67 161 40 121 50 59 40 379 99 310 84 41 122 76 6 4 30 22 59 58 30 59 30 59 29 59 N2 22 72 22 50 Intrauterine site % 47 57 Endometrium Endometrium 129 35 17 65 49 5 % 3 16 3 98 64 6 27 7.5 26 65 33 45 55 41 Endometrium 19 215 N3 41.6 41.6 41.5 53.3a 64.5a 83 8 27 24 41 7 16 0 8 1 13 23 27 0 13.6a 3.4 22a Endometrium 9.7% 9.7 Endometrium Placenta Amniotic fluid 27.6% 4 2 27.5 66 50 Amniotic fluid 2 2 N1 = no. of patients studied; N2 = no. of patients with positive culture at cervical/vaginal site; N3 = no. of patients with positive culture at intrauterine site. aP < 0.05. 376 D.A.Viniker Figure 1. Histology of endometrium from a 36-year-old lady with intermenstrual bleeding. This was initially assessed to be a possible focus of adenocarcinoma, but subsequently confirmed as focal intraendometrial abscess with regenerative atypia. Original magnification ×40. Scale bar = 20 µm. A 36-year-old lady was admitted for diagnostic laparoscopy to investigate pelvic pain. As there had been some recent intermenstrual bleeding, hysteroscopy, cervical dilatation and endometrial curettage were also performed. The histological assessment of the curettings initially indicated a possible focus of well-differentiated adenocarcinoma, but on reassessment the lesion proved to be focal intra-endometrial abscess with regenerative atypia (Figure 1). In another study, the characteristic subacute inflammation of the endometrium was described (Kundsin et al., 1986) and believed to be associated with Mycoplasma infection. The significance of endometriosis in relationship to pelvic pain is debated. Some women with extensive disease are asymptomatic, whereas others with minimal disease report severe pain. An endometriosis session held during ESHRE’s 7th Annual Meeting in 1991 concluded that “Endometriosis does not exist; all women have endometriosis” (Evers, 1994). Recent studies have shown that the symptoms of endometriosis are related to local inflammation (Taylor et al., 1997). It is possible that there may be micro-organisms in the pelvis that are causing symptoms without evidence of inflammation. Recently, we have been prescribing metronidazole and erythromycin around the time of laparoscopy, even in the absence of positive findings. Results have been encouraging, but larger patient numbers and controlled studies would be required before validated conclusions could be made. Bacteria endometrialis and unexplained infertility Infertility is a common problem that affects 15% of couples. Only half of these will eventually have a child (Templeton, 1995). Anovulation, pathology of the Fallopian tubes and male factor problems may be found in about 75% of couples investigated for infertility (Viniker, 1997b). Depending on the criteria employed, infertility remains unexplained in about 25% of cases. Ovulatory disorders can be successfully treated with simple orally administered ovulation-stimulation regimens (Viniker, 1996). Audit provides a means of improving the quality of care provided in fertility programmes (Viniker, 1999). Reviews of the literature have not given high priority to the concept that unexplained infertility might be due to sub-clinical intrauterine infection (Viniker, 1997b). Similarly, there has been little attention to the potential contribution of antibiotics for unexplained infertility (Viniker, 1997c). The role of infection as a cause of infertility is well recognized, but this is almost exclusively in relation to the Fallopian tubes. Cytokines secreted by endometrial leukocytes, which could be associated with infection, have been implicated in reproductive failure (Rutanen, 1993). A variety of micro-organisms in the endometrium have been associated with infertility (Emembolu, 1989). Important pathogens, such as Chlamydia trachomatis, can be present in the upper genital tract without evidence of pelvic infection being visible at laparoscopy and with negative cervical cultures (Lucisano et al., 1992). Several groups have concluded that C.trachomatis might be a cause of unexplained infertility in otherwise asymptomatic couples (Bjercke and Purvis, 1992; Berclaz et al., 1993; Greendale et al., 1993; Xiang and Chen, 1996). In contrast, no difference was found in the incidence of seropositive patients in infertile and fertile couples (Auroux et al., 1987). It was considered (Czernobilsky, 1978) that endometrial biopsy was an essential tool in the evaluation of infertility, the conclusion being that endometritis plays a significant role in the aetiology of subfertility. Plasma cells, which are evidence of chronic infection, and foci of lymphocytes in the endometrium may be the sole clue of Mycoplasma endometritis (Jones and Jones, 1981). Mycoplasma may be implicated in some cases of unexplained infertility, but a positive culture is difficult to obtain as the technique requires expertise (Kundsin et al., 1986). Endometrial washings taken from 59 women with unexplained infertility showed 16 patients (27%) to prove positive for Mycoplasma, and five (31%) of these conceived within a few months of a 7-day course of the antibiotic, doxycycline (Koren and Spigland, 1978). These authors concluded that endometritis due to Mycoplasma could be a cause of infertility. A further investigation was conducted in 161 consecutive infertility patients and 67 controls for U.urealyticum (Idriss et al., 1978). Of the patients with unexplained infertility, 55% had a positive culture from the endocervix compared with 32% of the controls (P = 0.03). In another study, Mycoplasma was isolated from 18% of infertile women, though patients with evidence of Mycoplasma in the endometrium had no specific histological feature (Bercovici et al., 1978). Sub-clinical bacteria of the endometrium A series of 379 women with infertility were examined for the presence of U.urealyticum and the results were compared with those from 40 controls (Stray-Pedersen et al., 1982). The organism was demonstrated in the cervix in about half of each group. There was a significant increase (P < 0.05) in positive cultures from the endometrium in the infertile group (26%) compared with the controls (7.5%) which could not be related to tubal pathology or male infertility; neither could results be attributed to contamination during sampling. In a study of 100 consecutive infertile middle and upper-class infertile couples with longstanding failure to conceive and who were previously under the care of others (Kundsin et al., 1986), genital Mycoplasma was cultured from 64 of the women, and antibodies to Chlamydia were present in 23. Endometrial histology showed evidence of Mycoplasma in 47 of the 86 patients biopsied, while 24 of the 39 negative biopsies yielded Mycoplasma on culture. Antibiotics for uterine bacteria in infertility Antibiotic therapy for infertility associated with infection specifically of the endometrium or cervix has received relatively scant attention. A search of the literature located no reference to sub-clinical infection as possible causation of infertility. Cervical infection has been implicated as a cause of infertility, and doxycycline was found to be efficient at reducing cervical mucus infection (Buvat et al., 1984). Following treatment there were nine pregnancies within four months in a group of 13 women with otherwise unexplained infertility of more than a year. A probable beneficial effect was found for couples with one partner having doxycycline for proven Mycoplasma infection (Hinton et al., 1979), but this could not be confirmed (Upadhyaya et al., 1983). In another study, tetracycline (500 mg, four times daily) was administered to both partners for 2 weeks if the endocervical culture of Mycoplasma taken at the time of postcoital test proved positive (Idriss et al., 1978). In the six months following treatment, the pregnancy rate was 42% compared with 32% without treatment, although this was not statistically significant. Others (Kundsin et al., 1986) administered oxytetracycline, if there was evidence of Mycoplasma, to their patients and partners for 10 days. If endometrial cultures were positive on repeat sampling 3 months later, a more prolonged and higher dose regimen was provided. There were 43 pregnancies among 58 patients (71%) who had been followed for one year after treatment. Seminal infection and infertility There has been only one reported study evaluating metronidazole in relation to semen quality (Gopalkrishnan et al., 1990). This group examined the semen from 1131 asymptomatic men, 50 of whom were found to have Trichomonas, and these were treated with metronidazole (400 377 mg, three times daily) for 10 days. Significant improvement in semen quality was demonstrated in 25 cases, the authors concluding that some cases of infertility in asymptomatic men could be due to Trichomonas, though they do not seem to have considered that the antibiotic might have been inadvertently treating other organisms. There is evidence that spermatozoa are vectors of disease (Profet, 1993) as, during mammalian insemination, bacteria from the male and female genital tract cling to sperm tails and are transported to the uterus. Indeed, an affinity of human spermatozoa for U.urealyticum has been demonstrated (Nunez-Calonge et al., 1998). Reactive oxygen species, which are produced by leukocytes, have an adverse effect on sperm function and are frequently observed in the semen of infertile men. Reactive oxygen species are frequently encountered with infections in general, and specifically in association with male accessory gland infection (Depuydt et al., 1996; Wang et al., 1997). Antioxidant treatment, however, probably provides little benefit (Martin-Du Pan and Sakkas, 1998). A Medline search of the literature found no evaluation of the possible benefit of erythromycin or metronidazole in relation to reactive oxygen species. The clinical significance of sub-clinical infection in the semen remains controversial (Eggert-Kruse et al., 1995). Pregnancy complications and sub-clinical bacteria of the endometrium Recurrent abortion and bacteria endometrialis Spontaneous abortion is estimated to occur in approximately 20% of pregnancies (Salat-Baroux, 1988). A standardized definition for recurrent abortion has not been universally adopted, but many apply the term to three consecutive miscarriages. Recurrent abortion affects between 0.4 and 0.8% of couples. A variety of factors have been implicated, including maternal congenital Mullerian anomalies, fibroids, intrauterine adhesions, cervical incompetence, fetal genetic disorders, elevated maternal luteinizing hormone (LH) concentrations, antiphospholipid syndrome, general maternal medical disorders, toxins and psychological disorders. The role of infectious agents remains to be elucidated (Carp, 1997). The one study that has investigated the possible relationship between BV and recurrent miscarriage (Llahi-Camp et al., 1996) demonstrated an increased incidence of BV with second-trimester miscarriage rather than first-trimester loss. The role of intrauterine colonization with organisms such as Mycoplasma has been difficult to evaluate (Salat-Baroux, 1988). U.urealyticum was found in the cervix of 41.6% of 310 normal pregnant women, in 53.3% of 122 women with spontaneous abortion, and in 64.5% of 76 women with recurrent spontaneous abortion (Naessens et al., 1987). The isolation of this organism was significantly higher in 378 D.A.Viniker association with spontaneous abortion (P < 0.05) and recurrent abortion (P < 0.005). In these studies, endometrial colonization by U.urealyticum was found to be more common in women with recurrent abortion (27.6%) than in normal fertile women (9.7%; P < 0.05). On administering doxycycline or erythromycin to women with recurrent abortion and genital Mycoplasma, the pregnancy loss rate was diminished with both antibiotics, although better results were recorded with erythromycin (Quinn et al., 1983). The course of subsequent pregnancies following doxycycline therapy after primary or recurrent spontaneous abortions in 254 couples was investigated (Toth et al., 1986), and the outcome found to be significantly better in the antibiotic-treated group, while the rate of spontaneous abortion recurrence was significantly lower (10% versus 38%). The incidence of premature rupture of the membranes was reduced after antibiotics (4% versus 30%), and there were significantly lower neonatal complications after antibiotics including prematurity, fetal distress, meconium, respiratory distress syndrome and neonatal infection. These authors concluded that spontaneous abortion might be caused by bacteria present in the genital tract at the time of conception. These bacteria might have an adverse effect on the course of pregnancy and result in increased maternal and fetal complications. A Medline search of the literature revealed no reference to the use of metronidazole as a treatment for recurrent miscarriage. Pregnancy complications and the antiphospholipid syndrome In a series of 933 consecutively booked pregnant women, 20 women (2%) had lupus anticoagulant or anticardiolipin (Pattison et al., 1993). These conditions were found in 15% of women with recurrent miscarriage (Rai et al., 1995). Seven women among a group of 43 (16%) presenting with severe pre-eclampsia before 34 weeks gestation had antiphospholipid antibodies (Branch et al., 1989). Women with positive results are at increased risk for recurrent pregnancy loss, intrauterine growth retardation, pre-eclampsia, placental abruption, perinatal morbidity and mortality (Pattison et al., 1993; Lakasing and Poston, 1997). The mechanism underlying these adverse outcomes has yet to be described (Lakasing and Poston, 1997). It is not known whether the antiphospholipids are a direct cause of these pregnancy complications, or an epiphenomenon (Rai and Regan, 1998). In the context of the hypothesis that micro-organisms within the uterus may be associated with pregnancy complications, it is intriguing that there is an association between a variety of infections, including Mycoplasma, and antiphospholipid antibodies (Snowden et al., 1990; Catteau et al., 1995). The most rewarding cause of recurrent abortion from the treatment point of view is the presence of lupus anticoagulant and anticardiolipin. Aspirin, heparin, steroids and immunoglobulins have been evaluated (Rai and Regan, 1998). Clinical trials have confirmed significant benefits with aspirin and combined aspirin and low-dose heparin. Antibiotics do not seem to have been considered. Some in-vitro fertilization (IVF) units, including our own, administer antibiotics at the time of egg collection to reduce the risk of pelvic infection, though it is possible that this would be treating undiagnosed BV or BE. We have commented previously on our low miscarriage rates, which we were unable to explain (Viniker et al., 1997). A national study has been commenced to compare outcomes between IVF units according to their antibiotic policy. Premature labour Premature delivery and placental insufficiency, with accompanying fetal oxygen deprivation (Viniker, 1979), are the leading causes of perinatal mortality and morbidity. Premature birth can be due to the culmination of a variety of interdependent factors, one of which is cervical length (Newman, 1998). Epidemiological trends in the prevalence of prematurity are difficult to interpret, as there are a number of confounding factors. Risk factors for premature delivery include previous low birth weight, maternal illness (including diabetes) and vaginal infections (Di Renzo et al., 1998). There is increasing evidence that BV is related to premature labour (Hauth et al., 1995; Di Renzo et al., 1998; McGregor and French, 1998). The vaginal fluid of pregnant women with BV has higher concentrations of mucolytic enzymes, including mucinases and sialidases (McGregor et al., 1994). It was suggested that these enzymes could adversely affect host defences during pregnancy, increasing the risk of sub-clinical intrauterine infection and premature delivery. In addition to premature labour, an association has been found with fetal growth retardation and stillbirth (Lamont et al., 1998). As a result of their investigations, these authors suggest that screening and treating for BV in pregnancy is likely to prove advantageous. Erythromycin and metronidazole can reduce the prevalence of premature delivery in women with bacterial vaginosis (Hauth et al., 1995). Treatment with topical vaginal antibiotics has proved less effective for prevention of premature delivery than oral antibiotics (Majeroni, 1998; McGregor, 1998). This would indicate that the micro-organisms responsible for premature labour have ascended out of reach of topically administered antibiotics. Warnings have been expressed (Brocklehurst, 1999) that trials of treatment to prevent premature delivery need to demonstrate a benefit for the baby; until this has been confirmed, such treatment should be seen as experimental. We have reviewed the records of 413 women with BV attending the Department of Sexual Health at Whipps Cross Hospital from 1994 (D.Viniker, K.Ihuoma and R.Melville, unpublished results). In total, 382 women were symptomatic Sub-clinical bacteria of the endometrium and were treated with metronidazole alone or in combination with tetracycline or with clindamycin; 31 were asymptomatic and not treated. Twenty-four of the women have attended our obstetric department. Twelve had treatment for BV before or during pregnancy and these all had uneventful outcomes, and 12 were found to have BV after pregnancy. One woman had a spontaneous abortion at 14 weeks gestation and two women had premature deliveries; nine had favourable outcomes. We believe that there is a tendency to cross district borders to attend departments of genitourinary medicine, so that our dataset is unlikely to be complete. The amniotic cavity should be sterile. Evidence linking intrauterine infection with long-term morbidity, including cerebral palsy, has been reviewed (Romero, 1998). Microorganisms were found in the fluid retrieved by amniocentesis in 25% of all premature births and in 10% of cases of premature labour with intact membranes, and there was also evidence of a role for proinflammatory cytokines in the onset of human labour. The frequency of sub-clinical intra-amniotic infection and cervical bacterial colonization in relation to preterm labour has been evaluated (Cararach et al., 1998). These authors looked at two cohorts—59 women with preterm labour and 30 controls who had amniocentesis to assess fetal lung maturity. There was a statistically increased incidence of Gram-negative vaginal bacteria in patients with premature labour, but no increase in intra-amniotic infection. A group of six women with uncontrollable preterm labour and intact membranes resulting in fetal loss between 20 and 28 weeks gestation were also investigated (Naessens et al., 1987). Intra-amniotic U.urealyticum was detected in only two women, the organism being isolated from the cervix in five cases and from the placenta in four. It would seem, therefore, that if U.urealyticum has a part in premature labour, it would be active extraamniotically rather than intra-amniotically. It has been suggested (Corfield et al., 1998) that the susceptibility of the cervical mucus to bacterial degradation may be a factor in premature delivery, and evidence was found of increased enzyme activity in patients with BV. The identification of genitourinary infections, most notably BV, can consistently reduce premature delivery and premature rupture of the membranes (McGregor and French, 1998). The recommended prophylactic treatment has been oral clindamycin or metronidazole. A relationship has been observed (Chen et al., 1996) between miscarriage and premature delivery; this would be predictable if BV is a cause for both such complications of pregnancy. Pre-eclampsia and bacteria endometrialis Pre-eclampsia is a complication of pregnancy characterized by hypertension, proteinuria and fluid retention. It may develop at an alarming speed with attendant risks of maternal mortality 379 and morbidity (Hunter et al., 1998). The causation of pre-eclampsia remains a mystery, and it seems probable that more than one factor is responsible. A summary of facts associated with pre-eclampsia and eclampsia that must be met by a hypothesis are summarized in Table III. With reference to Table III, if micro-organisms within the uterine cavity play a part in the aetiology of pre-eclampsia then it is possible that a state of immunization occurs to the dominant organism. This could explain the association with first pregnancy [Table III(A)]. Micro-organisms may be carried into the female genital tract attached to spermatozoa (Nunez-Calonge et al., 1998). A different strain of organism could be acquired from a new partner, resulting in a subsequent increased risk of pre-eclampsia [Table III(B)]. Table III. Facts associated with pre-eclampsia A. Occurs preferentially during first pregnancies B. Change of partner alters the incidence for later pregnancies C. Placental structural changes by early second trimester D. Increased incidence with hydatidiform mole E. Clinical presentation in late second trimester or in the third trimester F. Placental tissue required G. Extrauterine pathology including renal, hepatic and coagulation H. Increased incidence in multiple pregnancy and diabetes I. Familial tendency J. Ethnic variation It has been suggested (Toth et al., 1986; McGregor, 1998) that there might be endometrial infection at the time of implantation. The presence of micro-organisms within the uterine cavity could explain abnormal placentation [Table III(C)]. The uterus is full of trophoblastic tissue when there is a hydatidiform mole [Table III(D)]. There may be a complex relationship between the competence of the cervix and ascending genital tract infection (Jones et al., 1998). In early pregnancy the cervix is closed, providing a degree of protection against micro-organisms ascending. As term approaches, the cervix shortens and opens so that this mechanical protection is lost and more organisms could easily gain entrance into the uterine cavity [Table III(E)]. The micro-organisms would be removed when the uterus is completely emptied of placenta and membranes following delivery [Table III(F)]. In multiple pregnancy, the membranes and placental volume are larger and the cervix may shorten and start to dilate earlier [Table III(H)]. The placenta tends to be larger in diabetic pregnancy [Table III(H)]. BV can occur in virgins, and the organisms are known to be passed from mother to neonate, possibly contributing to familial tendency [Table III(I)]. Finally, women of black race have an increased incidence of BV during pregnancy (Goldenberg et al., 1996) and also pre-eclampsia (Knuist et al., 1998) [Table III (J)]. 380 D.A.Viniker Current theories on the pathogenesis of pre-eclampsia relate to changes in the renin–angiotensin–aldosterone system (Carr and Gant, 1983; Morgan et al., 1997), immunological susceptibility (Jenkins et al., 1978; Redman et al., 1978), disseminated intravascular coagulopathy (Roberts et al., 1989), prostacyclin and thromboxane (McParland et al., 1990) and maternal endothelial cell damage (Roberts et al., 1989). Each of these theories, however, relates to susceptibility or perpetuation of the syndrome rather than the triggering mechanism. According to one theory (Roberts et al., 1989), poorly perfused placental tissue releases one or more factors into the systemic circulation. This results in endothelial cell damage and a subsequent cascade of disseminated intravascular coagulation, vasoconstriction and body fluid redistribution. Endothelial cell dysfunction, which is a characteristic of the maternal syndrome of pre-eclampsia, might be caused by a factor released from the placenta into the maternal circulation. A fluoroimmunoassay has been developed (Knight et al., 1998) to measure syncytiotrophoblast microvilli (STBM) in the blood of pregnant women. These authors found that STBM are released into the circulation in greater amounts in pre-eclampsia than in normal controls. The concentrations of STBM in plasma correlated with endothelial anti-proliferative activity, indicating that STBM in the circulation may be involved in the endothelial dysfunction seen in pre-eclampsia. Monocytes were shown to phagocytose STBM, and this could possibly be linked to the release of cytokines and other toxins, which would enhance endothelial damage. There seems to be an already well-developed inflammatory response in normal pregnancy, so that the differences between normal pregnancy and pre-eclampsia are less apparent than those between normal pregnancy and the non-pregnant state. From these observations, it was concluded (Redman et al., 1998) that pre-eclampsia arises from a universal maternal intravascular inflammatory response to pregnancy, and there is then an additional stimulus or the maternal response is unduly strong. The current view is that endothelial cell dysfunction is a common end-point for a variety of causes of pre-eclampsia. Pre-eclampsia is associated with dyslipidaemia and overproduction of free radicals, which is reminiscent of atherosclerosis (Hubel, 1998). There is a change in placental morphology by 20 weeks gestation in patients likely to develop pre-eclampsia. There are restricted endovascular trophoblast invasion and atherotic lesions with local trophoblast destruction. An imbalance of maternal inflammatory cells and cytokine production may lead to the release of toxic factors and cytokines from the placenta, resulting in the clinical syndrome of pre-eclampsia (Pijnenborg, 1998). Early identification of a change in placental blood flow might provide a useful screening test for the subsequent development of pre-eclampsia (Campbell et al., 1986). Many screening tests have been developed, but lack of uniformity in the definition of pre-eclampsia confounds the analyses (Poulton et al., 1998). Hyperuricaemia and thrombocytopenia, as well as liver enzymes, provide a measure of some clinical value (Shennan, 1998), and ultrasound Doppler screening looks reasonably promising (Campbell, 1998). Vascular endothelial growth factor (VEGF), a protein released by damaged endothelium, is elevated in women with pre-eclampsia. Preliminary data have shown elevated levels of VEGF as early as 12 weeks gestation in women destined to develop pre-eclampsia, and there is a more than 3-fold increase by 20 weeks. This would suggest that VEGF may prove to be a useful screening test for pre-eclampsia (Hunter et al., 1998). It has been predicted (Redman et al., 1998) that there is no single gene disorder, single specific predictive test, nor a single preventive effective measure, though optimism has been expressed that the future of pregnancy screening will focus on ultrasound (Campbell, 1998); pregnant women will have an early dating scan, early mid-trimester anomaly scan, and Doppler ultrasound at 24 weeks to predict high risk for pre-eclampsia. There would appear to be a familial predisposition to pre-eclampsia and eclampsia, which is presumed to be genetic (Hubel, 1998). It is recognized that pre-eclampsia is a complex disorder reflecting the interaction of genes and ‘environmental’ factors (O’Shaughnessy, 1998). If there is an association between bacterial vaginosis and premature labour, as well as with pre-eclampsia, then there should be an increased incidence of premature labour with pre-eclampsia. This point has been confirmed (Chen et al., 1996), and it was also found that preterm births were almost twice as likely to occur in women with pregnancy-induced hypertension and more than four times more likely to occur in women with pregnancy-aggravated hypertension compared with normotensive women (Samadi and Mayberry, 1998). Possible modes of bacterial action in pre-eclampsia Chorioamnionitis is a term applied loosely to leukocyte infiltration of the placenta and extraplacental membranes. A high incidence of chorioamnionitis is found in cases of pre-eclamptic toxaemia, but this has been attributed to prolonged rupture of the membranes in these cases (Fox, 1973b). The most compelling argument to refute the hypothesis of BV-type organisms being a factor in pre-eclampsia is the poor correlation with chorioamnionitis. If intrauterine infection with BV-type organisms is the elusive ‘Scarlet Pimpernel’, then it could have been masked by the fact that evidence of an inflammatory response has not been forthcoming. To reiterate, BV is an ‘-osis’ rather than an ‘-itis’. How could bacteria endometrialis be responsible for the typical clinical features of pre-eclampsia and eclampsia? Two contending mechanisms would be exaggerated host response and bacterial toxins. Reproductive function, cytokines and exaggerated host response: Cytokines are potent polypeptides released from in- Sub-clinical bacteria of the endometrium flammatory cells as part of the host response. They have multiple paracrine and endocrine functions, and there are complex mechanisms for control and cessation of their actions (Galley and Webster, 1996). Cytokines have a central, but potentially, paradoxical role in the immunological and inflammatory processes. They are part of the local defence mechanism against infectious disease, but they are also implicated as mediators of the pathology of infectious disease (Henderson et al., 1996). The body contains an enormous number of microorganisms, which constitute the normal microflora; indeed, it has been estimated that there is a greater number of bacteria in the body than cells. Bacteria and viruses may produce molecules that induce inflammatory and anti-inflammatory reactions in the host. Endometrial receptors for cytokines have been demonstrated (Tabibzadeh, 1991). There is increasing evidence that cytokines, in addition to their role in the immune system, may have a significant role in the modulation of the neuroendocrine control of reproduction (Rutanen, 1993; Mathialagen and Roberts, 1994). Cytokines probably play a physiological part as local mediators of the actions of sex hormones on the endometrium (Tabibzadeh and Sun, 1992; Sharkey, 1998). Leukocyte numbers and cytokine expression are altered in some women with implantation defects (Stewart-Akers et al., 1998). Cytokines released by activated endometrial lymphocytes and macrophages can have a detrimental effect on embryo implantation and embryo development (Shaarawy and Nagui, 1997). A variety of conditions, not initially related to infection, can produce a state of shock that is indistinguishable from aspects of septic shock. It is, therefore, apparent that it is the host response to the severe insult, which is responsible rather than the initiating stimulus. This host response has become known as the systemic inflammatory response (SIRS) (Beal and Cerra, 1994). The immuno-inflammatory cascade can be initiated by toxins from micro-organisms. The main mediators of SIRS are the cytokines, particularly the interleukins (IL)-1 and IL-6 and tumour necrosis factor-α (TNFα). During the early stages of infection, the cytokines play an essential homeostatic role and their activities are mainly localized, with only minimal systemic effects. The inter-relationships between cytokines are complex. They may act synergistically so that their combined effects may be different from their individual actions (Offenstadt et al., 1993). Occasionally, there may be failure of cytokine control, leading to SIRS; when cytokine control has been lost, SIRS becomes self-sustaining with multiple organ dysfunction (Bone, 1991) [Table III (G)]. Toxins: The once popular concept of circulating toxin as the cause of hypertension and proteinuria in pregnancy fell into disrepute, apparently for lack of supportive evidence (Carey, 1974). The vaginal fluid of women with bacterial vaginosis contains significantly greater amounts of endotoxin than in controls (Sjoberg and Hakansson, 1991; Mattsby-Baltzer et al., 1998). BV-associated endotoxins are capable of inducing cyto- 381 kine production in pregnancy (Mattsby-Baltzer et al., 1998). An animal model for pre-eclampsia has been described (Faas et al., 1994) which involves ultra-low-dose endotoxin infusion in conscious, pregnant rats. The endotoxin, or saline for the controls, was infused for 1 h on the 14th day of gestation, and blood pressure, albuminuria and platelet counts were measured. A significant increase was found in blood pressure and albuminuria in the index group, while platelet coagulopathy and glomerular fibrinogen deposits were observed only in the endotoxin-treated group. The authors concluded that their model resembled the clinical and histopathological events of pre-eclampsia sufficiently to enable further investigation into the pathophysiological mechanisms of the condition. Subsequent studies have confirmed that pregnancy greatly influences the nature and strength of the inflammatory response following endotoxin administration (Faas et al., 1995). The results of infusing ultra-low-dose endotoxin into non-pregnant rats was also described (Faas, 1998). There was no effect with the first infusion, but if it was repeated 18–24 h later, a state typical of pre-eclampsia developed. When the endotoxin was administered to pregnant rats on just one occasion, the preeclampsia changes were seen. It was also suggested (Henderson et al., 1996) that understanding the control mechanisms of cytokine release induced by micro-organisms will advance anti-inflammatory therapy. If bacteria within the uterus are responsible for pre-eclampsia, there could be a role for antibiotics in the prophylaxis and treatment of this potentially lethal syndrome. A search of the literature using Medline has revealed no reference to the use of metronidazole or erythromycin in the prevention or management of pre-eclampsia. Conclusions In recent years, an unexpected infectious aetiology has been found for a variety of clinical conditions. In gynaecology, the role of the human papillomavirus in pre-malignant and malignant cervical disease has been confirmed and in obstetrics, we are learning of the relationship between BV and premature delivery, while Helicobacter pylori has become established as the cause of peptic ulceration. There is also a body of evidence implicating microbes in the pathogenesis of coronary heart disease, and although the bacteria Chlamydia pneumoniae (Jackson et al., 1997; Maass et al., 1998) and H.pylori appear to be the leading contenders, a number of viruses have also been suggested (Ellis, 1997). If bacteria in the endometrial cavity should be implicated as a significant, previously unrecognized, cause of obstetric and gynaecological disease, the place of appropriate antibiotic regimens would need careful consideration. Clinical trials involving pregnant women are not common. It is never easy to explain the concept of risk to patients, and this is compounded by the presence of an unborn child (Mohanna, 1998). The agents of choice for bacterial vaginosis are metronidazole for 382 D.A.Viniker the anaerobes, and erythromycin for Mycoplasma. Erythromycin is also effective against C.trachomatis. Pharmaceutical companies advise that metronidazole should only be given with caution during pregnancy, although several groups have reassuringly demonstrated no evidence of teratogenicity (Morgan, 1978; Roe, 1983; Piper et al., 1993; Burtin et al., 1995; Czeizel and Rockenbauer, 1998). Nevertheless, a policy of caution—particularly in the first trimester—and avoidance of high-dose regimens in pregnancy would seem appropriate. How can the clinical significance of BE be evaluated? We need to consider the classical principles of Koch, to review retrospectively appropriate databases, and to contemplate serological studies and perhaps trials of therapy. According to Koch’s principles, the micro-organism must: (i) be found in all cases of the disease; (ii) be isolated from the host and grown in pure culture; (iii) reproduce the original disease when introduced into a susceptible host; and (iv) be found in the experimental host so infected. The variety of organisms possibly implicated in ‘bacteria endometrialis’, the number of clinical conditions that could result, and their nature, are such that it is difficult to foresee how these principles could be tested. There is evidence that micro-organisms often contaminate the uterine cavity of some healthy women, although they are found more frequently in association with pathological processes. Bacteriological diagnosis of BV is not assisted by culture, and it is possible that this would also pertain for BE. Facilities to culture Mycoplasma are limited to a few centres. Developments in serological tests are enhancing our capability to detect evidence of infections in gynaecology. C.trachomatis heat shock proteins have been related to infertility (Witkin et al., 1998) and ectopic pregnancy (Sziller et al., 1998). Serological tests for Mycoplasma pneumoniae are used in current clinical practice, but not for M.hominis or U.urealyticum. Some research workers have started to evaluate new serological tests for U.urealyticum (Quinn et al., 1993), M.hominis (Shimada et al., 1998) and Mobiluncus (Schwebke et al., 1996). Finally, there is the option for a trial of therapy for patients with a presumptive diagnosis without laboratory confirmation. The antibiotics of choice—metronidazole and the macrolides such as erythromycin—are relatively innocuous. Nevertheless, antibiotics should be used with caution, as there are risks of bacterial resistance. In a study of induced abortion, women allocated to receive antibiotics had lower infection rates than those allocated to treatment only if screening was positive (Penney et al., 1998). With approval by our Ethics Advisory Committee, we have commenced a controlled study of erythromycin and metronidazole for all couples presenting with infertility and/or recurrent abortion. The hypothesis that one common factor, bacteria endometrialis, might provide a plausible explanation for a variety of obstetric and gynaecological enigmas is particularly intriguing. There is sufficient evidence to warrant further investigation. “We seek him here, we seek him there, Those Frenches seek him everywhere. Is he in heaven? – Is he in hell? That demmed, elusive Pimpernel”. The Scarlet Pimpernel, Baroness Orczy (1905) Acknowledgements Several colleagues have kindly provided invaluable advice. For their kindness and wise counsel, the author wishes to thank Mr R.Baldwin, Mr P.Breen (Latin teacher), Dr B.Chattopadhyay, Mrs A.Clemenson, Dr C.Jain, Mr R.Lamont, Professor J.MacVicar, Dr R.Melville, Mr E.E.Philipp, Professor L.Poston, Professor C.Rodeck, Professor S.K.Smith, Professor C.W.G.Redman, Mr G.Sadler, Professor P.Steer, Dr K.Thomas and Professor H.Youssef. References Amsel, R., Totten, P.A., Spiegel, C.A. et al. (1983) Diagnostic criteria and microbial and epidemiologic associations. Am. J. Med., 74, 14–22. Auroux, M.R., De Mouy, D.M. and Acar, J.F. (1987) Male fertility and positive Chlamydia serology. A study of 61 fertile and 82 subfertile men. J. Androl., 8, 197–200. Beal, A.L. and Cerra, F.B. (1994) Multiple organ failure syndrome in the 1990s. Systemic inflammatory response and organ dysfunction. J. Am. Med. Assoc., 271, 226–233. Berclaz, G., Hanggi, W., Birkhauser, M. et al. (1993) Chlamydia and Mycoplasma infections in men with involuntary sterility. Geburtshilfe. Frauenheilkd., 53, 539–542. Bercovici, B., Haas, H., Sacks, T. et al. (1978) Isolation of mycoplasmas from the genital tract of women with reproductive failure, sterility or vaginitis. Isr. J. Med. Sci., 14, 347–352. Bjercke, S. and Purvis, K. (1992) Chlamydial serology in the investigation of infertility. Hum. Reprod., 7, 621–624. Bone, R.C. (1991) The pathogenesis of sepsis. Ann. Intern. Med., 115, 457–469. Branch, D.W., Andres, R., Digre, K.B. et al. (1989) The association of antiphospholipid antibodies with severe pre-eclampsia. Obstet. Gynecol., 73, 541–545. Brocklehurst, P. (1999) Infection and preterm delivery. Br. Med. J., 318, 548–549. Bump, R.C. and Buesching W.J. (1988) Bacterial vaginosis in virginal and sexually active adolescent females: evidence against exclusive sexual transmission. Am. J. Obstet. Gynecol., 158, 935–939. Burtin, P., Taddio, A., Ariburnu, O. et al. (1995) Safety of metronidazole in pregnancy: a meta-analysis. Am. J. Obstet. Gynecol., 172, 525–529. Buvat, J., Buvat-Herbaut, M., Fourlinnie, J.C. et al. (1984) Results of treating infections of the cervical mucosa with doxycycline polyphosphate in 53 infertile women. J. Gynecol. Obstet. Biol. Reprod. (Paris), 13, 81–85. Campbell, S. (1998) Screening for pre-eclampsia I: The uterine notch. Paper presented at ‘The Problem with Prematurity II’, St Thomas’ Hospital, London, 7–9th September, 1998, The Parthenon Publishing Group. Campbell, S., Pearce, J.M., Haeckett, G. et al. (1986) Qualitative assessment of uteroplacental blood flow: early screening test for high-risk pregnancies. Obstet. Gynecol., 68, 649–653. Cararach, V., Gomez, O., Antolin, E. et al. (1998) Subclinical intraamniotic infection and cervical bacterial colonization related to preterm labor (Abstract). Prenat. Neonat. Med., 3 (Suppl. 2), 27. Carey, H.M. (1974) Toxaemias of pregnancy. In Donald, I. (ed.), Practical Obstetric Problems. 4th edn. Lloyd-Luke, pp. 239–271. Sub-clinical bacteria of the endometrium Carp, H.J.A. (1997) Investigation and treatment for recurrent pregnancy loss. In Rainsbury, P.A. and Viniker, D.A. (eds), Practical Guide to Reproductive Medicine. The Parthenon Publishing Group, New York, London, pp. 337–362. Carr, B.R. and Gant, N.F. (1983) The endocrinology of pregnancy-induced hypertension. Clin. Perinatol., 10, 737–761. Catteau, B., Delaporte, E., Hachulla, E. et al. (1995) Mycoplasma infection with Stevens-Johnson syndrome and antiphospholipid antibodies. Rev. Med. Interne., 16, 10–14. Chen, C.P., Wang, K.G., Yang, Y.C. et al. (1996) Risk factors for preterm birth in an upper middle class Chinese population. Eur. J. Obstet. Gynecol. Reprod. Biol., 70, 53–59. Christiansen, G., Jensen, L.T., Boesen, T. et al. (1997) Molecular biology of Mycoplasma. Wien. Klin. Wochenschr., 109, 557–561. Corfield, A.P., Wiggins, R., Kane, L. et al. (1998) Increased sialidase activity in women with bacterial vaginosis. Prenat. Neonat. Med., 3 (Suppl. 2), 20. Czeizel, A.E. and Rockenbauer, M. (1998) A population based case-control teratogenic study of oral metronidazole treatment during pregnancy. Br. J. Obstet. Gynaecol., 105, 322–327. Czernobilsky B. (1978) Endometritis and infertility. Fertil. Steril., 30, 119–130. Depuydt, C.E., Bosmans, E., Zalata, A. et al. (1996) The relation between reactive oxygen species and cytokines in andrological patients with or without male accessory gland infection. J. Androl., 17, 699–707. Dinsmoor, M.J., Ramamurthy, R.S. and Gibbs, R.S. (1989) Transmission of genital mycoplasmas from mother to neonate in women with prolonged membrane rupture. Pediatr. Infect. Dis. J., 8, 483–487. Di Renzo, G.C., Moscioni, P., Perazzi, A. et al. (1998) Risk factors of preterm delivery. Prenat. Neonat. Med., 3 (Suppl. 2), 12. Doderlein, A. (1892) Das Scheidensekret und Seine Bedeutung fur das Puerperalfieber. Besold: Leipzig. Cited in O’Dowd, M.J. and Philipp, E.E. (eds), The History of Obstetrics and Gynaecology. Parthenon Publishers, New York, London, 1994, p. 240. Donders, G.G.G. (1998) The flora of the female vagina: a review. Prenat. Neonat. Med., 3 (Suppl. 2), 13. Eggert-Kruse, W., Rohr, G., Strock, W. et al. (1995) Anaerobes in ejaculates of subfertile men. Hum. Reprod. Update, 1, 462–478. Ellis, R.W. (1997) Infection and coronary heart disease. J. Med. Microbiol., 46, 535–539. Emembolu, J.O. (1989) Endometrial flora of infertile women in Zaria, northern Nigeria. Int. J. Gynaecol. Obstet., 30, 155–159. Evers, J.L.H. (1994) Endometriosis does not exist; all women have endometriosis. Hum. Reprod., 9, 2206–2208. Faas, M.M. (1998) Ultra-low dose endotoxin infusion in pregnant and non-pregnant rats. Paper presented at ‘The Problem with Prematurity II’, St Thomas’ Hospital, London, 7–9th September, 1998, The Panthenon Publishing Group. Faas, M.M., Schuiling, G.A., Baller, J.F. et al. (1994) A new animal model for human preeclampsia: ultra low-dose endotoxin in pregnant rats. Am. J. Obstet. Gynecol., 171, 158–164. Faas, M.M., Schuiling, G.A., Baller, J.F. et al. (1995) Glomerular inflammation in pregnant rats after infusion of low dose endotoxin. An immunohistological study in experimental pre-eclampsia. Am. J. Pathol., 147, 1510–1518. Fox, H. (1973a) The normal and abnormal endometrium. In Fox, H. and Langley, F.A. (eds), Postgraduate Obstetrical and Gynecological Pathology. Pergamon Press, Oxford and New York, pp. 115–146. Fox, H. (1973b) The pathology of the placenta. In Fox, H. and Langley, F.A. (eds), Postgraduate Obstetrical and Gynecological Pathology. Pergamon Press, Oxford and New York, pp. 409–439. Galley, H.F. and Webster, N.R. (1996) The immuno-inflammatory cascade. Br. J. Anaesth., 77, 11–16. Gannon, H. (1993) Ureaplasma urealyticum and its role in neonatal lung disease. Neonatal Netw., 12, 13–18. Gardner, J.J. and Dukes, C.D. (1955) Haemophilus vaginalis vaginitis: a newly defined specific infection previously classified ‘non-specific’ vaginitis. Am. J. Obstet. Gynecol., 69, 962–976. Goldenberg, R.L., Klebanoff, M.A., Nugent, R. et al. (1996) Bacterial colonization of the vagina during pregnancy in four ethnic groups. Am. J. Obstet. Gynecol., 174, 1618–1621. 383 Gopalkrishnan, K., Hinduju, I.N. and Kumar, T.C. (1990). Semen characteristics of asymptomatic males affected by Trichomonas vaginalis. J. In Vitro Fertil. Embryo Transfer, 7, 165–167. Grattard, F., Soleihac, B., De Barbeyrac, B. et al. (1995) Epidemiologic and molecular investigations of genital mycoplasmas from women and neonates at delivery. Pediatr. Infect. Dis. J., 14, 853–858. Greendale, G.A., Haas, S.T., Holbrook, K. et al. (1993) The relationship of Chlamydia trachomatis infection and male infertility. Am. J. Public Health, 83, 996–1001. Hauth, J.C., Goldenberg, R.L., Andrews, W.W. et al. (1995) Reduced incidence of preterm delivery with metronidazole and erythromycin in women with bacterial vaginosis. N. Engl. J. Med., 333, 1732–1736. Hay, P. (1994) What is bacterial vaginosis? J. Sex. Health, 4, 6–7. Henderson, B., Poole, S. and Wilson, M. (1996) Microbial/host interactions in health and disease: who controls the cytokine network? Immunopharmacology, 35, 1–21. Hill, G.B. (1993) The microbiology of bacterial vaginosis. Am. J. Obstet. Gynecol., 169, 450–454. Hinton, R.A., Edgell, L.M., Andrews, B.E. et al. (1979) A double-blind cross-over study of the effect of doxycycline on mycoplasma infection and infertility. Br. J. Obstet. Gynaecol., 86, 379–383. Holst, E. (1990) Reservoir of four organisms associated with bacterial vaginosis suggests lack of sexual transmission. J. Clin. Microbiol., 28, 2035–2039. Hubel, C.A. (1998) Endothelium, lipids and oxidative stress in pre-eclampsia (Abstract). Prenat. Neonat. Med., 3 (suppl. 2), 16. Hunter, A.J., Aitkenhead, M., McCracken, D.J. et al. (1998) Vascular endothelial growth factor: a potential predictive marker for pre-eclampsia (Abstract) . Prenat. Neonat. Med., 3 (Suppl. 2), 28. Idriss, W.M., Patton, W.C. and Taymor, M.L. (1978) On the etiologic role of Ureaplasma urealyticum (T-mycoplasma) infection in infertility. Fertil. Steril., 30, 293–296. Jackson, L.A., Campbell, L.A., Kuo, C.C. et al. (1997) Isolation of Chlamydia pneumoniae from a carotid endarterectomy specimen. J. Infect. Dis., 176, 292–295. Jenkins, D.M., Need, J.A., Scott, J.S. et al. (1978) Human leucocyte antigens and mixed lymphocyte reaction in severe pre-eclampsia. Br. Med. J., 1, 542–544. Jones, G., Clark, T. and Bewley, S. (1998) The weak cervix: failing to keep the baby in or infection out? Br. J. Obstet. Gynaecol., 105, 1214–1215. Jones, H.W. and Jones, G.S. (1981) Infertility, recurrent and spontaneous abortion. In Novak’s Textbook of Gynecology. 10th edn. Williams & Wilkins, Baltimore, London, pp. 694–732. Khatamee, M.A. and Sommers, S.C. (1989) Clinicopathologic diagnosis of mycoplasma endometritis. Int. J. Fertil., 34, 52–55. Knight, M., Sargent, I. and Redman, C.W.G. (1998) The role of trophoblast microvilli in the maternal syndrome of pre-eclampsia (Abstract). Prenat. Neonat. Med., 3 (Suppl. 2), 17. Knuist, M., Bonsel, G.J., Zondervan, H.A. et al. (1998) Risk factors for preeclampsia in nulliparous women in distinct ethnic groups. A prospective cohort study. Obstet. Gynecol., 92, 174–178. Koren, Z. and Spigland, I. (1978) Irrigation technique for detection of Mycoplasma intrauterine infections in infertile patients. Obstet. Gynecol., 52, 588–590. Korn, A.P., Bolan, G., Padian, N. et al. (1995) Plasma cell endometritis in women with symptomatic bacterial vaginosis. Obstet. Gynecol., 85, 387–390. Kundsin, R.B., Falk, L., Hertig, A.T. et al. (1986) Mycoplasma, Chlamydia, Epstein–Barr, Herpes I and II, and Aids infections among 100 consecutive infertile female patients and husbands: diagnosis, treatment and results. Int. J. Fertil., 31, 356–359. Lakasing, L. and Poston, L. (1997) Adverse pregnancy outcome in the antiphospholipid syndrome: focus for future research. Lupus, 6, 681–684. Lamont, R.F., Petrou, S., Morgan, D.J. et al. (1998) BV and perinatal morbidity: a cost benefit analysis (Abstract). Prenat. Neonat. Med., 3 (Suppl. 2), 14. Larsson, P.G., Bergman, B., Forsum, U. et al. (1990) Treatment of bacterial vaginosis in women with vaginal bleeding complications or discharge and harboring Mobiluncus. Gynecol. Obstet. Invest., 29, 296–300. 384 D.A.Viniker Levett, P.N. (1989) Bacterial vaginosis. West Indian Med. J., 38, 126–132. Llahi-Camp, J.M., Rai, R., Ison, C. et al. (1996) Association of bacterial vaginosis with a history of second trimester miscarriage. Hum. Reprod., 11, 1575–1578. Lucisano, A., Morandotti, G., Marana, R. et al. (1992) Chlamydial genital infections and laparoscopic findings in infertile women. Eur. J. Epidemiol., 8, 645–649. Maass, M., Bartels, C., Engel, P.M. et al. (1998) Endovascular presence of viable Chlamydia pneumoniae is a common phenomenon in coronary artery disease. J. Am. Coll. Cardiol., 31, 827–832. MacDermott, R.I.J. (1995) Bacterial vaginosis. Br. J. Obstet. Gynaecol., 102, 92–94. Majeroni, B.A. (1998) Bacterial vaginosis: an update. Am. Fam. Physician, 57, 1285–1289. Martin-Du Pan, R.C. and Sakkas, D. (1998) Are anti-oxidants useful in the treatment of male infertility? Hum. Reprod., 13, 2984–2985. Mathialagan, N. and Roberts, R.M. (1994) A role for cytokines in early pregnancy. Indian J. Physiol. Pharmacol., 38, 153–162. Mattsby-Baltzer, I., Platz-Christensen, J.J., Hosseini, N. et al. (1998) IL-1beta, IL-6, TNFalpha, fetal fibronectin, and endotoxin in the lower genital tract of pregnant women with bacterial vaginosis. Acta Obstet. Gynecol. Scand., 77, 701–706. McGregor, J.A. Evidence based prevention of preterm birth/PROM: infection and inflammation. Paper presented at ‘The Problem with Prematurity II’, St Thomas’ Hospital, London, 7–9th September, 1998, Parthenon Publishing Group. McGregor, J.A. and French, J.I. (1998) Evidence based prevention of preterm birth/prom: infection and inflammation. (Abstract). Prenat. Neonat. Med., 3 (Suppl. 2), 14. McGregor, J.A., French, J.I., Jones, W. et al. (1994) Bacterial vaginosis is associated with prematurity and vaginal fluid mucinase and sialidase: results of a controlled trial of topical clindamycin cream. Am. J. Obstet. Gynecol., 170, 1048–1059. McParland, P., Pearce, J.M. and Chamberlain, G.V. (1990) Doppler ultrasound and aspirin in recognition and prevention of pregnancy-induced hypertension. Lancet, 335, 1552–1555. Mohanna, K. (1998) Giving and withholding consent to participate in clinical trials: decisions of pregnant women (Abstract). Prenat. Neonat. Med., 3 (Suppl. 2), 22. Moller, B.R., Kristiansen, F.V., Thorsen, P. et al. (1995) Sterility of the uterine cavity. Acta Obstet. Gynecol. Scand., 74, 216–219. Morgan, I. (1978) Metronidazole treatment in pregnancy. Int. J. Gynaecol. Obstet., 15, 501–502. Morgan, L., Crawshaw, S., Baker, P.N. et al. (1997) Functional and genetic studies of the angiotensin II type 1 receptor in pre-eclamptic and normotensive pregnant women. J. Hypertens., 15, 1389–1396. Naessens, A., Foulon, W., Cammu, H. et al. (1987) Epidemiology and pathogenesis of Ureaplasma urealyticum in spontaneous abortion and early preterm labor. Acta Obstet. Gynecol. Scand., 66, 513–516. Newman, R.B. (1998) The shortened cervix and premature birth. Abstract. Prenat. Neonat. Med., 3 (Suppl. 2), 14. Nunez-Calonge, R., Caballero, P., Redondo, C. et al. (1998) Ureaplasma urealyticum reduces motility and induces membrane alterations in human spermatozoa. Hum. Reprod., 13, 2756–2761. Offenstadt, G., Guidet, B. and Staikowsky, F. (1993) Cytokines and severe infections. Pathol. Biol. (Paris), 41, 820–831. O’Shaughnessy, K.M. (1998) The genetic basis of pre-eclampsia. Abstract. Prenat. Neonat. Med., 3 (Suppl. 2), 17. Pattison, N.S., Chamley, L.W., McKay, E.J. et al. (1993) Antiphospholipid antibodies in pregnancy: prevalence and clinical associations. Br. J. Obstet. Gynaecol., 100, 909–913. Penney, G.C., Thompson, M., Norman, J. et al. (1998) A randomised comparison of strategies for reducing infective complications of induced abortion. Br. J. Obstet. Gynaecol., 105, 599–604. Pijnenborg, R. (1998) The placenta and placental bed in preeclampsia. Prenat. Neonat. Med., 3 (Suppl. 2), 16. Piper, J.M., Mitchel, E.F. and Ray, W.A. (1993) Prenatal use of metronidazole and birth defects: no association. Obstet. Gynecol., 82, 348–352. Poulton, L.P.B., Chappel, L. and Sheenan, A.H. (1998) Criteria used to define pre-eclampsia. Prenat. Neonat. Med., 3 (Suppl. 2), 27. Profet, M. (1993) Menstruation as a defence against pathogens transported by sperm. Q. Rev. Biol., 68, 335–386. Quinn, P.A., Shewchuk, A.B. and Shuber, J. (1983) Efficacy of antibiotic therapy in preventing spontaneous pregnancy loss among couples colonized with genital mycoplasmas. Am. J. Obstet. Gynecol., 145, 239–244. Quinn, P.A., Li, H.C., Th’ng, C. et al. (1993) Serological response to Ureaplasma urealyticum in the neonate. Clin. Infect. Dis., 17 (Suppl. 1), S136–S143. Rai, R. and Regan, L. (1998) Antiphospholipid syndrome and pregnancy loss. Hosp. Med., 59, 637–639. Rai, R.S., Regan, L., Clifford, K. et al. (1995) Antiphospholipid antibodies and beta 2-glycoprotein-I in 500 women with recurrent miscarriage: results of a comprehensive screening approach. Hum. Reprod., 10, 2001–2005. Redman, C.W.G., Bodmer, J.G., Bodmer, W.F. et al. (1978) HLA antigens in severe pre-eclampsia. Lancet, ii, 397–399. Redman, C.W.G., Sacks, G.P. and Sargent, I.L. (1998) An inflammatory view of pre-eclampsia. Prenat. Neonat. Med., 3 (Suppl. 2), 17. Roberts, J.M., Taylor, R.N., Musci, T.J. et al. (1989) Preeclampsia: an endothelial cell disorder. Am. J. Obstet. Gynecol., 161, 1200–1204. Roe, F.J. (1983) Toxicological evaluation of metronidazole with particular reference to carcinogenic, mutagenic and teratogenic potential. Surgery, 93, 158–164. Romero, R. (1998) Intrauterine infection and preterm labor. Prenat. Neonat. Med., 3 (Suppl. 2), 13. Rutanen, E.M. (1993) Cytokines in reproduction. Ann. Med., 25, 343–347. Salat-Baroux, J. (1988) Recurrent spontaneous abortions. Reprod. Nutr. Dev., 28, 1555–1568. Samadi, A.R. and Mayberry, R.M. (1998) Maternal hypertension and spontaneous preterm birth among black women. Obstet. Gynecol., 91, 899–904. Schwebke, J.R., Morgan, S.C. and Hillier, S.L. (1996) Humoral antibody to Mobiluncus curtsii, a potential serological marker for bacterial vaginosis. Clin. Diagn. Lab. Immunol., 3, 567–569. Shaarawy, M. and Nagui, A.R. (1997) Enhanced expression of cytokines may play a fundamental role in the mechanisms of immunologically mediated recurrent spontaneous abortion. Acta Obstet. Gynecol. Scand., 76, 205–211. Sharkey, A. (1998) Cytokines and implantation. Rev. Reprod., 3, 52–61. Shennan, A. (1998) Screening for pre-eclampsia II: blood pressure and biochemical markers. Abstract. Prenat. Neonat. Med., 3 (Suppl. 2), 17. Shimada, M., Kotani, T., Sameshima, H. et al. (1998) Two patients with premature labour associated with Mycoplasma hominis. J. Med. Microbiol., 47, 179–182. Sjoberg, I. and Hakansson, S. (1991) Endotoxin in vaginal fluid of women with bacterial vaginosis. Obstet. Gynecol., 77, 265–266. Snowden, N., Wilson, P.B., Longson, M. et al. (1990) Antiphospholipid antibodies and Mycoplasma pneumonial infection. Postgrad. Med. J., 66, 356–362. Spiegel, C.A., Amsel, R. and Holmes, K.K. (1983) Diagnosis of bacterial vaginosis by direct gram stain of vaginal fluid. J. Clin. Microbiol., 18, 170–177. Stewart-Akers, A.M., Krasnow, J.S., Brekosky, J. et al. (1998) Endometrial leukocytes are altered numerically and functionally in women with implantation defects. Am. J. Reprod. Immunol., 39, 1–11. Stray-Pedersen, B., Bruu, A. and Molne, K. (1982) Infertility and uterine colonization with Ureaplasma urealyticum. Acta Obstet. Gynecol. Scand., 61, 21–24. Sziller, I., Witkin, S.S., Ziegert, M. et al. (1998) Serological responses of patients with ectopic pregnancy to epitopes of the Chlamydia trachomatis 60 kDa heat shock protein. Hum. Reprod., 13, 1088–1093. Tabibzadeh, S. (1991) Cytokine regulation of human endometrial function. Ann. N. Y. Acad. Sci., 622, 89–98. Tabibzadeh, S. and Sun, X.Z. (1992) Cytokine expression in human endometrium throughout the menstrual cycle. Hum. Reprod., 7, 1214–1221. Sub-clinical bacteria of the endometrium Taylor, R.N., Ryan, I.P., Moore, E.S. et al. (1997) Angiogenesis and macrophage activation in endometriosis. Ann. N. Y. Acad. Sci., 828, 194–207. Templeton, A. (1995) Infertility-epidemiology, aetiology and effective management. Health Bull. (Edinb.), 53, 294–298. Tindall, V. (1987) Jeffcoate’s Principles of Gynaecology. Butterworth, London. Toth, A., Lesser, M.L., Brooks-Toth, C.W. et al. (1986) Outcome of subsequent pregnancies following antibiotic therapy after primary or multiple spontaneous abortions. Surg. Gynecol. Obstet., 163, 243–250. Upadhyaya, M., Hibbard, B.M. and Walker, S.M. (1983) The role of mycoplasmas in reproduction. Fertil. Steril., 39, 814–818. Viniker, D.A. (1979) The fetal EEG (detection of oxygen deprivation). Br. J. Hosp. Med., 22, 504–510. Viniker, D.A. (1996) Late luteal phase dydrogesterone in combination with clomiphene or tamoxifen in the treatment of infertility associated with irregular and infrequent menstruation: enhancing patient compliance. Hum. Reprod., 11, 1435–1437. Viniker, D.A. (1997a) Menorrhagia and dysfunctional uterine bleeding. In Rainsbury, P.A. and Viniker, D.A. (eds), Practical Guide to Reproductive Medicine. The Parthenon Publishing Group, New York, London, pp. 431–444. Viniker, D.A. (1997b) Investigations for infertility management. In Rainsbury, P.A. and Viniker, D.A. (eds), Practical Guide to Reproductive Medicine. The Parthenon Publishing Group, New York, 385 London, pp. 93–110. Viniker, D.A. (1997c) Infertility treatment. In Rainsbury, P.A. and Viniker, D.A. (eds), Practical Guide to Reproductive Medicine. The Parthenon Publishing Group, New York, London, pp. 111–130. Viniker, D.A. (1999) Role of audit and computers in an ovulation induction program. In Sathanandan, S.M. and Jacobs, H. (eds), Practical Guide to Ovulation Induction. Imperial College Press, London (in press). Viniker, D.A., Matthews, T., Okuson, H. et al. (1997) The outcome of infertility treatment. In Rainsbury, P.A. and Viniker, D.A. (eds), Practical Guide to Reproductive Medicine. The Parthenon Publishing Group, New York, London, pp. 313–326. Wang, A., Fanning, L., Anderson, D.J. et al. (1997) Generation of reactive oxygen species by leukocytes and sperm following exposure to urogenital tract infection. Arch. Androl., 39, 11–17. Witkin, S.S., Askienazy-Elbhar, M., Henry-Suchet, J. et al. (1998) Circulating antibodies to a conserved epitope of the Chlamydia trachomatis 60 kDa heat shock protein (hsp60) in infertile couples and its relationship to antibodies to C.trachomatis surface antigens and the Escherichia coli and human HSP60. Hum. Reprod., 13, 1175–1179. Xiang, Y. and Chen, Q. (1996) Study on infertility caused by infections of Chlamydia trachomatis and Ureaplasma urealyticum. Chung Hua Fu Chan Ko Tsa Chih, 31, 223–255. Received on December 12, 1998; accepted on April 20, 1999