Chlamydia trachomatis and Genital Mycoplasms
David A. Eschenbach
Table Of Contents
David A. Eschenbach, MD
The organism that was eventually called Chlamydia trachomatis was first identified from the eye scrapings of adults with trachoma in 1907.1 The same organisms were also noted in the conjunctivae of neonates with nongonococcal ophthalmia neonatorum in 1909,2 and they were a well-established cause of neonatal conjunctivitis by the 1930s. Similar inclusions were recognized in the urethral discharge of men with nongonococcal urethritis by the 1940s, and this discharge was found to infect the cervix of pregnant female partners who in turn infected the eyes of neonates during delivery.3 However, at this date, the organism had not been cultured, and only smear identification was possible. Prior cytologic methods used to detect C. trachomatis in eye scrapings were too insensitive to :reliably identify genital tract C. trachomatis. Direct cytologic smears from the cervix, such as Giemsa stain, will detect only 20% to 40% of the cervical culture-positive women. The subsequent ability to culture chlamydiae allowed what has now become a tremendous expansion of knowledge about both the organism and the diseases it causes. In fact, it is now known that C. trachomatis causes a wide spectrum of disease (Table 1). The ability to culture the organism also led to new information about its physiology. It is now established that the organism is a small intracellular bacteria and not a virus as first believed. An understanding of its intracellular life cycle was an important step in explaining many of the clinical peculiarities of C. trachomatis, which are discussed later.
The genus Chlamydia contains two species, C. psittaci and C. trachomatis. Chlamydia psittaci is transmitted by birds and produces a respiratory infection, psittacosis, in humans and arthritis, conjunctivitis, and abortion in other mammals. Chlamydia trachomatis causes endemic and epidemic trachoma, which is still the leading cause of blindness throughout the world, and adult inclusion conjunctivitis, an infection often associated with a urethral discharge. Trachoma is usually caused by A-C immunotypes, and inclusion conjunctivitis is usually caused by D-K immunotypes.4 Chlamydia trachomatis also causes lymphogranuloma venereum (LGV), an ulcerative genital infection that is found in developing countries and in a small number of patients living in the southern United States. LGV is caused by the unique L immunotype. The remaining oculo-respiratory-genital infections in Table 1 are usually caused by D-K immunotypes,4 and they are spread by oculogenital contact. In fact, the spectrum of disease caused by C. trachomatis is virtually identical to that caused by Neisseria gonorrhoeae.
Chlamydiae are bacteria that contain both ribonucleic acid (RNA) and deoxyribonucleic acid (DNA). Like bacteria, they have a cell wall that is similar to that of gram-negative bacteria and they replicate by binary fission. However, similar to viruses, they are obligate intracellular organisms that require cell culture techniques for recovery.
Trachoma and oculo-respiratory-genital strains have similar in vitro virulence and usually attack only columnar or transitional epithelial cells; LGV strains are more pathogenic in tissue culture systems and in vivo attack even squamous epithelium. Chlamydia trachomatis causes infection of the urethra, cervix, endometrium, salpinx, rectum, or conjunctiva, and, except for LGV strains, chlamydiae do not infect vaginal or other squamous epithelium.
Chlamydiae have a slow developmental cycle, which explains the insidious onset of chlamydial infection. There are two forms of the organism. The elementary body (EB) is the infectious form. Once the EB attaches to the columnar epithelial cell, it is incorporated into the cell by a process of pinocytosis, in which the organism is surrounded with a phagosome membrane.5 Chlamydial particles remain within the phagosome membrane during their entire existence within the host cell. This membrane isolates the organism from the cellular immune system, protecting it from recognition and attack by usual cellular defense mechanisms. After being present for a few hours within the cell, the EB is transformed into a noninfectious reticular body (RB). Protected by the phagosome membrane from destruction, the RB directs virtually all of the cellular-produced adenosine triphosphate for its own growth. In this stage, the organism becomes metabolically active and it begins to actively synthesize protein. The RBs grow and eventually divide by binary fission, producing new infectious EBs. The multiplication process repeats itself until the chlamydial inclusion body becomes the dominant structure in the host cell.6 Numerous new infectious EBs are released back into the extracellular environment when the inclusion body ruptures, and the EBs infect adjacent epithelium while the host cell is destroyed.
The 24-hour replication process required for chlamydiae is rather slow compared with the 2- to 4-hour division time of most classic bacteria. Therefore, it is not surprising to find that there characteristically is a long latency period between the time chlamydiae are acquired and symptoms are produced. As an example, N. gonorrhoeae and C. trachomatis are often acquired at the same time in the urethra of males. Neisseria gonorrhoeae usually produces urethritis 3 to 7 days later. Gonococcal urethritis rapidly disappears when the patient is treated with penicillin. However, 2 to 6 weeks later, a second discharge (postgonococcal urethritis) occurs, caused by C. trachomatis. Owing in large part to the slow growth of the organism, chlamydial urethritis appears much later than gonococcal urethritis. In addition, chlamydial urethritis generally causes less leukocyte and tissue reaction than gonococcal urethritis. Thus, it is understandable that a second characteristic of chlamydial infection is the insidious onset of symptoms. In fact, often infection occurs without producing recognizable symptoms, and LGV and trachoma strains are known to exist for years while producing few or no symptoms.
Chlamydia trachomatis can be vertically acquired during birth from infected maternal genital secretions, or it can be acquired by sexual contact. Direct hand-to-eye spread also occurs in trachoma. Congenital chlamydial infection has not been documented. Neonatal infection is common, and it is directly related to maternal carriage rates. Approximately two thirds of infants exposed to maternal C. trachomatis will develop an antibody response (see Table 4). Microimmunofluorescent (MIF) antibodies are present in about 2% to 3% of pediatric patients under the age of 6, presumably as a result of vertical transmission from the mother.7 Exposure to chlamydiae subsequently occurs from genital contact as the population becomes sexually active. The prevalence of C. trachomatis antibody increases to 20% to 30% among youngsters between the ages of 9 and 15.8 From 20% to 40% of sexually active women (20-year-olds) have been exposed to chlamydiae based on the presence of chlamydial antibody.7,9
* Includes only infants who received silver nitrate eye prophylaxis.
Infants first acquire maternal C. trachomatis in the eyes and nasopharynx. The organism can be recovered from the eyes of infants who develop conjunctivitis, usually within the first 2 to 3 weeks of life.10,11 Among asymptomatic infants, chlamydiae are most frequently isolated from the nasopharynx, but cultures usually do not become positive until at least the fourth week of life.10 Thereafter, rectal cultures become positive. The organism is spontaneously cleared from the nasopharynx by 6 months of life,10 but longer periods of rectal colonization can occur.11
Active chlamydial infection demonstrated by culture has been present in up to 15% to 25% of selected sexually active adolescents.12,13,14 Several factors have been associated with an increased frequency of chlamydiae in teenagers. Since chlamydiae are sexually transmitted, it is not unexpected that infection has been present most often among women with multiple current partners,13 among women who had their first intercourse at a relatively young age, and among women with an increased number of years of sexual activity.12 Although some studies have indicated higher infection rates of women in lower socioeconomic groups,15 among adolescents, the organism is also prevalent in middle socioeconomic groups. For reasons that are not entirely clear, the organism is also associated with oral contraceptive use.12,14 Oral contraceptive use has been associated with the presence of cervical ectopy. It is theorized that women exposed to chlamydiae who have a larger cervical columnar cell surface area caused by the cervical ectopy may be at an increased risk of acquiring the organism.12 Chlamydiae are also frequently associated with other lower genital tract infections,12,15 particularly with gonorrhea. Chlamydia trachomatis has been isolated from about half of the women with gonorrhea attending sexually transmitted disease clinics. In fact, chlamydial infection rates usually parallel gonococcal rates. With the exception of women attending sexually transmitted disease clinics, chlamydial infection is usually three to four times more common than gonorrhea in sexually active female populations.12,15
Chlamydia trachomatis is a frequent infection among pregnant women. Although recovery rates vary widely, depending on the population studied, usually between 5% and 15% of pregnant women have C. trachomatis isolated. However, isolation rates of 30% to 40% have been reported (Table 2). As is true of nonpregnant women, pregnant women usually have C. trachomatis three to four times more frequently than N. gonorrhoeae.
* SES = Socioeconomic
‡ Second report, which included patients in ref 18.
§ p < 0.01.
º p < 0.025 comparing IgM-positive patients with Chlamydia-negative patients.
Female urethritis (acute urethral syndrome) is one of the newly described syndromes caused by C. trachomatis.32 Women who complain of acute dysuria and urinary frequency are usually considered to have a urinary tract infection. Bacteriuria in traditional levels (>105 organism/ml of urine) and in lower levels (usually 103 to 104 organism/ml detected by bladder tap) cause the symptoms in the majority of women. However, approximately 20% of women with these symptoms have sterile pyuria (>8 leukocytes/mm3 of urine) caused by chlamydiae. The following characteristics were associated with chlamydial urethritis: a long duration of symptoms, a high frequency of new sexual partners in the previous month, a low frequency of prior urinary tract infection, and a low rate of hematuria.32 Symptomatic women with sterile pyuria need to be scrutinized for chlamydial infection.
Chlamydiae cause a well-described cervicitis among nonpregnant women. The cervicitis can be detected either by finding a yellow cervical mucus or by finding ten or more polymorphonuclear leukocytes per high-power field on a cervical Gram stain.33 About half of the women with cervicitis had chlamydiae; conversely, yellow cervical mucus was present in 60% of women with C. trachomatis isolated.33 Ninety-five percent of the women with chlamydiae had either yellow cervical mucus or ten or more leukocytes. Although chlamydial cervicitis often causes a visible purulent cervical discharge, symptoms of an abnormal vaginal discharge or other symptoms are usually not present among women with chlamydiae, so symptoms are not closely correlated with the recovery of the organism.15 It is estimated that only one third of women with chlamydiae are symptomatic. The relative low sensitivity of symptoms and signs for chlamydial infection makes it difficult to use clinical criteria for detecting the organism. Additionally, the criteria of cervicitis have been developed for nonpregnant women only. The presence of C. trachomatis infection appears to be even more difficult to discern for pregnant than for nonpregnant women. Most pregnant women have a vaginal discharge, and a large proportion of normal pregnant women without C. trachomatis have both a cloudy mucus and white blood cells (WBCs) in the cervical mucus. During pregnancy, the presence of C. trachomatis has not been related to the presence of a cloudy cervical mucus or to WBCs, and the use of these clinical signs to detect women with chlamydiae is probably unreliable in the pregnant woman. However, a markedly purulent cervical discharge suggests either chlamydial infection or gonorrhea.
Among nonpregnant women, chlamydiae cause a well-recognized form of endometritis and salpingitis that is analogous to gonococcal infection. In pregnancy, chlamydial salpingitis has not been described to date, in contrast to rare but well-documented cases of gonococcal salpingitis. However, it is possible, although unproven, that chronic chlamydial endometritis could coexist with pregnancy.
Numerous cervicovaginal organisms have been associated with an adverse pregnancy outcome such as abortion, still-birth, premature labor, premature birth, and premature rupture of the membranes. Group B streptococci,34Ureaplasma urealyticum,35,36Mycoplasma hominis,37Trichomonas vaginalis,30 and bacterial vaginosis38,39 (formerly called Gardnerella vaginalis vaginitis or nonspecific vaginitis) have all recently been associated with one or more of these adverse pregnancy outcomes. Chlamydia trachomatis has also been associated with prematurity and premature rupture of the membranes.23,26 The potential impact of C. trachomatis is important because of the large number of pregnant women with chlamydiae. If there is a causal relationship between chlamydiae and pregnancy outcome, the mechanism of producing an adverse outcome is not known. Chlamydiae can attach to and grow in amniotic membrane cells,40 which raises the possibility that an ascent of cervical infection to the amnion could lead to amnion infection. As suggested, chlamydial decidual infection is also possible.
It is the current hypothesis that women at highest risk for an abnormal pregnancy outcome are those who have been recently exposed to chlamydiae. Women with chronic infection have usually manifested an IgG antibody response that may reduce the effect of chlamydial infection on these sites. This hypothesis was advanced by Harrison and co-workers,26 who found that an adverse pregnancy outcome occurred only among women who had both chlamydial infection and chlamydial IgM antibody. Women with chlamydial infection but only chlamydial IgG antibody, and women without chlamydiae, did not have an abnormal rate of pregnancy outcome.26
However, a true proportion of women with chlamydiae who deliver prematurely and the proportion of prematurity related to chlamydiae are still unclear. The pregnancy outcome of women with C. trachomatis infection is inconsistent from study to study. In some studies there has been a marked association between C. trachomatis and an adverse outcome,23 but there was no apparent relationship in other studies22,24,25 (see Table 2). Virtually all of the investigators have focused on a limited number of organisms and used too few women in the studies for definite conclusions to be reached. Therefore, we must await large studies in which not only chlamydiae but all of the organisms that have been associated with an adverse pregnancy outcome are simultaneously studied. In these reports, demographic, social, and other confounding factors associated with both organism prevalence and pregnancy outcome will also have to be controlled.
Since new data will not be available for several years, it is important to review the present data linking chlamydiae to an adverse pregnancy outcome. It has been demonstrated that experimental human C. trachomatis causes abortion and placentitis in cattle.41 Investigators conducting early studies on neonatal conjunctivitis going back to the 1930s retrospectively recorded a high rate of prematurity among the infants who developed conjunctivitis.42,43,44 The first prospective culture studies of chlamydiae during pregnancy in the 1970s did not report a difference in the prematurity rate of women when compared with women without chlamydiae.16,17,18,19,20,21 However, the women in these first studies were usually followed only during the third trimester, so second-trimester and early third-trimester pregnancy losses would have occurred before patients were studied (see Table 2). This potential bias was corrected by Martin and colleagues,23 who enrolled women into their study by the 18th week of pregnancy. They found that stillbirth, premature delivery, and perinatal death from prematurity were associated with chlamydiae. Most of the adverse pregnancy outcomes occurred in the second or early third trimester before the 30th week of pregnancy. The duration of pregnancy was less than 30 weeks for 5 of 18 women with and 7 (3%) of 238 women without chlamydiae (p < 0.01). Matching for variables associated with chlamydial infection did not change the results. However, only 18 women with chlamydiae were included in this study, and it should also be noted that the women with chlamydiae in this study were unusually young. Of those with chlamydiae, 60% were under 20 and 93% were under 25 years of age. In a larger prospective study of chlamydiae in over 1300 women, Harrison and associates26 found no overall increase in the rate of spontaneous abortion, stillbirth, prematurity, or premature rupture of the membranes among women with chlamydiae compared with women without chlamydiae. However, as previously mentioned, they further analyzed the pregnancy outcome of the 17 women who had chlamydial IgM antibody. These 17 women represented 24% of the 72 women with C. trachomatis recovered in this study. Women with both chlamydial infection and chlamydial IgM antibody were significantly more likely to have delivered a low-birth-weight infant (p < 0.025) and to have premature rupture of the membranes (p < 0.01) than either IgM-negative but culture-positive women or culture-negative women. The association between chlamydiae and these two outcomes could not be explained by the confounding variables of age, genital mycoplasmal infection, vaginal bleeding, or prenatal antibiotic treatment. Recently, Gravett and co-workers38,39 found that C. trachomatis was associated with a premature birth in both a case-controlled study and a cohort study of premature labor. It is of interest that the majority of women in the population studied by Harrison's group26 had only IgG antibody, and it is probable that the majority of pregnant women with C. trachomatis have chlamydial IgG but not IgM antibody. The presence of the IgG antibody may protect against chlamydial invasion, and these women may not be at risk for an adverse pregnancy outcome. However, women with a recently acquired C. trachomatis infection defined by the presence of chlamydial IgM antibody may be at risk for an adverse outcome caused by chlamydiae. Adolescent women may be at particularly high risk for an adverse pregnancy outcome from chlamydiae because of recently acquired infection.23
In contrast to these three studies, there are several other recent prospective studies28,29,30 in which there was not even a trend of an increased rate of adverse pregnancy outcome among women with chlamydiae (see Table 2). Some of these studies are relatively small, and many of the patients were enrolled after the time that some of the very early premature births or premature rupture of the membranes would have already occurred. However, the inconsistent findings between studies raise uncertainty about the magnitude with which chlamydiae are related to an adverse pregnancy outcome. In addition, the relative contribution of chlamydiae compared with other organisms must be further studied before more definite conclusions can be reached. At present, if there is an association between chlamydiae and these adverse pregnant outcomes, the relative risk is in the range of two to three, and perhaps the risk exists only among women who recently acquired the infection.
It has become well established that C. trachomatis is a major cause of salpingitis among nonpregnant women. Thus, it is not surprising to find an association between chlamydiae and postpartum endometritis. In fact, in the early retrospective studies made by ophthalmologists as early as the 1930s, it was noted that the mothers of infants with inclusion conjunctivitis had an unusually high rate of postpartum infection.31,44,45,46 Rees and co-workers47 first prospectively confirmed a high rate of developing postpartum endometritis among women with antenatal chlamydiae. Many of the infants born of these mothers also developed inclusion conjunctivitis. Because of the long latency period of chlamydial infection, the endometritis develops relatively late following delivery, usually between 2 and 6 weeks post partum.47 Wager and colleagues48 differentiated women who developed an early postpartum endometritis (within the first 48 hours post partum) from women who developed endometritis from 3 days to 6 weeks post partum. A retrospective medical chart review was made for a clinical diagnosis of postpartum endometritis. It was found that the majority of women who developed early endometritis delivered by cesarean section, in contrast to the women developing late endometritis, who usually had delivered vaginally. There was no association between chlamydiae and early endometritis,48 a finding subsequently prospectively confirmed.49 However, late endometritis developed in 22% of women with and 5% of women without antepartum chlamydiae.48 (Table 3). The association between chlamydiae and late postpartum endometritis remained when infected women were compared with controls matched for factors associated with postpartum endometritis. However, the persons in this retrospective study were not cultured for chlamydiae at the time they developed endometritis, and other organisms known to be associated with endometritis were not studied.
* Data are not provided.
In the previously mentioned large study of Harrison and co-workers,26 women with antepartum Mycoplasma hominis had a 7.3 times increased risk of postpartum endometritis compared with women without M. hominis, but there was only a weak association between postpartum endometritis and antepartum C. trachomatis. In two other studies,24,28 women with antepartum C. trachomatis were found to have higher rates of postpartum endometritis than Chlamydia-negative women. In the first prospective study in which C. trachomatis was cultured when endometritis was diagnosed, C. trachomatis was recovered from the cervix or endometrium of 60% and from the endometrium of 23% of the vaginally delivered women with late postpartum endometritis.50 This is the first direct culture evidence linking chlamydiae with late infection. Genital mycoplasmas were recovered from the endometrium of 33% of this group. Both chlamydiae and genital mycoplasmas may play a role in this clinically mild late postpartum infection. Few women with late endometritis were febrile, had a leukocytosis, or were seriously ill.50 The importance of this late postpartum infection has not been fully established. However, it is interesting to note that from one third to one half of infertile women have been previously pregnant. Multiple studies have found an association between distal tubal obstruction and chlamydial antibody titers.51 It is established that approximately 20% of the women with chlamydiae recovered from the cervix prior to abortion developed a postabortal salpingitis.52 Postabortal salpingitis was virtually confined to the women with chlamydiae. Therefore, it is possible that a significant proportion of infertile women developed their salpingitis from chlamydiae following a normal delivery or following abortion. At the present time, pregnant women who are known to have chlamydiae prior to delivery should receive therapy of the chlamydial infection prior to discharge from the hospital. In addition, if patients develop mild signs and symptoms of late postpartum endometritis, cultures for chlamydiae are warranted and therapy with erythromycin or, for non-breast-feeding women, tetracycline should be considered. A final word of caution is in order: a mild clinical infection does not necessarily mean a mild tubal infection. Consistently one half of women with fallopian tube infection severe enough to occlude the fallopian tube have never had a recognized clinical infection.51 Therefore, it seems likely that this infection is important to recognize and treat even if clinical manifestations are mild.
At least two neonatal infections have been well documented. As previously mentioned, the organism has long been associated with neonatal conjunctivitis, and it has recently been established that chlamydiae cause neonatal pneumonia. Congenital infection among infants who deliver by cesarean section with intact membranes has not been reported. Neonatal chlamydial infection results when the infant acquires the organism from contact with the infected maternal genital tract. Infants born vaginally to a woman with chlamydiae have a high rate of infection: from 50% to 70% develop either signs or symptoms and a positive culture or have asymptomatic culture or antibody evidence of chlamydial infection (Table 4). If the high rate of maternal chlamydiae generally ranges from 5% to 15%, between 3% and 11% of all vaginally delivered infants would be expected to develop chlamydial infection.
Conjunctivitis develops in about 40% (range 20%-40%) of the infants born of women with chlamydial infection (see Table 4). This widely variable attack rate includes both prospectively followed infants with minimally symptomatic disease and a smaller proportion of infants who presented for treatment because of overt symptoms. It is estimated that 20% of the infants develop overt symptoms of conjunctivitis and that the other 20% of infants can be detected only by close surveillance. The conjunctivitis usually occurs between 5 and 14 days after birth, which is a finding that has been used to differentiate chlamydial from gonococcal conjunctivitis.43 Gonococcal ophthalmia usually occurs within the first 3 days of life. However, there is considerable overlap in the time of onset between the two infections, and N. gonorrhoeae cultures must be performed to differentiate gonococcal from chlamydial conjunctivitis. Chlamydial conjunctivitis first causes conjunctival hyperemia followed by an inflammatory reaction characterized by a mucopurulent discharge and occasionally a pseudomembrane. Conjunctival swelling and the purulent discharge can cause difficulty for the infant in opening the eyes following sleep.
The differential diagnosis of neonatal ophthalmia includes the very destructive bacterial infections of gonorrhea and other bacteria such as Haemophilus influenzae and Pseudomonas. Until recently, it was believed that chlamydial conjunctivitis did not cause permanent eye damage, in contrast to adult trachoma, which causes pannus formation (corneal vascularization), corneal scarring, and blindness. However, it has now been demonstrated by careful slit lamp examination that minimal micropannus formation occurs following neonatal infection.46 There has been no decrease in visual acuity recognized from the initial chlamydial neonatal eye infection. Nevertheless, early treatment within 12 days of birth prevented the structural changes.
In 1977 Beem and Saxon described an interstitial pneumonitis among infants tinder the age of 6 months that was associated with the recovery of C. trachomatis.54 The pneumonitis usually presents between 4 and 8 weeks after birth,55 a finding that is consistent with the long latency period of chlamydial infections. Approximately 30% of all pneumonitis among infants under the age of 6 months are associated with chlamydiae.55
Infants have the gradual onset of a peculiar repetitive staccato-type cough that has usually persisted for several weeks. Nasopharyngeal congestion and tachypnea are common. Infants are usually afebrile and lack systemic symptoms or signs, but they often fail to gain weight. Interstitial infiltrates and hyperinflation of the lungs are present on chest film. Eosinophilia is common, and elevated IgM and other immunoglobulin levels are present.54 Untreated infants have persistent cough and pneumonia for several weeks to months. Patients receiving sulfisoxazole or erythromycin therapy have reduced length of cough and rapid resolution of the pneumonitis in comparison with patients who have received no therapy.56
It is not clear at the present time whether the pneumonia develops from material that has been aspirated into the lung at the time of delivery or from material that is aspirated into the lung postnatally from infected conjunctival and nasopharyngeal sites. Only about half of the infants with pneumonitis have had previous clinically symptomatic conjunctivitis, although perhaps an even larger number have nasopharyngeal chlamydial infection consisting of a mucus nasal discharge. Nasopharyngeal cultures are the most sensitive culture site. The differential diagnosis includes respiratory syncytial virus infection, adenovirus infection, and a variety of other bacterial pneumonias.55
Infants with pneumonitis commonly have evidence of otitis media. In fact, C. trachomatis has been isolated from the middle ear of infants with chlamydial pneumonia and simultaneous serous otitis media.57 However, among unselected children who were usually over the age of 1 year, C. trachomatis was not isolated from serous otitis fluid.58 In a retrospective serologic study,59 infants and children from 1 to 15 years of age with chlamydial antibody had increased rates of pneumonia and conjunctivitis in the first year of life compared with serum-negative children. However, following the first year there was no difference between the two groups in episodes of respiratory illness, otitis media, other chronic respiratory illness, or gastroenteritis. (Chlamydiae can be present on rectal culture beyond the first year.) Thus, there is no clear indication, at present, that a significant proportion of otitis media, gastrointestinal disease, or respiratory illness after the first year of life is attributed to chlamydia.
The possibility of chlamydial infection should always be entertained in those syndromes that have been highly associated with isolation of the organism. Thus, the presence of sterile pyuria in a woman complaining of acute dysuria,32 the presence of mucopurulent cervicitis in a patient complaining of an abnormal vaginal discharge,33 and the presence of lower abdominal pain among postpartum women50 are clinical conditions in which C. trachomatis should be suspected on a clinical basis. Chlamydial infection should also be suspected among women with gonorrhea, up to 50% of whom will also have C. trachomatis. Among neonates, the development of conjunctivitis 5 or more days following delivery and the development of pneumonia, particularly within the first 3 months of life, should also prompt a clinical suspicion of chlamydial infection. When chlamydial infection is suspected, both members of the infant-mother pair should be scrutinized.
For persons with conjunctivitis, chlamydiae can usually be found in a Giemsa stain. However, in the remaining syndromes, cultures or direct fluorescent antibody stains should be used to identify the organism. Infected material for culture should be obtained on Dacron swabs rather than cotton swabs because of the presence of a cytotoxic agent in the latter. Swabs used to obtain the cultures should be on either an aluminum or a plastic shaft. The presence of preservatives in the wood shafts also inhibits the growth of chlamydiae. The specimen is collected and temporarily stored in a transport vial that contains antibiotics to inhibit other bacteria and yeast. The transport vial should be refrigerated until the culture is obtained. The vial should be refrigerated again until the culture is inoculated. Inoculation should be done within 24 hours.
Chlamydiae are inoculated on a monolayer of either McCoy cells or HeLa cells that have been treated with iododeoxyuridine or with cycloheximide to depress cellular metabolism and prevent the replication of the host culture cells. Enhanced infectivity is obtained by centrifuging chlamydiae onto the monolayer of host cells. Chlamydial inclusions are identified 2 to 4 days after the incubation of the cultures. Traditionally, inclusions have been identified by iodine staining, but more recently fluorescein-tagged monoclonal chlamydial antibody has been used to identify chlamydiae on the culture plates.32
Direct smears of genital secretions have also been recently used to make a rapid diagnosis of chlamydial infection without the use of culture.32 The technique is useful for laboratories that do not have the facilities to culture chlamydiae. The direct slide method takes only a few hours to perform. A swab is used to place secretions on a glass slide, which is immediately fixed. Specimens can then be easily transported without the concern of organism viability. In the laboratory, fluorescein-tagged monoclonal chlamydial antibody is applied to the slide material. Fluorescent chlamydial elementary particles appear green under the fluorescent microscope. The sensitivity and specificity (both >90%) of the direct slide identification method appear to be equal to that of the culture among symptomatic persons. However, the sensitivity may be less among asymptomatic persons, so the direct method may not be as sensitive as the culture for screening purposes.
While 25% to 50% of women/infants with these syndromes will have chlamydial infection, only a proportion of women with chlamydiae will be detected by presenting with one or more syndromes. Nevertheless, general screening for C. trachomatis even during pregnancy cannot be advocated at present because of the high cost and general unavailability of the culture. Selected screening of pregnant women at high risk for C. trachomatis should be considered. Selection would include women with a purulent cervical discharge, gonorrhea and other sexually transmitted organisms, and adolescent women.
Although in vitro sensitivity testing of chlamydiae has not been well developed, chlamydial eradication in vivo has been roughly correlated with in vitro chlamydial inhibition. Antibiotics that appear to be most active against chlamydiae include tetracycline, erythromycin, sulfonamides, and rifampin. Clindamycin also appears to have activity against chlamydiae, with a minimal inhibitory concentration less than 2.0 μg/ml.60 The sensitivity of chlamydiae to penicillin is unclear at the present time. Penicillin is not highly active against Chlamydiae in vitro, and while penicillins inhibit the growth of chlamydiae, washing the meda and cells free of penicillin allows chlamydiae to reinstitute growth.61 However, ampicillin does appear to have some activity against chlamydiae present in cervical sites.62 Cephalosporins and aminoglycosides have virtually no in vitro or in vivo activity against chlamydiae.63 Agents that are particularly active against the organism should be chosen for treatment. Among pregnant women and breast-feeding women, erythromycin is the treatment of choice. Sulfonamides or trimethoprim/sulfamethoxazole can be used among asymptomatic patients.64 Non-breast-feeding postpartum women could be treated with tetracycline.
NEONATAL EYE PROPHYLAXIS.
It is apparent that silver nitrate prophylaxis used for gonococcal ophthalmia does not prevent chlamydial conjunctivitis. In fact, it is possible that silver nitrate damage of the conjunctival cells may increase the ability of chlamydiae to attack the conjunctiva. Erythromycin and perhaps tetracycline ophthalmic ointments placed in the infants' eyes appear to reduce the rate of neonatal chlamydial conjunctivitis. Neonatal conjunctivitis developed in 33% of infants randomly treated with silver nitrate and in none of 24 infants who received erythromycin ointment.20 However, erythromycin eye drops did not reduce nasopharyngeal colonization, and eye prophylaxis probably has no effect on the development of subsequent chlamydial pneumonia. In addition, unpublished reports indicate that erythromycin-treated infants can develop conjunctivitis.
Opinion is still divided on the treatment of neonatal chlamydial conjunctivitis. Traditional therapy has used topical sulfonamides, tetracyclines, and erythromycins four times daily for 3 weeks. Frequent recurrent chlamydial conjunctivitis has been noted after topical therapy. Recurrence may be due either to difficulty instilling eye drops or to noncompliance on the part of the mother. Because of a high failure rate following topical therapy, it is now recommended that oral erythromycin be instituted for 2 weeks. An added advantage of the oral erythromycin therapy is that nasopharyngeal colonization is eliminated and the possibility of pneumonitis is reduced.
Infant pneumonitis has been successfully treated with sulfisoxazole and with erythromycin for 2 weeks.56 The infants became culture negative and had no clinical relapses. Sulfisoxazole, 150 mg/kg/day, or erythromycin estolate, 40 mg/kg/ day, has been used.
The recommended regimens for treating maternal antepartum infection include erythromycin base, 500 mg, or erythromycin ethylsuccinate, 800 mg, both given four times daily for 7 days.65 However, gastrointestinal intolerance can occur at this dose, and the equivalent of 250 mg of the base four times daily has been effective for nonpregnant women. A recent report indicates that this lower equivalent dose (erythromycin ethylsuccinate, 1.6 g daily) is effective.66 This regimen eliminated maternal chlamydiae and prevented neonatal infection in more than 90% of pregnant women treated in the third trimester66 with a less than 5% incidence of side-effects. Pregnant women with chlamydiae should not receive tetracycline, of course, or erythromycin estolate, which has been associated with liver enzyme elevations during pregnancy.67
Postpartum women with asymptomatic cervical infection or with mild late postpartum endometritis should be treated with either erythromycin or tetracycline in doses of 500 mg four times daily for 10 days. An alternative regimen is doxycycline, 200 mg twice a day for 10 days. Trimethoprim/sulfamethoxazole (160 mg of trimethoprim and 80 mg of sulfamethoxazole) in 2 double-strength tablets twice daily for 5 days may also be used for minimally symptomatic or asymptomatic women.
It is apparent that since chlamydial infection is sexually transmitted, male contacts of women with chlamydiae require examination and possibly treatment. Asymptomatic male chlamydial carriage is common, but many asymptomatic males often have an increased number of WBCs in the urethral discharge. In addition, the physician may be asked to treat female sexual contacts of men with symptomatic nongonococcal urethritis or with proven chlamydial infection. Female contacts of men with chlamydial infection have a 60% to 70% incidence of infection themselves,68 and these females should be either cultured or treated.
In summary, chlamydial infection is common during pregnancy, usually occurring in 5% to 15% of women. It is particularly related to a young age, perhaps a lower economic status, the presence of other sexually transmitted organisms, and multiple sexual partners. The organism is also related to mucopurulent cervicitis and to sterile pyuria. There is a controversy whether or not the infection causes prematurity. However, it is well documented that neonatal conjunctivitis and pneumonia frequently occur when infants are born through an infected cervix. Late postpartum maternal endometritis also appears to be a consequence of antenatal infection. Recognition of the possibility of chlamydial infection with these clinical entities and a prompt diagnosis by a combination of clinical criteria, culture, or direct fluorescent antibody stain, together with the administration of the appropriate antibodies, should reduce the morbidity caused by this organism.
The role of genital mycoplasmas in the pathogenesis of genital and neonatal infections remains unsettled in part because these organisms are both exceedingly ubiquitous and have a relatively low virulence. Mycoplasmas are often present in the genital tract of women who have no symptoms or abnormalities; they are usually found together with other more virulent organisms, and they have been isolated from sterile sites without producing disease. They are of interest to the perinatologist because of their association with numerous perinatal infections. The two commonly isolated genital mycoplasmas are M. hominis and Ureaplasma urealyticum, formerly called T-strain (tiny) mycoplasma. Ureaplasma urealyticum is distinct from the other mycoplasmas because it is able to hydrolyze urea. Mycoplasma fermentens, an uncommon genital mycoplasma, and M. genitalium, a rare isolate, are not discussed.
Mycoplasmas were first isolated in 1898 from cattle with pneumonia, which led to the original term pleuropneumonia-like organisms (PPLO).69 Mycoplasmas were not isolated from humans until 1937, when a genital strain was isolated from a bartholinian abscess.70 Several oral strains of Mycoplasma were discovered between 1937 and 1962, but they were considered to be innocuous. Eaton and associates71 clearly established the pathogenicity of this organism as a cause of human pneumonia in 1962, when an oral strain, M. pneumoniae, was isolated from patients with cold-agglutin-positive atypical pneumonia. Interest in genital mycoplasmas increased when Shepard72 isolated T-strain mycoplasma from men with nongonococcal urethritis; U. urealyticum causes Chlamydia-negative nongonococcal urethritis in males with their first episode of urethritis.73
Genital mycoplasmas have been associated with low-birth-weight infants, stillbirths, and spontaneous abortions. Although infants usually come in contact with the organisms during labor, neonatal mycoplasmal infection has not been well established. There is increasing evidence that genital mycoplasmas are related to a proportion of postpartum endometritis infections.
Since the mycoplasmas are ubiquitous in the genital tract of females, their exact role in infections can be established only by using appropriately chosen control groups, avoiding lower genital tract flora contamination of sterile site culture, and obtaining data on the presence of coexisting bacterial infection. In many reports of possible Mycoplasma infection, these points have not been considered, and the role of genital mycoplasmas in several disease entities remains unsettled.
Mycoplasmas exist phytogenetically between bacteria and viruses. Mycoplasmas are different from bacteria in both structure and size. In place of the cell wall that exists around bacteria, mycoplasmas are surrounded by a nonrigid triple-layered membrane. This membrane accounts for the fragile, pleomorphic structure of the organism. Mycoplasmas have a wide range of sizes, but all are smaller than bacteria. They are the smallest free-living organisms that do not depend on the host cell for reproduction. Mycoplasmas contain both RNA and DNA. In contrast, viruses are distinguished by containing either RNA or DNA, and they depend on the host cell's metabolism for replication.
Mycoplasmas can easily be recovered from clinical specimens. A single culture is approximately 90% sensitive for the recovery of mycoplasmas.74 The recovery rate is higher from vagina samples than from samples obtained from other genital sites, including the cervix.
In the laboratory, mycoplasmas can be isolated from cell-free complex commercial broth that is supplemented with horse serum. They also grow on supplemented agar medium. On agar, M. hominis forms “fried-egg” colonies, which are easily observed without the aid of a microscope. Ureaplasma urealyticum forms smaller colonies, which can be distinguished from M. hominis by a drop of urease test reagent, which turns the Ureaplasma colonies brown. The small colonies can be seen with the aid of a dissecting microscope. Mycoplasmas have no cell wall and hence cannot be detected on Gram stain, and they are resistant to β-lactam antibiotics (penicillin or cephalosporin), which depend on inhibition of cell wall synthesis for effect. However, mycoplasmas are susceptible to antimicrobials that inhibit protein synthesis.
Genital mycoplasmas are infrequently isolated from prepubertal girls. However, the recovery rate increases dramatically following the onset of sexual intercourse. There is a strong direct relationship between the colonization rates of Ureaplasma and the number of sexual partners.75 Ureaplasma urealyticum was isolated from 6% of women who denied ever having intercourse, from 38% of women with one lifetime sexual partner, from 55% of women with two lifetime sexual partners, and from 75% of women with three or more previous sexual partners. While isolation rates of M. hominis do not have the same dramatic rise with increasing numbers of sexual partners, M. hominis has been isolated from only 1% of women without prior intercourse compared with 17% of women with three or more sexual partners. Differences in the number of sexual partners are often the most important determinant of the wide range of recovery rates observed in different populations. During pregnancy, isolation rates for U. urealyticum range from 40% to 95% and for M. hominis, 40% to 60%.
Genital mycoplasmas are isolated more frequently from women of lower socioeconomic groups than from patients who can afford private care and more frequently from black patients than from nonblack patients.76 However, the increased prevalence in women from lower socioeconomic groups and black patients has not been controlled for sexual exposure rates.
Vertical transmission to neonates occurs during birth, when the organism is acquired from the cervix or vagina. Infants acquire the organism either when cervicovaginal secretions are aspirated or from direct colonization of fetal skin and mucosal surfaces. Genital mycoplasmas can be cultured from the vagina and external auditory canals of approximately 25%, and from the throat of approximately 10%, of infants weighing over 2500 g. Neonates born by cesarean section have a low Mycoplasma colonization rate. Mycoplasmas have been detected 100 days after delivery, but the recovery rates then decline for the remainder of the first year of life. Few infants become colonized after the perinatal period, which further suggests that acquisition took place during birth.
The role of genital mycoplasmas in prematurity remains controversial. There are a number of studies in which women or infants with U. urealyticum,35,36,77,78M. hominis,77,78,79 or both80 delivered at either a significantly lower gestational age or lower birth weight than women without mycoplasmas. However, there are other studies in which no significant correlation was found between the presence of cervicovaginal mycoplasmas and prematurity. Many of the negative studies included too few women to provide valid data. However, in a large study of 1300 women,26 there was no increase in the prematurity rate of women with either U. urealyticum or M. hominis compared with women without mycoplasmas. In a recent study, 13% of women in premature labor had genital mycoplasmas isolated from the amniotic fluid.39 Bacteria/Candida were present in the amniotic fluid of 11% of the women in premature labor. Women with amniotic fluid bacteria delivered markedly prematurely, while the women in premature labor with genital mycoplasmas (usually U. urealyticum) delivered nearly at term.39 These data suggest a weak association, if any, between U. urealyticum and prematurity.
In contrast to these data there are several lines of evidence that suggest an association between genital mycoplasmas and prematurity. Abacteriuric women who received tetracycline were more likely to deliver infants weighing more than 2500 g (93%) than women who received placebo (83%).81 Women with cervicovaginal U. urealyticum who received erythromycin were more likely to deliver infants weighing more than 2500 g (97%) than women who received placebo (89%).36 These data suggest that prematurity can be reduced by the administration of an antimicrobial that is capable of inhibiting mycoplasmas and further implies that mycoplasmas may cause some prematurity. However, antimicrobial therapy probably does not rid the patient of cervicovaginal U. urealyticum, and the extensive effect of these broad-spectrum antimicrobials on other flora would be expected to treat other organisms associated with prematurity.
To date, the strongest data that associate genital mycoplasmas with prematurity are derived from placental cultures. Ureaplasma urealyticum was significantly more likely to be recovered from between the chorion and amnion of the placentas of infants who were stillborn, who died after delivery, or who were admitted to the neonatal intensive care unit (approximately 20%) than from placentas of infants who delivered at term (11%).82 Mycoplasma hominis was also found more frequently in the complicated (5%) than the control (0.5%) pregnancies.82 The recovery of other organisms such as coliforms or group B streptococci was not associated with these adverse neonatal outcomes. The association between prematurity and the isolation of genital mycoplasmas from between the membranes of the placenta has been confirmed.78,83,84 An association between placental mycoplasmas and histologic chorioamnionitis was also noted in these studies. Other bacteria were also associated with prematurity in two later studies.83,84 The :finding of significantly more genital mycoplasmas in the placentas of pregnancies that deliver prematurely than in those of infants born at term is strong but not irrefutable evidence that the organisms caused the premature labor. Alternatively, the organisms could colonize the membrane as a result of premature labor or rupture of the membranes.
The end result of these studies has been difficulty in the interpretation of the information on mycoplasmas and prematurity. The results have not been consistent between studies; in most of the studies the contribution of other organisms had not been adequately studied, and the ubiquity of the organism acts to magnify differences between groups. Treatment studies now in process may clarify the role of mycoplasmas in prematurity. At present, these disparate observations make it difficult for the physician to know whether or not to consider mycoplasmas a factor in prematurity.
SPONTANEOUS ABORTION AND STILLBIRTH.
The causal relationship between Mycoplasma and spontaneous abortion remains unclear. In 1964 Shepard72 isolated Ureaplasma from the decidua and placenta of a midtrimester abortion. In 1967, Jones85 reported the isolation of M. hominis from the lungs of 5 (8%) of 62 aborted fetuses whose mothers demonstrated an M. hominis serum antibody rise. In that same year, Harwick and co-workers86 reported the occurrence of a septic abortion in which M. hominis was isolated from the cervix, maternal blood, and fetal liver. Others have reported a higher cervical isolation rate of U. urealyticum87 or a higher prevalence of antibody88 among patients selected by history of previous abortion than among patients attending a routine prenatal clinic. However, other studies have reported that among women with a history of prior abortion, the rate of first-trimester abortion was not higher in women with U. urealyticum than in women without U. urealyticum. At the present time, no conclusions can be drawn regarding the association of U. urealyticum and spontaneous first-trimester abortion.
In cases of spontaneous second-trimester abortion, mycoplasmas were the most common organism isolated from both placental and fetal specimens.89 In contrast, mycoplasmas were infrequently isolated from induced midtrimester abortions. The M. hominis isolation rate among patients with febrile abortions is higher than the isolation rate both among patients who had an afebrile abortion and among patients who had a normal pregnancy. More febrile than afebrile patients who aborted had a significant rise in M. hominis antibody titer, which suggests that infection, and not simply colonization of the tissue, occurred.89
Mycoplasmas were most commonly isolated from the lungs of dead fetuses aborted in the second or third trimester, suggesting that the lungs were originally colonized by mycoplasmas from a maternal cervicovaginal site. The rate of fetal colonization increases in cases of premature rupture of membranes and prolonged bleeding. Mycoplasmas have also been recovered from fetal blood, brain, liver, and other tissues, suggesting that septicemia occurs most likely from the originally colonized lung site. Fever was more likely in women whose uteri contained mycoplasmas or other organisms than in women without such contamination. These data, however, do not prove or even suggest that mycoplasmas have a role in the initiation of midtrimester abortion. Although mycoplasmas are common isolates in febrile abortive tissue, research is not conclusive on whether their presence simply reflects the ability of mycoplasmas to colonize dead tissue because of their frequency in the genital tract or whether they are in some way associated with a causative role in the initiation of abortion.
There is now ample evidence that both M. hominis and U. urealyticum cause postpartum fever. Since PPLO organisms were isolated from the cervix and nine consecutive blood cultures of afebrile postpartum women in 1952,90 there have been individual case reports of Mycoplasma isolation from the blood and higher M. hominis isolation rates from the cervix,91 or of both organisms from the uterus/adnexae,92 among women who became febrile than women who remained afebrile. However, new systematically obtained data provide even stronger evidence of mycoplasmal involvement in postpartum fever.
Genital mycoplasmas were recovered from the blood of 15% of women within 2 minutes, 8% of women from 2 to 10 minutes, and 2% of women beyond 10 minutes of delivery.93 This study demonstrates the frequent showering of intrauterine organisms into the bloodstream of unselected women immediately after delivery. This mechanism of showering occurred nearly as frequently for bacteria (5%) as for mycoplasmas (8% overall). Organisms can also gain access to the blood from an infected postpartum uterus. Genital mycoplasmas were isolated from the blood of 13% of febrile women with postpartum endometritis.94 They were isolated alone from 9% of women and together with bacteria in an additional 4% of women. Because none of 60 afebrile control women cultured 24 hours after delivery had mycoplasmas in the blood, the presence of mycoplasma in the blood of these women undoubtedly resulted from uterine infection and not from the shower of organisms that occurs shortly after delivery.
Both genital mycoplasmas have also been implicated in postpartum endometritis by their recovery from the endometrium of febrile women using protected endometrial culture devices, which reduce cervical contamination. Ureaplasma urealyticum endometrial isolation was significantly more common among febrile women (76%) than from afebrile control women (28%) using the same transcervical technique. Of 76% of febrile women who had U. urealyticum isolated from the endometrium, it was present together with bacteria in 58%, but it was the sole isolate in 18% of women.95 Blood isolation of U. urealyticum occurred among women in this study who had only genital mycoplasmas isolated from the endometrium, which further supports the concept that U. urealyticum caused the infection in about 20% of the women with postpartum endometritis.
Mycoplasma hominis also seems to cause a proportion of postpartum endometritis. Mycoplasma hominis was isolated from the uterus of 18% of febrile women; it was the sole isolate in 4% of women.95 Mycoplasma hominis was not isolated from the endometrium of afebrile control women. In another report, a fourfold or greater M. hominis antibody rise occurred among 61% of 23 women with unexplained postpartum fever. A fever occurred in 39% of 36 women with and in 17% of 54 women without a fourfold or greater titer rise (p < 0.05).96 This observation suggested that M. hominis infection occurred in a substantial proportion of these women. Over two thirds of the unexplained fever occurred among women with low M. hominis antibody titer. Postpartum fever developed among 40% of 40 women with M. hominis antibody titers less than 1: 8 and among 14% of 50 women with antibody titers 1:8 or greater (p < 0.01).96
Genital mycoplasmas should also be considered when women have late postpartum endometritis. Mycoplasmas were isolated from the uterus of one third of women with late endometritis, and they were the only organisms isolated from 11%. Women with this infection were usually afebrile and responded to erythromycin.
There are no unique clinical features that distinguish women with genital mycoplasmas from women with a bacterial endometritis. Although there is considerable overlap in the clinical findings of women with bacterial and mycoplasmal early postpartum infection, women with genital mycoplasmas usually have a fever but a mild degree of abdominal pain and abdominal and uterine tenderness. The women have no other physical abnormalities, such as cervical pus, a foul odor, or painful mass.97 Most of the women with genital mycoplasmal infection respond to antibiotics, which do not inhibit U. urealyticum. However, a change in antibiotics perhaps indicating a limited response occurred more commonly among women with mycoplasmal septicemia than among women with bacterial or no septicemia.94 Women with persistent mycoplasmal septicemia usually have a low-grade fever and minimal physical findings, which may include a mildly tender uterus. Patients with these findings and no other source of infection should be treated with erythromycin. A prompt response occurs if genital mycoplasmal infection is present.
Genital mycoplasmal infection acquired by the liveborn infant during labor has generally not been associated with serious neonatal morbidity. Histologic evidence of chorioamnionitis was present in 18 of 54 infants (33%) from whom Ureaplasma was isolated as compared with only 30 of 178 (17%) without Ureaplasma isolated.98 Despite the high rate of histologic chorioamnionitis, none of the infants developed a clinical infection. There was no association between Mycoplasma and neonatal death in a large series in which Mycoplasma was isolated from the lung in only 1 of 103 consecutive infants who died.99 However, a U. urealyticum antibody response has been more frequent among infants who develop respiratory distress (55%) than among normal infants (4%) (p < 0.001),100 and in addition, M. hominis has occasionally been isolated from sites of infants with skin abscesses, conjunctivitis, meningitis, and pneumonia, and U. urealyticum has occasionally been isolated from infants with pneumonia.101 Thus, serious neonatal infection has been documented from these low-virulence organisms.
Genital mycoplasmas can be readily recovered on special media.72 A medium for their recovery from blood has also been devised.94 However, this medium is not generally available in the clinical microbiology laboratory. Except for research purposes, routine diagnosis and treatment of lower genital tract mycoplasmas are not advised for the antepartum patient. As mentioned, the recovery of genital mycoplasmas can explain about 20% of early postpartum endometritis. A clinical diagnosis of postpartum mycoplasmal infection should be considered for febrile women with minimal abdominal/uterine physical findings who do not respond to penicillins or cephalosporins. The culture of mycoplasmas to help diagnose neonatal infection is not necessary at this date until better data are available.
Both genital mycoplasmas are usually susceptible to tetracycline, although M. hominis resistance has been reported. Ureaplasma urealyticum (but not M. hominis) is susceptible to erythromycin. Mycoplasma hominis is susceptible to clindamycin. The mycoplasmas are not inhibited by penicillins or cephalosporins. Upper genital tract or sterile site eradication occurs when treatment is given with effective antimicrobials. Mycoplasma hominis is usually eradicated from the lower genital tract when effective antimicrobials are administered, but U. urealyticum eradication from the lower genital tract does not frequently occur even with effective antimicrobial administration.102
Supported by National Institutes of Health Program Project Grant AI-12191
4. Wang SP, Grayston JT, Kuo C-C et al: Serodiagnosis of Chlamydia trachomatis infections with the microimmunofluorescence test. In Hobson D, Holmes KK (eds): Nongonococcal Urethritis and Related Infection, p 237. Washington, DC, American Society of Microbiology, 1977
7. Schachter J, Dawson CR: Comparative efficiency of various diagnostic methods for chlamydial infection. In Hobson D, Holmes KK (eds): Nongonococcal Urethritis and Related Infections, 337 pp. Washington, DC, American Society of Microbiology, 1977
20. Hammerschlag MR, Chandler JW, Alexander ER et al: Erythromycin ointment for ocular prophylaxis of neonatal chlamydial infection. JAMA 244: 2291, 1980
24. Thompson S, Lopez B, Wong K-H et al: A prospective study of Chlamydia and Mycoplasma infections during pregnancy: Relation to pregnancy outcome and maternal morbidity. In Mardh P-A, Holmes KK, Oriel JD et al (eds): Chlamydial Infections, p 155. Elsevier Biomedical Press, 1982
47. Rees E, Tait AA, Hobson D et al: Chlamydia in relation to cervical infection and pelvic inflammatory disease. In Holmes KK, Hobson D (eds): Nongonococcal Urethritis and Related Infections, p 67. Washington, DC, American Society for Microbiology, 1977
60. Harrison HR, Riggins RM, Alexander ER et al: In vitro activity of clindamycin against strains of Chlamydia trachomatis, Mycoplasma hominis and Ureaplasma urealyticum isolated from pregnant women. Am J Obstet Gynecol 149: 477, 1984
64. Brunham RC, Kuo C-C, Stevens CE et al: Treatment of concomitant Neisseria gonorrhoeae and Chlamydia trachomatis infections in women: Comparison of trimethoprim/sulfamethoxazole with ampicillin-probenecid. Rev Infect Dis 4: 491, 1982
66. Schachter J, Sweet RL, Grossman M et al: Experience with the routine use of erythromycin for chlamydial infections in pregnancy. N Engl J Med 314: 276, 1986
67. McCormack WM, George H, Donner A et al: Hepatic toxicity of erythromycin estolate during pregnancy. Antimicrob Agents Chemother 12: 630, 1977
81. Elder HA, Santamarina BAG, Smith S et al: The natural history of asymptomatic bacteriuria during pregnancy: The effect of tetracycline on the clinical course and the outcome of pregnancy. Am J Obstet Gynecol 111: 441, 1971
84. Hillier SL, Martius J, Eschenbach DA: The prevalence of microorganism in placentas from women with term and preterm birth. Presented at the Infectious Disease Society Obstetrics and Gynecology, August 8–10. 1985, Vancouver, B.C.
95. Rosene K, Eschenbach DA, Tompkins LS et al: Polymicrobial early postpartum endometritis with facultatively anaerobic and anaerobic bacteria, genital mycoplasmas and Chlamydia trachomatis: Treatment with piperacillin or cefoxitin. J Infect Dis 153: 1028, 1986