Depression of the lymphocyte transformation response to microbial antigens and to phytohemagglutinin during pregnancy.

Lymphocyte transformation (LT) responses to Chlamydia trachomatis, to four other microbial antigens, and to phytohemagglutinin (PHA) were studied in 201 women during pregnancy and/or 3-18 wk postpartum. The LT responses to all stimulants tested were significantly depressed during pregnancy when compared with postpartum LT responses. This difference occurred whether LT assays were performed in autologous or pooled heterologous plasma collected from nonpregnant donors. Among women studied in the third trimester and again postpartum, the autologous LT stimulation index (LTSI) rose from 1.7 to 3.4 (P less than 0.001) with C. trachomatis elementary body antigen, from 3.7 to 7.9 (P less than 0.001) with Candida albicans cell wall extract, from 4.5 to 7.8 (P = 0.008) with streptokinase-streptodornase, from 1.7 to 3.0 (P = 0.007) with fluid tetanus toxoid, from 1.7 to 2.8 (P = 0.046) with mumps virus skin test antigen, from 35.5 to 87.0 (P less than 0.001) with PHA (2 micrograms/ml), and from 107.2 to 181.9 (P = 0.007) with PHA (10 micrograms/ml). LT responses to C. trachomatis were compared in 52 pregnant women and 58 nonpregnant women; all the women had C. trachomatis isolated at the time of LT assay. Using either plasma supplement, the mean LTSI with C. trachomatis antigen was significantly higher in nonpregnant women than in pregnant women, regardless of trimester (P less than 0.001). Among 12 women who were serially tested and remained culture positive for C. trachomatis throughout pregnancy and the postpartum period, the mean autologous LTSI rose from 1.9 in the third trimester to 7.8 postpartum (P = 0.0004). These data are the first to show that the immune response to an ongoing bacterial infection is depressed during pregnancy and to definitively document the depressed LT responses during human pregnancy.

New concepts of obstetric and gynecologic infection.

Multiple organisms are usually recovered from patients with acute pelvic inflammatory disease (PID). The sexually transmitted organisms Neisseria gonorrhoeae, Chlamydia trachomatis, and genital mycoplasmas frequently initiate infection. Although aerobic and anaerobic bacteria that constitute the normal vaginal flora are frequent secondary invaders following an initial sexually transmitted infection, these bacteria can also primarily cause PID. Early postpartum endometritis, the most common postpartum infection, amniotic fluid infection, and infection following hysterectomy are usually caused by the aerobic and anaerobic bacteria that constitute the normal vaginal flora. Because the organisms that cause the various obstetric-gynecologic infections are often found in combination, it is necessary to begin therapy with broad-spectrum antibiotics. Many of these infections are caused by anaerobic bacteria and it is often necessary to select antibiotics that inhibit anaerobes.

Nonspecific vaginitis. Diagnostic criteria and microbial and epidemiologic associations.

Numerous previous studies of nonspecific vaginitis have yielded contradictory results regarding its cause and clinical manifestations, due to a lack of uniform case definition and laboratory methods. We studied 397 consecutive unselected female university students and applied sets of well defined criteria to distinguish nonspecific vaginitis from other forms of vaginitis and from normal findings. Using such criteria, we diagnosed nonspecific vaginitis in up to 25 percent of our study population; asymptomatic disease was recognized in more than 50 percent of those with nonspecific vaginitis. A clinical diagnosis of nonspecific vaginitis, based on simple office procedures, was correlated with both the presence and the concentration of Gardnerella vaginalis (Hemophilus vaginalis) in vaginal discharge, and with characteristic biochemical findings in vaginal discharge. Nonspecific vaginitis was also correlated with a history of sexual activity, a history of previous trichomoniasis, current use of nonbarrier contraceptive methods, and, particularly, use of an intrauterine device. G. vaginalis was isolated from 51.3 percent of the total population using a highly selective medium that detected the organism in lower concentration in vaginal discharge than did previously used media. Practical diagnostic criteria for standard clinical use are proposed. Application of such criteria should assist in clinical management of nonspecific vaginitis and in further study of the microbiologic and biochemical correlates and the pathogenesis of this mild but quite prevalent disease.

Endometrial histopathology in patients with culture-proved upper genital tract infection and laparoscopically diagnosed acute salpingitis.

To define and quantitate histologic changes in the endometrium that best correlate with documented upper genital tract infection (UGTI) and laparoscopically diagnosed acute salpingitis, we studied endometrial biopsy specimens from 69 consecutive patients with clinically suspected acute pelvic inflammatory disease (PID) who underwent microbiological evaluation for UGTI and laparoscopic examination for acute salpingitis. Both UGTI and acute laparoscopically confirmed salpingitis were present in 37 patients (54%), UGTI without salpingitis in 1 (1%), salpingitis without UGTI in 11 (16%), and neither UGTI nor salpingitis in 20 (29%). Chlamydia trachomatis or Neisseria gonorrhoeae UGTI was found in 34 women, Escherichia coli in two patients, Peptococcus magnus in one woman, and with Streptococcus agalactiae in one woman. The following features were correlated both with UGTI and with salpingitis: presence of any neutrophils in the endometrial surface epithelium; neutrophils within gland lumens; dense subepithelial stromal lymphocytic infiltration; any stromal plasma cells; and germinal centers containing transformed lymphocytes. The simultaneous presence of five or more neutrophils per X 400 field in endometrial surface epithelium, together with one or more plasma cell per X 120 field in endometrial stroma, was the best predictor of UGTI plus salpingitis. This combination had a sensitivity of 92% and a specificity of 87% for predicting the diagnosis of both UGTI and laparoscopically confirmable acute salpingitis. Prospective studies are needed to assess the usefulness of these criteria.

The severity of fetal membrane infection and pregnancy complications.

The purpose of this report is to examine the association of histologic chorioamnionitis with microorganisms isolated from the fetal membranes and to evaluate whether microorganisms with or without inflammation are associated with labor characteristics and with pregnancy complications. Inflammation was more common among membranes that yielded pathogenic bacteria (47%, P = 0.002) or Urea-plasma urealyticum (34%, P = 0.03) than among membranes that yielded no growth or nonpathogenic bacteria (20%). Prolonged membrane rupture (P = 0.0001), infant birth weight less than 2500 g (P = 0.02), and intraamniotic infection (P = 0.001) occurred more frequently among those women whose membranes yielded pathogenic bacteria than among those whose membranes yielded no growth or nonpathogenic bacteria. Our findings suggest that placental membranes in which pathogenic bacteria are accompanied by inflammation are associated with the highest risk of pregnancy complications.

Histopathology of endocervical infection caused by Chlamydia trachomatis, herpes simplex virus, Trichomonas vaginalis, and Neisseria gonorrhoeae.

We determined the histologic correlates of clinically identified mucopurulent cervicitis, culture-proven cervical infection with Chlamydia trachomatis, Neisseria gonorrhoeae, herpes simplex virus (HSV), and vaginal infection with Trichomonas vaginalis by examining cervical biopsies from 83 women. Clinical mucopurulent cervicitis and culture-documented infection with one or more of these pathogens correlated histologically with intraepithelial neutrophils, reactive endocervical cells, edema, luminal neutrophils, and with several deeper tissue changes such as extensive and dense subepithelial inflammation, granulation tissue, and necrotic ulceration. Focal loss of surface columnar cells and spongiosis were also correlated with culture-confirmed infection. Well-formed germinal centers were seen in biopsies from 14 of 21 patients (67%) with C trachomatis infection alone, but in none of 17 patients with infections other than C trachomatis (P less than 0.001). A predominantly plasmacytic infiltrate was also significantly associated with chlamydial infection. Necrotic ulcers overlying a predominantly lymphocytic infiltrate were seen in six of nine patients (67%) with HSV infection alone but in only two of 40 patients (5%) with other infections (P less than 0.001). Marked inflammatory changes were not seen in the patients infected with N gonorrhoeae. The organism T vaginalis was not associated with any endocervical pathology. If these results are confirmed by prospective studies, they suggest that pathologists should alert clinicians to the possibility of recent or current infection with C trachomatis or HSV when cervical biopsies show the above changes. The loss of surface columnar epithelium with HSV, chlamydial, and gonococcal infection offers a possible explanation for the reported association of these infections with increased risk of acquiring human immunodeficiency virus infection.

Epidemiology and diagnosis of acute pelvic inflammatory disease.

Epidemiologic factors important in acute pelvic inflammatory disease (PID) are the pathogenic organisms Neisseria gonorrhoeae and Chlamydia trachomatis, less pathogenic mycoplasma, and endogenous aerobic and anaerobic bacteria. Other risk factors for PID include the number of sexual partners, previous PID, previous gonorrhea, and male gonorrhea. The patient's age, IUD use, and social factors may be dependent variables for the development of PID. The diagnosis is often difficult to establish; practical diagnostic methods include a careful history and physical examination, laboratory tests (particularly a cervical Gram stain), culdocentesis, and examination of the male. Laparoscopy should be used when the diagnosis is unclear.

Fatal perineal cellulitis from an episiotomy site.

Perineal cellulitis originating from an episiotomy incision resulted in 20% of the maternal mortality in King County, Washington, between 1969 and 1977. Necrotizing fasciitis was present in 2 of the cases, and clostridial myonecrosis was present in 1. These fatalities occurred because the practitioners were not aware that necrotizing fasciitis can occur in the fatty superficial fascia of the perineum and that resection of the necrotic tissue is necessary for successful therapy.

Recurrent and persistent vulvovaginal candidiasis: treatment with ketoconazole.

Forty-two women were administered ketoconazole, an orally absorbed antifungal agent. The symptoms of vulvar pruritus, burning, pain, and dysuria and the signs of vulvar erythema and fissures and vaginal erythema were associated with Candida albicans. Although ketoconazole relieved symptoms and signs seven days after therapy, a high recurrence rate occurred by 28 days after therapy.

Microbiology and treatment of late postpartum endometritis.

Eighteen vaginally delivered women who developed late postpartum endometritis seven to 42 days after delivery were prospectively studied. Multiple microorganisms were recovered from the endometrium, including Chlamydia trachomatis, genital mycoplasmas, and a wide variety of bacteria. Erythromycin therapy was successful in ten of the 13 women who were followed.

Early postpartum endometritis: the role of bacteria, genital mycoplasmas, and Chlamydia trachomatis.

To characterize the flora of early postpartum endometritis and the clinical features of women with specific organisms, endometrial cultures for facultative and anaerobic bacteria, genital mycoplasmas, and Chlamydia trachomatis were taken with a triple-lumen sampling device. More than one organism was recovered from 80% of the women. Over 60% of the women had Gardnerella vaginalis and/or anaerobes associated with bacterial vaginosis isolated from the endometrium; these women were more likely to have severe illness and to develop a wound infection than were other women. Genital mycoplasmas were isolated frequently, but specific antibiotic therapy was not required for clinical cure in the 10% of patients who had Ureaplasma urealyticum only. Chlamydia trachomatis was infrequently isolated, but C trachomatis commonly remained after therapy.

Development of wound infections among women undergoing cesarean section.

Sixty consecutive wound infections were studied among 1104 women undergoing cesarean section. Wound infections caused by cervical-vaginal flora were associated with prolonged labor, particularly with greater duration of fetal monitoring and number of vaginal examinations, and with organisms isolated from the endometrium at cesarean section. In contrast, women with wound infections caused by Staphylococcus aureus had neither prolonged labor nor S aureus isolated at cesarean section. The 25% of wound infections associated with S aureus represent potentially preventable conditions that presumably arise from exogenous sources.

Preterm labor associated with subclinical amniotic fluid infection and with bacterial vaginosis.

Maternal genital infection, particularly subclinical amniotic fluid infection, may cause preterm labor and a premature delivery. The prevalence of subclinical amniotic fluid infection was studied in 54 consecutive afebrile women in preterm labor with singleton gestations and intact fetal membranes. Microorganisms were recovered from the amniotic fluid by transabdominal amniocentesis in 13 (24%) of 54 patients. Bacteria or Candida albicans were recovered from six (11%), and genital mycoplasmas from seven (13%). Compared with women with sterile amniotic fluid, patients whose amniotic fluid contained bacteria or Candida organisms had a shorter interval from onset of preterm labor until delivery (0.6 versus 34.3 days, P less than .01), were less responsive to tocolytic therapy (0 versus 81% success rate, P less than .005), and more frequently developed subsequent intrapartum fever (83 versus 2.4%, P less than .005). In contrast, women whose amniotic fluid contained genital mycoplasmas did not differ in these parameters from those with sterile fluid. Also compared was cervical-vaginal infection among these patients in preterm labor with matched control subjects without preterm labor. In this analysis, bacterial vaginosis was identified in 43% of patients with and 14% of women without preterm labor (P = .02), yielding a relative risk of preterm labor for patients with bacterial vaginosis of 3.8. These data underscore the importance of amniotic fluid bacterial infections in preterm labor and premature delivery, and suggest that bacterial vaginosis is associated with prematurity.

Upper genital tract isolates at delivery as predictors of post-cesarean infections among women receiving antibiotic prophylaxis.

The introduction of antibiotic prophylaxis for cesarean delivery has decreased the risk of postpartum endometritis and wound infection, but factors that contribute to prophylaxis failure are not understood. To determine factors that might contribute to postpartum infections following antibiotic prophylaxis, we cultured amniotic fluid, decidua, and chorioamniotic membrane specimens for anaerobic and facultative bacteria and for genital mycoplasmas at cesarean delivery. Women were assessed daily for the development of infections, and if endometritis developed, a protected endometrial culture was obtained. Postpartum endometritis developed in 16 and wound infection in four of 102 women. Infection rates were similar for women receiving cefotetan (N = 50) or cefoxitin (N = 52) for prophylaxis. The isolation of group B streptococcus (P less than .001) or Enterococcus faecalis (P = .03) from the upper genital tract at delivery was significantly associated with postpartum endometritis. Antibiotic-resistant organisms (other than enterococci) were recovered uncommonly at delivery or with postpartum infections. Group B streptococcus was susceptible to the prophylactic agents used, suggesting that virulence factors other than antibiotic resistance are important for the development of postpartum endometritis. Group B streptococcus, E faecalis, and bacteria associated with bacterial vaginosis were recovered from the endometrium at the time of postpartum endometritis.

Evaluation of a rapid enzyme immunoassay test for detection of group B Streptococcus.

The purpose of this study was to determine the reliability of the ICON immunoassay in detecting vaginal group B Streptococcus in pregnant women. Vaginal cultures were obtained at the time of sterile speculum examinations from 300 laboring women. The frequency of vaginal group B streptococcal colonization was 10%. A duplicate vaginal swab was used in the immunoassay, which required less than 10 minutes for completion. The immunoassay could reproducibly detect 3+ to 4+ growth (greater than 10(5) colony-forming units per milliliter of vaginal fluid), which was considered heavy colonization. The overall sensitivity, specificity, positive predictive value, and negative predictive value were 33, 95, 43, and 93%, respectively, compared with culture. For heavy colonization (3+ to 4+ growth), the sensitivity, specificity, positive predictive value, and negative predictive value were 100, 95, 32, and 100%, respectively. The immunoassay is a highly sensitive and specific test for detection of heavy vaginal colonization by group B streptococci in obstetric patients, but cultures are required to detect light group B streptococcal colonization.