THE JOURNAL OF INFECTIOUS DISEASES. VOL. 135, NO.2. FEBRUARY 1977 1977 by the University of Chicago. All rights reserved.

©

Rectal Colonization with Group B Streptococcus: Relation to Vaginal Colonization of Pregnant Women From the Departments of Pediatrics, Obstetrics and Gynecology, and Clinical Pathology, University of Florida College of Medicine, Gainesville, Florida

Mohamad S. Badri, Sam Zawaneh, Amelia C. Cruz, Gonzalo Mantilla, Herman Baer, William N. Spellacy, and Elia M. Ayoub

higher rate of colonization of the anorectal site than of any of the other sites from which specimens were cultured. This finding suggests that the gastrointestinal tract may playa role in seeding of the genital tract in the female and in transmission of group B streptococci to the neonate.

Group B Streptococcus has emerged over the past decade as a common pathogen in the etiology of neonatal sepsis [1, 2]. Previous studies on the epidemiology of infections due to the group B Streptococcus have shown a high rate of colonization of the female genitourinary tract by this organism. Therefore, most investigators implicated the vaginal canal as the primary source of infection for neonates. Colonization of the anorectal area in women by the group B Streptococcus was reported by Franciosi et al. [3]. These investigators suggested that the presence of group B streptococci at that site represented contamination from the genital tract. Because the majority of nonstreptococcal agents associated with neonatal sepsis are derived from the gastrointestinal flora, we undertook the present study to assess the possible role of anorectal colonization in seeding of the genitourinary tract. Our studies revealed an impressively

Materials and Methods

Our study was performed in two different periods. The first period was from April to June of 1974 and included 260 pregnant women and 45 of their neonates. The second period was from October to December of 1975 and included 529 pregnant women and 377 neonates. The women were part of a group of women in the second and third trimesters of pregnancy who attended the prenatal clinic or were admitted for delivery at the Shands Teaching Hospital of the University Medical Center (Gainesville, Fla.). The mean age of these women was 21 years, and the racial distribution of women in the two groups was approximately equal. Specimens for culture were obtained in the clinic from 71 women during the second trimester of gestation, and specimens were subsequently obtained from these women during the third trimester or at the time of delivery. A woman was labeled as positive when any of the cultures yielded group B streptococci.

Received for publication April 26, 1976, and in revised form July 15, 1976. This work was supported in part by grants no. AI 09645 and AI 100341 from the National Institutes of Health. We acknowledge the technical assistance of Mr. Richard Schelm. Please address requests for reprints to Dr. Elia M, Ayoub, Department of Pediatrics, University of Florida College of Medicine, Gainesville, Florida 32610.

308

Downloaded from http://jid.oxfordjournals.org/ at Memorial University of Newfoundland on October 9, 2014

Anorect~l carriage as a possible primary source of vaginal colonization by group B Streptococcus was investigated. The study was performed during two separate periods and included 789 pregnant women and 422 neonates. Specimens from multiple sites were obtained for culture from all women and infants and were streaked onto blood agar plates containing 8 ILg of gentamicin sulfate /ml and 15 ILg of nalidixic acidj ml, which allow selective growth of streptococci. Cultures positive for group B streptococci were obtained from 162 (20.5%) of the pregnant women and from 50 (11.8%) of the neonates. Rectal cultures were positive for streptococci in 142 (17.9%) of the women, and vaginal cultures gave positive results in 81 (10.2%). The higher incidence of positive results in rectal as opposed to vaginal cultures (ratio of 2: I) was encountered during all phases of the study. This finding suggests that the gastrointestinal tract may be the primary site of colonization by group B Streptococcus and that vaginal colonization may represent contamination from this source.

Group B Streptococci of Pregnant T'f1ornell

309

Table 1.

Frequency of colonization with group B

Streptococcus in pregnant women and newborn infants. No. positive for group B streptococci (%) during Source of specimens

Study no. 1 *

Pregnant women No. cultured 260 No. with positive cultures 56 (21.5) Vaginal 27 (10.3) Rectal 46 (17.7) Newborn infants No. cultured 45 No. with positive cultures 8(17.4) Ear canal 3 (6.5) Throat 3 (6.5) Umbilical 7 (15.2) Rectal 7(15.2)

Study no.2t

Total no. positive (%)

529

789

106 (20.0) 54 (10.2) 96 (18.1)

162 (20.5) 81 (10.2) 142(17.9}

377

422

42 12 16 15 36

(11.1) (3.2) (4.2) (4.0) (9.5)

50 15 19 22 43

by Dr. Rebecca Lancefield, Rockefeller University, New York, N. Y. The selective medium consisted of a modification of that described by Vincent et al. [5]. Blood agar base (Difco, Detroit, Mich.j was autoclaved and cooled to 45 C. Defibrinated sheep's blood was then added to bring the concentration to 5%, and gentamicin sulfate and nalidixic acid were added next. The final concentrations of these antibiotics in the medium were 8 p,g of gentamicin sulfate /rnl and 15 p,g of nalidixic acid/ m!' Preliminary studies were performed for comparison of the efficacy of the selective medium for isolation of group B streptococcal colonies. A total of 468 duplicate cultures were made from samples from the 305 individuals involved in the first period of study. One specimen was plated onto regular blood agar, and the other was streaked onto a selective medium plate. 13hemolytic colonies, subsequently confirmed to be group B streptococci, were present in 78 rectal and 45 vaginal cultures. Of the 78 rectal cultures positive for group B streptococci, 76 yielded positive results on the selective medium, and 72 gave positive results on regular medium. Similarly, 44 of the 45 vaginal cultures were positive for group B streptococci on the selective medium, and only 39 gave positive results on the nonselective medium. Thus, use of the selective medium resulted in identification of a greater number of positive cultures. In addition, cultures performed on the selective medium yielded, in the majority of instances, a pure growth of 13hemolytic organisms. This pure growth contrasted with the overgrowth by coliforms that predominated in cultures on nonselective medium and made the isolation of f3-hemolytic streptococci difficult. For this reason, only the selective medium was used in the second period.

Results (11.8) (3.5) (4.5) (5.2) (10.1)

"Study no. 1 was done from April to June of 1974. tStudy no. 2 was done from October to December of 1975.

Group B streptococci were present almost twice as frequently in rectal as in vaginal cultures of samples obtained from pregnant women (table I). Among a total of 789 women from whom specimens were obtained during both study periods, 162 (20.5%) yielded cultures positive for group B streptococci on one or more occasions.

Downloaded from http://jid.oxfordjournals.org/ at Memorial University of Newfoundland on October 9, 2014

Vaginal and rectal specimens were procured from the women in the clinic, on admission to the labor area, or just prior to delivery. Vaginal or rectal samples were obtained by sterile cottontipped swabs inserted 1-2 inches into the vaginal canal or beyond the anal orifice, respectively. Throat specimens were obtained from pregnant womell during the initial part of the study. Because of the low yield of cultures positive for group B streptococci, throat cultures were not continued during the second part. Within 12 hr after admission of infants to the nursery, specimens for culture were obtained from ear canals, throat, umbilicus, and rectum, without prior preparation of the sites. All swabs for culture were first placed in a liquid medium that consisted of nutrient broth and were subsequently plated onto regular blood agar plates as well as onto selective media plates (see below). The plates were then incubated for 18 hr at 37 C. Representative f3-hemolytic colonies were picked from each plate and identified serologically as to group-specific and serotype reactivity by the Lancefield capillary precipitin technique [4]. Group-specific antiserum was furnished by the Center for Disease Control, Atlanta, Ga.; type-specific antisera were supplied

Badri et al.

310

Table 2.

four individuals. This yield represented a carriage rate of 1.5%. Throat cultures were not done during the second period. Results of cultures obtained from the 422 infants are also summarized in table 1. Fifty (11.8~~) yielded cultures positive for group B streptococci. Positive results from cultured specimens of the ear canal, throat, and umbilicus were obtained in 15, 19, and 22 neonates, respectively; rectal cultures were positive in 43 (10.1%) of the infants. A somewhat higher yield of positive cultures was obtained in infants seen during the first period. This higher incidence of positive results may be ascribed to a greater emphasis during the first study on procuring cultures from infants horn to mothers with known positive cultures. The serotype distribution of isolates of group B streptococci from mothers and neonates is shown in table 2. Types II and III predominated in the isolates from mothers and neonates; type I was isolated from >- 19% of all individuals cultured; and types II and III were isolated with about equal frequency from both mothers and infants. No significant difference in frequency of isolation of the three serotypes was found in specimens obtained from various sites on mothers, although type II was isolated more frequently from anorectal cultures than from vaginal cultures. Discussion

Suspicion that the gastrointestinal tract might playa primary role in the spread of group B Streptococcus was aroused by the fact that the majority of organisms associated with nonstreptococca1 sepsis are derived from the intestinal

Distribution of serotypes of isolates of group B Streptococcus from mothers and neonates. Infants Mothers

Serotype Ia Ib

Ie II III Total

No. with serotype

17 5 1 57 49 129

No. of isolates

No. of isolates Vaginal

Anorectal

No. with serotype

9 4 0 19 26 68

10 4 1 47 33 95

8 0 0 11 14 33

Throat

0 0 0 5 4 9

Anorectal

5 0 0 8 8 21

Umbilical

4 0 0 9 7 20

External ear canal

2 0 0 3 3 8

Total no. with serotype (%)

25 (15) 5 (3) 1 (0.6) 68 (42) 63 (39) 162

Downloaded from http://jid.oxfordjournals.org/ at Memorial University of Newfoundland on October 9, 2014

Vaginal cultures were pOSItIVe in 81 (10.2%) of these women, whereas rectal cultures were positive in 142 (17.9%). As can be seen in table 1, the percentages of vaginal and rectal cultures positive for group B streptococci in pregnant women were similar during the two study periods. Cultures were obtained for a second time from 71 women (during the third trimester of pregnancy). Cultures performed in the second trimester of pregnancy had been positive for group B streptococci for 15 of these women. All 15 women had had positive results in rectal cultures, and eight had had positive results in vaginal cultures. On repeat culture during the third trimester, specimens from 25 of the 71 women gave positive results, a finding which indicated a change in the carriage rate of from 21% during the second trimester to 35% in the third trimester. All 25 of these women had rectal cultures positive for group B streptococci, and 12 yielded positive vaginal cultures. Thus the increases in positive results from the second trimester to the third were from 21% to 35% for rectal cultures and from 11% to 17% for vaginal cultures. ~~-Ac:omparison of the relative frequency of positive rectal and vaginal cultures from group B streptococci revealed that, of the total of 142 women with positive results in rectal cultures, only 61 had positive results in vaginal cultures. In contrast, only 20 positive vaginal cultures were obtained from women who had negative rectal cultures. Thus, vaginal cultures were negative in 50% of 162 colonized women, whereas rectal cultures were negative in 12%. Cultures of throat specimens obtained from 260 women seen during the first period yielded positive cultures for group B streptococci in only

Group B Streptococci of Pregnant Women

ported differences in frequency of vaginal colonization. Despite these differences, the point to be emphasized is that a ratio of rectal vs. vaginal colonization of 2: 1 was obtained during both phases of our study, whether selective or nonselective medium was used. This ratio was maintained even when cultures were performed on 71 women during the second and third trimesters of pregnancy. Although the rate of vaginal colonization increased from 11% to 17%, a parallel increase in the rate of positive results in rectal cultures was encountered, with a rise from 21% to 35%. Thus, the 2: 1 ratio of positive rectal vs. positive vaginal cultures remained constant. These findings strongly suggest that the gastrointestinal tract may be the primary site for group B streptococcal colonization and that vaginal colonization reflects contamination from this source. The distribution of serotypes among isolates from pregnant women and neonates was similar; serotypes II and III predominated, whereas serotypes la, Ib, and Ie were less frequently isolated. This finding is analogous to that reported in other studies [11-13]. Not included in our data are the results from serotyping of 12 neonates admitted to this hospital with group B streptococcal sepsis. Ten of the strains recovered from these patients were serotype III, and two were serotype I. If serotype II is indeed as prevalent as type III in all sites cultured in pregnant women and healthy neonates, as has been shown in our study as well as in other studies [11-13], then the reason for the relative paucity of isolations of serotype II from patients with neonatal sepsis in these studies remains to be clarified. The high rate of recovery of group B Streptococcus from anorectal specimens cultured appears to elucidate as well as to raise questions regarding some aspects of neonatal sepsis due to this organism. Use of only vaginal specimens for culture would result in missing 50% of the potential carriers of the organism. If screening of neonates, pregnant women, nursing mothers, or nursery personnel for group B streptococci is to be carried out, cultures of rectal samples should be included. Second, although intrauterine and perinatal disease can reasonably be ascribed to acquisition of the organism from the mother's genital tract, this site is less likely to be the source

Downloaded from http://jid.oxfordjournals.org/ at Memorial University of Newfoundland on October 9, 2014

flora. Support for this suspicion was derived from the studies of Franciosi et al. [3], who found group B Streptococcus in cultures of anorectal specimens. The percentage of positive results in rectal cultures obtained by Franciosi et al. was 7.8 as compared with 4.6 for vaginal cultures. Despite the higher rate of positive results in rectal cultures, these investigators ascribed the positive rectal cultures to perineal skin contamination from the vagina. The possibility that the reverse was true, that contamination of the vagina was derived from a perineal source, was further suggested by the observation of Kexel and Beck [6], who found that 12.4% of samples cultured from the vulva were positive as opposed to 9.5% from the vagina and 6.7% from the cervix. These findings indicated that the lower rate of colonization in the genital tract was at the site furthest away from the perianal area. Our studies revealed a high rate of colonization by group B streptococci of the anorectal site in women and showed a consistently higher recovery of this organism from the anorectal site than from any other site studied. In addition, our preliminary observation of a limited number of nonpregnant females and of males disclosed a similar rate of rectal colonization (authors' unpublished observations). Although these findings incriminate the gastrointestinal tract as the probable reservoir for group B streptococci in humans, other factors that may contribute to the higher frequency of the microorganisms in rectal cultures should be considered. These factors include the possibility that group B streptococci prevail more at the anorectal junction than at the distal vagina because of differences in pH, local secretions, and nutrient factors that could enhance the growth of the organism in the anorectal area or inhibit its growth in the distal vagina. The frequency of recovery of group B streptococci from vaginal cultures in previous studies has ranged from 5% to 29% [6-11]. A rate of 10.2% for vaginal colonization obtained in our study is similar to that encountered in most of these earlier studies. Differences in culture technique and variations in the time of gestation at which the cultures were obtained [10, 11], together with seasonal [12] and geographic [13] factors, may have contributed to the previously re-

311

Badri et at.

312

7.

8.

9.

10.

11.

12. References 1. Eickoff, T. C., Klein, J. 0., Daly, A. K., Ingall, D., Finland, M. Neonatal sepsis and other infections due to group B beta-hemolytic streptococci. N. Engl. J. Med. 271:1221-1228,1964. 2. McCracken, G, H. Group B streptococci: the new challenge in neonatal infections. J. Pediatr. 82:703-706, 1973. 3. Franciosi, R. A., Knostman, J. D., Zimmerman, R. A. Group B streptococcal neonatal and infant infections. J. Pediatr. 82:707-718,1973. 4. Lancefield, R. C. A serological differentiation of human and other groups of hemolytic streptococci. J. Exp. Med. 57:571-595,1933. 5. Vincent, W. F., Gibbons, W. E., Gaafar, H. A. Selective medium for the isolation of streptococci from clinical specimens. Appl. Microbiol. 22:942-943, 1971. 6. Kexel, G., Beck, K. J. Untersuchungen tiber die Haufig-

13.

14.

15.

16.

keit del' B. Streptokokken im Wochcnbett. Geburtsh Frauenkeilk 25:1078-1085, 1965. Hood, M., Janney, A., Dameron, G. Beta hemolytic streptococcus group B associated with problems of the perinatal period. Am. J. Obstet. 82:809-818, 1961. Bergqvist, G., Hurvell, B., ThaI, E., Vaclavinkova, V. Neonatal infections caused by group B streptococci: relation between the occurrence in the vaginal flora of term pregnant women and infection in the newborn infant. Scand. J. Infect. Dis. 3:209-212, 1971. Bergqvist, G., Hurvell, B., Malmborg, A., Rylander, M., Tunell, R. Neonatal infections caused by group B streptococci. Scand. J. Infect. Dis. 3:157-162,1971. Abel', R. C., Facklam, R. R., Wilkinson, H. W., Wor· tham, R. C. Nosocomial group B streptococcal infections [abstract no. 33]. In Program and Abstracts of the 13th Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, D. C., September 19-21, 1973. American Society for Microbiology, Washington,D. C., 1973. Baker, C. J., Barrett, F. F., Yow, M. D. The influence of advancing gestation on group B streptococcal colonization in pregnant women. Am. J. Obstet. Gynecol. 122:820-823, 1975. Anthony, B. F., Okada, D. M., HobaI, R.C. Epidemiology of group B streptococci in pregnancy and the newborn. Clin. Res. 24:182A, 1976. Wilkinson, H. W., Facklam, R. R., Wortham, E. C. Distribution of serological type of group B streptococci isolated from a variety of clinical material over a five-year period (with special reference to neonatal sepsis and meningitis). Infee. Immun. 8:228235, 1973. Maher, E., Irwin, R. C. Group B streptococcal infection in infancy: a case report and review. Pediatrics 38:659-662,1966. Paredes, A., Wong, P., Mason, E. 0., Yow, M. D. Failure of penicillin to eradicate the carrier state of group B streptococci. Clin. Res. 24:71A, 1976. Hall, R. T., Barnes, W., Krishnan, L., Harris, D. T., Rhodes, P. G., Fayez, J., Miller, G. L. Antibiotic treatment of parturient women colonized with group B streptococci. Am. J. Obstet. Gyneco!. 124:630-634, 1976.

Downloaded from http://jid.oxfordjournals.org/ at Memorial University of Newfoundland on October 9, 2014

of infections acquired after delivery. The gastrointestinal tract would be a more likely source of contamination and transmission than the genital tract in cases where nosocomial infection is suspected. Finally, treatment with antibiotics for maternal carriers has been suggested as a prophylactic measure for the prevention of neonatal sepsis due to group B St-reptococcus [7, 9, 14]. The difficulty of complete eradication of any organism from the gastrointestinal tract raises serious questions with regard to the soundness of such an approach. This concern is supported by two recent studies [15, 16J that demonstrate the failure of antibiotics to reduce effectively carriage of group B streptococci in mothers and infants. In light of these findings, other methods will have to be developed if we are to reduce the incidence of this disease in newborn infants.