Incidence and Epidemiology of Streptococcus pseudoporcinus in the Genital Tract

Abstract

Streptococcus pseudoporcinus, a beta-hemolytic microorganism first isolated from the female gastrourinary tract in 2006, cross-reacts with serogrouping kits for group B Streptococcus (GBS) and could be misidentified in the laboratory. The epidemiologic characteristics of this species have not been reported previously, but this organism is thought to be rare. Paired vaginal and rectal samples were collected from 663 nonpregnant women enrolled in a phase II clinical vaccine trial of a GBS type III capsular polysaccharide-protein conjugate vaccine, and isolates initially identified as S. pseudoporcinus were collected for further testing. A total of 120 isolates of S. pseudoporcinus were recovered from 36 unique individuals with 5.4% of 663 women having this organism recovered at least once during follow-up. All of these isolates cross-reacted with a commercially available GBS serogrouping kit. Women colonized with isolates confirmed as S. pseudoporcinus by genotypic and phenotypic methodologies were compared to women who were not colonized to determine whether there were any significant factors associated with acquisition of S. pseudoporcinus. Acquisition of S. pseudoporcinus vaginally and/or rectally was 36 per 846.0 women-years of follow-up for an annual incidence of 4 per 100 woman-years of follow-up. Acquisition of S. pseudoporcinus was independently associated with black women, being 30 to 40 years of age, recent Trichomonas vaginalis infection, primary or recurrent genital herpes, having bacterial vaginosis by Nugent criteria, and having had two or more male sexual partners since the last visit. This study suggests that S. pseudoporcinus is not rare, especially among black women, and could be misidentified as GBS.

INTRODUCTION

Streptococcus pseudoporcinus is a facultative Gram-positive cocci usually characterized by a large zone of beta-hemolysis. It was initially thought to be Streptococcus porcinus, an organism in serological groups E, P, U, and V and found in the upper respiratory and genital tracts of swine (3), but in 2005, Duarte et al. demonstrated that strains of S. porcinus isolated from humans had unique pulsed-field gel electrophoresis profiles of the chromosomal DNA compared to the nonhuman strains and the majority of the human isolates belonged to serogroup NG1 (5). In 2006, Bekal et al. using 16S rRNA gene sequencing demonstrated that the organism found in humans had biochemical characteristics similar to those of S. porcinus but were genetically unique and proposed the name Streptococcus pseudoporcinus (1). Thompson and Facklam from the Centers for Disease Control and Prevention tested human strains of S. porcinus (serogroup NG1) with the group B reagents of 12 commercial Streptococcus test kits and reported that all of the kits evaluated cross-reacted with one or more of the isolates (17). Because S. pseudoporcinus cross-reacts with serogrouping kits for group B Streptococcus (GBS), S. pseudoporcinus may be misidentified as GBS in clinical settings.

S. pseudoporcinus has been isolated from multiple specimen types, including blood specimens, wounds, urine, skin, vaginal, rectal, cervical, and placenta specimens (1, 7, 10, 14). Currently, little is known about the epidemiology of this organism or its role in human disease.

The objectives of our study were to determine the prevalence and incidence of S. pseudoporcinus among sexually active nonpregnant women of reproductive age and to characterize the factors associated with the acquisition of this organism.

(A portion of this work was presented at the 35th Infectious Disease Society for Obstetrics and Gynecology annual meeting, Seattle, WA, 14 to 16 August 2008.)

MATERIALS AND METHODS

Study population.

Nonpregnant women who were 18 to 40 years old, who were sexually active and using effective birth control methods, and who intended to stay in the geographical area for the next 18 months were enrolled in a phase II clinical trial of a GBS type III capsular polysaccharide vaccine in Pittsburgh, PA, in Augusta, GA, and in Houston, TX. All women provided informed written consent. Women were excluded from the study if they were pregnant, had current vaginal bleeding, reported current use of systemic antimicrobials, or were colonized with GBS or S. pseudoporcinus in either the vagina or rectum at the screening visit, had received any vaccine in the previous 30 days, or planned to receive any vaccine in the next 30 days, had received the tetanus toxoid vaccine in the last 12 months, were nursing, or postmenopausal. Paired vaginal and rectal swabs, vaginal smears, and demographic and behavioral interviews were obtained by study personnel at the enrollment visit, as well as at 1 month, 2 months, and bimonthly for 18 months following enrollment.

Laboratory methods.

Vaginal and rectal swabs, collected in separate Amies transport medium (MML Diagnostics, Troutdale, OR), and smears were delivered to the Magee-Womens Research Institute in Pittsburgh, PA, for processing. The smears were Gram stained and the vaginal flora were evaluated according to the Nugent criteria (16). Upon receipt in the laboratory, the swabs were inoculated onto Columbia agar with 5% sheep blood, supplemented with colistin and nalidixic acid (CNA) (bioMérieux SA, Durham, NC) and into selective group B Strep broth (bioMérieux SA). The CNA plates were streaked into four quadrants, and 30-μg neomycin Sensi-Discs (Becton-Dickinson and Co., Sparks, MD) were placed in the first two quadrants to suppress the growth of susceptible organisms. CNA plates and the selective broth were incubated for 48 h at 37°C and 5 to 7% CO₂. After incubation, the group B strep broth was inoculated onto CNA agar, streaked for isolation, and incubated as described above to detect low levels of GBS. GBS and Lactobacillus were isolated and identified by methods that have been previously described (15). Catalase-negative, Gram-positive cocci with wide zones of beta-hemolysis that reacted with group B latex agglutination (PathoDx; Remel, Lenexa, KS) were further tested with API20Strep tests (bioMérieux SA). The first 500-bp portion of the 16S rRNA gene of representative isolates was sequenced. The sequences were compared for similarity with those found in GenBank using the Basic Local Alignment Search Tool (BLAST) 2.2.18.

Two clinical isolates of S. pseudoporcinus and GBS were grown on ChromID Strepto B agar (bioMérieux SA) to compare pigmentation.

Statistical analyses.

Statistical analyses were performed using Stata statistical software release 9.1 (Stata Corp., College Station, TX), and statistical tests were evaluated at the 0.05 two-sided significance level. Acquisition of S. pseudoporcinus was defined as the first positive vaginal and/or rectal culture following the culture-negative enrollment visit. Cox proportional-hazard models with time-varying covariates were used to evaluate the factors associated with acquisition of S. pseudoporcinus. The Efron method was used for handling tied failure times (6), and the variance estimates were calculated by the method of Lin and Wei (13). Variables were considered for inclusion in these models if the P value from the log-rank test for equality of survivor functions was less than 0.2. Forward stepwise regression was employed, and variables were retained in the model if the P value from the Wald chi-squared test statistic was 0.05 or less.

RESULTS

A total of 120 isolates of S. pseudoporcinus were isolated from vaginal and rectal swab samples obtained from 36 women at multiple visits. While most isolates had distinctly larger zones of beta-hemolysis on Columbia sheep blood agar compared to the zones observed with GBS, some isolates had hemolysis similar to that of GBS after 24 h of growth. However, when grown on ChromID Strepto B agar, the colonies of S. pseudoporcinus were indistinguishable from those of GBS. The API profile identified the isolates as S. porcinus, but the 500-bp sequences had only 94% similarity to S. porcinus, ATCC strain 42138 (American Type Culture Collection, Manassas, VA) with a 30-bp discordance. Our strains exhibited 100% sequence similarity to S. pseudoporcinus compared to the deposited sequences of S. pseudoporcinus in GenBank.

S. pseudoporcinus was recovered from the vagina and/or rectum from 36 women, giving a population prevalence of 5.4%. Demographic data from the 663 women enrolled in the study can be seen in Table 1. The majority of the women, 24 (67%), were colonized with S. pseudoporcinus at multiple visits, and the remaining 12 (33%) at a single visit. A total of 18 (50%) of the women were cocolonized in the vagina and the rectum with S. pseudoporcinus at the same visit. There were 120 isolates recovered from 90 visits with 67 (56%) from the vagina, 49 (41%) from the rectum, and 4 (3%) with no location recorded. Isolates were recovered 3% of the time at densities that we consider heavy colonization (growth in the 3rd or 4th quadrant). Most (67%) of the isolates were recovered from enrichment broth only. Cocolonization of GBS and S. pseudoporcinus was detected in 11 (31%) women.

Table 1.

Association of demographic characteristics with acquisition of Streptococcus pseudoporcinus (n = 663)^a

Demographic characteristic	n	No. of acquisition events/woman-yr	Acquisition rate/100 woman-yrs	HR (95% CI)b	P
Race
White	408	6/539.2	1	1.0
Black	230	30/274.5	11	10.1 (4.2–24.2)	<0.001
Other	25	0/32.3	0
Age (yrs)
18–29	519	22/663.4	3	1.0
30–40	144	14/182.6	8	2.3 (1.2–4.6)	0.014
No. of yrs of education
≤12 yrs	165	19/205.7	9	1.0
13–15 yrs	254	11/310.8	4	0.4 (0.2–0.8)	0.009
≥16 yrs	244	6/329.5	2	0.2 (0.1–0.5)	0.001
Marital status
Unmarried	536	34/678.8	5	1.0
Married	127	2/167.2	1	0.2 (0.1–1.0)	0.051

^an is the number of women.

^bHR, hazard ratio; 95% CI, 95% confidence interval.

There were 36 women who acquired S. pseudoporcinus in 846.0 woman-years of follow-up, for an incidence of 4 acquisitions per 100 woman-years. Black women were more likely to be colonized by S. pseudoporcinus than white women and women belonging to other races (P < 0.001). Other factors associated with colonization included older age, having less education, and being unmarried (Table 1).

Participants who used vaginal spermicides for contraception or who had a tubal ligation were more likely to be colonized by S. pseudoporcinus than women who did not use these contraception methods, while women who used oral contraceptives were significantly less likely to be colonized than women who did not use oral contraceptives (Table 2). Other methods of contraception were not associated with acquisition of S. pseudoporcinus, including the following: male condoms, abstinence, cervical ring, medroxyprogesterone acetate, an intrauterine device, levonorgestrel implant, a cervical cap, a diaphragm, emergency contraception, skin patch, the rhythm method, withdrawal, having a male partner with a vasectomy, or lesbian orientation (data not shown). The use of pads for menstrual protection was associated with acquisition of S. pseudoporcinus in the unadjusted analysis (Table 2), but this association was no longer significant after adjusting for race. Antibiotic use, over-the-counter vaginal product use, tampon use, and douching were not associated with acquisition of S. pseudoporcinus (data not shown).

Table 2.

Association of contraception methods, antimicrobial use, over-the-counter vaginal product use, sexual activity since the last visit, recent self-reported infections, and concurrent genital colonization by group B Streptococcus and Lactobacillus with acquisition of Streptococcus pseudoporcinus (n = 663)

Factor	No. of acquisition events/woman-yr	Acquisition rate/100 woman-yrs	HR (95% CI)a	P
Oral contraception
No	29/528.2	5	1.0
Yes	7/317.8	2	0.4 (0.2–0.9)	0.025
Spermicide
No	33/825.6	4	1.0
Yes	3/20.4	15	3.8 (1.1–12.6)	0.033
Tubal ligation
No	29/775.5	4	1.0
Yes	7/70.5	10	2.7 (1.2–6.0)	0.019
Pad use for menstrual protection
No	11/408.9	3	1.0
Yes	25/437.1	6	2.3 (1.1–4.6)	0.023
No. of male sex partners
None or one	32/813.6	4	1.0
Two or more	4/32.4	12	3.9 (1.4–11.1)	0.012
Concurrent Nugent score
Normal (0–3)	14/512.2	3	1.0
Intermediate (4–6)	2/127.2	2	0.6 (0.1–2.5)	0.452
Bacterial vaginosis (7–10)	20/205.1	10	3.6 (1.8–7.0)	<0.001
Trichomonas vaginalis
No	34/842.5	4	1.0
Yes	2/3.5	58	15.0 (4.2–54.3)	<0.001
Genital herpes (primary/recurrent)
No	34/829.2	4	1.0
Yes	4/16.8	24	6.7 (2.5–18.4)	<0.001
Concurrent group B Streptococcus colonization, vaginal or rectal
No	25/645.2	4	1.0
Yes	11/200.8	5	1.8 (0.9–3.6)	0.118
Concurrent vaginal Lactobacillus colonization
No	9/132.0	7	1.0
Yes	27/714.0	4	0.5 (0.3–1.1)	0.103

^aHR, hazard ratio; 95% CI, 95% confidence interval.

Women having two or more sexual partners since their last visit were significantly more likely to acquire S. pseudoporcinus than monogamous or abstinent women (Table 2). However, there was no increased or decreased risk for acquisition related to frequency of vaginal intercourse or reporting receptive anal intercourse (data not shown).

At each visit, participants reported whether they had any clinician-diagnosed sexually transmitted infections (STIs) since their last visit. Both Trichomonas vaginalis and genital herpes infection was associated with acquisition of S. pseudoporcinus (Table 2). There was no association of Chlamydia trachomatis, Neisseria gonorrhoeae, human papillomavirus, pelvic inflammatory disease, yeast infection, or a urinary tract infection with acquisition of S. pseudoporcinus (data not shown).

Women were evaluated at each visit for bacterial vaginosis (BV) based on the Nugent criteria. Women having a vaginal smear consistent with BV were significantly more likely to acquire S. pseudoporcinus during follow-up (Table 2), although the association between this organism and BV was observed only among black women (data not shown). Adjusted risks that were independently associated with acquisition of S. pseudoporcinus can be seen in Table 3.

Table 3.

Adjusted risks^a for acquisition of Streptococcus pseudoporcinus colonization

Characteristic	HR (95% CI)b	P
Black race	7.7 (3.1–19.2)	<0.001
Age 30 years or more	2.0 (1.0–3.9)	0.049
Trichomonas vaginalis since the last visit	4.6 (1.0–21.0)	0.050
Genital herpes (primary/recurrent)	8.5 (4.0–18.1)	<0.001
Bacterial vaginosis by Gram smear	2.0 (1.0–4.0)	0.045
Two or more male sex partners since the last visit	3.9 (1.2–12.8)	0.024

^aVariables included in the final multivariable proportional-hazard regression model.

^bHR, hazard ratio; 95% CI, confidence interval.

DISCUSSION

In this study, a population prevalence of 5.4% was observed for S. pseudoporcinus compared with a 54.8% prevalence of GBS (data not shown). Reasons for underreporting in the past may include the following: misidentification of the organism as GBS, failure to use broth enrichment during culture, failure to identify all of the beta-hemolytic Streptococcus present in a sample, and the use of molecular methods to detect GBS during routine screenings that will not detect S. pseudoporcinus. All 120 isolates evaluated in the present study cross-reacted with the PathoDx (Remel) GBS serogrouping kit. Cross-reaction to 12 other kits, including PathoDx was also reported by Thompson and Facklam (17). Consequently, clinical laboratories that rely on antigen-testing kits for identification may misidentify S. pseudoporcinus as GBS, and this could result in unnecessary administration of antibiotics during pregnancy. Therefore, colonies with complete beta-hemolysis and positive group B strep latex agglutination require further testing. Chromogenic agars, such as ChromID Strepto B agar (bioMérieux SA), are used as a fast reliable method for screening women and infants for GBS; however, it is reported in the company's technical data sheet for this agar that S. porcinus may exhibit pigmentation similar to that of GBS. Colonies of S. pseudoporcinus exhibited similar pigmentation to GBS in our study. If this agar is utilized for GBS, other screening methods to differentiate between GBS and S. pseudoporcinus are required.

In this study, S. pseudoporcinus colonization and its association with black women confirm the results of the study by Gaudreau et al. in 2007 (10). Novel findings from our study include increased S. pseudoporcinus colonization among women aged 30 to 40 years, women who ever reported having genital HSV infection, recent infection with T. vaginalis, or were diagnosed with recent BV by Nugent criteria, and having two or more sexual partners since their last visit, suggesting that this organism may have coepidemiology with STIs.

Possible explanations for differing colonization rates observed between black women and women of other races may be 2-fold. Previous studies have suggested that differences in vaginal bacterial communities exist among Caucasian and black women (8, 18). These data coupled with the known disparities that exist between black women and women of other races and their susceptibility to BV, herpes simplex virus (HSV), and other STIs, may account for the increased acquisition of S. pseudoporcinus observed (2, 4, 8, 9, 11, 12).

The limitations of the present study include the failure to perform nucleic acid amplification technique (NAAT) testing for STIs and the reliance on self-reported STI acquisition. There was insufficient sample size to delineate independent risk factors for vaginal versus rectal colonization by S. pseudoporcinus.

The present study used a combination of classical broth enrichment combined with sequencing to reveal that S. pseudoporcinus is emerging as a common organism colonizing the rectum and vagina of women. Its coepidemiology with increased numbers of sexual partners and sexually transmitted infections, as well as its reported association with adverse outcomes of pregnancy (14) suggests that further studies of this organism are warranted.