JOURNAL OF CLINICAL MICROBIOLOGY, June 1992, p. 1575-1577 Vol. 30, No. 6 0095-1137/92/061575-03$02.00/0 Copyright ©) 1992, American Society for Microbiology Involvement of Gardnerella vaginalis in Urinary Tract Infections in Men SHARON M. SMITH,l.2* TAJUDEEN OGBARA,3'4 AND ROBERT H. K. ENG3 4 Microbiology Section, Laboratory Service,' and Infectious Disease Section, Medical Service,- Department of Veterans Affairs Medical Center, East Orange, New Jersey 07018-1095, and Department of Laboratory Medicine and Pathology2 and Departmnent of Medicine,4 New Jersey Medical School/UMDNJ, Newark, New Jersey 07103 Received 30 December 1991/Accepted 5 March 1992

Fifteen male patients from whose urine samples Gardnerella vaginalis was isolated (clinical incidence of 0.1%) were evaluated for clinical signs and symptoms of urinary tract infection and modality of acquisition of the organism. Ten of 15 (67%) patients were symptomatic or had signs of inflammation as manifested by an increased number of urinary neutrophils. One patient had two bouts of infection caused by this organism which required two courses of antibiotic therapy. Colonies of diphtheroidlike organisms found in urine cultures should not be ignored as insignificant but should be further investigated to determine whether G. vaginalis is present.

Gardnerella vaginalis, formerly called Haemophilus vag- human blood (Becton Dickinson Microbiology Systems), inalis and Corynebacterium vaginale, is a facultative anaer- and plates were incubated at 35°C for 24 h. G. vaginalis obic, nonmotile, pleomorphic gram-negative to gram-vari- isolates will grow after overnight incubation on V agar and able rod (21). It is a well-recognized colonizer of the female will have beta-hemolysis (23). All such isolates were identi- genital tract (4, 10, 24). The most common disease that this fied as G. vaginalis by the RapID NH System (Innovation organism may cause is bacterial vaginosis (7, 9, 24), but Diagnostics, Inc., Atlanta, Ga.), a system which is used to serious diseases such as bacteremia (16, 20, 22, 25) and identify Neisseria and Haemnophillus species. The medical meningitis (2) have been reported. In males, the finding of records of the patients with G. vaginalis isolated were this organism has, up to now, not been considered clinically reviewed. significant (3, 5, 13) and has been only uncommonly reported During the 6-month period, G. vaginalis was isolated from as a cause of urethritis or prostatitis (1, 17) or urinary tract urine samples from 15 patients. Thirteen of the patients were infection (UTI) (12, 26). seen in the Medical, Genitourinary, or Hypertension Clinic Our laboratory has been reporting G. vaginalis at an and two patients were seen in the Infectious Disease Clinic. increasing frequency in the urine samples submitted by men Seventeen urine cultures from the 15 patients were positive for routine bacteriological culture. With the availability of for G. vaginalis, representing 0.1% (17 of 12,400) of the urine commercial systems capable of identifying G. vaginalis in 4 cultures submitted during the period of observation. All h, many organisms previously reported as diphtheroids are positive cultures had colony counts of more than or equal to now identified and reported as G. vaginalis (6). The signifi- 105 CFU/ml, and only G. vaginalis was isolated from the cance of a urine culture positive for G. vaginalis is unknown. positive cultures. The presence of organisms such as H. Whether the organism should be considered a pathogen or and only a colonizer in these patients and whether these patients influenzae, Chlamydia spp., anaerobes, mycobacteria, have a predisposition for this organism were investigated. however, cannot be ruled out. A summary of the 15 cases is For a period of 6 months, all reports on routine cultures provided in Table 1. submitted to the Microbiology Laboratory were reviewed, Ten of the 15 patients had complaints associatcd with UTI and a list of male patients in whom G. vaginalis grew was or signs of UTI such as dysuria, hematuria, frequency of kept. Urine specimens from patients who had been treated urination, and increased numbers of urinary leukocytes for UTIs, from patients with signs and symptoms of UTI, (small stream was excluded as an infection symptom since it from patients with underlying diseases associated with an could reflect only chronic structural changes). The ages of increased frequency of UTIs, and from older patients with the patients ranged from 30 to 89 years, with the majority obstructive prostate disease were submitted for culture. being older than 50 years. One of the 10 patients had no Urine cultures were performed by inoculation onto Trypti- sensation in the genitals because of an underlying neurologic case soy agar with 5% sheep blood and MacConkey and deficit. This patient and six symptomatic patients had ele- Columbia CNA agars (Becton Dickinson Microbiology Sys- vated numbers of leukocytes in their urine samples (>5 tems, Cockeysville, Md.), and plates were incubated at 35°C leukocytes per high-power field). Five had erythrocytes in in 5 to 10% CO2 for 48 h. Colony counts were determined by their urine samples. The urine pHs ranged from 5.0 to 6.5. quantitative subculture of 1 ,ul onto the blood-agar plate, Only two of the patients had peripheral blood leukocytosis. with isolates of 103 CFU/ml being further identified. Slowly No patient had an increased percentage of band forms. Only growing colonies of gram-variable rods generally requiring one patient (patient 9) had an elevated level of creatinine in 48 h for good growth were subcultured on V agar containing his serum (3.6 mg/dl). Five patients were noted to have histories of UTIs involving other bacteria. Four of the seven patients with urine leukocytosis were given antibiotics such * Corresponding author. as metronidazole, amoxicillin, doxycycline, ciprofloxacin, 1575 1576 NOTES J. CLIN. MICROBIOL.

TABLE 1. Summary of cases' Leukocytes Patient Age Underlying Sexual Signs and symptoms in urine Antimicrobial Course no. disease activity (no. per therapy HPF) 1 58 Past UTI, PUD, No Hematuria, prostate 5-10 None No reculture NIDDM, hypertension enlargement 2 53 Hypertension No None 0-5 None No reculture 3 44 Spinal cord injury, No None 0-5 None No reculture neurogenic bladder 4 54 Diverticulosis, NIDDM, Yes Frequency, inability 0-5 None No reculture hypertension to ejaculate 5 66 Bladder cancer No Urgency 5-10 None No reculture 6 64 Past UTI, abdominal No Hematuria, dysuria 10-20 SXT Responded; no reculture aneurism 7 39 Past UTI, past Yes Dysuria, frequency 20-30 Two courses of + Reculture once, gonorrhea amoxicillin responded once 8 45 HIV Ab positive, Yes Hematuria, dysuria 10-30 Metronidazole Responded; no reculture Trichomonas UTI 9 65 Meningioma Yes None >25 None No reculture 10 68 Hypertension, CVA No Small stream 0-5 Ciprofloxacin - Reculture 11 50 Diabetes mellitus, Yes None 1-5 None No reculture hypertension 12 73 Prostate enlargement, No Dysuria, frequency, Many Doxycycline, + Reculture, - reculture PUD flank pain ciprofloxacin 13 89 Hypertension, diabetes No Chronic frequency 1-5 Metronidazole Responded; no reculture mellitus 14 30 None Yes Dysuria 0-5 SXT Responded; no reculture 15 67 Hypertension No Small stream 1-5 None No reculture U Abbreviations: HPF, high-power field; PUD, peptic ulcer disease; NIDDM, non-insulin-dependent diabetes mellitus; HIV Ab, human immunodeficiency virus type 1 antibody positive; CVA, cerebral vascular accident; SXT, sulfamethoxazole-trimethoprim. Symbols: -, negative; +, positive.

and sulfamethoxazole-trimethoprim. All patients receiving who isolated G. vaginalis from bladder aspirates of patients therapy responded clinically (patients 6, 7, 8, 10, 12, 13, and who were pregnant and of patients with underlying renal 14). A urine culture from patient 7 (from the Infectious disease (8, 19). These factors may play a role in the lack of Disease Clinic) was positive only for G. vaginalis, and isolation of this organism from the urine of male patients. patient 7 had symptoms of dysuria and urgency. He was When the mode of acquisition of G. vaginalis was inves- treated for 10 days with oral amoxicillin (2 g/day), and the tigated, 6 of the 15 men (40%) reported sexual contact with symptoms resolved. Two months later, he had dysuria again, a woman within 48 h prior to collection of the urine cultures. and G. vaginalis was again recovered from the patient's These individuals may have acquired the organism during urine. He was again treated with oral amoxicillin, with direct sexual contact with the female partner. These data resolution of symptoms. Eight of 15 patients had become support the conclusions of earlier investigators in that the asymptomatic or had remained asymptomatic, and antibiotic acquisition of G. vaginalis does occur through sexual con- therapy was never initiated with these patients. Unfortu- tact (4, 16, 18). Whether sexual acquisition is important was nately, urine cultures to determine the length of G. vaginalis further studied by Holst (10). Male partners of females with colonization were not performed for any of these eight bacterial vaginosis and from whom G. vaginalis was initially patients. the urethra and who used condoms G. vaginalis has been known to preferentially colonize the isolated from during female genital tract (4, 10, 24). Two explanations have been sexual intercourse did not have G. vaginalis reisolated. offered. First, prostatic fluid contains high concentrations of Holst (10) concluded that G. vaginalis and Mobiluncus spp. the tracts of males as a zinc, which may be inhibitory, and second, prostatic lining transiently colonized genital healthy contains columnar or cuboidal epithelial cells that may resist result of continued passive acquisition from their female adherence of G. vaginalis (15). Nevertheless, carriage in the partners. On the other hand, 9 of 15 patients in our study genitals of healthy men has been reported at rates of 7.2% clearly denied any sexual contact with a female from 1 (13) and 11.4% (5) and carriage in seminal fluids has been month to several years prior to our obtaining the positive reported at a rate of 38% (11) when the specimens were urine cultures. In these cases, the organism may have been specifically cultured for this organism. In addition to the part of their own bacterial floras, although previous studies possible hostility of the male genital tract environment for have shown that this occurs infrequently (10). this organism, the urine itself may be bactericidal for the Our data indicate that G. vaginalis may play a role in organism. Lam et al. (14) found that G. vaginalis counts causing UTI in at least 67% of the male patients from whom declined by 99.9% in urine held at 37°C for 24 h at pH 7.0. it was isolated. Therefore, we encourage microbiology lab- Pyuria was not uniformly present in our patients (present in oratories to examine slowly growing diphtheroidlike organ- 7 of 15) and therefore could not be reliably used as an isms and to determine whether G. vaginalis is present. In indicator of infection. The lack of an inflammatory response order to clarify the causative role of G. vaginalis in UTIs in to G. vaginalis has also been reported by other investigators men, more clinical observations are needed. VOL. 30, 1992 NOTES 1577

REFERENCES infection in men. Br. J. Vener. Dis. 58:127-129. 1. Abercrombie, G. F., J. Allen, and R. Maskell. 1978. Corynebac- 14. Lam, M. H., and D. F. Birch. 1991. Survival of Gardnerella terium vaginale urinary tract infection in a man. Lancet i:766. vaginalis in human urine. Am. J. Clin. Pathol. 95:234-239. 2. Berardi-Grassias, L., 0. Roy, J. C. Berardi, and J. Furioli. 1988. 15. Lam, M. H., D. F. Birch, and K. F. Fairley. 1988. Prevalence of Neonatal meningitis due to Gardnerella vaginalis. Eur. J. Clin. Gardnerella vaginalis in the urinary tract. J. Clin. Microbiol. Microbiol. 7:406-407. 26:1130-1133. 3. Bowie, W. R., H. M. Pollock, P. S. Forsyth, J. F. Floyd, E. R. 16. Legrand, J. C., A. Alewaeters, L. Leenaerts, P. Gilbert, M. Alexander, S.-P. Wang, and K. K. Holmes. 1977. Bacteriology of Labbe, and Y. Glupczynski. 1989. Gardnerella vaginalis bacte- the urethra in normal men and men with nongonococcal urethri- remia from pulmonary abscess in a male alcohol abuser. J. Clin. tis. J. Clin. Microbiol. 6:482-488. Microbiol. 27:1132-1134. 4. Bump, R. C., and W. J. Buesching III. 1988. Bacterial vaginosis 17. Leopold, S. 1953. Heretofore undescribed organism isolated in virginal and sexually active adolescent females: evidence from the genitourinary system. U.S. Armed Forces Med. J. against exclusive sexual transmission. Am. J. Obstet. Gynecol. 4:263-266. 158:935-939. 18. Livengood, C. H., III, J. L. Thomason, and G. B. Hill. 1990. 5. Dawson, S. G., C. A. Ison, G. Csonkca, and C. S. F. Easmon. Bacterial vaginosis: diagnostic and pathogenetic findings during 1982. Male carriage of Gardnerella vaginalis. Br. J. Vener. Dis. topical clindamycin therapy. Am. J. Obstet. Gynecol. 163:515- 58:243-245. 520. 6. Eriquez, L. A., and N. E. Hodinka. 1983. Development of a test 19. McDowall, D. R. M., J. D. Buchanan, K. F. Fairley, and G. L. system for rapid differentiation of Neisseria and Haemophilus Anaerobic and other fastidious in spp. J. Clin. Microbiol. 18:1032-1039. Gilbert. 1981. microorganisms 7. Eschenbach, D. A., S. Hillier, C. Critchlow, C. Stevens, T. asymptomatic bacteriuria in pregnant women. J. Infect. Dis. DeRouen, and K. K. Holmes. 1988. Diagnosis and clinical 144:114-122. manifestations of bacterial vaginosis. Am. J. Obstet. Gynecol. 20. Patrick, S., and P. A. Garnett. 1978. Corynebacterium vaginale 158:819-828. bacteriaemia in a man. Lancet i:987-988. 8. Fairley, K. F., and D. F. Birch. 1983. Unconventional bacteria in 21. Piot, P. 1985. Gardnerella vaginalis, p. 874-876. In E. H. urinary tract disease: Gardnerella vaginalis. Kidney Int. 23:862- Lennette, A. Balows, W. J. Hausler, Jr., and H. J. Shadomy 865. (ed.), Manual of clinical microbiology, fourth ed. American 9. Gardner, H. L., and C. D. Dukes. 1955. Haemophilus vaginalis Society for Microbiology, Washington, D.C. vaginitis. A newly defined specific infection previously classi- 22. Reimer, L. G., and L. B. Reller. 1984. Gardnerella vaginalis fied as "nonspecific" vaginitis. Am. J. Obstet. Gynecol. 69:962- bacteremia: a review of thirty cases. Obstet. Gynecol. 64:170- 976. 172. 10. Hoist, E. 1990. Reservoir of four organisms associated with 23. Totten, P. A., R. Amsel, J. Hale, P. Piot, and K. K. Holmes. bacterial vaginosis suggests lack of sexual transmission. J. Clin. 1982. Selective differential human blood bilayer media for Microbiol. 28:2035-2039. isolation of Gardnerella (Haemophilus) vaginalis. J. Clin. Mi- 11. Ison, C. A., and C. S. F. Easmon. 1985. Carriage of Gardnerella crobiol. 15:141-147. vaginalis and anaerobes in semen. Genitourin. Med. 61:120- 24. Vontner, L. A., and D. A. Eschenbach. 1981. The role of 123. Gardnerella vaginalis in non-specific vaginitis. Clin. Obstet. 12. Josephson, S., J. Thomason, K. Sturino, R. Zabransky, and J. Gynecol. 24:439-460. William. 1988. Gardnerella vaginalis in the urinary tract: inci- 25. Wilson, J. A., and A. J. Barratt. 1986. An unusual case of dence of significance in a hospital population. Obstet. Gyneo- Gardnerella vaginalis septicaemia. Br. Med. J. 293:309. col. 71:245-250. 26. Woolfrey, B. F., G. K. Ireland, and R. T. Lally. 1986. Signifi- 13. Kinghorn, G. R., B. M. Jones, F. H. Chowdhury, and I. Geary. cance of Gardnerella vaginalis in urine cultures. Am. J. Clin. 1982. Balanoposthitis associated with Gardnerella vaginalis Pathol. 86:324-329.