- Email: [email protected]
Are seminal fluid microorganisms of significance or merely contaminants?
FERTILITY AND STERILITY威 VOL. 74, NO. 3, SEPTEMBER 2000 Copyright ©2000 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.
Are seminal fluid microorganisms of significance or merely contaminants? Evelyn Cottell, Ph.D.,a Robert F. Harrison, M.D.,a Mary McCaffrey, M.R.C.O.G.,a Tom Walsh, M.B.,a Eimear Mallon, M.R.C.G.P.,b and Carole Barry-Kinsella, M.D.a Royal College of Surgeons in Ireland and Human Assisted Reproduction Ireland, Rotunda Hospital, Dublin, Ireland
Objective: To determine the contribution of urethral and skin flora to seminal fluid cultures and the relation between bacteriospermia and seminal leukocytes. Design: Prospective study. Setting: IVF-ET unit at a university teaching hospital. Patient(s): Sixty men starting an IVF-ET program. Intervention(s): Culture of sequential first-catch urine, midstream urine, and semen samples with evaluation of seminal leukocytes. Main Outcome Measure(s): A comparison of microbes from first-catch urine, midstream urine, and semen samples and the correlations of seminal microbes, elevated leukocyte concentrations, and pregnancy. Result(s): Microorganisms were detected in 37% of first-catch urine samples, 27% of midstream urine samples, and 51% of semen samples. Most microorganisms were gram-positive microbes and were common to both urine and semen samples. Mean and median leukocyte concentrations were 0.98 ⫻ 106/mL and 0.10 ⫻ 106/mL, respectively. There was no correlation between seminal microbes and raised leukocytes or between leukocytospermia and/or bacteriospermia and pregnancy. Conclusion(s): Microorganisms are commonly found in insignificant quantities in the semen of asymptomatic men. The frequent isolation of gram-positive microbes common to both urine and semen and the absence of a correlation with raised leukocyte concentrations suggest that bacteriospermia most commonly represents contamination. (Fertil Steril威 2000;74:465–70. ©2000 by American Society for Reproductive Medicine.) Key Words: Semen, urine, microorganisms, leukocytes Received February 17, 2000; revised and accepted May 1, 2000. Supported by a research grant from the Royal College of Surgeons in Ireland, Dublin, Ireland. Reprint requests: Evelyn Cottell, Ph.D., Royal College of Surgeons in Ireland, Academic Department of Obstetrics and Gynaecology, Rotunda Hospital, Dublin 1, Ireland (FAX: 353-1-8727831; Email: [email protected]). a Academic Department of Obstetrics and Gynaecology, Royal College of Surgeons in Ireland, Rotunda Hospital. b Human Assisted Reproduction Unit, Rotunda Hospital. 0015-0282/00/$20.00 PII S0015-0282(00)00709-3
The presence of microorganisms in the semen of men in subfertile relationships is widely documented. However, conflicting reports on the effect of these microbes on semen quality have created difficulties in establishing their pathophysiologic role in male infertility (1–3). Discrepant results may be due to a number of factors. First, various sample-collection procedures have been used. Second, differing definitions of bacteriospermia are commonly used. Some reports have included all bacteria isolated in semen regardless of the bacterial strain, the colony count, or the relation with other bacteria, whereas other studies considered organisms only when present in concentrations of ⬎104/mL or potentially pathogenic microbes. Third, contamination of semen by nonpathogenic commensals of the skin, glans penis, or the lower urethra creates difficulties in identifying true pathology (4 – 6). Finally, seminal
microbes may be evaluated without considering the seminal leukocyte concentration. Because leukocytes may have an effect on the functional capacity of the spermatozoa (7), their contribution to the spermiogram is important. Semen is composed of spermatozoa from the epididymis and secretions from the seminal vesicles, prostate, and bulbourethral glands, and each region is normally sterile (8). It is therefore probable that seminal microbes result mainly from contamination by either urethral flora or skin commensals. Although commensal organisms may be of no significance in natural conception, contamination of a culture system by seminal microorganisms during assisted reproductive techniques may limit success (9, 10). This concern provides a new impetus for revisiting a previously recognized 465
problem. If urethral contamination does contribute significantly to bacteriospermia, midstream urine samples may have lower concentrations of microbes than first-catch urine samples, and collection of semen samples after passing urine may be an effective way of reducing microorganisms before IVF. This study first examined the microbial composition of first-catch urine, midstream urine, and semen samples from men in subfertile relationships to establish the contribution of commensal flora to seminal fluid cultures. Second, seminal microbes were evaluated in relation to leukocyte composition, and correlations were examined among microbes, elevated leukocytes, and pregnancy.
MATERIALS AND METHODS The Rotunda Hospital Ethics Committee granted approval for this study. Sixty men in subfertile relationships, about to start an IVF-ET program at this hospital, were recruited. Informed consent was obtained. Both verbal and written instructions were given to all men regarding abstinence, hygiene, and specimen production. Each man provided an early-morning first-catch urine sample (10 –20 mL), followed by a midstream urine sample (10 –20 mL), and finally a semen sample, taken by masturbation. Samples were taken at home into sterile containers provided (Sterilin; Abergaryoed, Gwent, United Kingdom) and were brought to the laboratory within 1 hour of collection. Before sample collection, the men were instructed verbally and in writing to abstain from ejaculation for 2–3 days. The men were asked to wash their hands and genital area with soap and water and to towel dry before taking the first urine sample (11). All men completed a questionnaire concerning histories of sexually transmitted disease (STD), genital tract surgery, urinary tract infection, dysuria, and circumcision. After liquefaction and under sterile conditions, the semen sample was mixed with a sterile pipette. Aliquots were removed for semen analysis, immunocytochemical staining of leukocytes, and microbiologic evaluation. Semen analysis was performed with a Horwell Fertility counting chamber (A. R. Horwell, London, United Kingdom) and a phase contrast microscope. Seminal fluid leukocyte concentrations were determined by an immunocytochemical stain with an anti–HLe-1 monoclonal antibody (Becton Dickinson, San Jose, CA) with the use of an alkaline phosphatase anti– alkaline phosphatase staining technique (12). Leukocytospermia was defined as the presence of ⱖ1 ⫻ 106 leukocytes per milliliter of semen (11). An aliquot of the first-catch urine, midstream urine, and liquefied semen samples was reserved for culture of aerobic and facultative anaerobic microorganisms. For aerobic cultures, blood agar plates and chocolate plates for Gonococcus species culture were inoculated with standard loops (0.001 466
Cottell et al.
Urogenital flora and leukocytospermia
mL) of specimen. The plates were cultured for 48 hours in 5% CO2 in air at 37°C. Cultures were identified with the commercial kits Streptex and Candida albicans (Murex Diagnostics Inc., Norcross, GA), Staphlase (Oxoid; Unipath, Basingstoke, Hampshire, United Kingdom), and API 20 E (bioMe´rieux, Marcy l’Etoile, France). Diphtheroids and Escherichia coli were identified directly from cultures. For anaerobic cultures, blood agar plates were inoculated with standard loops and cultured for 48 hours at 37°C under anaerobic conditions, achieved with a commercial gas-generating system (Oxoid). Anaerobic microbes were identified by a Mastring ID kit (Mast Diagnostics, Bootle, Merseyside, United Kingdom). Aerobic and anaerobic cultures were considered positive when ⱖ103 organisms per milliliter, equivalent to one observed colony, were noted. Significant bacteriospermia was defined as ⱖ104 nonpathogenic organisms per milliliter of semen or ⱖ103 potentially pathogenic organisms per milliliter. Nonpathogenic organisms included Staphylococcus epidermidis, nonhemolytic streptococci, diphtheroids, and ␣hemolytic streptococci; potential pathogens included E. coli, -hemolytic streptococci, anaerobes, Proteus species, coliforms, and mixed microbes. Colony counts of ⱖ100 in midstream urine samples were considered clinically significant. First-catch urine and semen samples also were analyzed for the primary genital mycoplasmas, Mycoplasma hominis and Ureaplasma (u.) urealyticum, with the use of a mycoplasma kit, Mycofast “All-In” (International Mycoplasma, Toulon, France). Mycoplasma counts of ⱖ104 colony-forming units (cfu) were considered significant. The remainder of the first-catch urine was analyzed for Chlamydia with an amplified enzyme immunoassay (IDEIA Chlamydia; DAKO Diagnostics Ltd., Ely, Cambridgeshire, United Kingdom). Urine was transferred to conical Falcon tubes (Becton Dickinson Ltd., Plymouth, England) and centrifuged at 3,000 ⫻ g for 15 minutes. The supernatants were discarded, and the deposits were resuspended in 1 mL of supplied workingstrength Chlamydia transport medium in a small screwcapped glass bottle. Samples were stored at ⫺70°C and analyzed in batches for Chlamydia antigen. The effect of significant bacteriospermia and/or leukocytospermia on the pregnancy rate (PR) was examined. A clinical pregnancy was defined as the presence of a gestational sac at 5 weeks after ET. Relation between significant bacteriospermia and leukocytospermia, circumcision, and pregnancy; between leukocytospermia and pregnancy; and between bacteriospermia combined with leukocytospermia and pregnancy were examined by calculating odds ratios (ORs) and their confidence intervals (CIs). Female age and oocyte number were compared in the pregnant and nonpregnant groups with Student’s t-tests at the 95% confidence level. Statistical analysis was performed with a Stata Release 6 package at the Department of Psychology, Royal College of Surgeons in Ireland. Vol. 74, No. 3, September 2000
TABLE 1
TABLE 2
Percentage isolation rate of microorganisms from firstcatch urine, midstream urine, and semen samples from 60 men in subfertile relationships.
Colony counts in 60 semen samples.
Isolation rate (%) Microorganism
Microorganism
Firstcatch urine
Nonpathogens Staphylococcus epidermidis 10.0 (6) Nonhemolytic streptococci 1.7 (1) Diphtheroids 3.3 (2) Ureaplasma urealyticum 1.7 (1) ␣-Hemolytic streptococci 1.7 (1) Potential pathogens -Hemolytic streptococci 13.3 (8) Escherichia coli 6.7 (4) Proteus species 1.7 (1) Coliforms 1.7 (1) Mixed microbes 3.3 (2) Anaerobes 0.0 (0)
Midstream urine
Semen
Semen (strict criteriaa)
11.7 (7) 1.7 (1) 6.6 (4) 0.0 (0) 3.3 (2)
25.0 (15) 11.7 (7) 16.7 (10) 15.0 (9) 11.7 (7) 10.0 (6) 6.7 (4) 1.7 (1) 3.3 (2) 3.3 (2)
6.7 (4) 5.0 (3) 1.7 (1) 0.0 (0) 0.0 (0) 0.0 (0)
20.0 (12) 20.0 (12) 3.3 (2) 3.3 (2) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1)
Note: Values in parentheses represent number of samples. a Only ⱖ104 nonpathogenic organisms per milliliter or ⱖ103 potentially pathogenic organisms per milliliter are reported. Cottell. Seminal fluid microorganisms. Fertil Steril 2000.
RESULTS Microorganisms were detected in 22 (37%) of the firstcatch urine samples, 16 (27%) of the midstream urine samples, and 33 (54%) of the semen samples. The majority of isolates were gram-positive bacteria, including staphylococci, streptococci, and diphtheroids. There were low frequencies of isolation of potentially pathogenic gram-negative bacilli, anaerobes, and mycoplasmas, and no isolation of Chlamydia or Gonococcus (Table 1). Bacterial isolation rates in semen are presented first as any presence of bacteria and second as a significant presence, on the basis of strict criteria described in Materials and Methods. Table 2 presents the mean colony counts, SDs, and ranges for each organism isolated. In 20 men (33.3%), cultures on all three samples were negative. In a further 14 men (23.3%), both urine samples were sterile. However, corresponding semen cultures were positive. Nine men (15%) had triple growth of organisms from first-catch urine, midstream urine, and semen samples. The remaining 17 men (28.3%) had microbes isolated from either first-catch urine and/or midstream urine and/or semen. Of the 22 men with positive cultures for first-catch urine, 10 (45.5%) had negative midstream urine cultures. Twelve men (20%) had microbes grown from both urine samples, of whom 2 had identical colony counts for both samples and 6 had lower colony counts of the same organism in the midstream urine sample. One man had a higher microbial count FERTILITY & STERILITY威
No. of Colony count samples with (⫻ 103 orgs/mL) microbe Mean ⫾ SD Range isolated
Nonpathogens Staphylococcus epidermidis Nonhemolytic streptococci Diphtheroids ␣-Hemolytic streptococci Ureaplasma urealyticum
15 10 7 2 4
10.4 ⫾ 12.5 1–50 43.7 ⫾ 41.3 5–100 27.9 ⫾ 20.2 6–56 8.0 ⫾ 5.7 4–12 5.5 ⫾ 5.2 1–10 (⫻ 103 cfua) (⫻ 103 cfua)
Potential pathogens -Hemolytic streptococci Escherichia coli Proteus species Coliforms Mixed microbes Anaerobes
12 2 1 1 1 1
26.6 ⫾ 30.3 48.0 ⫾ 59.4 12.0 11.0 ⬎100 80
2–100 6–90 – – – –
a
Ureaplasma urealyticum was expressed in colony-forming units (cfu) as opposed to organisms per milliliter. Cottell. Seminal fluid microorganisms. Fertil Steril 2000.
in the midstream urine sample and, on three occasions, different combinations of microbes were isolated from the two urine samples. Of the 44 men who had negative midstream urine samples, their corresponding semen cultures were positive in 20 cases (45%) and negative in 24 cases (55%). The mean concentration, median value, and range for seminal leukocytes were 0.98 ⫻ 106/mL, 0.10 ⫻ 106/mL, and 0 –14.0 ⫻ 106/mL, respectively. No correlation was found between elevated seminal leukocytes (leukocytospermia) and the significant presence of microorganisms (OR 1.25, 95% CI 0.35– 4.4). Of the 11 men who had leukocyte concentrations of ⱖ1 ⫻ 106/mL, 5 had negative seminal fluid cultures. The remaining 6 men with leukocytospermia had significant bacteriospermia. These included 2 men with low growth (2 ⫻ 103 and 7 ⫻ 103 organisms/mL) and 1 man with moderate growth (31 ⫻ 103 organisms/mL) of -hemolytic streptococci. The remaining 3 men had heavy growth of mixed microbes (⬎105 organisms/mL), mixed growth of nonhemolytic streptococci (99 ⫻ 103/mL) and -hemolytic streptococci (23 ⫻ 103/mL), and moderate growth of U urealyticum (104 cfu/mL). Data from the questionnaire showed that one man had a history of an STD (urethritis due to C trachomatis) and two men had dysuria. Seven men (11.7%) had been circumcised. ORs and CIs showed no correlation between significant concentrations of seminal microbes and circumcision (OR 1.33, 95% CI 0.30 –5.85). There were no correlations observed between either raised concentrations of microbes or leukocytospermia and a history of infection or dysuria, although the low number of cases in these two groups prevents meaningful statistical analysis. 467
Of the 60 couples who had an IVF-ET cycle, 19 (31.7%) were successful in attaining a clinical pregnancy, all resulting in live births. The mean female age and number of oocytes collected were 34.1 and 34.5 years and 10.0 and 9.5 eggs, respectively, for the pregnant and nonpregnant groups. No statistically significant difference was found between the groups for either variable (Student’s t-test at the 95% level). Although the odds of pregnancy were slightly lower in the presence of leukocytospermia (OR 0.77, 95% CI 0.2–3.1), this was not statistically significant. The PR was slightly higher in the presence of significant bacteriospermia (OR 2.8, 95% CI 0.9 – 8.5), although this difference did not reach statistical significance. Likewise, when the combined factors of leukocytospermia and significant bacteriospermia were examined in relation to the PR, no statistically significant difference was observed between the presence or absence of this combined condition (OR 0.39, 95% CI 0 –2.8).
DISCUSSION The prevalences of bacteriospermia in men of both fertile and subfertile relationships range from 10% to 100% in the published literature (8). This wide range may reflect the variety of semen-collection protocols used. The World Health Organization (11) recommended that special precautions be taken to avoid microbial contamination when collecting semen samples for culture. They advise strict hygiene and stress the importance of first passing urine before producing a semen sample by masturbation. Boucher et al. (13) reported a significant reduction in bacterial count and number of species and an increase in the percentage of sterile cultures when direct verbal counseling, as opposed to written instructions only, was given to men before specimen production. In this microbial investigation of urogenital fluids, both written and verbal instructions were given to each man, and all semen samples were obtained by masturbation immediately after passing urine. However, despite these precautions, seminal microorganisms were recovered in 54% of the cases. Common microorganisms were isolated from first-catch urine, midstream urine, and semen samples, although a higher prevalence was found in semen. The seminal fluid isolation rate of 6.7% for U urealyticum is lower than that found elsewhere (14). Urethral colonization by ureaplasmas is common (15), and the low isolation rate from semen may reflect a washout effect in the urethra during the passing of urine immediately before producing the semen sample. Although only one first-catch urine sample was positive for U urealyticum, higher detection rates may have been found if centrifuged deposits of urine samples were examined. The positivity rate for Chlamydia, with use of the same amplified ELISA technique as in this study, among men and women attending STD clinics in Dublin ranges from 7% to 9%. Therefore, the negative results for Chlamydia in this 468
Cottell et al.
Urogenital flora and leukocytospermia
study most likely reflect a low incidence of active chlamydial infection in our infertile population, but also may be due to insensitivity of the enzyme immunoassay used. However, this amplified ELISA has shown sensitivities comparable to cell culture of 400 endocervical and urethral swabs (unpublished data, Virus Reference Centre, Ardmore, Dublin, Ireland). Moreover, a centrifuged deposit of a first-catch earlymorning urine sample may be an acceptable noninvasive sample for detecting Chlamydia in the male urethra (16, 17). Although quantitative assessment of microbial growth was determined for all samples, it was valid to compare colony counts only in first-catch and midstream urine samples because of differences in volume from urine to semen. More than one third of all men had microbes isolated from their first-catch urine samples, with colony counts of ⬎100 in only two cases. Microorganisms were either absent or reduced in number in the corresponding midstream urine samples of 73% of these men, probably reflecting a dilution of microorganisms out of the distal urethra with the firstcatch urine. The colony count in midstream urine samples was never of clinical significance. Among the 44 cases in which midstream urine cultures were negative, 20 (45%) of the corresponding semen samples had positive cultures. The probability that these culture results represent infection is unsupported by clinical, bacteriologic, and immunologic considerations. However, limitations to the sensitivity of the specimen types examined preclude dismissal of prostatitis and urethritis in our study group. Criteria for the diagnosis of male accessory gland infection must take into consideration the type of specimen analyzed, the organism detected, the colony count, the presence of ⬎106 leukocytes per milliliter of semen, and a history or clinical signs of genital tract infection (18). Although bacterial infection of the prostate is usually associated with irritative voiding symptoms and varying degrees of perineal or suprapubic discomfort (14), chronic bacterial prostatitis may be asymptomatic. Diagnosis can be made only with quantitative segmental cultures of first and second voided urine samples, expressed prostatic secretions, and finally a third voided urine sample (Meares and Stamey test) (19). The urethral swab has been the traditional specimen to test for urethritis. Semen and unconcentrated first-catch urine samples are probably not sensitive enough to detect urethritis, and centrifuged deposits from first-catch urine samples may have been a better sample type because they have proved as sensitive as swabs for Chlamydia detection when the same test is applied (16). Of the 20 men with negative midstream urine and positive seminal fluid cultures, 2 reported pain on passing urine. The first man had had a urinary tract infection treated 6 months before the study. Only nonhemolytic streptococci (22 ⫻ 103/mL) were cultured from his semen, and the seminal leukocyte concentration was not raised (0.6 ⫻ 106/mL). Vol. 74, No. 3, September 2000
Culture results for the second man showed U urealyticum, S epidermidis, and diphtheroids, all with low colony counts. No leukocytes were detected in this semen sample. It is unlikely that these men had prostatitis because these infections are rarely caused by more than one organism, and the infecting bacterium can usually be isolated persistently and in large concentrations from either expressed prostatic secretions or seminal fluid. However, a Meares and Stamey test (19) is required to clearly exclude prostatitis. Neither the two men with dysuria nor the man with a history of an STD had seminal microbes or leukocytes consistent with infection of the accessory glands, although the low numbers prevent meaningful statistical analysis. It is therefore likely that the seminal organisms cultured from their samples and from those of the other 18 men with negative midstream urine samples were contaminants from the urethra, glans penis, or hands during sample collection. This is supported by the work of Willen et al. (20), who found that 71% of strains colonizing the coronal sulcus were also present in the distal urethra and that urethral flora may then contaminate the semen. The different rheologic properties of semen and urine passing through the urethra, the different hydrostatic pressure conditions in the urethra during voiding and masturbation, and the manipulations to the penis during masturbation, with potential contamination by cutaneous bacteria adjacent to the urethral meatus, may explain the higher isolation rate of organisms in semen (21). These findings are supported by a small study by Fowler and Mariano (6), who reported greater contamination of expressed prostatic secretions and seminal fluid than of voided urine by the aerobic bacterial flora of the urethra. Microorganisms were detected frequently in semen, and although the majority were nonpathogenic commensals or contaminants, their presence may be of greater significance in IVF-ET, where natural defenses in the female genital tract (22), including antimicrobial secretion of lysozyme and lactoferrin, together with phagocytic cell production of immunoglobulin, are bypassed. Contamination of the culture system with seminal microbes may lead to suboptimal fertilization rates or impaired embryonic development (23) if adequate seminal processing techniques with antibiotic-rich culture medium are not used (10). In this study, common organisms were detected in urine and semen samples from the same individual. Most were gram-positive microbes normally associated with the urethra. Despite efforts to minimize microbial contamination of semen by first passing urine and adhering to instructions for genital hygiene (11), we isolated seminal fluid microbes in ⬎50% of samples. Although both fluids share a final common pathway in the urethra, semen appears to be more susceptible than urine to microbial contamination from the urethra, glans penis, or hands. Even if urethral contamination could be eliminated, the precise origin of microbes in semen FERTILITY & STERILITY威
could not be determined because of the different sources of seminal fluid components. Therefore, the role of infection in male infertility might be analyzed best with respect to the infecting microorganism rather than to the suspected site of infection. Positive seminal fluid cultures must be interpreted with caution, taking into account the presence of normal or abnormal urethral flora, raised colony counts of single isolates, and seminal leukocyte concentrations. These variables should be evaluated with a thorough history and physical examination to uncover symptoms and signs of infection. When both raised colony counts and leukocytes are observed, further investigation with urethral swabs and prostatic massage should ensue. Only then can physicians avoid the common misdiagnosis of genital tract infection, based on the presence of seminal microbes, and the possible unnecessary treatment with antibiotics.
Acknowledgments: The authors thank the laboratory staff, nurses, and clinicians at the Human Assisted Reproduction Ireland unit; the staff of the microbiology laboratory, Rotunda Hospital for their cooperation with this study; and Ronan Conroy of the Department of Psychology, Royal College of Surgeons in Ireland for statistical analyses.
References 1. Harrison RF, Hannon K, Butler M. Aerobic and anaerobic culture of human semen and the relationship to characteristics in the spermiogram. Fortschr Androl 1983;8:46 –54. 2. Busolo F, Zanchetta R, Lanzone E, Cusinato R. Microbial flora in semen of asymptomatic infertile men. Andrologia 1984;16:269 –75. 3. Huyser C, Fourie FLeR, Oosthuizen M, Neethling A. Microbial flora in semen during in vitro fertilisation. J In Vitro Fert Embryo Transfer 1991;8:260 – 4. 4. Desai S, Cohen MS, Khatamee M, Leiter E. Ureaplasma urealyticum infection: does it have a role in male infertility? J Urol 1980;124:469 – 71. 5. Granouillet R, Gaudin OG, Laurent JC, Rousset H, Moulin A. Study of the spermatic bacterial flora in infertile males. Pathol Biol 1982;30: 22– 6. 6. Fowler JE, Mariano M. Difficulties in quantitating the contribution of urethral flora to prostatic fluid and seminal fluid cultures. J Urol 1984;132:471–3. 7. Hill JA, Haimovici F, Polish JA, Anderson DJ. Effects of soluble products of activated lymphocytes and macrophages (lymphokines and monokines) on human sperm motion parameters. Fertil Steril 1987;47: 460 –5. 8. Fowler JE. Infections of the male reproductive tract and infertility: a selected review. J Androl 1981;3:121–31. 9. Cottell E, McMorrow J, Lennon B, Fawzy M, Cafferkey M, Harrison RF. Microbial contamination in an in vitro fertilisation-embryo transfer system. Fertil Steril 1996;66:776 – 80. 10. Cottell E, Lennon B, McMorrow J, Barry-Kinsella C, Harrison RF. Processing of semen in an antibiotic-rich culture medium to minimize microbial presence during in vitro fertilization. Fertil Steril 1997;67: 98 –103. 11. World Health Organization. WHO laboratory manual for the examination of human semen and sperm-cervical mucus interaction. 4th ed. Cambridge: Cambridge University Press, 1999. 12. Cordell JL, Falini B, Erber WN, Gosh AK, Abdulaziz Z, MacDonald S, et al. Immunoenzymatic labelling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP) complexes. J Histochem Cytochem 1984;32: 219 –29. 13. Boucher P, Lejeune H, Pinatel MC, Gille Y. Spermoculture: improve-
469
14. 15. 16.
17. 18.
ment of the bacteriological quality of samples by direct verbal counselling before semen collection. Fertil Steril 1995;64:657–9. Megory E, Zuckerman H, Shoman Z, Lunenfeld B. Infection and male infertility. Obstet Gynecol Surv 1987;42:283–90. Taylor-Robinson D. Evaluation of the role of Ureaplasma urealyticum in infertility. Pediatr Infect Dis 1986;5:262–5. Hay PE, Thomas BJ, Gilchrist C, Palmer HM, Gilroy CB, TaylorRobinson D. The value of urine samples from men with non-gonococcal urethritis for the detection of Chlamydia trachomatis. Genitourin Med 1991;67:124 – 8. Paul ID, Caul EO. Evaluation of three Chlamydia trachomatis immunoassays with an unbiased, noninvasive clinical sample. J Clin Microbiol 1990;28:220 –2. Rowe PJ, Comhaire FH, Hargreave TB, Mellows HJ. WHO manual for
470
Cottell et al.
Urogenital flora and leukocytospermia
19. 20. 21. 22. 23.
the standardized investigation and diagnosis of infertile couples. Cambridge: Cambridge University Press, 1993. Meares EM, Stamey TA. The diagnosis and management of bacterial prostatitis. Br J Urol 1972;44:175–9. Willen M, Holst E, Myhre EB, Olsson AM. The bacterial flora of the genitourinary tract in healthy fertile men. Scand J Urol Nephrol 1996; 30:387–93. Stamey TA. Prostatitis. Monogr Urol 1981;2:131. Hewitt J, Cohen J, Fehilly CB, Rowland G, Steptoe P, Webster J, et al. Seminal bacterial pathogens and in vitro fertilization. J In Vitro Fert Embryo Transfer 1985;2:105–7. Guillet-Rosso F, Fari A, Taylor S, Forman R, Belaisch-Allart J, Testart J, et al. Systematic semen culture and its influence on IVF management. Br J Obstet Gynaecol 1987;94:543–7.
Vol. 74, No. 3, September 2000
Are seminal fluid microorganisms of significance or merely contaminants? Evelyn Cottell, Ph.D.,a Robert F. Harrison, M.D.,a Mary McCaffrey, M.R.C.O.G.,a Tom Walsh, M.B.,a Eimear Mallon, M.R.C.G.P.,b and Carole Barry-Kinsella, M.D.a Royal College of Surgeons in Ireland and Human Assisted Reproduction Ireland, Rotunda Hospital, Dublin, Ireland
Objective: To determine the contribution of urethral and skin flora to seminal fluid cultures and the relation between bacteriospermia and seminal leukocytes. Design: Prospective study. Setting: IVF-ET unit at a university teaching hospital. Patient(s): Sixty men starting an IVF-ET program. Intervention(s): Culture of sequential first-catch urine, midstream urine, and semen samples with evaluation of seminal leukocytes. Main Outcome Measure(s): A comparison of microbes from first-catch urine, midstream urine, and semen samples and the correlations of seminal microbes, elevated leukocyte concentrations, and pregnancy. Result(s): Microorganisms were detected in 37% of first-catch urine samples, 27% of midstream urine samples, and 51% of semen samples. Most microorganisms were gram-positive microbes and were common to both urine and semen samples. Mean and median leukocyte concentrations were 0.98 ⫻ 106/mL and 0.10 ⫻ 106/mL, respectively. There was no correlation between seminal microbes and raised leukocytes or between leukocytospermia and/or bacteriospermia and pregnancy. Conclusion(s): Microorganisms are commonly found in insignificant quantities in the semen of asymptomatic men. The frequent isolation of gram-positive microbes common to both urine and semen and the absence of a correlation with raised leukocyte concentrations suggest that bacteriospermia most commonly represents contamination. (Fertil Steril威 2000;74:465–70. ©2000 by American Society for Reproductive Medicine.) Key Words: Semen, urine, microorganisms, leukocytes Received February 17, 2000; revised and accepted May 1, 2000. Supported by a research grant from the Royal College of Surgeons in Ireland, Dublin, Ireland. Reprint requests: Evelyn Cottell, Ph.D., Royal College of Surgeons in Ireland, Academic Department of Obstetrics and Gynaecology, Rotunda Hospital, Dublin 1, Ireland (FAX: 353-1-8727831; Email: [email protected]). a Academic Department of Obstetrics and Gynaecology, Royal College of Surgeons in Ireland, Rotunda Hospital. b Human Assisted Reproduction Unit, Rotunda Hospital. 0015-0282/00/$20.00 PII S0015-0282(00)00709-3
The presence of microorganisms in the semen of men in subfertile relationships is widely documented. However, conflicting reports on the effect of these microbes on semen quality have created difficulties in establishing their pathophysiologic role in male infertility (1–3). Discrepant results may be due to a number of factors. First, various sample-collection procedures have been used. Second, differing definitions of bacteriospermia are commonly used. Some reports have included all bacteria isolated in semen regardless of the bacterial strain, the colony count, or the relation with other bacteria, whereas other studies considered organisms only when present in concentrations of ⬎104/mL or potentially pathogenic microbes. Third, contamination of semen by nonpathogenic commensals of the skin, glans penis, or the lower urethra creates difficulties in identifying true pathology (4 – 6). Finally, seminal
microbes may be evaluated without considering the seminal leukocyte concentration. Because leukocytes may have an effect on the functional capacity of the spermatozoa (7), their contribution to the spermiogram is important. Semen is composed of spermatozoa from the epididymis and secretions from the seminal vesicles, prostate, and bulbourethral glands, and each region is normally sterile (8). It is therefore probable that seminal microbes result mainly from contamination by either urethral flora or skin commensals. Although commensal organisms may be of no significance in natural conception, contamination of a culture system by seminal microorganisms during assisted reproductive techniques may limit success (9, 10). This concern provides a new impetus for revisiting a previously recognized 465
problem. If urethral contamination does contribute significantly to bacteriospermia, midstream urine samples may have lower concentrations of microbes than first-catch urine samples, and collection of semen samples after passing urine may be an effective way of reducing microorganisms before IVF. This study first examined the microbial composition of first-catch urine, midstream urine, and semen samples from men in subfertile relationships to establish the contribution of commensal flora to seminal fluid cultures. Second, seminal microbes were evaluated in relation to leukocyte composition, and correlations were examined among microbes, elevated leukocytes, and pregnancy.
MATERIALS AND METHODS The Rotunda Hospital Ethics Committee granted approval for this study. Sixty men in subfertile relationships, about to start an IVF-ET program at this hospital, were recruited. Informed consent was obtained. Both verbal and written instructions were given to all men regarding abstinence, hygiene, and specimen production. Each man provided an early-morning first-catch urine sample (10 –20 mL), followed by a midstream urine sample (10 –20 mL), and finally a semen sample, taken by masturbation. Samples were taken at home into sterile containers provided (Sterilin; Abergaryoed, Gwent, United Kingdom) and were brought to the laboratory within 1 hour of collection. Before sample collection, the men were instructed verbally and in writing to abstain from ejaculation for 2–3 days. The men were asked to wash their hands and genital area with soap and water and to towel dry before taking the first urine sample (11). All men completed a questionnaire concerning histories of sexually transmitted disease (STD), genital tract surgery, urinary tract infection, dysuria, and circumcision. After liquefaction and under sterile conditions, the semen sample was mixed with a sterile pipette. Aliquots were removed for semen analysis, immunocytochemical staining of leukocytes, and microbiologic evaluation. Semen analysis was performed with a Horwell Fertility counting chamber (A. R. Horwell, London, United Kingdom) and a phase contrast microscope. Seminal fluid leukocyte concentrations were determined by an immunocytochemical stain with an anti–HLe-1 monoclonal antibody (Becton Dickinson, San Jose, CA) with the use of an alkaline phosphatase anti– alkaline phosphatase staining technique (12). Leukocytospermia was defined as the presence of ⱖ1 ⫻ 106 leukocytes per milliliter of semen (11). An aliquot of the first-catch urine, midstream urine, and liquefied semen samples was reserved for culture of aerobic and facultative anaerobic microorganisms. For aerobic cultures, blood agar plates and chocolate plates for Gonococcus species culture were inoculated with standard loops (0.001 466
Cottell et al.
Urogenital flora and leukocytospermia
mL) of specimen. The plates were cultured for 48 hours in 5% CO2 in air at 37°C. Cultures were identified with the commercial kits Streptex and Candida albicans (Murex Diagnostics Inc., Norcross, GA), Staphlase (Oxoid; Unipath, Basingstoke, Hampshire, United Kingdom), and API 20 E (bioMe´rieux, Marcy l’Etoile, France). Diphtheroids and Escherichia coli were identified directly from cultures. For anaerobic cultures, blood agar plates were inoculated with standard loops and cultured for 48 hours at 37°C under anaerobic conditions, achieved with a commercial gas-generating system (Oxoid). Anaerobic microbes were identified by a Mastring ID kit (Mast Diagnostics, Bootle, Merseyside, United Kingdom). Aerobic and anaerobic cultures were considered positive when ⱖ103 organisms per milliliter, equivalent to one observed colony, were noted. Significant bacteriospermia was defined as ⱖ104 nonpathogenic organisms per milliliter of semen or ⱖ103 potentially pathogenic organisms per milliliter. Nonpathogenic organisms included Staphylococcus epidermidis, nonhemolytic streptococci, diphtheroids, and ␣hemolytic streptococci; potential pathogens included E. coli, -hemolytic streptococci, anaerobes, Proteus species, coliforms, and mixed microbes. Colony counts of ⱖ100 in midstream urine samples were considered clinically significant. First-catch urine and semen samples also were analyzed for the primary genital mycoplasmas, Mycoplasma hominis and Ureaplasma (u.) urealyticum, with the use of a mycoplasma kit, Mycofast “All-In” (International Mycoplasma, Toulon, France). Mycoplasma counts of ⱖ104 colony-forming units (cfu) were considered significant. The remainder of the first-catch urine was analyzed for Chlamydia with an amplified enzyme immunoassay (IDEIA Chlamydia; DAKO Diagnostics Ltd., Ely, Cambridgeshire, United Kingdom). Urine was transferred to conical Falcon tubes (Becton Dickinson Ltd., Plymouth, England) and centrifuged at 3,000 ⫻ g for 15 minutes. The supernatants were discarded, and the deposits were resuspended in 1 mL of supplied workingstrength Chlamydia transport medium in a small screwcapped glass bottle. Samples were stored at ⫺70°C and analyzed in batches for Chlamydia antigen. The effect of significant bacteriospermia and/or leukocytospermia on the pregnancy rate (PR) was examined. A clinical pregnancy was defined as the presence of a gestational sac at 5 weeks after ET. Relation between significant bacteriospermia and leukocytospermia, circumcision, and pregnancy; between leukocytospermia and pregnancy; and between bacteriospermia combined with leukocytospermia and pregnancy were examined by calculating odds ratios (ORs) and their confidence intervals (CIs). Female age and oocyte number were compared in the pregnant and nonpregnant groups with Student’s t-tests at the 95% confidence level. Statistical analysis was performed with a Stata Release 6 package at the Department of Psychology, Royal College of Surgeons in Ireland. Vol. 74, No. 3, September 2000
TABLE 1
TABLE 2
Percentage isolation rate of microorganisms from firstcatch urine, midstream urine, and semen samples from 60 men in subfertile relationships.
Colony counts in 60 semen samples.
Isolation rate (%) Microorganism
Microorganism
Firstcatch urine
Nonpathogens Staphylococcus epidermidis 10.0 (6) Nonhemolytic streptococci 1.7 (1) Diphtheroids 3.3 (2) Ureaplasma urealyticum 1.7 (1) ␣-Hemolytic streptococci 1.7 (1) Potential pathogens -Hemolytic streptococci 13.3 (8) Escherichia coli 6.7 (4) Proteus species 1.7 (1) Coliforms 1.7 (1) Mixed microbes 3.3 (2) Anaerobes 0.0 (0)
Midstream urine
Semen
Semen (strict criteriaa)
11.7 (7) 1.7 (1) 6.6 (4) 0.0 (0) 3.3 (2)
25.0 (15) 11.7 (7) 16.7 (10) 15.0 (9) 11.7 (7) 10.0 (6) 6.7 (4) 1.7 (1) 3.3 (2) 3.3 (2)
6.7 (4) 5.0 (3) 1.7 (1) 0.0 (0) 0.0 (0) 0.0 (0)
20.0 (12) 20.0 (12) 3.3 (2) 3.3 (2) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1) 1.7 (1)
Note: Values in parentheses represent number of samples. a Only ⱖ104 nonpathogenic organisms per milliliter or ⱖ103 potentially pathogenic organisms per milliliter are reported. Cottell. Seminal fluid microorganisms. Fertil Steril 2000.
RESULTS Microorganisms were detected in 22 (37%) of the firstcatch urine samples, 16 (27%) of the midstream urine samples, and 33 (54%) of the semen samples. The majority of isolates were gram-positive bacteria, including staphylococci, streptococci, and diphtheroids. There were low frequencies of isolation of potentially pathogenic gram-negative bacilli, anaerobes, and mycoplasmas, and no isolation of Chlamydia or Gonococcus (Table 1). Bacterial isolation rates in semen are presented first as any presence of bacteria and second as a significant presence, on the basis of strict criteria described in Materials and Methods. Table 2 presents the mean colony counts, SDs, and ranges for each organism isolated. In 20 men (33.3%), cultures on all three samples were negative. In a further 14 men (23.3%), both urine samples were sterile. However, corresponding semen cultures were positive. Nine men (15%) had triple growth of organisms from first-catch urine, midstream urine, and semen samples. The remaining 17 men (28.3%) had microbes isolated from either first-catch urine and/or midstream urine and/or semen. Of the 22 men with positive cultures for first-catch urine, 10 (45.5%) had negative midstream urine cultures. Twelve men (20%) had microbes grown from both urine samples, of whom 2 had identical colony counts for both samples and 6 had lower colony counts of the same organism in the midstream urine sample. One man had a higher microbial count FERTILITY & STERILITY威
No. of Colony count samples with (⫻ 103 orgs/mL) microbe Mean ⫾ SD Range isolated
Nonpathogens Staphylococcus epidermidis Nonhemolytic streptococci Diphtheroids ␣-Hemolytic streptococci Ureaplasma urealyticum
15 10 7 2 4
10.4 ⫾ 12.5 1–50 43.7 ⫾ 41.3 5–100 27.9 ⫾ 20.2 6–56 8.0 ⫾ 5.7 4–12 5.5 ⫾ 5.2 1–10 (⫻ 103 cfua) (⫻ 103 cfua)
Potential pathogens -Hemolytic streptococci Escherichia coli Proteus species Coliforms Mixed microbes Anaerobes
12 2 1 1 1 1
26.6 ⫾ 30.3 48.0 ⫾ 59.4 12.0 11.0 ⬎100 80
2–100 6–90 – – – –
a
Ureaplasma urealyticum was expressed in colony-forming units (cfu) as opposed to organisms per milliliter. Cottell. Seminal fluid microorganisms. Fertil Steril 2000.
in the midstream urine sample and, on three occasions, different combinations of microbes were isolated from the two urine samples. Of the 44 men who had negative midstream urine samples, their corresponding semen cultures were positive in 20 cases (45%) and negative in 24 cases (55%). The mean concentration, median value, and range for seminal leukocytes were 0.98 ⫻ 106/mL, 0.10 ⫻ 106/mL, and 0 –14.0 ⫻ 106/mL, respectively. No correlation was found between elevated seminal leukocytes (leukocytospermia) and the significant presence of microorganisms (OR 1.25, 95% CI 0.35– 4.4). Of the 11 men who had leukocyte concentrations of ⱖ1 ⫻ 106/mL, 5 had negative seminal fluid cultures. The remaining 6 men with leukocytospermia had significant bacteriospermia. These included 2 men with low growth (2 ⫻ 103 and 7 ⫻ 103 organisms/mL) and 1 man with moderate growth (31 ⫻ 103 organisms/mL) of -hemolytic streptococci. The remaining 3 men had heavy growth of mixed microbes (⬎105 organisms/mL), mixed growth of nonhemolytic streptococci (99 ⫻ 103/mL) and -hemolytic streptococci (23 ⫻ 103/mL), and moderate growth of U urealyticum (104 cfu/mL). Data from the questionnaire showed that one man had a history of an STD (urethritis due to C trachomatis) and two men had dysuria. Seven men (11.7%) had been circumcised. ORs and CIs showed no correlation between significant concentrations of seminal microbes and circumcision (OR 1.33, 95% CI 0.30 –5.85). There were no correlations observed between either raised concentrations of microbes or leukocytospermia and a history of infection or dysuria, although the low number of cases in these two groups prevents meaningful statistical analysis. 467
Of the 60 couples who had an IVF-ET cycle, 19 (31.7%) were successful in attaining a clinical pregnancy, all resulting in live births. The mean female age and number of oocytes collected were 34.1 and 34.5 years and 10.0 and 9.5 eggs, respectively, for the pregnant and nonpregnant groups. No statistically significant difference was found between the groups for either variable (Student’s t-test at the 95% level). Although the odds of pregnancy were slightly lower in the presence of leukocytospermia (OR 0.77, 95% CI 0.2–3.1), this was not statistically significant. The PR was slightly higher in the presence of significant bacteriospermia (OR 2.8, 95% CI 0.9 – 8.5), although this difference did not reach statistical significance. Likewise, when the combined factors of leukocytospermia and significant bacteriospermia were examined in relation to the PR, no statistically significant difference was observed between the presence or absence of this combined condition (OR 0.39, 95% CI 0 –2.8).
DISCUSSION The prevalences of bacteriospermia in men of both fertile and subfertile relationships range from 10% to 100% in the published literature (8). This wide range may reflect the variety of semen-collection protocols used. The World Health Organization (11) recommended that special precautions be taken to avoid microbial contamination when collecting semen samples for culture. They advise strict hygiene and stress the importance of first passing urine before producing a semen sample by masturbation. Boucher et al. (13) reported a significant reduction in bacterial count and number of species and an increase in the percentage of sterile cultures when direct verbal counseling, as opposed to written instructions only, was given to men before specimen production. In this microbial investigation of urogenital fluids, both written and verbal instructions were given to each man, and all semen samples were obtained by masturbation immediately after passing urine. However, despite these precautions, seminal microorganisms were recovered in 54% of the cases. Common microorganisms were isolated from first-catch urine, midstream urine, and semen samples, although a higher prevalence was found in semen. The seminal fluid isolation rate of 6.7% for U urealyticum is lower than that found elsewhere (14). Urethral colonization by ureaplasmas is common (15), and the low isolation rate from semen may reflect a washout effect in the urethra during the passing of urine immediately before producing the semen sample. Although only one first-catch urine sample was positive for U urealyticum, higher detection rates may have been found if centrifuged deposits of urine samples were examined. The positivity rate for Chlamydia, with use of the same amplified ELISA technique as in this study, among men and women attending STD clinics in Dublin ranges from 7% to 9%. Therefore, the negative results for Chlamydia in this 468
Cottell et al.
Urogenital flora and leukocytospermia
study most likely reflect a low incidence of active chlamydial infection in our infertile population, but also may be due to insensitivity of the enzyme immunoassay used. However, this amplified ELISA has shown sensitivities comparable to cell culture of 400 endocervical and urethral swabs (unpublished data, Virus Reference Centre, Ardmore, Dublin, Ireland). Moreover, a centrifuged deposit of a first-catch earlymorning urine sample may be an acceptable noninvasive sample for detecting Chlamydia in the male urethra (16, 17). Although quantitative assessment of microbial growth was determined for all samples, it was valid to compare colony counts only in first-catch and midstream urine samples because of differences in volume from urine to semen. More than one third of all men had microbes isolated from their first-catch urine samples, with colony counts of ⬎100 in only two cases. Microorganisms were either absent or reduced in number in the corresponding midstream urine samples of 73% of these men, probably reflecting a dilution of microorganisms out of the distal urethra with the firstcatch urine. The colony count in midstream urine samples was never of clinical significance. Among the 44 cases in which midstream urine cultures were negative, 20 (45%) of the corresponding semen samples had positive cultures. The probability that these culture results represent infection is unsupported by clinical, bacteriologic, and immunologic considerations. However, limitations to the sensitivity of the specimen types examined preclude dismissal of prostatitis and urethritis in our study group. Criteria for the diagnosis of male accessory gland infection must take into consideration the type of specimen analyzed, the organism detected, the colony count, the presence of ⬎106 leukocytes per milliliter of semen, and a history or clinical signs of genital tract infection (18). Although bacterial infection of the prostate is usually associated with irritative voiding symptoms and varying degrees of perineal or suprapubic discomfort (14), chronic bacterial prostatitis may be asymptomatic. Diagnosis can be made only with quantitative segmental cultures of first and second voided urine samples, expressed prostatic secretions, and finally a third voided urine sample (Meares and Stamey test) (19). The urethral swab has been the traditional specimen to test for urethritis. Semen and unconcentrated first-catch urine samples are probably not sensitive enough to detect urethritis, and centrifuged deposits from first-catch urine samples may have been a better sample type because they have proved as sensitive as swabs for Chlamydia detection when the same test is applied (16). Of the 20 men with negative midstream urine and positive seminal fluid cultures, 2 reported pain on passing urine. The first man had had a urinary tract infection treated 6 months before the study. Only nonhemolytic streptococci (22 ⫻ 103/mL) were cultured from his semen, and the seminal leukocyte concentration was not raised (0.6 ⫻ 106/mL). Vol. 74, No. 3, September 2000
Culture results for the second man showed U urealyticum, S epidermidis, and diphtheroids, all with low colony counts. No leukocytes were detected in this semen sample. It is unlikely that these men had prostatitis because these infections are rarely caused by more than one organism, and the infecting bacterium can usually be isolated persistently and in large concentrations from either expressed prostatic secretions or seminal fluid. However, a Meares and Stamey test (19) is required to clearly exclude prostatitis. Neither the two men with dysuria nor the man with a history of an STD had seminal microbes or leukocytes consistent with infection of the accessory glands, although the low numbers prevent meaningful statistical analysis. It is therefore likely that the seminal organisms cultured from their samples and from those of the other 18 men with negative midstream urine samples were contaminants from the urethra, glans penis, or hands during sample collection. This is supported by the work of Willen et al. (20), who found that 71% of strains colonizing the coronal sulcus were also present in the distal urethra and that urethral flora may then contaminate the semen. The different rheologic properties of semen and urine passing through the urethra, the different hydrostatic pressure conditions in the urethra during voiding and masturbation, and the manipulations to the penis during masturbation, with potential contamination by cutaneous bacteria adjacent to the urethral meatus, may explain the higher isolation rate of organisms in semen (21). These findings are supported by a small study by Fowler and Mariano (6), who reported greater contamination of expressed prostatic secretions and seminal fluid than of voided urine by the aerobic bacterial flora of the urethra. Microorganisms were detected frequently in semen, and although the majority were nonpathogenic commensals or contaminants, their presence may be of greater significance in IVF-ET, where natural defenses in the female genital tract (22), including antimicrobial secretion of lysozyme and lactoferrin, together with phagocytic cell production of immunoglobulin, are bypassed. Contamination of the culture system with seminal microbes may lead to suboptimal fertilization rates or impaired embryonic development (23) if adequate seminal processing techniques with antibiotic-rich culture medium are not used (10). In this study, common organisms were detected in urine and semen samples from the same individual. Most were gram-positive microbes normally associated with the urethra. Despite efforts to minimize microbial contamination of semen by first passing urine and adhering to instructions for genital hygiene (11), we isolated seminal fluid microbes in ⬎50% of samples. Although both fluids share a final common pathway in the urethra, semen appears to be more susceptible than urine to microbial contamination from the urethra, glans penis, or hands. Even if urethral contamination could be eliminated, the precise origin of microbes in semen FERTILITY & STERILITY威
could not be determined because of the different sources of seminal fluid components. Therefore, the role of infection in male infertility might be analyzed best with respect to the infecting microorganism rather than to the suspected site of infection. Positive seminal fluid cultures must be interpreted with caution, taking into account the presence of normal or abnormal urethral flora, raised colony counts of single isolates, and seminal leukocyte concentrations. These variables should be evaluated with a thorough history and physical examination to uncover symptoms and signs of infection. When both raised colony counts and leukocytes are observed, further investigation with urethral swabs and prostatic massage should ensue. Only then can physicians avoid the common misdiagnosis of genital tract infection, based on the presence of seminal microbes, and the possible unnecessary treatment with antibiotics.
Acknowledgments: The authors thank the laboratory staff, nurses, and clinicians at the Human Assisted Reproduction Ireland unit; the staff of the microbiology laboratory, Rotunda Hospital for their cooperation with this study; and Ronan Conroy of the Department of Psychology, Royal College of Surgeons in Ireland for statistical analyses.
References 1. Harrison RF, Hannon K, Butler M. Aerobic and anaerobic culture of human semen and the relationship to characteristics in the spermiogram. Fortschr Androl 1983;8:46 –54. 2. Busolo F, Zanchetta R, Lanzone E, Cusinato R. Microbial flora in semen of asymptomatic infertile men. Andrologia 1984;16:269 –75. 3. Huyser C, Fourie FLeR, Oosthuizen M, Neethling A. Microbial flora in semen during in vitro fertilisation. J In Vitro Fert Embryo Transfer 1991;8:260 – 4. 4. Desai S, Cohen MS, Khatamee M, Leiter E. Ureaplasma urealyticum infection: does it have a role in male infertility? J Urol 1980;124:469 – 71. 5. Granouillet R, Gaudin OG, Laurent JC, Rousset H, Moulin A. Study of the spermatic bacterial flora in infertile males. Pathol Biol 1982;30: 22– 6. 6. Fowler JE, Mariano M. Difficulties in quantitating the contribution of urethral flora to prostatic fluid and seminal fluid cultures. J Urol 1984;132:471–3. 7. Hill JA, Haimovici F, Polish JA, Anderson DJ. Effects of soluble products of activated lymphocytes and macrophages (lymphokines and monokines) on human sperm motion parameters. Fertil Steril 1987;47: 460 –5. 8. Fowler JE. Infections of the male reproductive tract and infertility: a selected review. J Androl 1981;3:121–31. 9. Cottell E, McMorrow J, Lennon B, Fawzy M, Cafferkey M, Harrison RF. Microbial contamination in an in vitro fertilisation-embryo transfer system. Fertil Steril 1996;66:776 – 80. 10. Cottell E, Lennon B, McMorrow J, Barry-Kinsella C, Harrison RF. Processing of semen in an antibiotic-rich culture medium to minimize microbial presence during in vitro fertilization. Fertil Steril 1997;67: 98 –103. 11. World Health Organization. WHO laboratory manual for the examination of human semen and sperm-cervical mucus interaction. 4th ed. Cambridge: Cambridge University Press, 1999. 12. Cordell JL, Falini B, Erber WN, Gosh AK, Abdulaziz Z, MacDonald S, et al. Immunoenzymatic labelling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP) complexes. J Histochem Cytochem 1984;32: 219 –29. 13. Boucher P, Lejeune H, Pinatel MC, Gille Y. Spermoculture: improve-
469
14. 15. 16.
17. 18.
ment of the bacteriological quality of samples by direct verbal counselling before semen collection. Fertil Steril 1995;64:657–9. Megory E, Zuckerman H, Shoman Z, Lunenfeld B. Infection and male infertility. Obstet Gynecol Surv 1987;42:283–90. Taylor-Robinson D. Evaluation of the role of Ureaplasma urealyticum in infertility. Pediatr Infect Dis 1986;5:262–5. Hay PE, Thomas BJ, Gilchrist C, Palmer HM, Gilroy CB, TaylorRobinson D. The value of urine samples from men with non-gonococcal urethritis for the detection of Chlamydia trachomatis. Genitourin Med 1991;67:124 – 8. Paul ID, Caul EO. Evaluation of three Chlamydia trachomatis immunoassays with an unbiased, noninvasive clinical sample. J Clin Microbiol 1990;28:220 –2. Rowe PJ, Comhaire FH, Hargreave TB, Mellows HJ. WHO manual for
470
Cottell et al.
Urogenital flora and leukocytospermia
19. 20. 21. 22. 23.
the standardized investigation and diagnosis of infertile couples. Cambridge: Cambridge University Press, 1993. Meares EM, Stamey TA. The diagnosis and management of bacterial prostatitis. Br J Urol 1972;44:175–9. Willen M, Holst E, Myhre EB, Olsson AM. The bacterial flora of the genitourinary tract in healthy fertile men. Scand J Urol Nephrol 1996; 30:387–93. Stamey TA. Prostatitis. Monogr Urol 1981;2:131. Hewitt J, Cohen J, Fehilly CB, Rowland G, Steptoe P, Webster J, et al. Seminal bacterial pathogens and in vitro fertilization. J In Vitro Fert Embryo Transfer 1985;2:105–7. Guillet-Rosso F, Fari A, Taylor S, Forman R, Belaisch-Allart J, Testart J, et al. Systematic semen culture and its influence on IVF management. Br J Obstet Gynaecol 1987;94:543–7.
Vol. 74, No. 3, September 2000