Effectiveness and limits of antimicrobial treatment on seminal leukocyte concentration and related reactive oxygen species production in patients with male accessory gland infection

E. Vicari

Andrology Centre, Department of Internal Medicine, University of Catania, Via A.Diaz 15, 95125 Catania, Italy


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
To evaluate whether bacteriological cure, sperm outcome, spontaneous pregnancy rate and white blood cell (WBC)-related reactive oxygen species (ROS) production were related to the extent of the infection and to an intermittent and repetitive antimicrobial treatment, 122 patients with bacterial [>105 colony-forming units (CFU)/ml] male accessory gland infections (MAGI) were studied. According to ultrasound criteria, patients had prostatitis (PR, n = 52), prostatovesiculitis (PV, n = 32) or prostatovesiculoepididymitis (PVE, n = 38). Each group was further subdivided into two subsets: one subset (PR, n = 40; PV, n =20; PVE, n = 25) was given ofloxacin or doxycycline for 14 consecutive days per month for 3 months; the other subset (PR, n = 12; PV, n = 12; PVE, n = 13) received no treatment. The female partners were also treated. All patients were evaluated before, during (1 and 3 months) and after (3 months) treatment. The bacteriological cure rate was the highest (92.5%) after the third antibiotic course in PR, followed by PV (70.4%), and the lowest in PVE (52.0%). At 3 months after therapy discontinuation, some sperm parameters, seminal WBC concentration and ROS generation (assessed in the 45% Percoll fraction) were ameliorated in PR and PV, whereas no improvement occurred in patients with PVE, except for the percentage of coiled tails. Antibiotic treatment in PR and PV patients led to positive effects on sperm output and spontaneous pregnancy rate (40%) by removing pro-oxidant noxae (microbial and/or WBC-related ROS production). The persistent infertility, dyspermia and sperm-derived ROS overproduction in PVE may relate to a significant percentage of antibiotic-independent re-infection and/or to low antioxidative epididymal properties, which persisted following antimicrobial treatment.

Key words: antimicrobial treatment/male accessory gland infections/ROS production/seminal white blood cells/sperm parameters


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A number of controversial results have been published with regard to the effects of antimicrobial therapy on sperm quality and pregnancy rate in male accessory gland infections (MAGI) (Homonnai et al., 1975Go; Nikkanen et al., 1979Go; McGowan et al., 1981Go; Giamarellou et al., 1984Go; Comhaire et al., 1986Go; Rhemrev et al., 1987Go; Soffer et al., 1990Go; Weidner et al., 1991Go; Krause, 1994Go; Liversedge et al., 1996Go; Omu et al., 1998Go). The following might be considered the reasons for this discrepancy: (i) treatment of a heterogeneous MAGI population, in which only men with a chronic active infection with the presence of microorganisms and a limited degree of damage may benefit from a suitable therapy, as suggested previously (Purvis and Christiansen, 1995Go) and subsequently proved (Weidner et al., 1998Go) in patients with chronic bacterial prostatitis; (ii) antibacterial therapy may have been inadequate in terms of dosage and/or duration in chronic and often asymptomatic MAGI patients, probably including infection of one or more sexual accessory glands; and (iii) a white blood cell (WBC) and/or sperm-generated production of reactive oxygen species (ROS) may persist even after antimicrobials in some patients affected by complicated MAGI, for example prostatovesiculoepididymitis (PVE) (Vicari et al., 1997Go).

Recently, a strong association has been demonstrated between sexual gland post-inflammation damage found by an abnormal ultrasound and increased WBC concentration and ROS production, particularly in patients affected by PVE compared with values seen in patients with prostatitis (PR) or prostatovesiculitis (PV) (Vicari, 1999Go). Thus, a prospective, randomized, controlled trial was carried out on 122 asymptomatic, infertile males with a high bacterial count [>105 colony-forming units (CFU)/ml] and ultrasound evidence of MAGI limited to the prostate (n = 52), extended to seminal vesicles (n = 32), or also to the epididymis (n = 38), to evaluate whether bacteriological cure, sperm outcome, spontaneous pregnancy rate and WBC-related ROS production were related to the extent of the infection and to an intermittent and repetitive treatment with two different antibiotics, ofloxacin and doxycycline. These antibiotics were chosen because they possess certain general characteristics (broad antibacterial spectrum against relevant Gram-negative, Gram-positive pathogens as well as Chlamydia trachomatis or Ureaplasma urealyticum; excretion primarily by the kidneys with minimal metabolism; sufficient drug penetration into the inflamed sexual glands; low rate of adverse sperm effects following short-term treatment) and specific in-vitro susceptibility. Doxycycline was proven to penetrate into the normal prostate due to acidity of the secretions and ionic trapping, but does not penetrate well into the diseased prostate because of the alkaline pH of the gland's secretions (Schramm, 1986Go; Cunha and Garabedian-Ruffalo, 1990Go; Naber et al., 1993Go).


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Subjects
Between March 1993 and March 1999, a total of 1187 male infertile patients was screened, and 122 married men (median age 32 years; range 25–43 years) presenting with primary infertility (median 6.2 years duration; range 2 to 10 years) with diagnosis of MAGI made according to World Health Organization clinical and laboratory criteria (WHO, 1993) were enrolled in the study. No female infertility factor was apparently present, as all female partners were ovulating regularly, based on biphasic basal body temperature and a follicular ultrasound scan, luteal phase progesterone concentrations, or endometrial biopsy. Furthermore, tubal patency was assessed by hysterosalpingogram or laparoscopy in all women. A complete reproductive and sexual history was also obtained and a physical examination performed on each male patient.

Rather, a male infertility factor might be postulated since: (i) both the results of the post-coital test and in-vitro bovine cervical mucus sperm penetration capacity were repeatedly negative. The former was true in 68.0% of cases [83/122 patients exhibiting <10 spermatozoa/high-power field (HPF) with directional motility in the endocervical specimen], and the latter in 73.8% of cases (90/122 samples exhibiting penetration <16 mm); and (ii) all patients had to exhibit oligo-, astheno- or teratozoospermia, as suggested by conventional criteria (WHO, 1993).

Clinical characteristics of infected categories
All infertile patients were affected by presumptive MAGI, suspected upon the basis of oligozoospermia (sperm concentration <20x106/ml), asthenozoospermia (<50% spermatozoa with forward progression, categories a and b), or teratozoospermia (<30% spermatozoa with normal oval form), and fulfilled conventional criteria (WHO, 1993). Eligibility criteria (clinical, laboratory and ultrasound examination) and exclusion criteria (i.e. history of smoking, alcoholic consumption, occupational chemical exposure) were as previously reported (Vicari, 1999Go). In particular, all patients had one or two consecutive cultures with significant bacteriospermia (>105 CFU/ml), C. trachomatis or U. urealyticum detected in cultures of urethral swabs obtained after prostatic massage. A cut-off value of >105 CFU/ml was used; this was much higher than the value cited earlier (WHO, 1992) because it was based on the strong association with ultrasound abnormalities, as reported recently (Vicari, 1999Go).

The clinical diagnosis of epididymitis was suspected when, at the physical examination of the epididymis, there were painful palpation, swelling and/or microcysts or micronodularity in the caput or caudal region from one or both sides. These clinical findings suggested that the ultrasound investigation to be second mandatory additional proof, even in order to distinguish accurately the site (PR, PV and PVE) of MAGI. To this end, ultrasound findings considered indicative of chronic inflammation were used, as reported recently (Vicari, 1999Go).

Study design and treatments
Before initiating antibiotic treatment, sperm and microbial assessments were performed on the three major categories (PR, PV, PVE) of the overall infected population (n = 122). Patients of each category were allocated at random into three subsets which were assigned either to antimicrobial therapy (ofloxacin or doxycycline) or to no therapy (controls). Randomization was achieved using a computer program. The first random set of numbers was assigned to treatment with ofloxacin, the second set to treatment with doxycycline, and the third set to controls. In particular, the first two random sets of numbers, including together 40 PR, 20 PV and 25 PVE, received ofloxacin (200 mg orally every 12 h; Flobacin®, SigmaTau, Italy) or doxycycline (100 mg orally once daily; Bassado®, Poli, Italy) intermittently for 14 days each month over a 3-month period. The third random set (12 PR, 12 PV and 13 PVE patients) received no treatment. All patients provided written informed consent; they completed the entire trial, including treatment and follow-up examinations.

In order to eradicate infection from each couple, the female partner was treated for 8 days/cycle (days 5 to 12) with the same antibiotic as used in the male partner. During this time, sexual intercourse using condoms was recommended until bacteria were eradicated in both partners.

The duration of treatment for each group was 14 consecutive days per month, with a subsequent 2-week washout period for a 3-month period. This treatment course was chosen arbitrarily on the basis of two considerations: (i) modern fluoroquinolones are recommended to be administered for a minimum of 2–4 weeks (Weidner, 1992Go; Bjerklund Johansen et al., 1998Go); and (ii) it was reported previously that, in some normozoospermic and most oligoasthenoteratozoospermic (OAT) infected patients, prolonged treatment with antimicrobials (>2 weeks) was accompanied by an antibiotic-independent `stress tubular pattern' (STP), characterized by reduced sperm concentration and increased percentages of both spermatozoa with abnormal morphology and sperm precursors (Vicari and Sidoti, 1991Go). This pattern was reversible between 3–9 months after discontinuation of therapy. Since in our experience the potential risk of STP was greater in patients with OAT, most patients in the present study might incur this adverse side effect and thus misrepresent the usefulness of antimicrobials, which in MAGI have a rational application.

Pre-treatment microbial and sperm characteristics of the different MAGI categories
The microbiological results of the three major clinical categories of MAGI are reported in Table IGo.


View this table:
[in this window]
[in a new window]
 
Table I. Bacterial isolates and sperm variables performed in overall infected population (n = 122) before the start of the therapeutical trial. Sperm data are expressed as median values and the 10th and 90th percentiles. See text for abbreviations
 
The same microorganism was always isolated in the two semen specimens cultured in the pre-treatment. From 122 samples, 71 (58.2%) aerobes (enterobacteria in 42 specimens, Gram-positive bacteria in 29 specimens) and 13 (10.7%) anaerobes were classified as a monoinfection, and there was no mixed infection. C. trachomatis or U. urealyticum were detected in 25 (20.5%) and 13 (10.7%) cultures of urethral swabs obtained after prostatic massage respectively. Enterobacteria were the dominant original causative organisms in all of the major clinical conditions, both complicated (PV or PVE groups) and uncomplicated (PR group) MAGI categories. Although the rates of occurrence of aerobic strains or U. urealyticum were not significantly different among infected MAGI categories, anaerobic isolates occurred at a higher (P < 0.05) rate in the PV patients, and the rate of occurrence of C. trachomatis was significantly (P < 0.01) higher in the PVE group (36.8%) than that found in the PR (19.2%) and PV (3.1%) patient groups (Table IGo).

Isolated bacterial strains exhibited a high susceptibility to both ofloxacin and doxycycline in the present study. In terms of MIC (minimum inhibitory concentration) values of the isolates, the random antibiotic (ofloxacin or doxycycline) chosen had lower MIC values (<1 mg/l) than any of the other antimicrobials tested in vitro.

Results of conventional semen analyses in the patients with PR and PV showed median values significantly (P < 0.01) higher than those found in patients with PVE. The values were not significantly different between the PR and PV patients (Table IGo). Seminal WBC concentrations in the PVE group were significantly (P < 0.01) higher than those found in PR patients, but not higher than those of PVE patients (Table IGo).

Sampling management
In the overall population, sperm and bacteriological effectiveness was assessed by semen analysis and culture tests based on CFU/ml measurements in baseline or pre-treatment condition (T0) or tests for Chlamydia and Mycoplasma in cultures of urethral swabs obtained following a prostate massage. Two semen specimens were collected by masturbation into a sterile polystyrene jar, after 3–5 days of sexual abstinence, spaced 5–10 days apart. Patients were requested to wash their hands and genital region carefully with bactericidal and antifungal soap, and to rinse the soap away with isotonic solution. All patients agreed to the above control analyses either during treatment, and two observations obtained 7 and 11 days after temporary withdrawal of the therapy at month 1 (T1) and month 3 (T3), and 90 days (range 90–103 days) after completion of therapy (T6).

After liquefaction, semen samples were analysed for sperm concentration (C; x106/ml), total sperm number (TSN; x106/ejaculate), forward motility percentage (M; % grade a+b, after 1 h), sperm morphology [percentages of both normal oval forms (Ov) and pathological coiled tails (Ct)] and seminal leukocyte concentration (WBC; x106/ml) according to WHO guidelines (WHO, 1992). In particular, conventional immunocytochemical staining (WHO, 1992) was used to assess seminal WBC concentration, according to previously published methods (Vicari, 1999Go). Furthermore, sperm preparation by a two-step Percoll 45/90% discontinuous gradient separation technique as well as determination of WBC-specific ROS production were performed using methods and WBC chemoattractant peptide fMPL (formyl-leucyl-phenylalanine; Sigma Chemical Co., St Louis, MO, USA) dissolved in dimethyl sulphoxide to measure the maximal WBC-specific ROS production, as reported previously (Vicari, 1999Go). The ROS measurements were made on 400 µl aliquots of cell suspension. ROS measurements were reported to a standard concentration of 2.5x106 spermatozoa/ml to reduce the number of WBC in the aliquots and consequently the number of `overflow', particularly in patients with PVE. The same sample was measured three times when overflow (>1000x1000 c.p.m.) values were detected the first time: in the presence of 2/3 counts >1000 (x 1000 c.p.m.), the final value assigned to the subject was arbitrarily 1000 (x 1000 c.p.m.).

An aliquot from all samples was also cultured aerobically and anaerobically after 1:2 dilution in saline solution. Standard bacteriological methods were used to quantify and identify all organisms according to previously published methods (Vicari et al., 1986Go). Primary infection agents were considered to be bacteria isolated in a concentration >105 CFU/ml from semen specimens.

In addition, all spermiograms were performed by the same investigator in a blinded fashion, thus minimizing methodological error.

Assessment of efficacy
The bacteriological response of all major MAGI categories to antimicrobial (ofloxacin or doxycycline) treatment was rated as a cure if a negative (0–103 CFU/ml) culture was present at the time when treatment was discontinued (TI or T3), or after completion of therapy (T6). Relapse (same species) or re-infection (different species) were assumed if significant bacteriospermia was present at any time point during treatment or follow-up.

Assessment of safety
The safety assessment included a medical history, a physical examination, haematology and serum chemistry profile at all visits, and monitoring for drug-related adverse events by indirect questioning and patient diaries.

Statistical analysis
In every phase of the trial (pre-treatment, treatment and follow-up examinations) sperm results were the mean of two consecutive semen specimens, collected 3–5 days apart and registered in each patient. Sperm results are reported as median, with the 10th and 90th percentile in parentheses. Within each subset, one-way variance analysis (ANOVA) followed by Duncan's multiple range test was employed to evaluate differences between assessment intervals and pre-treatment conditions for each parameter, throughout the study. A {chi}2 test and Fisher's exact test were used throughout the study to compare: (i) the pre-treatment percentages of microorganisms isolated of each MAGI category; (ii) the change in the proportion of patients who tested negative for seminal culture at T1, T3 or T6 within three subsets for each major infected category as well as between varying groups at T1, T3 or T6 respectively; and (iii) pregnancy rates (and pregnancy rate per cycle) in treated versus untreated cases. Comparisons between treated subsets and matched control groups were carried out using the non-parametric, Mann–Whitney U-test for unpaired data.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Safety assessment
Antibiotic regimens were generally well tolerated. No laboratory abnormalities were observed. Some mild and transient drug-related adverse effects (dizziness, headache, nausea, vomiting, abdominal pain or diarrhoea) were observed in 10 patients.

Bacteriological effects
In all control groups (no treatment), seminal cultures always remained positive with a bacteriological cure rate of 0% over time, which was significantly different compared with patients who became negative throughout the study (Table IIGo). There was no significant difference between treatment groups in all categories (data not shown). Bacteriological cure rates were significantly different among the different MAGI groups (PR > PV > PVE) in a time-dependent manner (Table IIGo). Indeed, at T1 and T3, bacteriological cure rates were 75.0% and 92.5% for PR patients, 40.9% and 70.4% for PV patients, and 16.0% and 52.0% for PVE patients. At T6, the bacteriological cure rates remained significantly different among infected categories, being 87.5%, 50.0% and 36.0% in PR, PV and PVE respectively. The decrease in the bacteriological cure observed at T6 was explained by a negligible relapse with the same bacterial species (growth of >105 E. coli or Proteus spp.) which occurred both in the treated PR subset on the one hand, and a significant percentage of antibiotic-independent re-infection with a different bacterial species which occurred in PV and PVE patients on the other hand. It is noteworthy that only Gram-positive bacteria caused re-infection (growth of >105 Enterococcus spp. or Staphylococcus aureus) that was resistant to both drugs (MIC 8–16 mg/l).


View this table:
[in this window]
[in a new window]
 
Table II. Bacteriological cure (BC, = CFU 0–103/ml) in response to antimicrobials (ofloxacin or doxycycline) according to the extent of the infection
 
In the female partners, a single treatment for 8 days/cycle (days 5 to 12), using the same antibiotic as in the male partner and sexual intercourse protected by condoms, was always sufficient to eradicate the pathogenic microorganisms in the cervical mucus specimens.

When the bacteriological cure was obtained, 24/85 (28.2%) patients treated with antimicrobials (17/40 with PR, 7/20 with PV, but 0/25 patients with PVE) and 2/37 (5.4%) patients not treated (both affected by PR) achieved a spontaneous pregnancy. The difference between the pregnancy rate in treated versus untreated cases was significant ({chi}2 test, P = 0.0097). It is noteworthy that while most patients (20/24; 76.9%) impregnated their partners following the third course of treatment, the remaining four patients (all with PR) became fertile following the first course of treatment, while the two patients with untreated PR fathered after the third month of follow-up. Thus, the pregnancy rate per cycle was also significantly different between treated versus untreated patients with PR or PV, particularly after one and three courses of treatment. In the former group, the pregnancy rate was respectively 10% (4/40) and 12% (13/108) versus 0% and 6% (2/36), while in the latter group it was 0% and 12% (7/60) versus 0% in both follow-up examinations. Finally, the pregnancy rate per cycle was not significantly different between treated versus untreated patients with PVE, being always 0% in both subsets.

Sperm outcome
Sperm outcomes, exhibiting significant changes throughout the trial, are reported in Tables III and IVGoGo. No statistical significant difference was found in the three subsets after randomization at T0. Sperm outcome did not vary in MAGI patients without treatment. However, following antibiotic treatment, only some sperm parameters showed significant changes in the three treated subsets throughout the trial (P < 0.01, ANOVA). The significant differences were always registered at T3 and/or at T6 versus T0 and T1 (P < 0.05, Duncan's multiple range test) within each category. Particularly, in comparison with pre-treatment sperm values, the percentages of forward motility showed a significant (P < 0.05) change in both the PR and PV treated subsets; total sperm number showed a significant (P < 0.05) response in the PV treated subset, while the percentages of pathological coiled tails were significantly (P < 0.05) lowered in all patient groups (Table IIIGo). The percentage of normal oval forms also remained unaffected during antibiotic treatment in all patient groups.


View this table:
[in this window]
[in a new window]
 
Table III. Effect of different antimicrobial (ofloxacin or doxycycline, pooled) treatment on sperm parameters during the trial in prostatitis, prostatovesiculitis and prostatovesiculoepididymitis subsets
 

View this table:
[in this window]
[in a new window]
 
Table IV. Changes in the seminal leukocyte (WBC) concentrations and (basal and fMLP-stimulated WBC-specific) reactive oxygen species (ROS) generation in the 45% Percoll fraction, in response to antimicrobials (ofloxacin or doxycycline, pooled) according to the extent of the infection.
 
With regard to the seminal WBC concentrations, while WBC values remained unmodified in matched controls, these values were significantly (P < 0.05) lowered in all treated subsets (Table IVGo), and in time-dependent manner. Furthermore, in the PR treated subsets, median seminal WBC concentrations at T3 and T6 were also significantly (P < 0.01, U-test) lower than those in PV and PVE treated subsets.

Although in all patient groups the 45% Percoll fraction WBC-specific ROS overproduction registered in the pre-treatment was significantly (P < 0.05) reduced in time-dependent manner, at T3 and T6 the chemiluminescent ROS signals were also significantly (P < 0.01) different among treated subsets (group P < PV < PVE). In particular, baseline (non-fMLP-stimulated) ROS counts in the PVE group remained significantly (P < 0.01) and markedly higher than those found in both the PR and PV groups (Table IVGo).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
MAGI, abnormal semen quality, abnormal ultrasound and no female infertility factor occurred in 122 out of 1187 (10.3%) infertile couples evaluated during a 6-year period. This relative frequency is comparable with that reported in literature (Andreeßen et al., 1993Go), but higher than the value of 1.6% among 2871 couples registered by others (Comhaire et al., 1986Go). This apparent difference might be explained in part through a non-random counselling of infertile couples from the area around our Andrology Centre, with a high frequency of just male factor, since diagnosis of MAGI in part had been previously made in other settings, and/or specific andrological clinical counselling was requested from us following previous IVF/embryo transfer programme failures.

Our data demonstrated that patients with bacterial MAGI are a heterogeneous composition of different clinical entities (PR, PV and PVE), with varying expression on the sperm output. Because of the well-defined homogeneity of the three patient groups, the efficacy of the antimicrobial treatment can be proven without a placebo-controlled trial. Indeed, an important reason for performing a placebo-controlled study is the clinical heterogeneity of patients under treatment (O'Donovan et al., 1993Go; Schill, 1995Go).

In comparison with matched (untreated) control groups, although occasional pregnancy occurred after the third month of follow-up in two patients who had untreated PR and showed both bacteriospermia and leukocytospermia that remained elevated and unmodified throughout the trial, only antibiotic therapy (ofloxacin or doxycycline) was efficacious in eradicating microorganisms and reducing the WBC concentration in sperm specimens. In contrast, antimicrobial treatment was also required because patients were selected suffering from andrological infection (>105 CFU/ml) and persistent leukocytospermia (Fedder, 1996Go), and not bland colonization (<104 CFU/ml) (Andreeßen et al., 1993Go).

Following antimicrobial treatment, although the bacteriological results were related to two factors—namely, the extent of infection (bacteriological cure in the PR group > PV group > PVE group) and duration of therapy (better at T3 than at T1)—sperm outcome (including WBC concentrations and ROS analysis) and the spontaneous pregnancy rate reflected both the above factors and the varying antibiotic-related antioxidant balance. During the trial, the sperm analysis was characterized by significant changes in three parameters (total sperm number, forward motility and coiled forms) in PV patients and by two parameters (forward motility and coiled forms) in PR patients, but by only one parameter (coiled forms) in PVE patients. During follow-up, these differences in sperm results were maintained at the same statistical significance among groups and compared with the pre-treatment period. The pre-treatment sperm parameters, as well as the post-antimicrobial sperm outcome detected in the three MAGI groups, could not be explained clearly. However, they could in part reflect varying pathogenically effective mechanisms, especially involving the complicated MAGI (PV and PVE) and the degree of their reversibility following antimicrobial treatment: direct effects of the microorganisms on spermatozoa (Gnarpe and Friberg, 1973Go; Swenson et al., 1979Go), ductal stenosis (Schoysman, 1981Go) or subclinical orchitis (Nilsson et al., 1968Go), secretory malfunction of the epididymis or/and the accessory sex glands (Cooper et al., 1990Go; Wolff et al., 1991Go; Gonzales et al., 1992Go), induction of sperm phagocytosis (Berger et al., 1982Go) or immunological processes, WBC response in terms of cytokines (Bar-Chama and Fisch, 1993Go) and ROS overproduction (Wang et al., 1997Go; Vicari, 1999Go). ROS produced during infections of the testis and epididymis are especially harmful to spermatozoa due to the longer contact time and the lack of antioxidant protection (Ochsendorf, 1999Go).

Our sperm findings are in agreement with the suggestion that the function of the seminal vesicles (Gonzales et al., 1992Go) or epididymis is important to prevent a deleterious effect of WBC on semen output or semen quality, through the above pathogenic mechanisms. Indeed, in agreement with other experimental models (Depuydt et al., 1996Go; Wang et al., 1997Go), we found varying changes of WBC and WBC response, in terms of ROS production, which was markedly enhanced in the presence of infectious (and probably cytokine-stimulating) factors in the three major MAGI groups. Following antibiotic treatment, although seminal WBC concentrations as well as ROS production (basal and not fMLP-stimulated) were lowered in patients with PR or PV, a clear juxta-sperm (basal and not fMLP-stimulated ROS hyperproduction) oxidative stress was persistent in patients with PVE. The varying antioxidative seminal balance registered among the three (treated or untreated) infected categories was directly associated with the spontaneous pregnancy rate that occurred in these patients, being 42.5%, 35% and 0% in treated patients with PR, PV or PVE respectively. On the other hand, among untreated (matched control) infected patients, only two (16.6%) with untreated PR impregnated their partners, while no patients with untreated PV or PVE were able to impregnate.

Thus, in patients with PVE a quantitatively persistent dyspermia and infertility could be secondary to a hostile epididymal microenvironment due to the residual presence of pro-oxidant factors (microorganisms, expression of some cytokines) (Omu et al., 1998Go) and/or inadequate antioxidant activity (epididymal secretory dysfunction) (Ochsendorf, 1999Go).

The present results are in agreement with those previously reported by our group, i.e. in patients with treated epididymitis, an unmodified dyspermia was associated with juxta-sperm ROS overproduction in the 90% Percoll fraction and in the seminal plasma, and this persisted, even following antimicrobial treatment (Vicari et al., 1997Go).

We believe that our study raises some important issues with respect to MAGI and infertility. In particular, in order to point out some important inflammatory damages from acting directly on semen and accessory sex glands, we included two types of patient: (i) those with >105/ml bacteria in their semen cultures, approximating to the study of spermatozoa infected experimentally with E. coli (106 CFU/ml) in which a reduction in sperm motility was demonstrated (Auroux et al., 1991Go); and (ii) those with abnormal ultrasound signs to reinforce the hypothesis suggested by some authors (Purvis and Christiansen, 1993Go; Vicari, 1999Go) that MAGI may contribute indirectly to infertility through combined abnormal structural echopattern and biochemical inflammatory damage (Vicari, 1999Go), and not simply through inflammatory biochemical damage, with decrease of the specific markers of the epididymis and/or the accessory sex glands, as reported in the literature (Cooper et al., 1990Go; Wolff et al., 1991Go; Gonzales et al., 1992Go).

In the present study, a comprehensive microbiological ejaculate analysis, including Chlamydia and U. urealyticum in cultures of urethral swabs obtained after prostatic massage, allowed us to detect significant aerobic bacterial growth in 58.2% of semen specimens, with Chlamydia (mainly in PVE patients) in 20.5% and U. urealyticum in 10.7%, while anaerobic growth was seen in the remaining 10.7%. It is worth noting that proof of PVE or PR and not just urethritis was two-fold: these patients were asymptomatic and had had no presence of Chlamydia or U. urealyticum in cultures of urethral swabs obtained without prostatic massage.

In comparison with the higher (42%) anaerobic colonization rate reported earlier (Eggert-Kruse et al., 1995Go), our low anaerobic colonization might in part be related to methodology used (our seminal plasma samples were cultured after diluting 1:2 in saline; Comhaire et al., 1980) and to different selection criteria of patients, since others (Eggert-Kruse et al., 1995Go) identified an asymptomatic infertile group without `clinical signs' of genital tract infections.

From the therapeutic point of view, several points may be underlined about bacteriological outcome. First, that bacteriological outcome reflected a significantly different positive balance between the percentage of cured patients among categories, and the percentage of failure (with relapse or re-infection). Since these results are time- and infected gland-dependent (PR group > PV group > PVE group), we propose at least two possible explanations. First, some microorganisms could require a long-term therapy because of their strong adhesion to the mucous glandular membrane, e.g. Gram-positive microorganisms have a tendency to be encapsulated, protected either by a glycocalix matrix formed by themselves (Nickel and Costerton, 1993Go) and by an inflamed, hyperviscous vesicular microenvironment. Thus, these mechanisms predispose the host for intracanicular spread to the rest of the reproductive tract. Furthermore, the bacterial re-infections in patients with PV or PVE might in part be explained through a stratified bacterial infection, in which the bacterial glycocalix exhibiting a strong adhesion on one side directly to the glandular mucous membrane could enhance a weaker adhesion to the fimbrias of the Gram-negative organisms on the opposite side. This weak adhesion could be more easily removed by antibiotic treatment, masking the re-infection due to Gram-positive organisms, as seen in our patients. The second point is that, under these especially expressed conditions of MAGI, it is important to emphasize that in all control groups seminal cultures always remained positive with a bacteriological cure (defined as CFU 0–103/ml) of 0%, as if a natural history towards spontaneous resolution of bacteriospermia in these patients was not present. A spontaneous resolution of bacteriospermia described in some patients with MAGI (Comhaire et al., 1986Go) could be due to temporary inflammatory episodes or to a low number of sexual gland abnormalities detected by ultrasound (Vicari, 1999Go), probably through a low degree of bacterial adhesion to infected glands. The third point is that, although twice-daily ofloxacin or once-daily doxycycline, when given orally for 14 days each month over three consecutive months, seem equipotent in the treatment of PR, PV and PVE, and covering the range of causative organisms, antimicrobial treatment showed limited bacteriological cure in patients suffering from PV, and was even less effective in those patients with PVE.

Data reported about the effects of fluoroquinolones on sperm output in infertile MAGI patients with complicated (PV or PVE) infections are difficult to compare because not all investigators used the same diagnostic criteria, and there was a considerable range in the dosage, duration of treatment and follow-up (Schramm, 1986Go; Weidner et al., 1987Go; Malinverni and Glauser, 1988Go; Naber, 1991Go; Andreeßen et al., 1993Go; Stamm and Hoton, 1993Go; Vicari and Mongioì, 1993Go, 1994Go; Vicari et al., 1997Go). However, integrating our results with bacteriological rates reported with one of the newer oral fluoroquinolones (ciprofloxacin, enoxacin, fluoroxacin, lomefloxacin, norfloxacin, ofloxacin or pefloxacin) in the treatment of cystitis (94–99%) (Del Rio et al., 1996Go) and uncomplicated prostatitis (70–92%) (Weidner et al, 1987Go; Andriole, 1991Go), one can postulate that in MAGI categories, bacteriological rate is related to the extent of the infectious process, being highest in the PR group receiving antibiotics (92.5%), followed by the PV group (70.4%), and lowest in the PVE group (52.0%). Thus, the proposed rationale would be to further treat PVE patients with antimicrobials alone, or in combination with antiphlogistic and/or antioxidant drugs. However, the incomplete bacteriological efficacy of even repetitive courses of antimicrobials in PVE patients might have an additional explanation. Indeed, in this group the prolonged pre-ejaculatory spermatozoa–WBC interaction could also progressively reduce the antioxidative properties of the epididymis, through low and non-antibiotic-reversible seminal levels of some functional seminal markers (e.g. carnitine, alpha-glucosidase) as was detected preliminarily in some of our patients (data not shown) following antimicrobial treatment.

Since the pro- and antioxidative properties of therapeutic drugs as part of anti-infectious therapies (Ochsendorf, 1999Go) are a promising strategy to improve human reproductive function, so that a re-equilibrium between leukocytes and/or sperm ROS production and anti/pro-oxidant scavenger efficacy in seminal plasma can be reached (for reviews, see Ford and Whittington, 1998; Geva et al., 1998; Lenzi et al., 1998; Tarín et al., 1998), patients affected by PVE may represent a valid clinical model for the rational use of antioxidant scavengers, since they showed persistent infertility as well as a juxta-sperm oxidative stress even following antimicrobial treatment, which may play a causative role in their infertility (Sikka et al., 1995Go).


    Acknowledgments
 
The author would like to thank Dr Giuseppe Aranzulla, Division of Radiology, Garibaldi Hospital, Catania and Dr Francesco Caccamo, Institute of Microbiology, University of Catania, for their skilful technical assistance.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Andreeßen, R., Sudhoff, F., Borgmann, V. and Nagel, R. (1993) Results of ofloxacin therapy in andrologic patients suffering from therapy-requiring asymptomatic infections. Andrologia, 25, 377–383.[ISI][Medline]

Andriole, V.T. (1991) Use of quinolones in treatment of prostatitis and lower urinary tract infections. Eur. J. Clin. Microbiol. Infect. Dis., 10, 342–350.[ISI][Medline]

Auroux, M.R., Jacques, L., Mathieu, D. et al. (1991) Is the sperm bacterial ratio a determining factor in impairment of sperm motility: an in-vitro study in man with Escherichia coli. Int. J. Androl., 14, 264–270.[ISI][Medline]

Bar-Chama, N. and Fisch, H. (1993) Infection and pyospermia in male infertility. World J. Urol., 11, 76–81.[ISI][Medline]

Berger, R.E., Karp, L.E., Williamson, R.A. et al. (1982) The relationship of pyospermia and seminal fluid bacteriology to sperm function as reflected in the sperm penetration assay. Fertil. Steril., 37, 557–564.[ISI][Medline]

Bjerklund Johansen, T.E., Gruneberg, R.N., Guibert, J. et al. (1998) The role of antibiotics in the treatment of chronic prostatitis. A consensus statement. Eur. Urol., 34, 457–466.[ISI][Medline]

Comhaire, F., Verschraegen, G. and Vermeulen, L. (1980) Diagnosis of accessory gland infection and its possible role in male infertility. Int. J. Androl., 3, 32–45.[ISI][Medline]

Comhaire, F.H., Rowe, P.J. and Farley, T.M.M. (1986) The effect of doxycycline in infertile couples with male accessory gland infection: a double blind prospective study. Int. J. Androl., 9, 91–98.[ISI][Medline]

Cooper, T.G., Weidner, W. and Nieschlag, E. (1990) The influence of inflammation of the human male genital tract on secretion of the seminal markers alpha-glucosidase, glycerophosphocholine, carnitine, fructose and citric acid. Int. J. Androl., 13, 329–336.[ISI][Medline]

Cunha, B.A. and Garabedian-Ruffalo, S.M. (1990) Tetracyclines in urology: current concepts. Urology, 36, 548–556.[ISI][Medline]

Del Rio, G., Dalet, F., Aguilar, L. et al. (1996) Single-dose rufloxacin versus 3-day norfloxacin treatment of uncomplicated cystitis: clinical evaluation and pharmacodynamic considerations. Antimicrob. Agents Chemother., 40, 408–412.[Abstract]

Depuydt, C.E., Bosmans, E., Zalata, A. et al. (1996) The relation between reactive oxygen species and cytokines in andrological patients with or without male accessory gland infection. J. Androl., 17, 699–707.[Abstract/Free Full Text]

Eggert-Kruse, W., Rohr, G., Strock, W. et al. (1995) Anaerobes in ejaculates of subfertile men. Hum. Reprod. Update, 1, 462–478.[Abstract]

Fedder, J. (1996) Nonsperm cells in human semen: with special reference to seminal leukocytes and their possible influence on fertility. Arch. Androl., 36, 41–45.[ISI][Medline]

Ford, W.C.L. and Whittington, K. (1998) Antioxidant treatment for male subfertility: a promise that remains unfulfilled. In: Debate `Is antioxidant therapy a promising strategy to improve human reproduction?' Hum. Reprod., 13, 1415–1424.[Free Full Text]

Geva, E., Lessing, J.B., Lerner-Geva, L. and Amit, A. (1998) Free radicals, antioxidants and human spermatozoa: clinical implications. In Debate `Is antioxidant therapy a promising strategy to improve human reproduction?' Hum. Reprod., 13, 1415–1424.[Free Full Text]

Giamarellou, H., Tympanidis, K., Bitos, N.A. et al. (1984) Infertility and chronic prostatitis. Andrologia, 16, 417–423.[ISI][Medline]

Gnarpe, H., and Friberg, J. (1973) T-mycoplasmas on spermatozoa and fertility. Nature, 245, 97–98.[ISI][Medline]

Gonzales, G.F., Kortebani, G. and Mazzolli, A.B. (1992) Leukocytospermia and function of the seminal vesicles on seminal quality. Fertil. Steril., 57, 1058–1065.[ISI][Medline]

Homonnai, T.Z., Sasson, S., Paz, G. et al. (1975) Improvement of fertility and semen quality in men treated with a combination of anticongestive and antibiotic drugs. Int. J. Fertil., 20, 45–49.[ISI][Medline]

Krause, W. (1994) Prostatitis and male infertility. In Weidner, W., Madsen, P.O. and Schiefer, H.G. (eds), Prostatitis Etiopathology, Diagnosis and Therapy. Springer, Berlin, pp. 91–109.

Lenzi, A., Gandini, L. and Picardo, M. (1998) A rationale for glutathione therapy. In Debate `Is antioxidant therapy a promising strategy to improve human reproduction?' Hum. Reprod., 13, 1419–1422.[ISI][Medline]

Liversedge, N.H., Jenkins, J.M., Keay, S.D. et al. (1996) Antibiotic treatment based on seminal cultures from asymptomatic male partners in in-vitro fertilization is unnecessary and may be detrimental. Hum. Reprod., 11, 1227–1231.[Abstract]

Malinverni, R. and Glauser, M.P. (1988) Comparative studies of fluoroquinolones in the treatment of urinary tract infections. Rev. Infect. Dis., 10 (Suppl.1), S153–S163.[Medline]

McGowan, M.P., Burger, H.G., Baker, H.W.G. et al. (1981) The influence of non-specific infection in the semen in fertile and subfertile males. Int. J. Androl., 4, 657–662.[ISI][Medline]

Naber, K.G. (1991) The role of quinolones in the treatment of chronic bacterial prostatitis. Infection, 19 (Suppl. 3), S170–S177.[ISI][Medline]

Naber, K.G., Kinzig, M., Sorgel, F. et al. (1993) Penetration of ofloxacin into prostatic fluid, ejaculate and seminal fluid. Infection, 21, 98–100.[ISI][Medline]

Nickel, J.C. and Costerton, J.W. (1993) Bacterial localization in antibiotic-refractory chronic bacterial prostatitis. Prostate, 23, 107–114.[ISI][Medline]

Nikkanen, V., Gronroos, M., Suominen, J. et al. (1979) Silent infection in male accessory organs and male infertility. Andrologia, 11, 236–241.[ISI][Medline]

Nilsson, S., Obrant, K.O. and Persson, P.S. (1968) Changes in the testis parenchyma caused by acute nonspecific epididymitis. Fertil. Steril., 19, 748–757.[ISI][Medline]

Ochsendorf, F.R (1999) Infections in the male genital tract and reactive oxygen species. Hum. Reprod. Update, 5, 399–420.[Abstract/Free Full Text]

O'Donovan, P.A., Vandekerchhove, P., Lilford, R.J. and Hughes, E. (1993) Treatment of male infertility: is it effective? Review and meta-analyses of published randomised controlled trials. Hum. Reprod., 8, 1209–1222.[Abstract]

Omu, A.E., al-Othman, S., Mohamad, A.S., et al. (1998) Antibiotic therapy for seminal infection. Effect on antioxidant activity and T-helper cytokines. J. Reprod. Med., 43, 857–864.[ISI][Medline]

Purvis, K. and Christiansen, E. (1993) Infection in the male reproductive tract. Impact, diagnosis and treatment in relation to male infertility. Int. J. Androl., 16, 1–13.

Purvis, K. and Christiansen, E. (1995) The impact of infection in sperm quality. J. Br. Fertil. Soc., 1, 31–41.

Rhemrev, J.P.T., Vermeiden, J.P.W., Bermardus, R.E. et al. (1987) Antibiotic treatment improves the number of motile spermatozoa harvested by Percoll density gradient centrifugation. Hum. Reprod., 2 (Suppl. 1), 40–45.

Schill, W.B. (1995) Survey of medical therapy in andrology. Int. J. Androl., 18, 56–62.[ISI][Medline]

Schoysman, R. (1981) Epididymal causes of male infertility: pathogenesis and management. In Bollack, C.B. and Clavert, A. (eds), Progress in Reproductive Biology, Pathology and Pathophysiology of the Epididymis. Karger, Basel, pp. 102–109.

Schramm, P. (1986) Ofloxacin: concentration in human ejaculate and influence on sperm motility. Infection, 14 (Suppl. 4), 274–275.

Sikka, S.C., Rajasekaran, M. and Hellstrom, W.J.G. (1995) Role of oxidative stress and antioxidants in male fertility. J. Androl., 16, 464–468.[Abstract/Free Full Text]

Soffer, Y., Ron-El, R., Golan, A. et al. (1990) Male genital mycoplasmas and Chlamydia trachomatis cultures: its relationship with accessory gland function, sperm quality, and autoimmunity. Fertil. Steril., 55, 331–336.

Stamm, W.E. and Hoton, T.M. (1993) Management of urinary tract infections in adults. N. Engl. J. Med., 329, 1328–1334.[Free Full Text]

Swenson, C.E., Toth, A. and O'Leary, W.M. (1979) Ureaplasma urealyticum and human fertility: the effect of antibiotic therapy on semen quality. Fertil. Steril., 31, 660–665.[ISI][Medline]

Tarín, J.J., Brines, J. and Cano, A. (1998) Antioxidants may protect against infertility. In Debate `Is antioxidant therapy a promising strategy to improve human reproduction?' Hum. Reprod., 13, 1415–1416.[Free Full Text]

Vicari, E. (1999) Seminal leukocyte concentration and related specific reactive oxygen species production in patients with male accessory gland infections. Hum. Reprod., 14, 2025–2030.[Abstract/Free Full Text]

Vicari, E. and Mongioì, A. (1993) Problems in the therapy of infertile patients with infections of reproductive tract: effects of different quinolones in the sperm quality. Int. J. Feto-Matern. Med., 6, 83–88.

Vicari, E. and Mongioì, A. (1994) Further studies on male accessory glands microbiological infections: different antimicrobial outcome in relation to the several degree of pretreatment bacterial adherence. Arch. AIDS/STD Res., 5, 167–175.

Vicari, E. and Sidoti, G. (1991) Gli sperm precursors nel varicocele e nelle flogosi uro-genitali croniche batteriche: significato clinico in un follow-up longitudinale. In Pietri, P. (ed.), XI Congr. Naz. Soc. Ital. Endocrinochirurgia. Monduzzi ed., Bologna, Italy, pp. 659–663.

Vicari, E., Mongioì, A., Speciale, A. et al. (1986) Enhancing detection of gonococcus in ejaculates of adult males using sperm dilution. Arch. Androl., 16, 19–23.[ISI][Medline]

Vicari, E., Barone, N. and D'Agata, R. (1997) Infertility or complete IVF failure in presence of persistent detection of high reactive oxygen species (ROS) sperm production in a large number of patients affected by epididymitis previously treated with antimicrobials. In Ambrosini, A., Melis, G.B., Dalla Pria, S. and Dessole, S. (eds), International Meeting of Infertility and Assisted Reproductive Technology (From Research to Therapy). Monduzzi ed., Bologna, Italy, pp. 99–103.

Wang, A., Fanning, L., Anderson, D.J. and Loughlin, K.R. (1997) Generation of reactive oxygen species by leukocytes and sperm following exposure to urogenital tract infection. Arch. Androl., 39, 11–17.[ISI][Medline]

Weidner, W. (1992) Prostatitis: diagnostic criteria, classification of patients and recommendations for therapeutical trials. Infection, 20 (Suppl. 3), 227–231.[ISI][Medline]

Weidner, W., Schiefer, H.G. and Dalhoff, A. (1987) Treatment of chronic bacterial prostatitis with ciprofloxacin. Am. J. Med., 82 (Suppl. 4A), 280–283.[ISI][Medline]

Weidner, W., Jantos, C., Schieffer, H.G. et al. (1991) Semen parameters in men with and without proven chronic prostatitis. Arch. Androl., 26, 173–183.[ISI][Medline]

Weidner, W., Ludwig, M. and Miller, J. (1998) Therapy in male accessory gland infection – what is fact, what is fiction? Andrologia, 30, 87–90.[ISI][Medline]

Wolff, H., Bezold, G., Zebhauser, M. et al. (1991) Impact of clinically silent inflammation on male genital tract organs as reflected by biochemical markers in semen. J. Androl., 12, 331–334.[Abstract]

World Health Organization (1992) WHO Laboratory Manual for the Examination of Human Semen and Semen-Cervical Mucus Interaction. 3rd edn. Cambridge University Press, Cambridge, UK.

World Health Organization (1993) WHO Manual for the Standardized Investigation and Diagnosis of the Infertile Couple. Cambridge University Press, Cambridge, New York.

Submitted on February 24, 2000; accepted on September 8, 2000.