Acute urethritis in men could be caused by various microorganisms, including bacteria, viruses, and protozoa.1–3 Haemophilus spp. have been isolated from men with acute urethritis.4–6 A recent study suggested that H. influenzae might be detected more often in men with nongonococcal urethritis (NGU) than in control men without NGU.3 In our previous study, we also isolated Haemophilus influenzae from the urethra of men with urethritis symptoms and reported that the prevalence of H. influenzae was 5.2% in men with acute urethritis, 7.4% in those with NGU, and 14.3% in those with nonchlamydial NGU.2 Because there have been few studies on pathogenic roles of H. influenzae in acute urethritis, H. influenzae is yet to be defined as a pathogen-causing acute urethritis. In addition, there have been few studies on antimicrobial chemotherapies for H. influenzae–positive NGU and on antimicrobial susceptibilities of H. influenzae clinical strains isolated from the urogenital tract. It is uncertain, therefore, whether the regimens of antimicrobial agents recommended for the treatment of acute urethritis, in particular, those for chlamydial NGU and nonchlamydial NGU, are effective on H. influenzae–positive NGU.
In this retrospective study, we retrieved clinical data from the medical records of men with H. influenzae–positive urethritis who were treated by the regimens recommended for treatment for acute urethritis in Japan.7 We determined microbiological responses of H. influenzae to the antimicrobial chemotherapies in men with H. influenzae–positive urethritis. We investigated the pathologic roles of H. influenzae in acute urethritis by analyzing the association of clinical outcomes with its microbiological responses to the antimicrobial chemotherapies. We also determined minimum inhibitory concentrations (MICs) of the agents, which were commonly prescribed for treatment of urethritis, for clinical strains of H. influenzae that were isolated from men with acute urethritis and stored after isolation.
MATERIALS AND METHODS
Patients and Isolates of H. Influenzae
A total of 1518 men with urethritis symptoms visited the iClinic, Sendai, Japan, between January 2009 and December 2015. They underwent genital examinations, and their Gram-stained urethral smears were examined for the presence of leukocytes and Gram-negative diplococci. Their first-voided urine (FVU) specimens (first 20 to 30 mL of the initial flush of urine after not having urinated for at least 3 hours) were examined for quantification of leukocytes with automated quantitative urine particle analyzers (UF-1000i, Sysmex Corporation, Tokyo, Japan) according to the manufacturer's instructions. They were also examined for Neisseria gonorrhoeae and Chlamydia trachomatis by APTIMA Combo2 assay (Hologic, Inc., Bedford, Mass) and for Mycoplasma genitalium, Mycoplasma hominis, Ureaplasma parvum, and Ureaplasma urealyticum by InvaderPlus® assay as reported previously.8 Urethral swabs collected from each subject underwent culture for N. gonorrhoeae, H. influenzae, and Streptococcus pneumoniae. To isolate H. influenzae, the specimens were inoculated onto chocolate agar. The colonies that were presumptively regarded as Haemophilus spp. on the chocolate agar were inoculated onto Haemophilus ID Quad with Growth Factors (Japan Becton, Dickinson and Company, Tokyo, Japan), and the isolates growing on Quadrant-III and -V were subjected to the HN-20 Rapid system identification test (Nissui, Tokyo, Japan) to identify H. influenzae. The H. influenzae isolates were serotyped by slide agglutination assay with antisera purchased from Denka Seiken (Tokyo, Japan). The isolates were stored frozen at −70°C for later antimicrobial susceptibility testing.
The men who were diagnosed as having urethritis were treated by one of the antimicrobial regimens recommended by the Japanese guidelines for the treatment of urethritis that were published in 2006 and 2011.7 In this retrospective study, no protocols for antimicrobial regimens for the treatment of acute urethritis were decided, but the choice of a regimen depended on the physicians in charge of the subjects. According to the Japanese guidelines for clinical research of antimicrobial agents on urogenital infections, the men with gonococcal urethritis and those with NGU were asked to return for reexamination 5 to 9 days and 14–28 days after the completion of treatment, respectively, and told to remain sexually abstinent until the follow-up visit,9 at which their symptoms and signs were examined. In brief, their FVU was subjected to leukocyte quantification, and their FVU and/or urethral swab were subjected to microbiological examinations as performed at the first visits, including culture for H. influenzae.
We retrospectively retrieved the clinical data of men who had culture-positive results for H. influenzae at their first visits. Microbiological outcomes of antimicrobial chemotherapies were based on the results of posttreatment culture for H. influenzae adopted as a test of cure. Clinical outcomes were assessed in the men with NGU who were positive only for H. influenzae at their first visits. Their clinical outcomes were defined based on self-reported urethritis symptoms (dysuria, urethral irritation, and urethral discharge) after treatment as complete response, complete disappearance of symptoms; partial response, alleviation but incomplete disappearance of all symptoms; and failure, no alleviation or exacerbation of symptoms. The leukocyte counts in the FVU specimens after treatment were compared with those before treatment.
Antimicrobial Susceptibility Testing
For the pretreatment and posttreatment isolates of H. influenzae that survived storage and were recultured, MICs of ceftriaxone, levofloxacin, sitafloxacin, azithromycin, tetracycline, and doxycycline were determined by the broth dilution method recommended by the Clinical and Laboratory Standards Institute (CLSI).10
Comparison of the clinical outcomes with the microbiological outcomes was performed by χ2 contingency table analysis. Comparisons of the leukocyte counts in the pretreatment FVU with those in the posttreatment FVU were performed by Mann-Whitney U test and by Wilcoxon signed-rank test. Multiple comparisons of the chronological MIC distributions of the antimicrobial agents were performed by Kruskal-Wallis nonparametric 1-way analysis of variation and Mann-Whitney U test with Bonferroni correction. The Mann-Whitney U test was used to compare the distributions of azithromycin MICs between the pretreatment isolates eradicated by the treatment and those persisting after the treatment. All statistical analyses were conducted with JMP Version 5.01J (SAS Institute Japan, Tokyo, Japan) with significance set at P < 0.05. Between-group differences analyzed by the Mann-Whitney U test with Bonferroni correction were considered to be statistically significant at a Bonferroni-corrected P < 0.05/number of tests.
This retrospective study was approved by the Institutional Review Board of the Graduate School of Medicine, Gifu University, Gifu, Japan (reference number 20-94). However, the consent of the subjects was not obtained when this study was conducted because they could not be traced. Therefore, we sufficiently anonymized their information.
Patients and Isolates of H. Influenzae
In 77 (5.1%; 95% confidence interval [CI], 4.0–6.2%) of the 1518 men complaining of urethral symptoms, including 22 men involved in our previous study,2 the isolates cultured from their urethral swabs were identified as H. influenzae (Fig. 1). All of the isolates of H. influenzae were nontypeable by serotyping. One man with acute prostatitis, 6 men with acute epididymitis, and 1 man with condyloma acuminatum on the glans found on genital examination were excluded from this study, as was one man treated by an antimicrobial regimen that was not recommended by the Japanese guidelines for the treatment of acute urethritis.7 Of the remaining 68 men (median age, 35 years) with H. influenzae–positive urethritis, 1 had coinfection with N. gonorrhoeae, 1 with C. trachomatis, 1 with S. pneumoniae, 5 with U. urealyticum, 4 with U. urealyticum and M. hominis, and 2 with U. parvum, whereas 54 were positive only for H. influenzae. All of the men had symptoms and signs compatible with acute urethritis and pyuria of ≥5 leukocytes per high-power (×1000) microscopic field in their Gram-stained urethral smear and/or pyuria of ≥10 leukocytes per 1 μL of their FVU.
Microbiological Outcomes of Antimicrobial Chemotherapies
The 68 men were treated with antimicrobial regimens recommended by the Japanese guidelines for the treatment of urethritis (Table 1). Fourteen men were excluded because of no follow-up visit (Fig. 1). The remaining 54 men returned to the clinic between 7 and 33 days (median, 21 days) after treatment. However, 18 men did not return within the designated periods of time after treatment.9 The microbiological responses of H. influenzae to the regimens were assessed in the 36 men who returned within the designated periods and in the 54 men who returned to the clinic between 7 and 33 days. All men treated with the ceftriaxone, levofloxacin, or sitafloxacin regimens had negative culture results for H. influenzae after treatment regardless of the assessment periods (Table 1). In the men treated with the extended-release azithromycin (azithromycin-SR) regimen, H. influenzae was eradicated in 19 of 21 men (90.5%; 95% CI, 86.8–94.2%) assessed between 14 and 28 days and in 29 of 34 men (85.3%, 95% CI, 73.4–97.2%) assessed between 7 and 33 days after treatment. The 3 men with positive culture results for H. influenzae after treatment underwent a test of cure at their follow-up visits prior to the designated periods. All 5 of the men in whom treatment to eradicate H. influenzae failed had H. influenzae–positive monomicrobial urethritis.
Association of Clinical Outcomes With Microbiological Responses to Chemotherapies
Eight men who were positive for other pathogens at their first visits were excluded from the 54 men who returned to the clinic between 7 and 33 days after treatment (Fig. 1). The remaining 46 men with H. influenzae–positive monomicrobial urethritis were assessed for clinical outcomes based on pretreatment and posttreatment differences in clinical symptoms and leukocyte counts in FVU. Of the 41 men from whom H. influenzae was eradicated, 32 (78.0%; 95% CI, 65.3–90.7%) and 8 (19.5%; 95% CI, 7.4–31.6%) had complete and partial responses to the chemotherapies, respectively, in terms of urethritis symptoms, whereas 2 of 5 men with persistence of H. influenzae after treatment had symptoms as seen originally and were judged as having treatment failures (Table 2). There was a significant difference in clinical responses between the men with success and those with failure in eradicating H. influenzae (P = 0.026).
There was no significant difference in the distribution of leukocyte counts in FVU before treatment between the men with success and those with failure in eradicating H. influenzae (P = 0.332) (Table 2). After treatment, however, the leukocyte counts in men with successful eradication of H. influenzae were significantly lower than those in men with failure in eradicating H. influenzae (P = 0.002), and 29 (70.7%; 95% CI, 56.8–84.6%) of the 41 men microbiologically cured had less than 10 leukocytes in 1 μL of FVU. Figure 2 shows the pretreatment and posttreatment changes of leukocyte counts in the FVU of each subject. The leukocyte counts in FVU in men with success in eradicating H. influenzae significantly decreased after treatments (P < 0.001), though there was no significant difference between pretreatment and posttreatment leukocyte counts in FVU in the men with failure in eradicating H. influenzae (P = 0.312).
Two of the 5 men with treatment failure in eradicating H. influenzae were subsequently treated with the sitafloxacin or levofloxacin regimen at their second visits and were clinically and microbiologically cured by the second treatment. The remaining 3 men, who were asymptomatic despite a significant pyuria at the second visit, did not undergo further treatment. However, one of them returned to the clinic 39 days later complaining of the recurrence of urethritis symptoms, whereupon he was treated with the sitafloxacin regimen and was clinically and microbiologically cured. The other 2 men were lost to follow-up.
Fifty-nine of the pretreatment isolates of H. influenzae from the 68 men survived storage and were recultured. For these isolates, MIC ranges and MICs of the 6 agents that inhibited 50% and 90% of them are summarized in Table 3. Based on the MIC breakpoints recommended by the CLSI,11 all of the isolates were susceptible to ceftriaxone (MIC ≤2 μg/mL), levofloxacin (MIC ≤2 μg/mL), and azithromycin (MIC ≤4 μg/mL). The 58 isolates (98.3%; 95% CI, 95.0–100%), except for the one isolated in 2015 that exhibited a tetracycline MIC of 16 μg/mL and a doxycycline MIC of 2 μg/mL, were susceptible to tetracycline (MIC ≤2 μg/mL). The MICs of sitafloxacin for the isolates were lower than those of levofloxacin. The isolates that were susceptible to levofloxacin could also be considered susceptible to sitafloxacin, and those that were susceptible to tetracycline could also be considered susceptible to doxycycline. Multiple comparisons of the chronological MIC distributions of the antimicrobial agents across the years of isolation of H. influenzae showed that there were significant differences in the MIC distributions of sitafloxacin (P = 0.007), azithromycin (P < 0.001), and tetracycline (P = 0.038) (Table 3). The sitafloxacin MICs for the isolates in 2014–2015 increased significantly compared with those for the isolates in 2009–2011. The azithromycin MICs for the isolates in 2014–2015 increased significantly compared with those for the isolates in 2009–2011 and in 2012–2013. For tetracycline, however, there were no significant differences in its MICs between the groups over the years of storage.
Association of Microbiological Outcomes With Antimicrobial Susceptibilities
The ceftriaxone MICs for 2 isolates eradicated by the ceftriaxone regimen were 0.12 and 0.25 μg/mL. The levofloxacin MICs for 3 isolates eradicated by the levofloxacin regimen were all 0.016 μg/mL. Thirteen of 15 isolates eradicated by the sitafloxacin regimen were recovered from storage, and the sitafloxacin MICs for these isolates were 0.002 or 0.004 μg/mL. Twenty-five of 29 pretreatment isolates eradicated by the azithromycin-SR regimen and 4 of 5 pretreatment isolates not eradicated by the regimen were also recovered from storage and subjected to MIC determination. The azithromycin MICs for the 25 pretreatment isolates, whose eradication was confirmed after treatment, ranged from 0.12 to 1 μg/mL, whereas those for one and 3 pretreatment isolates persisting after treatment were 0.25 and 1 μg/mL, respectively (Table 4). The azithromycin MIC distribution for the persistent isolates was not significantly different from that for the eradicated isolates (P = 0.607). Fifteen of 18 isolates with an azithromycin MIC of 1 μg/mL were eradicated by the azithromycin-SR regimen, but the azithromycin MICs for 3 of 4 isolates persisting after the azithromycin-SR were 1 μg/mL (Table 4).
The proportion of H. influenzae isolates with an azithromycin MIC of 1 μg/mL increased chronologically (Table 5). Sixteen (88.9%; 95% CI, 74.4–100%) of 18 isolated in 2014–2015, including all 10 isolates in 2015, had an azithromycin MIC of 1 μg/mL.
Of 5 posttreatment isolates, 2 were recovered from storage. The azithromycin MICs of 1 μg/mL for these isolates were the same as those for the respective pretreatment isolates.
Based on the culture results, all of the patients with H. influenzae–positive NGU treated with the ceftriaxone or levofloxacin regimens were microbiologically cured, although the numbers of patients were limited. The eradication rate of H. influenzae by the sitafloxacin regimens was 100% in 15 patients, whereas that by the azithromycin-SR regimen was 85.3% (95% CI, 73.4–97.2%) in 34 patients. Thus, the regimens recommended for treatment for acute urethritis appeared to be microbiologically effective on H. influenzae–positive NGU, but the azithromycin-SR regimen appeared to be inferior to other regimens.
To detect C. trachomatis in cases of NGU, culture is not used; rather, nucleic acid amplification tests (NAATs), which target bacterial DNA or rRNA, are commonly used in clinical practice. As a test of cure for C. trachomatis infections, NAATs might yield a false-positive test result earlier after treatment because of the detection of DNA or rRNA from residual unviable bacteria. Therefore, the Japanese guideline for clinical research of antimicrobial agents on urogenital infections recommends that the microbiological efficacies of antimicrobial chemotherapies for NGU be evaluated between 14 and 28 days after the completion of treatment.9 In this study, the 3 men with positive culture results for H. influenzae underwent a test of cure at the follow-up visit between 7 and 13 days after the azithromycin-SR regimen. Although the sensitivity to detect H. influenzae by culture might be less compared with NAATs, the positive culture results of the test of cure could not be false-positive results for H. influenzae in these cases. Therefore, the microbiological outcomes evaluated between 7 and 33 days after treatment could reflect accurate microbiological efficacies of the chemotherapies.
In this study, the disappearance or alleviation of urethritis symptoms and the decreases in leukocyte counts in FVU were significantly associated with the eradication of H. influenzae after treatment. Additionally, such an association was also observed when the 3 men with failure in eradicating H. influenzae with the azithromycin-SR regimen were treated by second fluoroquinolone regimens. The associations of the clinical outcomes of the chemotherapies with their microbiological outcomes suggested that H. influenzae could play pathogenic roles in causing acute urethritis in men.
Among the 59 pretreatment isolates that were recovered from storage and analyzed for antimicrobial susceptibility, all but one isolate with decreased susceptibility to tetracyclines were susceptible to ceftriaxone, levofloxacin, sitafloxacin, tetracycline, doxycycline, and azithromycin based on the MIC breakpoints of the CLSI.11 However, these breakpoints would not be applicable for acute urethritis. In this study, 5 isolates persisted after treatment with the azithromycin-SR regimen, and 3 of them exhibited an azithromycin MIC of 1 μg/mL. The men who were unsuccessfully treated with the azithromycin-SR regimen might have been reexposed to and reinfected by H. influenzae. In our previous study, however, the azithromycin MIC breakpoint associated with the failure to eradicate N. gonorrhoeae in men with gonococcal urethritis appeared to be 1 μg/mL when the same azithromycin-SR regimen was administered.12 Therefore, the azithromycin MIC of 1 μg/mL might also be the breakpoint in the treatment for H. influenzae–positive NGU with the azithromycin-SR regimen. The increase in the prevalence of H. influenzae with an azithromycin MIC of 1 μg/mL that was observed in this study might threaten the efficacies of the regimens, including the conventional 1-g azithromycin single-dose regimen,7,13–15 on H. influenzae–positive NGU.
In Japan, nationwide surveillance of antimicrobial susceptibilities of H. influenzae clinical strains isolated from respiratory tract infections was conducted from 2006 to 2010.16,17 The azithromycin MIC ranges, MICs inhibiting growth of 50% of the isolates, and MICs inhibiting growth of 90% of them were ≤0.06–2, 0.5, and 1 in 2006 and ≤0.06–4, 1, and 2 μg/mL in 2010. Although all H. influenzae isolates from the respiratory tract infections were susceptible to azithromycin according to the breakpoint of the CLSI,11 they exhibited higher azithromycin MICs compared to those isolated from the urethra in the present study. In our previous study, all men with H. influenzae–positive NGU practiced unprotected oral sex, but none of the men who reported vaginal sex only without oral sex contracted H. influenzae infections.2 Other studies also reported that urethritis caused by Haemophilus spp. was characterized by transmission of the pathogens through oral sex.1,6 From now on, therefore, H. influenzae with decreased susceptibility to azithromycin in the pharynx could be transmitted more often to the male urethra to cause NGU so that it could become more difficult to treat H. influenzae–positive NGU with azithromycin regimens.
Limitations of this retrospective study include the small number of patients, less sensitive detection of H. influenzae by culture, lack of control men without NGU to analyze differences in the prevalence of H. influenzae between men with and without NGU, no protocols for the randomized choice of the regimens, difficulties in ruling out the possibility of reinfection by H. influenzae in the men unsuccessfully treated, and lack of use of other regimens such as the 100-mg doxycycline twice-daily 7-day regimen and the conventional 1-g azithromycin single-dose regimen that are recommended for the treatment of NGU.7,13–15 Although the MIC breakpoint for azithromycin in the treatment of acute urethritis was suggested in this study, those of the other antimicrobial agents could not assessed. Nevertheless, this study provides comprehensive data to understand and manage H. influenzae–positive NGU, which has rarely been studied.
In conclusion, H. influenzae showed good responses to the regimens recommended for the treatment of NGU in Japan, especially the fluoroquinolone regimens, in men with acute urethritis. The significant associations of the clinical outcomes of the chemotherapies with their microbiological responses suggested that H. influenzae could play pathogenic roles in causing acute urethritis. H. influenzae isolated from men with acute urethritis was susceptible to the antimicrobial agents commonly prescribed for the treatment of acute urethritis. However, the azithromycin MIC of 1 μg/mL might be associated with the risk of treatment failure for H. influenzae–positive urethritis even with the azithromycin-SR regimen. The chronologically increasing prevalence of H. influenzae with an azithromycin MIC of 1 μg/mL might threaten the efficacies of azithromycin regimens on H. influenzae–positive urethritis. Further studies are needed to confirm our findings and to establish appropriate treatments for H. influenzae–positive NGU.
1. Bradshaw CS, Tabrizi SN, Read TR, et al. Etiologies of nongonococcal urethritis: bacteria, viruses, and the association with orogenital exposure. J Infect Dis 2006; 193:336–345.
2. Ito S, Hanaoka N, Shimuta K, et al. Male non-gonococcal urethritis: From microbiological etiologies to demographic and clinical features. Int J Urol 2016; 23:325–331.
3. Frølund M, Lidbrink P, Wikström A, et al. Urethritis-associated pathogens in urine from men with non-gonococcal urethritis: a case-control study. Acta Derm Venereol 2016; 96:689–694.
4. Sturm AW. Haemophilus influenzae
and Haemophilus parainfluenzae
in nongonococcal urethritis. J Infect Dis 1986; 153:165–167.
5. Hall GD, Washington JA 2nd. Haemophilus influenzae
in genitourinary tract infections. Diagn Microbiol Infect Dis 1983; 1:65–70.
6. Deza G, Martin-Ezquerra G, Gómez J, et al. Isolation of Haemophilus influenzae
and Haemophilus parainfluenzae
in urethral exudates from men with acute urethritis: a descriptive study of 52 cases. Sex Transm Infect 2016; 92:29–31.
7. Japanese Society for Sexually Transmitted Diseases. Guidelines for the diagnosis and treatment of sexually transmitted diseases 2006. Jpn J Sex Transm Dis 2006:17 Suppl:35–67.
8. Takanashi M, Ito S, Kaneto H, et al. Development and clinical application of an InvaderPlus® assay for the detection of genital mycoplasmas. J Infect Chemother 2015; 21:516–519.
9. Yasuda M, Takahashi S, Kiyota H, et al. Japanese guideline for clinical research of antimicrobial agents on urogenital infections: The first edition. J Infect Chemother 2011; 17:579–594.
10. Clinical and Laboratory Standards Institute. Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically; Approved Standard-Eight Edition M07-A8. CLSI, Wayne, PA, USA, 2009.
11. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing; Twentieth Informational Supplement M100-S20. CLSI: Wayne, PA, USA, 2010.
12. Yasuda M, Ito S, Kido A, et al. A single 2 g oral dose of extended-release azithromycin for treatment of gonococcal urethritis. J Antimicrob Chemother 2014; 69:3116–3118.
13. Workowski KA, Bolan GA, Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines, 2015. MMWR Recomm Rep 2015; 64:1–137.
14. Bignell C, Unemo M, European STI Guidelines Editorial Board. 2012 European guideline on the diagnosis and treatment of gonorrhoea in adults. Int J STD AIDS 2013; 24:85–92.
16. Niki Y, Hanaki H, Yagisawa M, et al. The first nationwide surveillance of bacterial respiratory pathogens conducted by the Japanese Society of Chemotherapy. Part 1: A general view of antibacterial susceptibility. J Infect Chemother 2008; 14:279–290.
17. Yanagihara K, Kadota J, Aoki N, et al. Nationwide surveillance of bacterial respiratory pathogens conducted by the surveillance committee of Japanese Society of Chemotherapy, the Japanese Association for Infectious Diseases, and the Japanese Society for Clinical Microbiology in 2010: General view of the pathogens' antibacterial susceptibility. J Infect Chemother 2015; 21:410–420.