Role of Mycoplasma genitalium and Ureaplasma

O R I G I N A L
A R T I C L E urealyticum in non-gonococcal urethritis in Hong
Kong
John THT Yu
William YM Tang Objective To determine the association of Mycoplasma genitalium and
KH Lau Ureaplasma urealyticum in symptomatic male patients presenting
LY Chong with non-gonococcal urethritis in a sexually transmitted infection
KK Lo clinic in Hong Kong.
Carlos KH Wong Design Cross-sectional study.
MY Wong Setting A sexually transmitted infection clinic, Department of Health,
Centre for Health Protection, Hong Kong.
Patients A cohort of consecutive new male patients attending the
government sexually transmitted infection clinic.
Main outcome measures Prevalence of Mycoplasma genitalium and Ureaplasma
urealyticum among symptomatic male patients with non-
gonococcal urethritis and asymptomatic patients without non-
gonococcal urethritis.
Results Specimens of 22 and 10 patients tested positive by polymerase
chain reaction for Ureaplasma urealyticum and Mycoplasma
genitalium respectively, among the symptomatic non-gonococcal
urethritis group (n=98). In the asymptomatic control group
(n=236), corresponding patient numbers whose specimens
tested positive were 47 and 5. There was no statistically significant
difference between the two groups, in terms of the proportion
of patients infected with Mycoplasma genitalium (P=0.799) or
Ureaplasma urealyticum (P=0.535).
Conclusions In our study, demonstration of Mycoplasma genitalium and
Ureaplasma urealyticum by polymerase chain reaction was not
associated with symptomatic non-gonococcal urethritis in male
patients attending a Hong Kong government clinic for sexually
transmitted infections.
Key words
Mycoplasma infections; Polymerase chain
reaction; Sexually transmitted diseases;
Ureaplasma infections; Urethritis Introduction
Hong Kong Med J 2008;14:125-9
Urethral inflammation is the most common presentation among male patients attending
government sexually transmitted infection (STI) clinics in Hong Kong. In 2006, among
male attendees of Social Hygiene Clinic (SHC), 9572 were newly diagnosed to have
Yau Ma Tei Dermatology Centre, Social
an STI; among these 1413 (14.8%) and 4540 (47.4%) patients respectively, were found
Hygiene Service, Hong Kong
JTHT Yu*, FHKCP, BM BCh (Oxon) to have gonorrhoea and non-gonococcal urethritis (NGU).1 Urethritis is classified as
KH Lau, FRCP, FHKAM (Medicine) gonococcal or non-gonococcal depending on the presence or absence of Neisseria
LY Chong, FRCP, FHKAM (Medicine)
gonorrhoeae isolated as the aetiological agent. In the past, gonorrhoea was the
Tuen Mun Social Hygiene Clinic, Social
Hygiene Service, Hong Kong commonest STI in Hong Kong, but more recently, it has been replaced by NGU. The
WYM Tang, FRCP, FHKAM (Medicine) most frequent cause of NGU is Chlamydia trachomatis (CT),1 whilst other possible
Cheung Sha Wan Dermatology Clinic,
associations are identified less frequently and include Mycoplasma genitalium (MG),2-9
Social Hygiene Service, Hong Kong
KK Lo, FRCP, FHKAM (Medicine) Ureaplasma urealyticum (UU),10,11 Trichomonas vaginalis,12 Neisseria meningitidis, Herpes
Department of Mathematics, Hong Kong simplex virus, Candida species, bacterial urinary tract infection, urethral stricture, and
University of Science and Technology,
foreign bodies.
Clear Water Bay, Kowloon, Hong Kong
CKH Wong, BSc Chlamydia trachomatis has been well established as a pathogen responsible for NGU
MY Wong, MSc, PhD
in Hong Kong, but other aetiological agents are not well defined. The objective of this
* Current address: St Teresa’s Hospital, 327 study was to determine the association of MG and UU in patients presenting with NGU
Prince Edward Road, Kowloon, Hong Kong in a government STI clinic in Hong Kong, by comparing the proportion of patients testing
Correspondence to: Dr JTHT Yu
positive for these organisms among symptomatic patients with NGU and asymptomatic
E-mail: [email protected] controls (without NGU).

Hong Kong Med J Vol 14 No 2 # April 2008 # www.hkmj.org 125

 # Yu et al #

urethral discharge was specifically noted. When there

was no visible discharge, the patient would be asked
to massage the urethra and express any possible
exudate. For every patient, the first author obtained
the history and conducted the physical examination.
Symptomatic patients were defined as those who
presented with either: frequency, dysuria, discharge,
urethral discomfort, or testicular discomfort or they
had an observable urethral discharge upon physical
examination with or without ‘milking’. Asymptomatic
patients were defined as those who did not complain
of either of these symptoms, and had no observable
urethral discharge upon physical examination.
For each recruit, the time they had last voided
urine was noted. A urethral smear was obtained by
a sterile inoculation loop and directly plated onto
the specimen slide for Gram stain and Thayer-Martin
medium for gonococcal culture. Only two designated
research assistants at Yung Fung Shee SHC performed
the microscopy in order to limit the inter-operator
variability. The number of polymorphonuclear
leukocytes (PMLs) with or without intracellular
gonococci under a high-power (x1000) microscopic
field over an average of five fields were recorded as
<1, <5, 5-9, and 10-29. Patients were defined as having
NGU if their (i) urethral smear for Gram stain showed
five or more PMLs per high-power microscopic
field, and (ii) their smears were negative for Gram-
negative diplococci and cultures negative for N
gonorrhoeae. Whereas, gonococcal urethral infection
Methods was diagnosed if Gram staining of the urethral smear
revealed intracellular Gram-negative diplococci or
A cross-sectional study was conducted in a were culture positive for N gonorrhoeae.
government STI clinic (Yung Fung Shee SHC,
Kowloon), where consecutive new male patients were Subsequently, a thin cotton-tipped swab was
recruited to participate in the study from November inserted into the urethra to collect further exudates,
2005 to May 2006 (7 months). The study was carried and the swab was then shaken vigorously in the transit
out with the approval from the ethics committee of medium, which was then subjected to nucleic acid
the Department of Health and informed consent of amplification testing for Chlamydia (COBAS Amplicor,
the patients. Patients under the age of 18 years, those Roche Diagnostic Systems, Inc, Branchburg, NJ).
who had received antibiotic therapy within 1 month A first catch urine sample was obtained and tested
prior to presentation, or had structural abnormality for MG and UU by a technique involving real-time
of the urogenital system were excluded. Those who polymerase chain reaction (PCR). The DNA was
were homosexual, or unable to establish effective extracted from 1 mL of urine using the NucliSens mini-
and meaningful communication due to cognitive MAG (Magnetic extraction) System (bioMérieux, The
impairment, psychiatric illness, language barrier, or Netherlands). Two real-time PCR assays were used to
physical disability were also excluded. differentially detect UU and MG. Real-time PCR assay
of UU was specifically designed to target the urease
A questionnaire was used to record the history gene by using two TaqMan minor groove binder
and in particular symptoms of urethral discharge, primers and probe sets. Then MG was detected by
dysuria, penile irritation, frequency, and testicular a TaqMan minor groove binder real-time PCR assay,
discomfort. In addition, the patient’s age, ethnicity, targeting a conserved region of the MgPa adhesion
age of first sexual intercourse, as well as history of gene. Exogenous internal positive control (Applied
prior STIs, casual sex in the past 12 months, history Biosystems, CA) was added into both real-time PCR
of new sexual partner in the past 3 months, and assays for monitoring PCR inhibition. All recruited
frequency and correct use of condoms during casual patients were also tested for syphilis and human
sex were recorded. immunodeficiency virus. If STIs were detected, the
A full urogenital examination was performed patients were managed accordingly with appropriate
to detect any STIs and the presence or absence of treatment protocols and health counselling.

126 Hong Kong Med J Vol 14 No 2 # April 2008 # www.hkmj.org

 # Non-gonococcal urethritis in Hong Kong #

TABLE 1. Demographic and sexual behaviour data for the symptomatic non-gonococcal urethritis (NGU) group and asymptomatic controls
Symptomatic NGU group Asymptomatic control group P value
(n=98) (n=236)
Mean age (range) [years] 38 (18-75) 41 (18-87) 0.048*
Age of sexual debut (years) 20 21 0.027*
No. with a history of sexually transmitted infections 60 (61%) 59 (25%) <0.001†
No. with a history of casual sex in the past 12 months 80 (82%) 173 (73%) 0.106†
No. with a history of new partner(s) in the past 3 months 69 (70%) 127 (54%) 0.005†
Always used a condom during casual sex in the past 3 months 43% 57% 0.064†
No. of correct use of condom 31 (32%) 90 (38%) 0.260†
*
t test
†
χ2 test

TABLE 2. Numbers of patients with organisms detected by polymerase chain reaction in the non-gonococcal urethritis (NGU) group and asymptomatic
controls
Organisms Symptomatic NGU group (n=98) Asymptomatic control group (n=236) P value (with logistic regression)
Chlamydia trachomatis 49 (50%) 14 (6%) <0.001
Ureaplasma urealyticum 22 (22%) 47 (20%) 0.535
Mycoplasma genitalium 10 (10%) 5 (2%) 0.799

By comparing the numbers and proportions of and 138 patients respectively had casual sex within
persons testing positive for MG and UU among NGU the past 3 months. Among the patients who had such
patients and those without, the significance of these casual sex in the past 3 months, 31 (43%) within the
two factors as causative agents in NGU could be NGU group and 78 (57%) within the asymptomatic
explored. Data were entered and stored in Microsoft controls used a condom consistently. A summary
Access and analysed using the Statistical Package for of the demographic data and sexual behaviour
the Social Sciences (Windows Version 13, SPSS Inc, pertaining to the two groups is shown in Table 1.
Chicago, US). Proportions were compared using Within the symptomatic NGU group, the
Pearson’s Chi squared test. Logistic regression with number of patients testing positive by PCR for CT,
confounding factors adjustment was used to test the UU, and MG were 49, 22, and 10 respectively. Among
association of CT, MG, and UU with NGU. Unpaired these patients, 10 were co-infected with CT and UU
independent t tests were utilised for comparing and three were co-infected with CT and MG. Among
continuous variables. the asymptomatic controls, the number of patients
testing positive by PCR for CT, UU, and MG were 14,
Results 47, and 5 respectively; and among these, seven were
co-infected with CT and UU, one was co-infected with
A total of 507 patients were recruited into the study CT, MG, and UU. There was no statistically significant
from 543 consecutive patients. Among the patients difference in the proportion of patients infected with
who were excluded, 20 had taken antibiotics in the MG (P=0.799) or UU (P=0.535) in the symptomatic
previous month, five were homosexuals, six had NGU and asymptomatic control groups (Table 2).
indeterminate PCR results secondary to the presence
of inhibitors, four posed a language barrier, and one
had bladder extrophy. Discussion
The mean age of the patients was 40 (median Chlamydia trachomatis is well established as a cause
38; range, 18-87) years; 99% of whom were Chinese. of NGU and this was confirmed again in the current
There were 98 symptomatic patients (with NGU) and study in which there was a significant difference
236 asymptomatic controls; their mean ages were 38 between the symptomatic NGU and asymptomatic
and 41 years respectively. The mean age at first sexual control groups (P<0.001). However, the relationship
intercourse for the symptomatic NGU group was 20 and association of NGU with UU and MG is still
years, whereas for the asymptomatic controls, it was 21 debated. In 1977, two volunteers self-inoculated UU
years. In the symptomatic NGU group, a significantly onto their urethra and this produced symptomatic
greater proportion of patients than asymptomatic urethritis with PML in the urine and provided
controls had a history of prior STIs and admitted to evidence that UU was pathogenic.13 In the current
having a new partner in the past 3 months. In all, 72 study, among 98 patients with symptomatic NGU, UU

Hong Kong Med J Vol 14 No 2 # April 2008 # www.hkmj.org 127

 # Yu et al #

TABLE 3. Summary of recently published articles on the association of Ureaplasma urealyticum (UU) with non-gonococcal urethritis (NGU)
Paper Symptomatic NGU cases Asymptomatic controls Detection of UU (symptomatic
cases vs controls)
Present study, 2007, Hong Kong 98, ≥5 PML* on smear 236, <5 PML on smear 22% vs 20% (P=0.535)
Morency et al,3 2001, Central African Republic 136, ≥5 PML on smear 100 44% vs 48% (P>0.05)
Totten et al, 2001, Seattle, US
14
121, ≥5 PML on smear 117, <5 PML on smear 35% vs 42% (P=0.34)
Pepin et al, 2001, West Africa
4
659 with urethral discharge 339 26% vs 28% (P>0.05)
Deguchi et al,10 2004, Japan 317, ≥5 PML on smear 141, <5 PML on smear 16% vs 8% (P=0.025)
Bakare et al, 2002, Nigeria
15
154 with NGU Not specified 14% of cases (P<0.001)
*
PML denotes polymorphonuclear leukocytes per high-power field

TABLE 4. Summary of recently published articles on the association of Mycoplasma genitalium (MG) with non-gonococcal urethritis (NGU)
Paper Symptomatic NGU cases Asymptomatic controls Detection of MG (symptomatic
cases vs controls)
Present study, 2007, Hong Kong 98, ≥5 PML* on smear 236, <5 PML on smear 10% vs 2% (P=0.799)
Leung et al, 2006, UK
9
302, ≥5 PML on smear 378 11% vs 1% (P<0.001)
Mena et al,2 2002, New Orleans, US 97, ≥5 PML on smear 184 21% vs 8% (P<0.002)
Morency et al, 2001, Central African Republic
3
136, ≥5 PML on smear 100 42% vs 15% (P<0.001)
Totten et al, 2001, Seattle, US
14
121, ≥5 PML on smear 117, <5 PML on smear 22% vs 4% (P<0.01)
Pepin et al,4 2001, West Africa 659, urethral discharge 339 10% vs 9% (P=0.05)
Johannisson et al, 2000, Sweden
5
115, ≥5 PML on smear 118, <5 PML on smear 15% vs 1% (P<0.001)
Busolo et al,6 1997, Italy 52, urethritis 44 12% vs 0% (P=0.03)
Janier et al, 1995, Paris
7
100, with discharge 96, no discharge 29% vs 9% (P<0.001)
Horner et al,8 1993, UK 103, ≥5 PML on smear 53, <5 PML on smear 23% vs 6% (P=0.006)
*
PML denotes polymorphonuclear leukocytes per high-power field

was found in 22%, whereas in the 236 asymptomatic One limitation of the current study was the
patients without NGU, UU was found in 20%. More considerable difference in the size of the symptomatic
recently, findings from a number of studies3,4,10,14,15 NGU and asymptomatic control groups, which was
investigating the possible association of UU with due to consecutive patient recruitment (the strategy
NGU using PCR have been reported; their results adopted to reduce selection bias). It may have been
were conflicting, but could not be compared directly more advantageous to compare groups of similar size
as the selection of controls differed (Table 3). and with similar sexual demographics. In addition,
Mycoplasma genitalium is the smallest free- a larger study population would be more likely
living organism, widespread in nature and was first to compensate for several possible confounding
isolated in 1980 from two of 13 men with NGU.16 Since factors. For example, the patients in the symptomatic
the development of sensitive and specific PCR assays NGU group were younger and had an earlier age at
in the early 1990s, around the world a number of first sexual intercourse. They also tended to engage
clinical studies2-9,14 have shown a strong association in more risky sexual behaviour, being more likely to
between MG and NGU, independent of CT (Table have a new partner in the preceding 3 months and a
4). In the current local study, five (2%) of the 236 history of prior STIs. In a sense, these differences were
asymptomatic patients without NGU tested positive only to be expected, as this study involved a selected
for MG, compared to 10 (10%) of the 98 symptomatic group of symptomatic patients with confirmed NGU
patients with NGU. However, this difference was not to compare with asymptomatic patients without
statistically significantly different when they were NGU. Men who had sex with men were not included
compared by logistic regression analysis (P=0.799; in the current study to avoid further complicating the
Table 2). As the prevalence of MG was relatively issues.
low in Hong Kong, a type II error may have been In conclusion, the current study did not
inadvertently introduced, so a larger study might demonstrate an association between MG or UU and
help to clarify any possible association of MG with NGU among symptomatic male patients attending a
NGU. government STI clinic in Hong Kong. Nevertheless,

128 Hong Kong Med J Vol 14 No 2 # April 2008 # www.hkmj.org

 # Non-gonococcal urethritis in Hong Kong #

the use of condom which gives significant protection investigate NGU among male patients and their
against other STIs like CT should be emphasised female sexual partners are needed in the future to
to the at-risk population.17 Larger local studies to better define the pathogenicity of MG and UU.

References

1. Chan KT. Non-gonococcal urethritis and non-specific 9. Leung A, Eastick K, Haddon LE, Horn CK, Ahuja D, Horner
genital infection. In: Lo KK, Chong LY, Tang YM, Ho KM, PJ. Mycoplasma genitalium is associated with symptomatic
editors. Handbook of dermatology and venereology. Vol 2. urethritis. Int J STD AIDS 2006;17:285-8.
3rd ed. Hong Kong: Social Hygiene Service, Department of 10. Deguchi T, Yoshida T, Miyazawa T, et al. Association of
Health; 2001:316-23. Ureaplasma urealyticum (biovar 2) with nongonococcal
2. Mena L, Wang X, Mroczkowski TF, Martin DH. Mycoplasma urethritis. Sex Transm Dis 2004;31:192-5.
genitalium infections in asymptomatic men and men with 11. Maeda S, Deguchi T, Ishiko H, et al. Detection of
urethritis attending a sexually transmitted diseases clinic in Mycoplasma genitalium, Mycoplasma hominis, Ureaplasma
New Orleans. Clin Infect Dis 2002;35:1167-73. parvum (biovar 1) and Ureaplasma urealyticum (biovar 2)
3. Morency P, Dubois MJ, Gresenguet G, et al. Aetiology of in patients with non-gonococcal urethritis using polymerase
urethral discharge in Bangui, Central African Republic. Sex chain reaction-microtiter plate hybridization. Int J Urol
Transm Infect 2001;77:125-9. 2004;11:750-4.
4. Pepin J, Sobela F, Deslandes S, et al. Etiology of urethral 12. Krieger JN, Jenny C, Verdon M, et al. Clinical manifestations
discharge in West Africa: the role of Mycoplasma genitalium of trichomoniasis in men. Ann Intern Med 1993;118:844-9.
and Trichomonas vaginalis. Bull World Health Organ 13. Taylor-Robinson D, Csonka GW, Prentice MJ. Human intra-
2001;79:118-26. urethral inoculation of ureaplasmas. Q J Med 1977;46:309-
5. Johannisson G, Enstrom Y, Lowhagen GB, et al. Occurrence 26.
and treatment of Mycoplasma genitalium in patients visiting 14. Totten PA, Schwartz MA, Sjöström KE, et al. Association of
STD clinics in Sweden. Int J STD AIDS 2000;11:324-6. Mycoplasma genitalium with nongonococcal urethritis in
6. Busolo F, Camposampiero D, Bordignon G, Bertollo G. heterosexual men. J Infect Dis 2001;183:269-76.
Detection of Mycoplasma genitalium and Chlamydia 15. Bakare RA, Oni AA, Umar US, Kehinde AO, Fayemiwo
trachomatis DNAs in male patients with urethritis using the SA, Fasina NA. Ureaplasma urealyticum as a cause of non-
polymerase chain reaction. New Microbiol 1997;20:325- gonococcal urethritis: the Ibadan experience. Niger Postgrad
32. Med J 2002;9:140-5.
7. Janier M, Lassau F, Casin I, et al. Male urethritis with and 16. Tully JG, Taylor-Robinson D, Cole RM, Rose DL. A newly
without discharge: a clinical and microbiological study. Sex discovered mycoplasma in the human urogenital tract.
Transm Dis 1995;22:244-52. Lancet 1981;1:1288-91.
8. Horner PJ, Gilroy CB, Thomas BJ, Naidoo RO, Taylor- 17. Niccolai LM, Rowhani-Rahbar A, Jenkins H, Green S, Dunne
Robinson D. Association of Mycoplasma genitalium with DW. Condom effectiveness for prevention of Chlamydia
acute non-gonococcal urethritis. Lancet 1993;342:582-5. trachomatis infection. Sex Transm Infect 2005;81:323-5.

Hong Kong Med J Vol 14 No 2 # April 2008 # www.hkmj.org 129