Value of detecting leukocytospermia in the

diagnosis of genital tract infection in

subfertile men

Abstract
Objectives:Toevaluatewhetherdetectionofleukocytospermiainaroutinesemenanalysisisof
diagnosticvalueinselectingmenwithanactualmicrobialinfectionandtoassesstheassociation
betweenleukocytospermiaandahistoryofbacterialandviralinfections.
Design:Prospectiveclinicalstudy.
Setting:InfertilityclinicattheCenterforReproductiveMedicine,AcademicMedicalCenter,
Amsterdam,theNetherlands.
Patient(s):Onehundredeightyfourmenamongsubfertilecouplesattendingourinfertilityclinic.
Intervention(s):Thenumberofleukocyteswasassessedinthreesemensamples.Serologictestswere
performed,aswastransurethralcultureafterdigitalprostaticmassage.
MainOutcomeMeasure(s):Diagnosisofactualbacterialandviralinfectionsinrelationtoseminal
leukocyteconcentrations.Theassociationofahistoryofsexuallytransmitteddiseaseswithseminal
leukocyteconcentration.
Result(s):Anactualbacterialinfectionwaspresentin39%ofmen,and11%ofmenhadanactualviral
infection.Theareaunderthereceiveroperatingcurve,whichwasusedtodeterminewhetherdetection
ofleukocytospermiawasofdiagnosticvalueinidentifyingmenwithactualbacterialorviralinfections,
was0.55and0.56forbacterialandviralinfection,respectively.ApastinfectionwithN.gonorrhoeae
wasassociatedwiththepresenceofleukocytospermia.Apastviralinfectionwasnotassociatedwith
leukocytospermia.
Conclusion(s):Detectionofleukocytospermiaappearstobeofnodiagnosticvalueforselectionofmen
withactualmicrobialinfections,butleukocytospermiaisassociatedwithahistoryofgonorrhea.
Leukocytospermia,definedbytheWorldHealthOrganizationas>1106whitebloodcells(WBCs)
permLofsemen,hastraditionallybeenassociatedwithsubclinicalgenitaltractinfection(1,2,3,4).
Microorganismsthatmayleadtosubclinicalgenitaltractinfectionthatinducesurethritis,
prostatovesiculitis,andepididymitisareUreaplasmaurealyticum,Mycoplasmahominis,and
Chlamydiatrachomatis(5,6,7).Themostreliablewaytodetecttheseorganismsisbyculturinga
urethralswabafterdigitalprostaticmassage(8,9).However,thistestisexpensiveandtimeconsuming
andisreportedasanegativeexperiencebymostmen.
Toourknowledge,onlyonestudyhasexaminedtherelationshipbetweengenitaltractinfectionand
leukocytospermiatoassesswhetherdetectingleukocytospermiaisofdiagnosticvaluefordetermining
activemicrobialinfection(10).Thispreliminarystudyamongpotentialsemendonorsconcludedthat
seminalperoxidasepositivecellsarenotanadequateindicatorofasymptomaticurethralgland
infection.However,therewasonlyasmallnumberofsubjectswith>1106WBCs/mLofsemen.The
possibilitythatleukocytospermiamaybecausedbysexuallytransmissiblevirusesotherthanhuman
immunodeficiencyvirusandhepatitisBvirushasneverbeenstudied(11).

ThepurposeofthisstudywastoevaluatewhetherdetectionofWBCsinroutinesemenanalysescould
beusedasadiagnostictestforselectingmenwithapositivebacterialcultureorviralinfection;men
withbacterialinfectionscouldthenbetreated.Wealsoevaluatedwhetherleukocytospermiacanbe
regardedasachronicinflammatoryresponseinpatientswithahistoryofsexuallytransmitteddiseases
(STDs).

Materialsandmethods
BetweenApril1994andSeptember1996,200randomlychosenmenwhowere>18yearsofageand
whopresentedtotheCenterforReproductiveMedicineattheAcademicMedicalCenterinAmsterdam
wereaskedtoparticipateinthestudy.Coupleswerereferredbecauseofprimaryorsecondary
infertility.Thestudywasapprovedbyourhospitalsinstitutionalreviewboard.
Afterpatientsgaveinformedconsent,adetailedhistoryofSTDswastaken.Threesemensampleswere
obtainedfromallmen;therewasanintervalofatleast2weeksbetweeneachsample.Semenwas
collectedbymasturbationafteratleast72hoursofabstinence.Semensampleswereassessedwithuse
ofstandardizedtechniques,andtheresultswereanalyzedaccordingtoWorldHealthOrganization
criteriaforthefollowingparameters:volume,acidity,concentration,motility,morphology,fructose,
andantispermantibodies(1).
Thenumberofleukocytesineachsemensamplewasassessedwiththeperoxidaseassay(12).Brown
stainingroundcellswereclassifiedasperoxidasepositive.ThenumberofWBCswascalculatedby
multiplyingthenumberofroundcellsbythepercentageofperoxidasepositivecells.
Allmenunderwentdigitalprostaticmassage,afterwhichtwocottonurethralswabspecimenswere
obtained;standardmethodswereusedtodeterminewhetherbacteriawerepresent.Agramstainofa
urethralsmearwasexaminedforthepresenceofNeisseriagonorrhoeae.Onecottonswabwasplaced
inStuartmediumandusedforbacterialculture.CultureforN.gonorrhoeaewasperformedwithuseof
amodifiedThayerMartinplate.Inaddition,achocolateagarplatewasusedforroutineculture.
CultureofU.urealyticumandM.hominiswasperformedwithuseofaShepardbrothandaShepard
plate.ThepresenceofC.trachomatiswasinvestigatedwithDNARNAhybridization(GenProbe,San
Diego,CA).Apositivebacterialcultureorchemiluminescencesignalindicatedbacterialinfection.
Todetectanactiveorpreviouscytomegalovirus(CMV),EpsteinBarrvirus(EBV),orhepatitisBvirus
(HBV)infectionandapastC.trachomatisinfection,serumplasmawastestedforthepresenceof
antibodies.IgGandIgMantibodiestoCMVweredetectedwiththeImXsystem(AbbottLaboratories,
Chicago,IL).IgMpositiveserawereconfirmedintheVidassystem(bioMrieux,MarcylEtoile,
France).AntibodiestoEBVweredetectedbyindirectimmunofluorescencewithuseofIgGandIgM
antibodiestoEBVviralcapsidantigenandbyanticomplementimmunofluorescencewithuseofthe
EpsteinBarrvirusassociatednuclearantigen(GullLaboratories,SaltLakeCity,UT).Serologyfor
HBV(hepatitisBsurfaceantigen,antibodiestohepatitisBcoreantigen,andifappropriate,hepatitisB
eantigen)wasperformedwiththeImXsystem.IgGandIgAantibodiestoC.trachomatiswere
detectedwithuseofanenzymeimmunoassayfromLabSystems(Helsinki,Finland)basedona
speciesspecificsyntheticpeptide.ThistestallowsthescreeninganddiagnosingofC.trachomatis
infectionswithoutinterferenceofC.pneumoniaeantibodies.
ActualinfectionswerediagnosedbyhightitersofIgMtoCMV,ahightiterofIgMtoEBVinthe
absenceofantibodiestoEpsteinBarrvirusassociatednuclearantigen,andthepresenceofhepatitisB
surfaceantigenandhepatitisBeantigen.PastinfectionswithCMVandEBVwerediagnosedby
positiveIgGandnegativeIgMserology.PastHBVinfectionwasdiagnosedbydetectionofantibodyto
hepatitisBcoreantigenwithoutdetectionofhepatitisBsurfaceantigen.PastC.trachomatisinfection
wasdiagnosedbypositiveIgGand/orIgAserology.

Statistical analysis
The concentration of leukocytes in the first and second semen
sample were compared by calculating an intraclass correlation
coefficient. Differences in the number of leukocytes in sperm
between patients with a positive and negative history of bacterial or
viral infections and serologic findings of a previous bacterial or viral
infection were assessed with use of the Mann-Whitney U test. A P
value of <.05 was defined as statistically significant.
The diagnostic value of detecting leukocytes in sperm for identifying
men with a positive culture for C. trachomatis, U. urealyticum, or M.
hominis was assessed with a receiver operating characteristic curve.
We assessed the diagnostic performance of leukocytes in two ways.
First, we determined the concentration of leukocytes in the first
sample that indicated an active bacterial infection. Second, the
highest leukocyte concentration that was determined in the three
samples was assessed. The area under the receiver operating curve,
which was used to determine the performance of the test, was
calculated. The same was done to identify men whose positive
serology indicated an active CMV, EBV, or HBV infection.
Results
Of the 200 men that were asked to participate in the study, 12 men
refused participation and 4 men were excluded because of
hypergonadotropic hypogonadism. Therefore, 184 men were
available for analysis. The mean age (SD) was 34.7 6.7 years.
One hundred thirty-five patients (74%) had primary infertility. One
hundred thirty-six patients (74%) had oligoasthenospermia. Seven
female partners (3.8%) had subfertility due to a tubal factor,
whereas two female partners (1.1%) had polycystic ovary syndrome.
At the first semen analysis, 53 men had a WBC count of >1
106/mL. Among these 53 men, 36 (68%) had a WBC count of >1
106/mL at the second analysis. Of the 131 men with a WBC count of
<1 106/mL at first analysis, 22 had a WBC count of >1 106/mL
at the second analysis. The intraclass correlation coefficient for the
number of WBCs in the first and the second semen sample was 0.38
(95% CI, 0.270.55). Since the intraclass correlation coefficient
between the number of WBCs in the first and second semen sample
was low, the WBC concentration in the third sample was not taken
into account.
The prevalences of actual bacterial and viral infections in the male
genital tract are shown in Table 1. The association between a
positive history for gonorrhea, C. trachomatis infection, or syphilis
and leukocytospermia is shown in Table 2. The concentration of
WBCs in the first semen sample was significantly higher in patients

with a history of gonorrhea than in patients without a history of

gonorrhea (P = .001).
TABLE 1Prevalence of active bacterial and viral genital tract
infections in 184 men.
No. of
Organism
Prevalence (%)
infections
Ureaplasma
67
36
urealyticum
Mycoplasma
14
7.6
hominis
Chlamydia
1
0.5
trachomatis
Any bacterial
72
39
infection
Cytomegalovirus
16
8.7
Epstein-Barr virus
1
0.5
Hepatitis B virus
4
2.2
Any viral infection
20
10.9
TABLE 2Association between a history of a sexually transmitted
disease and leukocytospermia. legend
No. of
No. of
Mean no.
Mean no.
Sexually
men
men
P
of WBCs
of WBCs
transmitted
with a
without
valu
(106/mL)
(106/mL)
disease
history
a history
e
(range)
(range)
of STDs
of STDs
.
Gonorrhea
26
2 (013)
158
0.5 (040)
0013
Chlamydia
trachomatis
15
0.5 (07)
169
0.5 (040) .42
infection
Syphilis
2
0.5 (01)
182
0.5 (040) .38
legendSTDs = sexually transmitted diseases; WBCs = white blood
cells.
In the first semen sample.
The association between a positive serology indicating a previous
infection with CMV, EBV, HBV, or C. trachomatis and
leukocytospermia is shown in Table 3. Previous viral infections were
not associated with the concentration of WBCs in semen (P>.05).
Serologic markers indicating a past C. trachomatis infection were
not associated with the concentration of WBCs in semen (P>.05).
TABLE 3Association between a positive serology indicating a
previous infection with Epstein-Barr virus, cytomegalovirus,
hepatitis B virus, or Chlamydia trachomatis and
leukocytospermia.legend
Positive
No. of
Mean no.
No. of
Mean no.
P

serology

men
men
of WBCs
of WBCs
with a
without a
valu
6
(10 /mL)
(106/mL)
history
history
e
(range)
(range)
of STDs
of STDs

Epstein-Barr
178
0.5 (040)
6
0 (02)
.64
virus
Cytomegalovir
87
0.5 (020)
97
0.5 (040)
.39
us
Hepatitis B
22
0.5 (07)
162
0.5 (040)
.34
virus
Chlamydia
64
0.5 (013)
120
0.5 (040)
.38
trachomatis
legendSTDs = sexually transmitted diseases; WBCs = white blood
cells.
In the first semen sample.
The receiver operating curve for the number of WBCs in sperm that
were indicative of an active bacterial or viral infection are shown in
Figure 1. The area under the receiver operating curve was 0.55
(95% CI, 0.460.63) for a positive culture indicating an actual
bacterial genital tract infection. The area under the receiver
operating curve was 0.56 (95% CI, 0.440.78) for a positive serology
indicating an actual viral genital tract infection. The sensitivity and
specificity of leukocytospermia for detecting M. hominis infection
were 93% (95% CI, 8897) and 10% (95% CI, 421), respectively.
The sensitivity and specificity of leukocytospermia for detecting U.
urealyticum infection were 65% (95% CI, 5774) and 40% (95% CI,
2854), respectively. The sensitivity and specificity of
leukocytospermia for detecting C. trachomatis infection could not be
calculated because of the low prevalence of this organism.

FIGURE 1
ReceiveroperatingcurvesforvaryingseminalWBCconcentrationsininfertilemenwithabacterial
genitaltractinfection()oraviralgenitaltractinfection().

View Large Image | Download PowerPoint Slide

Discussion
This study shows that leukocytospermia is of no diagnostic value in
selecting patients with an actual bacterial or viral infection and
colonization of the genital tract. In this infertility cohort, the overall
prevalence of silent genital tract infections due to U. urealyticum, M.

hominis, or C. trachomatis was 39%. The overall prevalence of

infections with CMV, EBV, or HBV was 11%. It is not clear whether
serologic evidence of a recent infection reflects the presence of
these viruses in the genital tract.
Detection of sexually transmitted bacterial genital tract infections in
men is important because these infections, like C. trachomatis
infection, may cause severe damage in the genital tract of their
female partners (13). There have been few studies on the long-term
consequences of persistent U. urealyticum, M. hominis, and C.
trachomatis infections of the male genital tract that suggest that
these infections may lead to subfertility (14, 15, 16). Studies should
be performed to determine whether the presence of U. urealyticum
and M. hominis in bacterial cultures reflects colonization of the distal
urethra rather than an active infection. In view of the above and
considering the fact that there is no relation between
leukocytospermia and actual bacterial infection, cultures should be
performed for detection of genital infections.
Most men who have a transurethral swab test after digital prostatic
massage report that it is a negative experience. Since U.
urealyticum, M. hominis, and C. trachomatis are sexually
transmitted, obtaining a cervical swab specimen for culture from
their female partners as part of the routine fertility work-up may be
a suitable alternative for diagnosing a silent genital tract infection in
men. Apart from the poor correlation between leukocytospermia and
genital tract infection, leukocytospermia itself is episodic and
disappears over time, as is demonstrated by an intraclass
correlation coefficient of only 0.38 (95% CI, 0.270.55) between the
number of WBCs in the first and second semen sample. In contrast
with other investigators (10), we found an association between a
history of gonorrhea and a persistent immunologic response caused
by peroxidase-positive cells. Further research must elucidate the
role of these WBCs in the reproductive process.
In conclusion, silent genital tract infection or colonization with U.
urealyticum and M. hominis is a frequent finding in populations of
infertile men. The quantification of WBCs in semen is of no
diagnostic value in selecting patients with an active genital tract
infection and therefore cannot replace culture. The role of WBCs in
semen in predicting reproductive performance warrants further
study.