international journal of andrology, 17: 127- 134 (1994)

Peroxidase-positiveround cells and microorganisms in human

semen together with antibiotic treatment adversely influence
the outcome of in-vitro fertilization and embryo transfer

CH. DE GEYTER', M. DE GEYTER~,H. M. BEHRE~,H. p. G.

SCHNEIDER' and E. NIESCHLAG2

'Woman's Hospital of the University, Albert-Schweitzer StraRe 33, D-48129 Miinster and
21nstitute of Reproductive Medicine, Steinfurter StraBe 107, D-48149 Munster, Germany

Summary
Human semen contains not only spermatozoa but also other cells routinely
differentiated as being peroxidase-positive (e.g. leucocytes) and peroxidase-negative
(eg. immature germ cells and lymphocytes) cells. Considerable uncertainty exists about
their role in male fertility. To assess the clinical value of both parameters, and of
microorganisms in semen, 391 treatments with in-vitro fertilization and embryo transfer
were analysed retrospectively, and the concentrations of both peroxidase-positive and
-negative cells, together with the presence of microorganisms in semen, were compared
with both the fertilization and pregnancy rates. The data indicate that the results of
treatment were affected only by excessively elevated concentrations of peroxidase-
positive cells (>6 x 10h/ml) and only marginally by the presence of microorganisms in
the semen. The pregnancy rate after in-vitro fertilization and embryo transfer was not
improved by antibiotic treatments preceding gamete recovery by several weeks. The
increased presence of peroxidase-negative cells (e.g. germ cells) in semen was not
associated with a significant change in the pregnancy rate. However, the concentration
of peroxidase-negative cells in semen correlated significantly with sperm numbers
(p<O.Ol), sperm concentration (p<O.Ol), and normal morphology rates (p<O.Ol). It is
concluded that short-term antibiotic treatment of asymptomatic patients before assisted
reproduction should be handled with caution. The widespread view that peroxidase-
negative cells in semen are harmful is rejected.

Keywords: semen analysis, round cells, in-vitro fertilization, leucocytes in semen.

Introduction
Human semen contains not only spermatozoa but also a
heterogeneous group of non-spermatozoa1 cells, consisting
of leucocytes, lymphocytes, immature germ cells and epithe-
lial cells from the efferent duct and the accessory glands.
Male infertility has been associated with increased preva-
lence of leucocytes in semen (Wolff & Anderson, 1988),
orignating from subclinical infections of the epididymis
Correspondence: Prof. Dr E. Nieschlag, Institute of Reproductive
Medicine, Steinfurter Strafle 107, D-48149 Miinster, Germany.
(El Deniiry et al., 1986), the prostate (Conihaire ef a/.,
1980) or the seminal vesicles (Gonzales et al., 1992).
Seminal leucocytes may interfere with fertilization a5
demonstrated experimentally in the hamster zona-free
oocyte penetration test (Berger et al., 1982; Maruyania et a/.,
1985; Vogelpoel et nl., 1991) and the presence of leucocytes
in senien is associated with reduced in-vitro fertilization
rates (Cohen et a/., 1985; Talbert et a / . , 1987).
To distinguish leucocytes from other non-spermatozoa1
cells in the semen, peroxidase staining was introduced into

128 Ch. Da Geyter ef a/.
routine semen analysis (WHO, 1987). Leucocytes are iden-
tified by the peroxidase staining of cytoplasmic granules
(Nahoum & Cardozo, 1980). The presence of peroxidase-
positive and -negative cells in senien frequently causes
uncertainties in the assessment of sperm function, the role of
which in assisted reproduction has been evaluated intensive-
ly (De Geyter et al., 1988, 1992). Since assisted reproduction
provides an opportunity to have more factors under control
than during natural intercourse, the results of in-vitro ferril-
ization and embryo transfer (IVFIET) were used to investi-
gate the influence of non-spermatozoa1 cells in semen on
fertilization and pregnancy rates.
Material and methods
Diagnostic and therapeutic measures in preparationfor
assisted fertilization
Before treatment all couples were subjected to an inten-
sive diagnostic investigation. All women received a c o n -
plete endocrinological work-up, including early follicular
phase serum gonadotrophins, midluteal progesterone, serum
androgens, serum prolactin, and thyroid hormones (De
Geyter et a/., 1993). Furthermore, the immune status for
rubella, HIV, syphilis, and hepatitis B antigens was checked.
Possible infections of the cervix and urethra with Chlamydia
trachomatis were excluded using both serology and cell cul-
ture techniques. In addition to a routine physical examina-
tion and midcycle vaginosonography, both laparoscopy and
hysteroscopy were performed to examine pelvine anatomy
and tubal function.
The husbands were also subjected to a thorough physical
examination, endocrine screening (serum testosterone,
luteinizing hormone, follicle-stimulating hormone, pro-
lactin) and sonography of the scrota1 organs. At least two
semen analyses were performed ( W H O , 1987). Sperm func-
tion was further assessed using an homologous sperm-cervi-
cal mucus penetration test in vitro (WHO, 1987; De Geyter
et ai., 1988). The immune status of the husband for HIV and
hepatitis B antigens was checked.
Approximately 5 weeks before IVF, microbiological ana-
lyses consisting of standard aerobic and anaerobic bacterial
cultures were perfornied in the Institute of Microbiology of
the University on 281 out of 391 patients. If > 1 x 10h/ml
peroxidase-negative cells or > 8 x 106/ml peroxidase-nega-
tive cells were detected, an additional specific search for
Mycoplasma hominis and Ureaplasnia urealyticum was ini-
tiated. If after microbiological identification more that 1000
bacteria/ml senien were found, antibiotic sensitivity was
determined and both partners were treated with appropriate
antibiotics before IVF. All patients included in the analysis
were free of symptoms of genital infections.
Patients' data
The data from 391 consecutive treatment cycles in 199
couples participating in our IVF/ET programme were
13652605, 1994, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2605.1994.tb01231.x by Test, Wiley Online Library on [13/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
analysed. The data were collected over a period of 19
months. The median number of treatments per couple was
1.96 (range 1-5). The median duration of infertility was 8
years (range 1-23). The most common indication for
IVF/ET was bilateral tubal occlusion or loss of tubal func-
tion (75.6% of all couples). Tuba1 infertility was found in
combination with abnormal sperm quality in 53 couples
(26.4%). Other indications included male infertility only
(14.9%), unexplained infertility (7%), and secondary amen-
orrhoea (2.5%). Male infertility was diagnosed if two con-
secutive conventional sperm analyses, performed following
the W H O guidelines, revealed subnomial values (WHO,
1987). Female patients with causes of infertility other than
tubal occlusion were selected for IVF/ET only after four
treatment cycles with gonadotrophin stimulation of the
ovaries followed by intrauterine insemination. Positive
M A R tests for IgG were found in 64 treatment cycles
(16.4%) and for IgA in 19 treatment cycles (4.9%).
Semen analysis and processing of spermatozoa
Semen analysis was performed according to WHO
guidelines ( W H O , 1987) and in all patients the standard
swim-up procedure was used for sperm preparation (De
Geyter et al., 1992):Semen was obtained in the clinic by
masturbation after a period of abstinence ranging from 2 to
7 days. Semen was allowed to liquefy at 37OC for 30 min.
Motility of spermatozoa in seminal plasma was assessed at
4 0 0 ~niagnification. For morphological assessment of sper-
matozoa an air-dried smear was stained by the Papanicolaou
method and sperm morphology was evaluated at 4 0 0 mag- ~
nification using phase-contrast microscopy.
Various quality control measures in the andrology labo-
ratory were performed during data collection. Details have
been described elsewhere (Cooper et a)., 1991, 1992). The
coefficients of variation were calculated based on the
monthly mean values of all patients attending the andrology
laboratory (ranging from 102 to 207 every month), and was
<18% for sperm concentration, <lo% for spemi (ini)niotil-
ity, and <20% for normal sperni niorphology.
Leucocytes were identified by peroxidase staining of the
somatic cells in semen. An aliquot of 100 p1 of the liquefied
ejaculate was mixed with 1 nil toluidine blue solution
(Nahouni & Cardozo, 1980; W H O , 1987). The peroxidase-
positive (e.g. leucocytes) and the peroxidase-negative cells
were counted in a haeniocytonieter. The coefficient of vari-
ation was calculated based on the monthly mean values of all
patients attending the andrology laboratory and was <2'%,for
peroxidase-positive cells and < 18% for peroxidase-negative
cells.
Protocol f o r IVF/ET
All patients were treated following the same protocol
described in detail elsewhere (De Geyter et al, 1992, 1993).
Ovarian stimulation was performed with human
menopausal gonadotrophin and/or human urinary FSH

(Pergonal and Fertinorm, Serono) after previous pituitary
suppression with a long-acting G n R H analogue (3.75 mg
triptorelin, Decapeptyl Depot, Ferring, Kiel, Germany).
Oocytes were cultured in small drops of 100 pl Ham's F10
medium supplemented with 3 mg/nll BSA under paraffin
oil. Approximately 7 h after collection of the oocytes
8000-12 000 motile spermatozoa were added to up to five
oocytes. If, from previous treatment trials, the fertilization
rate was known to be reduced, the number of oocytes
inseminated was increased to seven and the number of sper-
matozoa added to the oocytes was increased to 30000. The
occurrence of fertilization was identified by the visualization
of two pronuclei and/or normal embryo cleavage. The fer-
tilized oocytes were subsequently cultured in 100 p1 Ham's
F10 supplemented with 7.5% (v/v) heat-inactivated human
serum under paraffin oil. Embryo transfer into the uterine
cavity or transvaginally into the fallopian tube of the patient
was performed the next day. All normally fertilized oocytes
were transferred. The h e a l phase was supported with
repeated injections of human chorionic gonadotrophin
(hCG, Pregnesin, Serono). Pregnancy was confirmed by
repeated determination of the serum hCG concentration
and by the vaginosonographic visualization of a pregnancy
sack.
Statistical analysis
Statistical analysis was performed with Statgraphics
(STST Rockville, MD, USA). Both the fertilization and
pregnancy rates were compared and assessed using chi-
squared analysis. Differences in sperm concentration, pro-
gressive motility and normal morphology rates were
analysed using the KruskaU-Wallis analysis by ranks. Simple
and multiple regression analyses were performed after loga-
rithmic transformation of the non-normally distributed data
or after arcsine transformation of percentage values.
Results
IVF/ET resulted in 87 clinical pregnancies (21.2%),
including 13 twin pregnancies (15.7%), four triplets (4.8%),
and one quadruplet (1.2%). In the patients suffering from
tuba1 infertility the pregnancy rate was 22.496, those with
male infertility 12.2%, those with unexplained infertility
24.196, and those with hypogonadotrophic ovarian failure
22.2%. The miscarriage rate was 31.3%.
Efect o f the concentration ofperoxiduse-positive cells in
s e m m on the outcome o f I V F / E T
The concentration of peroxidase-positive cells in semen
varied between 0 and 10 x lO'/rnl (up to 32.4%)of the cells
in semen). The effect of the number of peroxidase-positive
cells in semen on the fertilization and pregnancy rate in
IVF/ET is presented in Fig. 1. Both the number of oocytes
fertilized and the number of pregnancies were reduced in
the presence of >4 x 10' peroxidase-positive cells/nil
13652605, 1994, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2605.1994.tb01231.x by Test, Wiley Online Library on [13/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Human semen component and IVF 129
semen, but the differences were not statistically significant.
In the presence of >6 x 10' peroxidase-positive cells/ml
semen, only four out of 17 oocytes in three patients were
fertilized and no pregnancy occurred.
The effect of the ratio of peroxidase-positive cells to the
number of spermatozoa in semen was also analysed with
regard to the outcome of IVF/ET (Fig. 2). In the presence
of >8% of peroxidase-positive cells the fertilization rate was
reduced significantly (p<O.Ol), whereas the pregnancy rate
was not significantly lower.
Concentration of peroxidase-negative cells in semen relat-
ed to the outcome o f I V F / E T
The concentration of peroxidase-negative cells usually
exceeded the concentration of peroxidase-positive cells and
vaned between 0 and 68 x 10h/ml) (up to 59.8%)of the cells
in semen). The effect of the number of peroxidase-negative
cells on the outcome of IVF/ET is presented in Fig. 3. The
number of oocytes fertilized in vitro was not affected by the
concentration of peroxidase-negative cells in semen. In the
presence of high numbers of round cells in semen (>6 x
10h/ml) the pregnancy rate after IVF/ET was higher,
although the difference did not reach statistical significance.
Correlation between the concentration ofperoxidase-posi-
tive and -negative cell numbers with other semen vari-
ables
The concentration of peroxidase-positive cells was not
correlated with any semen sperm parameter (Table 1). In
contrast, the concentration of peroxidase-negative cells cor-
related significantly with the concentration of spermatozoa
and with the percentage of spermatozoa with normal mor-
phology (p<0.01). Multiple regression analysis of the con-
centration of peroxidase-negative cells with the various
morphological abnormalities of sperm head, midpiece and
tail revealed an inverse relationship only between the num-
ber of peroxidase-negative cells and spermatozoa with an
'amorphous head' (r=-0.017, ~ - 4 . 0 5 ,and r=-0.015,
p<0.05 respectively, not shown in Table 1).
Evaluation of the presence of microorganisms in semen
and o f previous antibiotic treatment on the outcome of
IVF/ET
The prevalence of the most common bacterial isolates in
the semen of asymptomatic patients treated with IVF/ET is
listed in Table 2. Five weeks before gamete recovery for
IVF, a semen sample was obtained to check for bacterial
contamination in 281 cases (Table 3). In 120 cases (42.7%)
n o contamination was found, whereas in 99 cases (35.2%)
<lo00 bacteria/ml were found, and this was judged to need
no specific treatment. In 62 cases (22.1%) definite contami-
nation (>lo00 bacteria/nd semen) was diagnosed and treat-
ment with antibiotics was started. The antibiotics used were
erythromycin in 31 cases (SO%), vibramycin in 15 cases
(24.2%), cotrimaxazol in nine cases (14.5%))and penicillin in
 13652605, 1994, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2605.1994.tb01231.x by Test, Wiley Online Library on [13/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
130 Ch. De Geyter et al.

32
66/103
33/104
-

_.
24

-
19/44 - w
3
2 Figure 1. Relationship between the concentra-
3 16
CI tion of peraxidasepositive cells in semen and
a
the fertilization rate (a) and pregnancy rate (b)
B
L.
H in IVF/ET. In panel a the number of oocytes fer-
tilized divided by the total number of oocytes
0
inseminated i s indicated at the top of each col-
10.- umn. In panel b the number of pregnancies
divided by the number of treatment cycles is

0- - 0 - indicated above each column. The differences

0 >4 <2 2-(4 >4 were not statistically significant (chi-squared
concentration (milI./ml) concentration (mill./ml) analysis).

9/ze
-
2/14

60
76/13?
39/73

-3
w
50

l6
3/15

2 40 0

-
21/63 -
2/13
3." 30
c1

2
2 20

10

0 - 1-<2 -
<1 142 !-<4 4-<0 >8 !-<4 4-<0 >0
relative number of perox.pos. cells ( X ) relative number of perox.pos. cells ( X )
Figure 2. Relationship between the relative concentration (%) of peroxidasepositive cells compared
to the concentration of spermatozoa in semen and the fertilization rate (a) and pregnancy rate (b) in
IVF/ET. In panel a the number of oocytes fertilized divided by the number of oocytes inseminated is
indicated above each column, whereas in panel b the number of pregnancies divided by the number
of treatment cycles i s shown. *Means significantly different (pc0.01,chi-squared analysis).

seven cases (11.3%).In those patients treated with antibiotics
prior to IVF/ET both the fertilization and pregnancy rates
were lower, although the differences were not statistically
significant. However, preceding antibiotic treatment of
asymptomatic infections of the accessory glands was associ-
ated with significantly reduced values for sperm concentra-
tion @<0.05)and progressive motility (p<0.05),but not for
normal sperm morphology nor of the concentration of per-
oxidase-negative round cells. The concentration of peroxi-
dase-positive cells in the semen sample was not influenced
by preceding treatment with antibiotics.
O n the day of gamete recovery for IVF, 91 examinations
 13652605, 1994, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2605.1994.tb01231.x by Test, Wiley Online Library on [13/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Human semen component and IVF 131

17/1M 1o/a

18/74

lB/Bl
- -
17/84

.9/110

I
<2
I
2-<4 4-<6 6-<12 >12
concentration (mill./ml)
<2
1
2-<4 4-<6 6-<12 >12
concentration (mill./ml)
Figure 3. Relationship between the concentration of peroxidasenegative cells in semen and the fer-
tilization rate [a) and pregnancy rate (b) in IVF/ET. In panel a the number of oocytes fertilized divid-
ed by the total number of oocytes inseminated i s indicated at the top of each column. In panel b the
number of pregnancies divided by the number of treatment cycles i s indicated above each column.
The differences were not statistically significant (chi-squared analysis).

Table 1. Relationship between the presence of peroxidaseposi- Table 2. Prevalence of the most common bacterial isolates in the
tive and -negative cells in semen with other semen variables. The semen of asymptomatic patients treated with IVF/ET
correlation (r) between the various sperm properties with each
~~

type of round cells was assessed by simple regression analysis Microorganism Number of Percentage
after logarithmic or arcsine transformation of the data positive cultures of patients

Semen Peroxidasepositive Peroxidase-negative Enterococcus 66 20.2

variable cells cells Staphylococcus epidermidis 57 18.4
Streptococcus viridans 51 16.3
Abstinence -0.090 0.002
Staphylococcus coagulasenegative 47 14.2
Volume 0.079 0.01 1
Ureaplasma urealyticum 29 9.3
PH 0.049 -0.075
Escherichia coli 15 1.5
Sperm concentration -0.067 0.290* *
Diphtheroid 13 4.2
Total sperm -0.057 0.244**
Proteus mirabilis 7 2.1
Progressive motility 0.045 -0.022
Mycoplasma hominis . 7 2.1
Normal morphology 0.072 0.147*
~~~~~~~~~~~~~~~~~~
Streptococcus group 6 6 1.8
* p<O.Ol, * * p ~0.001(all other regressions were non-significant/ Enterobacter 5 1.5

for bacterial contamination were performed, either because
of the presence of > I x 106/ml peroxidase-positive cells or
because of the presence of >8 x 106/ml peroxidase-negative
cells. In 37 patients (40.7%) no bacterial contamination was
found, whereas in 54 cases (59.3%) one or more microor-
ganisms were detected. Seven pregnancies occurred after
IVF/ET despite bacterial contamination (18.9% of all exani-
ined cases), whereas 14 pregnancies occurred in patients
with no bacterial contamination (25.9%). The difference
was not statistically significant.
 13652605, 1994, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2605.1994.tb01231.x by Test, Wiley Online Library on [13/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
132 Ch. De Geyter et a/.

Table 3. Influence of preceding antibiotic treatment on the outcome of IVF/ET. Group A included patients in whom no microorganisms
were found in the semen 5 weeks before IVF/ET. In group B <lo00 microorgonisms/ml semen were found, needing no antibiotic treat-
ment. Group C included patients in whom >lo00 microorganisms/ml semen led to treatment with antibiotics prior to IVF/ET

Patient group
Difference
between
Parameter A 6 C groups

Number of cycles 120 99 62

Number of oocytes inseminated 579 468 308
Number of oocytes fertilized 368 301 168
Fertilization rate 66.7% 64.3% 54.5% NS"
Number of pregnancies 31 (25.8%) 24 (24.2%) 7 (1 1.3%) NS"
Number of miscarriages 1 1 (35.5%) 7 (29.2%) 2 (28.6%) NS"
Sperm concentration (1 06/mI)* 83.1 (68.9-97.4) 91.9(78.7-1
05.1) 64.4(49.9-78.8) ~<0.05~
Progressive motility (%)* 55.5 (53.3-57.7) 54.3(51.6-57.0) 49.1 (45.2-53.1) p<0.05
Normal sperm morphology (%)* 45.2 (42.8-47.6) 44.2 (41.3-47.1) 39.3(35.3-43.3) NSb
Concentration of peroxidase
negative cells (1o6/mI)* 0.52(0.33-0.71
) 0.53(0.25-0.80) 0.51 (0.25-0.77) NSb
Concentration of peroxidase-
positive cells (1 06/mI)* 5.36 (4.31-6.41) 5.95(4.44-7.46) 4.75(3.63-5.87) NSb

*Data expressed as average (with 5 9 5 % confidence interval).

:Chiquared analysis.
Kruskall-Wallis analysis .by ranks.

Discussion enger superoxide disniutase (Alvarez et a/., 1987), other less

In agreement with earlier studies (Cohen et al., 1985;
Talbert et al., 1987), the presence of peroxidase-positive
cells had a negative impact on both the fertilization and the
pregnancy rate in IVF/ET. However, the outcome of
IVF/ET was affected only in the presence of exceedingly
high concentrations of peroxidase-positive cells, or in oligo-
zoospemiic samples containing few sperniatozoa in the pres-
ence of high numbers of leucocytes (Maruyama et a/., 1985;
Eggert-Kmse et al., 1992). In contrast to some studies (Van
der Ven et a!., 1987; Wolff et GI., 1990; Eggert-Kruse et a/.,
1992), but in agreement with others (El-Deniiry et a/., 1986;
Schobel et a/., 1989), the presence of peroxidase-positive
cells in semen was not related to other semen parameters.
Seminal leucocytes niay interfere with sperni function
through several mechanisms. They may damage spermato-
zoa by liberating cytokines (Hill ct al., 1987) o r by nianu-
facturing free oxygen radicals (Aitken & West, 1990;
Kovalski et al., 1992). Through liberation of reactive oxygen
radicals, leucocytes may damage spermatozoa, as evidenced
by reduction in niotility in the presence of leucocyte con-
centrations as low as 0.6 X 106/nd (Kovalski e t a / . , 1992). O n
the one hand, the potential toxicity of this system is reduced
by various semen properties such as the oxygen radical scav-
defined factors in seminal plasma (Kovalski et a/., 1992), and
specific properties of semen leucocytes themselves (D'Cruz
et a]., 1992). O n the other hand, the concentration of leu-
cocytes after sperni preparation is reduced greatly and is well
below the toxic range. Spermatozoa may be damaged or
destroyed by phagocytic activity of neutrophils and
macrophages in seminal plasma (Vogelpoel & Verhoef,
1985). T h e biological significance of this phagocytosis of
spemiatozoa by leucocytes in senien is unclear (Barratt ct nl.,
1990). Phagocytosis of sperniatozoa by leucocytes in semen
niay be directed specifically at morphologically abnormal
spermatozoa (Tonilinson et a/., 1992a) and therefore
improve spemi function.
The present data illustrate that sperni function is impaired
only in the presence of substantial numbers of leucocytes in
semen. Although the peroxidase-staining test is a reliable
and feasible method to screen semen for the presence of leu-
cocytes (Wolff et al., 1992; Politch ct a/., 1993), its specifici-
ty for granulocytes only causes some underestimation of the
real number of white blood cells in semen (Politch ct a/.,
1993). However, even if a niore complete identification of
the leucocytes in senien can be achieved with the aid of
monoclonal antibodies, the present results demonstrate

clearly that limited numbers of leucocytes in semen are not
associated with a detectable reduction in sperm function.
Peroxidase-positive cells in semen are associated with geni-
tal infections (Berger et al., 1982) and bacteria in seminal
plasma may impair sperm function (Riedel ei al., 1983).
However, in agreement with other studies (Stoval ef al.,
1993), positive bacterial cultures in semen did not signifi-
cantly depress the outcome of in-vitro fertilization.
All possibly damaging effects of seniinal leucocytes and
microorganisms involve close contact between leucocytes
and the spermatozoa and act over short distances. Lowering
the concentration of peroxidase-positive cells and pathogen-
ic microorganisms in semen by antibiotic treatment of the
husband before in-vitro fertilization may improve the preg-
nancy rates (Wolff et al., 1990; Maruyama et a/., 1985).
Therefore, 5 weeks before IVF/ET many husbands were
screened for the presence of microorganisms in their semen.
However, in the present analysis both the fertilization and
pregnancy rates, together with the concentration and pro-
gressive motility of spermatozoa in semen, were lower after
antibiotic treatment immediately before IVF/ET, whereas
in the presence of microorganisms in semen the pregnancy
rate was only affected marginally. The present data therefore
suggest that any treatment with antibiotics shortly before
assisted fertilization does not improve the chances of preg-
nancy. Various antibiotics have been claimed to interfere
with male fertility (Schlegel et al., 1991).
The majority of the non-spermatozoa1 cells in semen are
peroxidase-negative cells: Some of these may be lympho-
cytes or epithelial cells, but the majority of them are imnia-
ture germ cells (Eggert-Kruse et al., 1992). Their function in
References
Aitken, R. J. & West, K. M. (1990) Analysis of the relationship
between reactive oxygen species production and leukocyte
infiltration in fractions of human semen separated on Percoll
gradients. I~iferriatiorialjoirrrialq f f A wdroloyy, 13, 433-451.
Alvarez, J. G., Touchstone, J. C, Blasco, L. & Storey, T. S. (1987)
Spontaneous lipid peroxidation and production of hydrogen
peroxide and superoxide in hunian spermatozoa. Superoxide
dismutase as major enzyme protectant against oxygen toxicity.
Jottrrrai cf A ridrolqy, 8, 338-348.
Barratt, C. L. F., Bolton, A. E. & Cooke, I . D. (1990) Functional
significance of white blood cells in the male and fenmale repro-
ductive tract. Hirniari Rcprodrrtfion, 5 , 639-648.
Berger, R. E., Karp, L. E., Williamson, R.A., Koehler, J., Moore,
D. E. & Holnies, K. K. (1982). The relationship of pyosperniia
and seniinal fluid bacteriology to sperm function as reflected in
the sperm penetration assay. Fertility arid S t e d i f y , 37, 557-564.
Cohen, J., Edwards, R., Fehilly, C.. Fishel, S., Hewitt, J., Purdy,
J., Rowland, G., Steptoe, P. & Webster, J. (1985) In vitro fer-
tilization: a treatment for male infertility. Fertility atrd Sterility,
43, 422-432.
Conihaire, F., Verschraegen, G. & Vernieulen, L. (1980) Diagnosis
of accessory gland infection and its possible role in male infer-
gf Aridroloxy, 3, 32-45.
tility. Zrrterriatiotin/~~~irnra/
13652605, 1994, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2605.1994.tb01231.x by Test, Wiley Online Library on [13/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Human semen component and IVF 133
fertility is unclear. Because they are considered clinically
harmful, it is widespread practice to treat patients with high
numbers of peroxidase-negative cells with antibiotics or
anti-inflammatory drugs. In contrast to common practice,
and in disagreement with on previous communication,
(Tomlinson et a/., 1992b), the pregnancy rates in the present
study were not decreased by the presence of many peroxi-
dase-negative cells in semen. As the concentration of perox-
idase-negative cells correlated significantly with normal
sperm morphology rates, with sperm concentration, and
with total sperm number, the presence of many peroxidase-
negative cells in the senien may reflect the degree of testic-
ular spermatogenesis, rather than epididymal function. More
specific methods for the identification of ininiature germ
cells (for example the method suggested by Honiyk et a/.,
1990) may be introduced into semen analysis to examine the
relationship between the presence of immature germ cells
and testicular spermatogenic activity. In view of the harmful
effect of antibiotics on male fertility suggested by the present
data, high numbers of peroxidase-negative cells in semen
should clearly not be treated with antibiotics.
Acknowledgments
W e thank Heidi Beering, Susanne Willms, Sigrid Haring,
Cordula Stetskamp, Jutta Salzig, Gisela Schlingniann and
Martina Niemeyer for technical assistance, and Susan
Nieschlag MA for language editing. This study has been
supported by the Deutsche Forschungsgenieinschaft (Ni
130/11).
Cooper, T. G., Jockenhovel, F. & Nieschlag, E. (1991) Variations
in semen parameters from fathers. Hun~arr Reprodicrfiotr, 6,
859-866.
Cooper, T. G., Neuwinger, J.. Bahrs, S . & Nieschlag, E. (1992)
Internal quality control of senien analysis. Fertility arrd Sferility,
58, 172-178.
D’Cruz, 0.J.. Wang, B. L. & Haas, G. G. (1992) Phagocytosis of
immunoglobulin G and C3-bound human sperm by hutnan
polymorphonuclear leukocytes is not associated with the release
of oxidative radicals. Biok,gy Pf Repmditrtiorr, 46, 721-732.
De Geyter, Ch., Bals-Pratsch, B., Doeren, M., Ycung, C. H.,
Grunert, J. H., Bordt, J., Schneider, H. P. G. & Nieschlag, E.
(1988) Human and bovine cervical mucus penetration ac a test
of sperm function for in-vitro fertilization. Httrtrarr Rrprodirttiori,
3,948-954.
De Geyter, Ch., De Geyter, M., Schneider, H. P. G. & Nieschlag,
E. (1992) Subnormal sperm paranieters in conventional setnen
analysis cause discrepancy between fertilization and pregnancy
rates in in vitro fertilization and embryo transfer. Irifcrtrafiotinl
Joirmal qf Aridrchgy, 15, 485-497.
De Geyter, Ch., De Geyter, M., Castro, E., Bals-Pratsch, M.,
Hanker, J. P., Schlegel, W., Nieschlag, E. & Schneider, H. 1’.
G. (1 993) Predictive paranieters for ovarian response to hyper-

134 Ch. De Geyter ef a/.
stimulation with exogenous gonadotropins after suppression of
gonadotropin secretion of the pituitary using a long-acting
GnRH agonist. European journal .f Obstetrics, Gynaecology and
Reproductive Biology, 51, 139-147.
Eggert-Kruse, W., Bellmann, A,, Rohr, G., Tilgen, W. &
Runnebaum, B. (1992) Differentiation of round cells in semen
by means of monoclonal antibodies and relationship with male
fertility. Fertility and Sterility, 58, 1046-1055.
El-Demiry, M. I. M., Young, H., Elton, R. A,, Hargreave, T. B.,
James, K. & Chisholm, G. D. (1986) Leukocytes in the ejacu-
late from fertile and infertile men. British journal of Urology, 58,
7 15-720.
Gonzales, G. F., Kortebani, G. & Mazzolli, A. B. (1992)
Leukocytospermia and function of the seminal vesicles on sem-
inal quality. Fertility and Sterility, 57, 1058-1065.
Hill, J. A., Haimovici, F., Politch, J. A. &Anderson, D. J. (1987)
Effects of soluble products of activated lymphocytes and
macrophages (lymphokines and monohnes) on human sperm
motion parameters. Fertility atid Sterility, 47, 460-465.
Homyk, M., Anderson, D. J., Wolf€, H. & Herr, J. C. (1990)
Differential diagnosis of immature germ cells in semen utilizing
monoclonal antibody MHS-10 to the intra-acrosomal antigen
SP-10. Fertility and Sterility, 53, 323-330.
Kovalski, N. N., de Lamirande, E. & Gagnon, C. (1992) Reactive
oxygen species generated by human neutrophils inhibit sperm
motility: protective effect of seminal plasma and scavengers.
Fertility and Sterility, 58, 809-816.
Maruyama Jr., D. K., Hale, R. W. & Rogers,'B. J. (1985) Effects
of white blood cells on the in vitro penetration of zona-free
hamster eggs by human spermatozoa. journal of Andrology, 6 ,
127-135.
Nahoum, C. R. D. & Cardozo, D. (1980) Staining for volumetric
count of leukocytes' in semen and prostate-vesicular fluid.
Fertility and Sterility, 34, 68-69.
Politch, J. A,, Wol& H., H a , J. A. &Anderson, D. J. (1993)
Comparison of methods to enumerate white blood cells in
semen. Fertility and Sterility, 60, 372-375.
Riedel, H. H., Baukloh, V. & Mettler, L. (1983) Die Bedeutung
der Sperma-Qualitat fur dle in vitro-Fertilisation - Ergebnisse
von Untersuchungen im Humansystem und im Zona pelluci-
da-freien Hamstereizellsystem.Andrologia, 15, 595-604.
Schlegel, P. N., Chang, T. S. K. & Marshall, F. F. (1991)
Antibiotics: potential hazards to male fertility. Fertility and
Sterility, 55, 235-242.
Schobel, W. A,, Schieferstein, G . & Uchanska-Ziegler, B. (1989)
Inimunocytochemical characterization of round cells in human
semen using monoclonal antibodies and the APAAP-tech-
13652605, 1994, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2605.1994.tb01231.x by Test, Wiley Online Library on [13/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
nique. Andrologia, 21, 370-376.
Stovall, D. W., Bailey, L. E. & Talbert, L. M. (1993) The role of
aerobic and anaerobic semen cultures in asymptomatic couples
undergoing in vitro fertilization: effects on fertilization and
pregnancy rates. Fertility and Sterifity, 59, 197-201.
Talbert, L. M., Hammond, M. G., O'Rand, M., Fryer, J. G. &
Ekstrom, R. D. (1987) Semen parameters and fertilization of
human oocytes in vitro: a multivariable analysis. Fertility and
Sterility, 48, 270-762.
Tomlinson, M. J., White, A., Barratt, C. L. R., Bolton, A. E. &
I Cooke, I. D. (1992a) The removal of morphologcally abnor-
mal sperm forms by phagocytes: a positive role for seminal
leukocytes? Human Reproduction, 7, 517-522.
Tomlinson, M. J., Barratt, C. L. R., Bolton, A. E., Lenton, E. A.,
Roberts, H. B. & Cooke, I. D. (1992b) Round cells and spemi
fertilizing capacity: the presence of immature germ cells but not
seminal leukocytes are associated with reduced success of in
vitro fertilization. Fertility and Sterility, 58, 1257-1 259.
Van der Ven, H. H., Jeyendran, R. S., Perez-Pelaez, M., Al-
Hasani, S., Diedrich, K. & Krebs. D. (1987) Leukospermia and
the fertilizing capacity of spermatozoa. European Journal of
Obstetrics, Gynecology and Reproductive Biology, 24, 49-52.
Vogelpoel, F. R. & Verhoef, J. (1985) Activation of polynior-
phonuclear leukocytes by spermatozoa. Archives qf Andrology,
14, 123-131.
Vogelpoel, F. R., van Kooij, R. J., te Velde, E. R. & Verhoef, J.
(1991) Influence of polymorphonuclear granulocytes on the
zona-free hamster oocyte assay. Human Reproduction, 6 ,
11 0 4 1107.
Wolff, H. & Anderson, D. J. (1988) Immunohistologic characteri-
zation and quantitation of leukocyte subpopulations in human
semen. Fertility and Sterility, 49, 497-504.
Wolf€, H., Politch, J. A., Martinez, A,, Haimovici, F., Hill, J. A. &
Anderson, D. J. (1990) Leukocytospermia is associated with
poor semen quality. Fertility and Sterility, 53, 528-536.
Wolff, H., Panhans, A., Zebhauser, M. & Meurer, M. (1992)
Comparison of three methods to detect white blood cells in
semen: leukocyte esterase dlpstick test, granulocyte elastase
enzymeimmunoassay, and peroxidase cytochemistry. Fertility
and Sterility, 58, 1260-1262.
World Health Organization (1987) Laboratory Manual .fir the
Examination of Human Semen and Semen-Cervical Mucus
Interaction, 2nd ed. Cambridge University Press, Cambridge,
1987.
Received 2 3 J u l y 1993; accepted 5 January 1994