Document 139916

Conventional treatments of male infertility in the age of
evidence-based andrology
A.Kamischke and E.Nieschlag1
Institute of Reproductive Medicine of the University (WHO Collaborating Centre for Research in Human
Reproduction), Domagkstrasse 11, D-48129 Minister, Germany
lr
To whom correspondence should be addressed
Introduction
The past 20 years have witnessed rapid growth in
health care information. Every month, clinicians,
policy makers and patients face a deluge of evidence published in 20 000 journals catering to the
health care community. These reports are now so
numerous and dispersed that it is unreasonable to
expect people to read and retain all the information
they contain. Because of this mass of information,
most health care providers rely on reviews, which
claim to provide an overview of results of primary
research. Unfortunately, the quality of at least some
reviews leaves much to be desired (AntczakBouckoms, 1995).
62
The most important factor influencing the ability
of reviews and practised medicine to enhance
personal experiences and the readers' own views
is the quality of the primary basic and clinical
studies reviewed. Although many of the studies
were conducted with the best of intentions, some
did more harm than good. In experimental designs,
randomized, controlled prospective clinical trials
are regarded as the most reliable method of proving
the effectiveness of diagnostic and therapeutic
strategies. Too often, however, the results of highquality experimental trials and high-quality, observational, quantitative epidemiological studies are
ignored. There are several reasons for this, and
important among them are the simple lack of
awareness of trial results and the confusion which
arises when several trials of the same topic provide
inconclusive or conflicting results.
These problems are common to all clinical
disciplines. Andrology, and the question of male
infertility treatment in particular, present further
problems. One reason is that the pathogenesis
of many fertility disturbances has not yet been
elucidated and therefore rational approaches to
treatment are lacking. In ~30% of cases no obvious
cause for abnormal seminal parameters can be
found and the condition is classified as 'idiopathic
infertility' (Nieschlag and Behre, 1997). Another
large proportion of infertile men present with
varicoceles, leukocytes in seminal plasma or sperm
antibodies. Whether such findings are, however,
causative of impaired fertility or just coincidental
has not yet been convincingly proven. The second
problem of conventional treatment of male infertility is the definition of the endpoints of therapy.
Although the induction of pregnancy and the birth
© European Society for Human Reproduction & Embryology
Human Reproduction Volume 13 Supplement 1 1998
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
The field of reproductive medicine is characterized by the rapid development of assisted fertilization. Today, fertilization can be achieved by
cytoplasmic microinjection of single spermatozoa or spermatids into oocyte cytoplasm. In
view of these new reproductive techniques, conventional diagnosis and treatment of male infertility seem to be rather outdated and limited.
However, the final goal of reproductive medicine
should be to improve reproductive function in
order to provide each couple with the chance
of conceiving offspring naturally. In this paper
we analyse the effectiveness of current conventional treatments of male infertility in the light of
high-quality studies and emphasize that future
therapies must be evaluated in properly conducted studies.
Key words: andrology/clinical trials/hormonal
treatment/male infertility
Conventional treatments of male infertility
of a child are clearly defined endpoints, they
require the interaction of two individuals. Therefore, factors outside the andrological realm influence the success of andrological therapy and have
to be considered and evaluated in couple diagnosis
and treatment.
Basis of good clinical trials
The basis of good clinical experimental trials is
summarized in Table I.
Diagnostic methods
Applied diagnostic procedures should be established in an appropriate spectrum of patients who
are subjected to an independent, blind comparison
of diagnostic tests, thus avoiding any kind of bias
or systematic error (Jaschke et al, 1994). High
sensibility and specificity of the. applied methods
provide the basis for good laboratory practice,
which is concerned with the organizational process
and the conditions under which laboratory studies
are planned, performed, monitored, recorded and
reported [Organisation for Economic Co-Operation
and Development (OECD), 1992].
Proper andrological diagnosis
Clinical problem solving must rely on an understanding of the underlying pathophysiology.
Clinical experience and the development of clinical
instincts (particularly with respect to diagnosis)
are crucial and necessary skills of a competent
physician. Together with careful history taking and
physical examination, these tools provide much,
and often the best, evidence for diagnosis and
direct treatment decisions. Therefore, the more the
experienced clinicians can dissect the processes
used in diagnosis, the greater the benefit for the
patients and clinical trials (Campbell, 1987; Evidence-Based Medicine Working Group, 1992).
In addition to semen analysis, a comprehensive
history and physical examination with particular
emphasis on detection of any genital abnormality,
63
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Good clinical trials: implications for evidencebased andrology
No valid laboratory model exists for male infertility, and conventional therapeutic strategies often
rely on speculative concepts and clinical observations. Even when derived from pathophysiological
laboratory findings, therapeutic concepts must be
confirmed or challenged by systematic observations
in clinical trials. Clinical trials should be conducted
in agreement with the rules of good clinical practice, which provide an international ethical and
scientific quality standard for designing, conducting, recording and reporting trials that involve
the participation of human subjects [World Health
Organization (WHO), 1993b; International Conference on Harmonisation (ICH), 1996]. Good clinical
practice rules obligate the medical community to
lead the way to appropriate clinical evaluation of
diagnostic tests and trials of clinical treatment
procedures.
The concept of evidence-based medicine provides this appropriate tool, which de-emphasizes
intuition, unsystematic clinical experience and
pathophysiological rationale as sufficient grounds
for clinical decision making and stresses the examination of evidence from clinical research (Evidence-Based Medicine Working Group, 1992). The
concept of evidence-based medicine is rooted in
five interlinked ideas: (i) Clinical decisions should
be based on the best available scientific evidence,
(ii) The clinical problem (rather than habits or
protocols) should determine the type of evidence
to be sought, (iii) Identifying the best evidence
means using epidemiological and biostatistical
ways of thinking, (iv) Conclusions derived from
identifying and critically appraising evidence are
useful only if put into action in managing patients
or making health care decisions, (v) Performance
should be constantly evaluated (Davidoff et al,
1995). Therefore, literature-based evidence, and
not personal skills and authority born of experience,
is the major determining factor(s) in the outcome
of research-oriented medicine when therapeutic
decisions must be taken.
Compared with other fields of medicine, andrology was especially late in applying the paradigm
of evidence-based medicine in clinical practice
(Olive, 1986). Evidence-based andrology is, notwithstanding the exponential increase in the
number of published randomized, controlled clinical trials concerning infertility treatment over the
past decade, only in its infancy (O'Donovan et al,
1993; Vandekerckhove et al, 1993; Leifke and
Nieschlag, 1996).
A.Kamischke and E.Nieschlag
Table I. Basis of good clinical experimental trials
Criteria which should be fulfilled
Diagnostic methods
High sensibility and high specificity of applied methods
Avoiding bias or systematic error
Valid internal and external quality control
Diagnosis
Clinical experience
Careful history taken and physical examination
Design
Results should provide scientific and clinical relevance (outcome research)
Fulfilling rules of good clinical practice (WHO, 1993a,b)
Prevention of any systematic bias by
Inclusion of control groups
Randomized allocation to the groups
Double blinding and placebo medication
High precision and external validity
Statistical power estimation a priori
such as maldescended testes, absence of the vas
deferens, occlusions, phimosis or varicocele,
should be included in the patient's work-up. Further
evaluations should routinely include ultrasonography of the scrotal organs, as infertile patients
often present various abnormalities. Endocrinological evaluation of reproductive hormones helps to
identify the cause of infertility and is mandatory
in the diagnosis of obstructive azoospermia and
differentiation between hyper- and hypogonadal
hypogonadism (e.g. Klinefelter syndrome versus
Kallmann syndrome; ESHRE Capri Workshop,
1996; Behre et al, 1997).
Because of the subjective nature of much of
andrology, it is difficult, but feasible, to implement
quality guidelines into the sperm assessment
methods, as has already been done for hormones
(WHO, 1993a). A continuous record of technician
performance is necessary and is the purpose of
internal quality control (Cooper et al, 1992). It is
equally obvious that, in addition to internal quality
control, external quality control schemes which
aim to ensure agreement between different centres
are essential if men with the same problems are
not to receive different diagnosis and treatment in
different centres or countries as a result of different
diagnoses (Cooper, 1996). Internal and external
quality control can be achieved by standardizing
the methods of sperm preparation and examination
(WHO, 1992a) and use of equipment to set absolute
values for sperm concentration (flow cytometry)
64
and motility (computer-assisted sperm analysis,
CAS A). For sperm morphology, absolute values
do not yet exist, but internal and external quality
control could be achieved by repeated assessment
of the same morphology slide by different technicians and centres, as a recent enquiry among the
Training Centres for Andrology of the European
Academy of Andrology (EAA) has shown
(Cooper, 1996).
Apart from the minimum requirements of andrological patient work-up, various standardized and
unstandardized diagnostic sperm function tests
have been implemented in andrology without any
diagnostic advance. Therefore, urgent consideration must be given to the development of sperm
function tests that are able to provide a likelihood
of pregnancy for given diagnostic situations within
certain timeframes (ESHRE Andrology Special
Interest Group, 1996). In general, useful diagnostic
fertility tests can be classified into three categories:
(i) abnormal test results that are consistently correlated with impaired fertility. In each case, when the
test result is unequivocally abnormal, fertility is
unarguably impaired without therapy; (ii) abnormal
test results that are not consistently correlated
with impaired fertility. For these diagnostic tests,
abnormal results are frequently associated with
subsequent fertility without therapy; (iii) abnormal
test results that do not appear to be correlated with
impaired fertility. For these diagnostic tests, either
there are data that confirm the lack of a correlation
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Topic
Conventional treatments of male infertility
with pregnancy, or such follow-up studies do not
exist (ESHRE Capri Workshop, 1996).
Good laboratory practice and proper application
of the diagnostic procedures are the tools of good
clinical practice, which is the substantial basis for
good clinical trials and evidence-based medicine/
andrology (WHO, 1993a,b; ICH, 1996).
Providing evidence by experimental treatment
studies
Aim and validity of a study: Precise definition of
the question to be answered forms the basis for
the ensuing clinical trial (WHO, 1993b; ICH,
1996). The patient's disease, its treatment and
outcome as well as the clinical context, e.g. the
existence of ancillary treatment, co-morbidity and
male-independent female factors, should be denned
before patients are assigned to treatment groups
Control groups, randomization and crossover
designs: 'Controlled' studies compare one or more
treatments to one or more control groups. Control
groups are essential to exclude performing bias.
Especially in conditions such as severe oligozoospermia, in which variable semen parameters are
observed, spontaneous changes are likely to be
mistaken for a genuine therapeutic effect. Furthermore, when performing a study, it should be taken
into account that no medical intervention is without
an effect (Kleijnen et al, 1994). In order to
discriminate between
intervention-associated
specific and unspecific effects and avoid biases,
untreated control groups should be included in
clinical studies.
Randomization of treatment reduces selection
bias by ensuring balance between known and
unknown covariables and between comparison
groups. The ideal process of allocation to the
treatment groups is achieved if an assignment
schedule generated using true randomization is
administered by someone who is not responsible
for recruiting participants (Altman and Dore,
1990). After the randomized assignment to the
treatment groups has been revealed, no alteration
of the assignment or decision about eligibility
is possible.
'Crossover' refers to a study design in which
more than one intervention was administered in a
definite order, one following another. As at least
the second treatment is not assignment randomized
to the participant, crossover studies are not truly
randomized. For example, the evaluation of drugs
can be biased by carry-over effects of prolonged
action. This is particularly pertinent when considering the duration of spermatogenesis and the
factors influencing it. In addition, if pregnancy is
the ultimate goal of therapy, responders will leave
65
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Considerations for design of good clinical trials
Providing evidence
Regardless of the aim of the clinical trial, it should
provide evidence of scientific and clinical relevance
(outcome research). The quality of the study and
the resulting evidence for its conclusion depend
on the quality of the diagnostic methods, proper
diagnosis and on the design of the study. The
quality of a trial can be defined as the extent to
which its design and conduct are likely to have
prevented systematic bias. Mainly, there are four
kinds of bias to avoid: (i) selection bias (systematic
differences in the comparison groups), (ii) performance bias (systematic differences in the care provided apart from the intervention being evaluated),
(iii) exclusion bias (systematic differences in withdrawals from the trial) and (iv) detection bias
(systematic differences in outcome assessment).
In addition to the quality of the clinical trial,
there are two other important aspects: precision
(i.e. to what extent the clinical trial avoids the
likelihood of random errors) and external validity
(i.e. to what extent the outcome results are generalizable or applicable). These rules for highquality studies apply to both experimental treatment and observational studies. Some of the general
principles and limits concerning the design of
such studies are discussed below in the section
'Conventional male infertility treatment'.
(inclusion criteria). The broader the clinical context, the more generalizable will be the results
(external validity) of the trial; however, this will
also render it more likely that a possible effect
will be masked or biased by unknown confounding
factors. Homogenization by precise inclusion or
exclusion criteria, rules for drop outs and stratification help to avoid this problem, but will also limit
the validity of the results and narrow the spectrum
of patients or clinical context.
A.Kamischke and E.Nieschlag
the first part of the study and thus bias the second
part by dropping out.
Statistical considerations and multicentre studies
The significance of investigations depends on the
incidence of outcome parameters. Statistical power
estimation (internal validity) prior to the recruitment phase should provide an estimate of the
number of patients required to exclude for a low
probability of false-positive (type I error) as well
as for low false-negative (type II error) errors
(Freiman et al, 1978). The 'minimal important
benefit' defines the smallest relevant difference in
function that a patient and clinician agree to be
worth the risks and trouble of offering and
accepting the treatment (Sackett and Cook, 1993;
WHO, 1993b). Cumulative pregnancy rates on the
basis of life table calculation are the best parameter
for outcome research of male infertility treatment,
although they may not always be a practicable
and therefore reliable endpoint because of low
pregnancy rates among infertile couples, interference by additional female factors and violations
of the protocol by resorting to performance of
assisted reproduction techniques.
Because pregnancies among infertile couples
occur relatively infrequently, the number of patients
that has to be included in a trial on the basis of a
power estimate is usually larger than the number
66
Meta-analysis
Meta-analysis is a method of summarizing the
results of a number of randomized trials. This
statistical approach is no doubt useful if applied
correctly. Meta-analyses could provide evidence if
the results of several experiments are contradictory
or inconclusive. In addition, even if the results of
single high-quality studies are consistently nonsignificant, a meta-analysis might show an overall
significant effect. A meta-analysis is only as good
as the included trials (Whitehead and Jones, 1994).
Meta-analysis of results from trials of variable
quality can result in false-positive or false-negative
conclusions if the less-rigorous studies that are
included are biased towards overestimating or
underestimating the effectiveness of the intervention being evaluated (Detsky et al, 1992).
Results of meta-analysis should be presented as
odds ratios or relative risks. Odds ratios and relative
risks greater than one (risk differences less than
zero) always indicate that treatment is better. Where
the 95% confidence interval is entirely above or
below 1, the treatment group did significantly (at
the 5% level) better or worse than the placebo
control group (for more details, see Morris and
Gardner, 1988). If confidence intervals for the
results of the included studies overlap, differences
between single studies should be considered statistically insignificant and therefore homogeneous
(Walker et al, 1988). In cases of homogeneity,
combined odd ratios for every treatment can be
calculated (example in Figure 3).
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Blinding and placebo effects: In a 'blinded' study,
the participants and/or the investigators (single/
double blind) evaluating its outcome are deliberately kept unaware of the group to which the
participants have been assigned. The purpose of
blinding is to minimize the bias that might result
from knowing which intervention a participant is
receiving (detection and performance bias; Gotzsche, 1994). Trials which blind outcome assessors
regarding treatment allocation should be rated
more highly than trials which do not include any
blinding. The placebo effect is the difference in
outcome between a placebo-treated group and an
untreated control group in an unbiased experiment
(non-specific effect). In clinical trials, placebo
treatment versus experimental treatment reduces
detection and performance bias and should be
provided in a double blind manner (Kleijnen
etal, 1994).
available in a given centre at a certain period and
the number of patients that can be handled by
the investigators in a reasonable time. Therefore,
multicentre studies provide larger sample sizes and
more clinical judgement than single-centre studies.
Unfortunately, multicentre studies are often complicated by systematic inhomogeneity between the
centres, and they depend on the assumption that
the differences between the compared groups are
constant from centre to centre (performing and
detection bias). Because of the general lack of
internal and external quality control in andrology
laboratories, performing and detection bias are
further common difficulties in multicentre studies.
Conventional treatments of male infertility
Interim conclusion
Because of the above-mentioned considerations,
the statistically valid, (placebo) controlled, randomized, double-blind clinical trial is today
regarded as the most reliable method of proving
the effectiveness of diagnostic and therapeutic
strategies. This kind of study provides an excellent
tool to show whether as yet unproven therapies
are superior to no or existing treatments and
are therefore recommended by drug and health
authorities to evaluate the effectiveness of new
drugs.
Conventional male infertility treatment
In this section, examples of therapies for male
infertility are discussed in view of high-quality,
properly designed, truly randomized, placebo-controlled, at least single-blind studies (with the exception of varicocele) which have pregnancy as the
main outcome measure. Crossover studies were
only included in this analysis if pregnancy rates
could be separately calculated from the overall
outcome for the first treatment phase. In areas
where no high-quality studies could be identified
or strong observational evidence existed, treatment
options were discussed on the basis of the trials
available. The studies were identified by computerized library searches (Medline and Cochrane) under
the keywords comprising the special issues listed
in the following subheadings. In addition, it is
strongly recommended that optimization of the
female reproductive functions, adequate deposition
of the ejaculate and appropriate timing of sexual
intercourse in relation to ovulation are the basis of
all male infertility treatments (Wilcox et ah, 1995).
Secondary hypogonadism
Testosterone serum concentrations, testicular
growth and appearance of spermatozoa in the
ejaculate are clearly definable outcome parameters
in secondary hypogonadism, where specific
supplementation of what the endocrine system
endogenously lacks will lead to dramatic reconstitution of male reproductive functions. Therefore,
observational studies were adequate to demonstrate
the effectiveness of gonadotrophin-releasing hormone (GnRH) or gonadotrophin therapy (Black,
1996), and evidence for their effectiveness has
been provided.
To date, however, it has not been possible
to show conclusively whether human chorionic
gonadotrophin/human menopausal gonadotrophin
(HCG/HMG),
-recombinant
gonadotrophins
(Kliesch et ah, 1995) or GnRH are superior in
achieving paternity in patients with idiopathic
hypogonadotrophic hypogonadism (IHH; Schopohl
et ah, 1991; Kliesch et ah, 1994). This would
67
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Providing evidence by observational studies
Observational designs are not in contrast to experimental studies or evidence-based andrology, but
they have to be seen as complementary approaches.
The principal observational epidemiological
methods are non-randomized trials, cohort studies
(prospective and retrospective) and case control
studies (Black, 1996). Whether an experimental or
observational approach to the problem is preferable
depends mainly on the question to be answered.
Observational studies are adequate to demonstrate effectiveness when the intervention is
dramatic and unknown confounding factors can be
ignored, as in hormone substitution therapy in
diabetes mellitus, hypothyroidism or hypogonadotrophic hypogonadism. Observational studies
might also be preferable in evaluating interventions
designed to prevent very rare events or in detecting
rare adverse events as well as when the outcome
of the intervention is far in the future (Black,
1996). For example, because of limitations in study
size, the influence of therapy for maldescended
testes in early infancy on subsequent development
of testicular cancer and infertility might only be
detectable by observational studies. Observational
studies might sometimes also be more appropriate
than experimental studies if the patient's beliefs
and wishes and the clinician's attitudes, experiences
and beliefs are crucial to determining the success
of the intervention. A special problem with the
placebo effect arises in experimental surgery
studies, because placebo (sham) operations (i.e.
for the treatment of varicocele) are unethical in
humans (Johnson, 1994). Observational studies
may sometimes provide a rationale for performing
randomized controlled trials (Hennekens and
Buring, 1993).
A.Kamischke and E.Nieschlag
require controlled randomized trials comparing the
therapeutic entities. As the number of patients
with IHH is small and its clinical appearance is
heterogeneous, studies providing sufficient statistical power are almost impossible to perform in a
single institution; thus treatment of IHH would be
appropriate for prospective, randomized, controlled
multicentre trials.
Varicocele
Varicoceles are the most frequent physical finding
in infertile men (WHO, 1992c). Based on these
observational data and the speculative concept that
varicoceles may cause testicular and epididymal
damage while elevating intrascrotal temperature,
occlusion of the spermatic vein by ligation or
embolization has long been accepted as the treatment of choice (Takihara et al, 1991). Most studies
have assessed treatment success by the physical
disappearance of the Valsava-positive varicocele
68
Figure 1. Cumulative pregnancy rates over 12 months in 48
untreated patients (solid line) and 47 treated patients (dotted
line) with varicocele (Nieschlag et al, 1995).
and the purported improvement of male fertility
has hardly been assessed by controlled clinical
trials aiming to appraise pregnancy rates. In addition, in most studies, female factors were not
clearly defined by inclusion or exclusion criteria.
Recent studies that corrected for these possible
biases have suggested significant benefits for male
fertility after surgical varicocele repair compared
with unoperated controls (Madgar et al, 1994;
Hargreave, 1995). In these latter studies, the control
groups remained completely untreated, not even
counselled, and thus the existence of unspecific
(placebo) effects associated with any medical intervention was ignored. In contrast to these studies,
a proper randomized, controlled, prospective trial
recently showed no significant difference in the
cumulative pregnancy rate over a 12-month period
between the control and treatment groups when
the control group was counselled throughout the
follow-up time (Figure 1; Nieschlag et al, 1995).
This study again emphasizes the importance of
properly designed control groups and the significance of bias. In order to evaluate the magnitude
of counselling effects, a design with three groups,
i.e. untreated, counselled and surgically or radiologically treated varicoceles, would be preferable,
but a sham-operated group cannot be included
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Testicular maldescent
A history of maldescended testes is frequent among
infertile patients, although only a few cases
attending an infertility clinic present with testes
that are still in an abnormal position (Nieschlag
et al, 1997). For prevention of infertility and to
reduce the risk of testicular cancer, it was usual in
the past to treat maldescended testes in advanced
childhood or in puberty. In the course of the past
20 years it has become accepted that the position
of the testes should be corrected by the end of the
first year of life. There is evidence that descent can
be achieved in uncomplicated cases by intranasal
GnRH or intramuscular HCG therapy, although
truly randomized controlled trials have not been
performed. If endocrine therapy fails twice, an
orchidopexy should be performed [Deutsche
Gesellschaft fur Endokrinologie (DGE), 1991].
Whether treatment of maldescent earlier in childhood, as is practised today, is superior to later
treatment is unclear, as the patients treated earlier
are only beginning their reproductive phase of life
and high-quality studies on this topic are lacking
(Nieschlag, 1993). For treatment of maldescentrelated infertility in general, the same treatment
options as in idiopathic male infertility have to be
considered.
Conventional treatments of male infertility
for ethical reasons. Since it remains questionable
whether interventive treatment is superior to no
treatment or to counselling in terms of fertility,
intervention can no longer be recommended.
Immunological male infertility
The concept of immunological infertility is not yet
well established. The term refers mainly to the
detection of sperm antibodies in the seminal fluid
of men whose infertility is otherwise unexplained.
Findings based mainly on laboratory tests support
the relevance for male fertility of the presence of
immunoglobulin (Ig)G and IgA antibodies on the
sperm surface (Eggert-Kruse et al, 1991;
Mahmoud et al, 1996). A prospective and controlled study demonstrated significantly lower
sperm motility and pregnancy rates when antisperm antibodies of the IgG, and particularly of
the IgA, class were present in semen and yielded
a positive mixed agglutination reaction (MAR) test
>30% (Eggert-Kruse et al, 1995), while other
studies showed no effects of anti-sperm antibodies
on fertility (Collins et al, 1993).
Immunosuppressive agents such as gluco-
Idiopathic male infertility
One third of our patients, representing the largest
group of those consulting for infertility, suffer from
'idiopathic infertility', which is characterized by
subnormal semen parameters and possibly also by
pathological hormone concentrations whose cause
is unknown (Nieschlag and Behre, 1997). Despite
the absence of clear pathophysiological concepts,
several treatments have been recommended.
Non-hormonal therapies
Kallikrein is a glycoprotein involved in the enzymatic activation of kininogens that has been shown
to have a positive effect on sperm motility in vitro
and on stimulation of sperm metabolism and to
improve cervical mucus penetration (Schill, 1978).
Kallikrein administered orally has been widely
used in andrological practices for idiopathic male
infertility. Sixteen randomized trials have been
performed with this drug, but only four were truly
randomized and included a placebo-treated control
group (Bedford and Elstein, 1989; Izzo et al,
1984; Glezerman et al, 1993; Keck et al, 1994).
Two single studies have shown a significant positive effect of kallikrein on pregnancy rates (Schill,
1978; Micic et al, 1985). Unfortunately, both were
methodologically impaired. The first (Schill, 1978)
was only pseudorandomized and recorded pregnancies retrospectively over a post-treatment period
69
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Infections
Prevention of infections of the reproductive organs
plays a more important role than treatment. Before
antibiotics were introduced, occlusions of the efferent ducts as sequelae of venereal diseases were the
most common reasons for male infertility and
still prevail in some areas in sub-Saharan Africa
(Nieschlag, 1993). The aetiology and evidence of
effective antibiotic treatments in leukocytospermia
and subclinical genital tract infections remain
unclear. Randomized controlled studies have
shown a high resolution rate in the placebo group.
These studies result in the recommendation that
patients with a single positive leukocytospermia
test should not to be given antibiotics without
further evidence of a specific genital tract bacterial
infection (Comhaire et al, 1986; Yanushpolsky
et al, 1995). Because of the low number of
good quality studies, the non-efficacy of antibiotic
treatment of subclinical genital infections should
be investigated further in truly randomized trials
that have pregnancy or at least a convincing
increase in sperm quality as the outcome.
corticoids have been tried as treatment on the
grounds of immunological considerations. Most
studies concerning this issue were uncontrolled or
not truly randomized, and only a few investigated
pregnancy as an outcome (O'Donovan et al, 1993;
Howards, 1995). Qualitatively adequate studies of
corticosteroid treatment (Figure 3) have claimed
both efficacy (Haas and Manganiello, 1987) and
lack of efficacy (Bals-Pratsch et al, 1992).
As the clinical results are at least conflicting
and it is still unclear whether and to what extent
anti-sperm antibodies are causative for fertility
problems in men, well-designed clinical studies
should readdress this issue. Such studies require
that laboratory findings concerning special subgroups of antibodies or their titre within seminal
plasma should be better defined and that randomized treatments should be performed after a certain
class or titre of antibodies is found in semen.
A.Kamisehke and E.Nieschlag
10
8
6
.D
HCG/HMG:
placebo:
4
[106]
2
0
-2
-4
Figure 2. Difference between the means of pretreatment and post-treatment results for total number of motile, normally
formed spermatozoa (volume + concentration + percentage of motile percentage of normally formed spermatozoa) in two
treatment groups. Individual values of patients (placebo, n = 20; human chorionic gonadotrophin (HCG)/human menopausal
gonadotrophin (HMG), n = 17) are matched between groups according to the magnitude of the differences. Asterisks indicate
lack of appropriate counterpart due to differences in group sizes (Knuth et al., 1987).
of 12 months, which is too long for pregnancies
to be ascribed only to kallikrein, which has a very
short half-life. The second study (Micic et al,
1985) was not double blinded. The higher quality
studies (Figure 3) in contrast could not demonstrate
any benefits of kallikrein on sperm parameters or
pregnancy rates (Bedford and Elstein, 1989; Izzo
et al, 1984; Glezerman et al, 1993; Keck et al,
1994). The implication of these studies for clinical
andrology clearly highlights the importance of
statistical and outcome-valid controlled studies and
questions the benefit of kallikrein for idiopathic
infertility (for review, see Vandekerckhove et al,
1995).
Other non-hormonal therapies for idiopathic
male infertility include pentoxifylline (Micic et al.,
1988) and the antioxidants vitamin C (Dawson
et al, 1992) and vitamin E (Moilanen and Hovatta,
1995). Existing data for pentoxifylline (Micic et al,
1988) are not conclusive, because of uncertain
randomization and insufficient data. Despite a
known existing pathophysiological role for the
antioxidants, most published studies are not truly
randomized. The truly randomized studies with
vitamin C yielded only marginal improvements,
with low statistical power and low external validity
(Dawson et al, 1992). Moreover, the vitamin E
study of highest quality has uncertain inclusion
70
(e.g. concomitant medication) and exclusion
criteria (Moilanen and Hovatta, 1995) and a statistical power too low to provide sufficient evidence.
As efficacy cannot be ruled out on the basis of
existing data, the antioxidants should be evaluated
in good quality trials.
Hormonal treatment
Since GnRH, gonadotrophins and testosterone are
required for normal testicular function (for review,
Weinbauer et al, 1997) and are effective in the
treatment of hypogonadism, their application has
also been tried for treatment of idiopathic male
infertility, although no endocrine defect has been
demonstrated as a cause of this type of infertility.
It was suggested that oligoasthenoteratozoospermia in patients with elevated follicle stimulating
hormone (FSH) might be caused by too infrequent
GnRH pulses and hence that sperm parameters
might be improved by pulsatile GnRH therapy
(Wagner and Warsch, 1984). This could not be
confirmed by a longitudinal uncontrolled study
where patients were treated with GnRH pulses
(Bals-Pratsch etal, 1989). Because of the negative
result compared with pretreatment semen values,
a controlled study appears not to be necessary, as
it would have been if the result had been positive.
Gonadotrophins were applied for the treatment
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
matched pairs of patients
Conventional treatments of male infertility
Kallikrein treatment
COR
Bedford and Elstein, 1981 n=23a
Izzo etal., 1984 n=30
Keck et al.. 1994 n=91
A?=144
HCG/HMG treatment.
Knuth etal., 1987 n=38
Mauss, 1974 n=209
Aafjessfa/., 1983 n=598
Scottish Infertility Group, 1984 n=328
WHO, 1989 n=105
Pusch, 1989 n=57
Comhaire, 1990 n=24 a
Androgen treatment
COR n=834
Antiestrogenic treatment
COR n=244
Ronnberg, 1990 n=29
Torok, 1985 n=54
Sokolefa/., 1988 n=20
WHO, 1992 n=141
Immunological infertility
COR n=54
»-
Bals-Pratschefa/., 1992 n=19
Haas and Manganiello, 1987 n=35
Nieschlag ef a/., 1995 n=95
0.01
0.10
0.50 1
2
5
10
50
Combined odds ratios
Figure 3. Combined odds ratios (COR) of randomized, placebo-controlled clinical trials on male infertility with the outcome
expressed as pregnancies per patient. Results were presented as COR with 95% confidence interval (CI) of all single studies
mentioned on the right side of the figure. aIf pregnancies could be clearly allocated to treatment phases in crossover studies,
data were calculated only for the first treatment phase. bNo pregnancy in one of the groups; one was added to all groups for
statistical purpose.
of idiopathic infertility on the hypothesis that
elevation of gonadotrophin values may lead to
stimulation of spermatogenesis. The HCG/HMG
treatment was used for many years. In a review,
39 uncontrolled studies were summarized that
reported average pregnancy rates of 8-14% (Schill,
1986). In contrast to these studies, a randomized,
double-blind, placebo-controlled study of HCG/
HMG treatment for normogonadotrophic oligoasthenoteratozoospermic men could not demonstrate any beneficial effect on sperm parameters or
pregnancy rate (Figure 2; Knuth et al, 1987).
This study provides evidence that the HCG/HMG
treatment schedule did not improve semen
parameters beyond random fluctuations.
In monkeys it was shown that FSH treatment
alone was able to stimulate spermatogenesis (van
Alphen et al, 1988). Previous uncontrolled
attempts to treat patients with idiopathic infertility
with HMG, i.e. FSH alone, failed (Pescosolido
et al, 1985), which was claimed to be due to a
loss of biopotency of serum FSH and of residual
HCG activity. The availability of recombinant FSH
and positive results in an observational study with
pure FSH have led to reconsideration of FSH
treatment for male infertility (Acosta et al, 1992;
Simoni and Nieschlag, 1995). Other groups have
also confirmed a benefit from pure FSH treatment
in various uncontrolled trials. As no improvements
in classical semen parameters in the treatment
group were observed in some trials, in a different
uncontrolled trial the efficacy of rhFSH was considered to be due to the better fine structure of the
sperm head subcellular organelles (Bartoov et al,
1994). As all conducted studies published so far
were not properly controlled, the conclusion that
FSH is really efficacious in male infertility has
still not been convincingly proven. This stresses the
need for double-blind, placebo-controlled studies,
which are now under way (Baker et al, 1992;
Simoni and Nieschlag, 1995).
The requirement of testosterone for normal spermatogenesis has also led to the use of androgens
in treatment of idiopathic male infertility. Mesterolone was used for more than 20 years in the
treatment of idiopathic infertility on the basis of
71
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Varicocele
A.Kamischke and E.Nieschlag
Conclusions
The above-mentioned examples of male infertility
treatment highlight the importance of high-quality
studies in andrology. As such studies are complex
and time consuming, these problems should be
tackled by certain centres and institutions which
have developed a reputation for the quality of their
work over time. Such centres (of excellence, as
they are sometimes called) or organizations (such
as WHO) have a specific responsibility to maintain
the quality of their work (Leifke and Nieschlag,
1996) and focus their efforts on achieving new
therapeutic regimes for male infertility.
New therapeutic strategies for male infertility
are urgently required. These might be derived from
increasing knowledge about testicular paracrinology (for review, see Spiteri-Grech and
Nieschlag, 1993) and the biology of spermatogenesis (e.g. Blendy et al, 1996) These factors, in
addition to endocrine regulation, may offer a great
opportunity for successful infertility treatment and
72
should be thoroughly explored with the help of
high-quality studies.
Acknowledgements
A. Kamischke is a recipient of a fellowship from
the Federal Health Ministry (Bonn) and the Klinische
Forschungsgruppe fur Reproduktionsmedizin e.V
(Miinster). In addition, our own recent work reported here
was supported by the Deutsche Forschungsgemeinschaft
(DFG). The authors wish to thank Susan Nieschlag,
M.A. for language editing.
References
Aafjes, J., VanDer Vijer, J. and Brugman, F. (1983)
Double blind cross-over treatment with mesterolone
and placebo of subfertile men. Value of testicular
biopsy. Andrologia, 15, 531-535.
Acosta, A.A., Khalifa, E. and Oehninger, S. (1992) Pure
human follicle stimulating hormone has a role in
the treatment of severe male infertility by assisted
reproduction: Norfolk's total experience. Hum.
Reprod., 7, 1067-1072.
Altman, D.G. and Dore, C.J. (1990) Randomisation and
baseline comparisons in clinical trials. Lancet, 335,
149-153.
Antczak-Bouckoms, A. (ed.) (1995) Appendix 7. Hand
searching manual. In Cochrane Collaboration
Handbook, section V. The Cochrane Collaboration,
Oxford.
Baker, G., Brindsen, P., Jouannet, P. et al. (1992)
Treatment of severe male infertility with pure follicle
stimulating hormone (FSH): the need for a properly
controlled multicentre trial. Hum. Reprod., 8,500-501.
Bals-Pratsch, M , Knuth, U.A., Honigl, W. et al. (1989)
Pulsatile GnRH-therapy in oligozoospermic men does
not improve seminal parameters despite decreased
FSH levels. Clin. EndocrinoL, 30, 549-560.
Bals-Pratsch, M., Doren, M., Karbowski, B. et al. (1992)
Cyclic corticoid immunosuppression is unsuccessful
in the treatment of sperm antibody-related male
infertility: a controlled study. Hum. Reprod., 7, 99104.
Bartoov, B., Eltes, F. and Lunenfeld, E. (1994) Sperm
quality of subfertile males before and after treatment
with follicle-stimulating hormone. Fertil. Steril, 61,
727-734.
Bedford, N. and Elstein, M. (1981) Effects of kallikrein
on male subfertility—a double blind cross-over study.
In Insler, V. and Bettendorf, G. (eds), Advances in
Diagnosis and Treatment of Infertility. Elsevier North
Holland, New York, pp. 339-343.
Behre, H.M., Yeung, C.H. and Nieschlag, E. (1997)
Diagnosis of male infertility and hypogonadism. In
Nieschlag, E. and Behre, H.M. (eds), Andrology:
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
uncontrolled studies. Truly randomized, placebocontrolled, double-blind, studies, however, clearly
showed the ineffectiveness (Figure 3) of mesterolone and oral testosterone undecanoate treatment in
controlled studies (for review, see Vandekerckhove
et al, 1996).
It is well known that oestrogens can have a
negative feedback on hypophyseal gonadotrophin
secretion. Anti-oestrogenic compounds competitively blocking oestrogens at the receptor site might
reduce the inhibitory feedback and lead to elevation
of gonadotrophin and testosterone serum concentrations. Assuming that an increase in luteinizing
hormone and FSH may improve sperm production,
anti-oestrogens (in particular tamoxifen) have been
widely used. None of the truly randomized, doubleblind and placebo-controlled studies demonstrated
any significant beneficial effect of tamoxifen
therapy on fertility (Figure 3) (for review, see Rolf
et al, 1996). Nor could any significant therapeutic
effect be demonstrated for clomiphene, which is
the other widely used anti-oestrogenic compound
(WHO, 1992b). Therefore, anti-oestrogenic treatment has to be considered obsolete, especially in
view of its potentially toxic side effects.
Conventional treatments of male infertility
the practice of medicine. J. Am. Med. Assoc, 268,
2420-2425.
Freiman, J.A., Thomas, A.B., Chalmers, C. et al. (1978)
The importance of beta, the type II error and sample
size in the design and interpretation of the randomized
control trial. N. Engl. J. Med., 299, 690-694.
Gerris, J., Comhaire, E, Hellemans, P. et al. (1991)
Placebo controlled trial of high dose Mesterolone
treatment of male infertility. Fertil. Steril., 55, 603607.
Glezerman, M., Lunenfeld, E., Potashnik, G. et al.
(1993) Efficacy of kallikrein in the treatment of
oligozoospermia and asthenozoospermia: a doubleblind trial. Fertil. Steril, 60, 1052-1056.
Gotzsche, PC. (1994) Is there logic in the placebo?
Lancet, 344, 925-926.
Haas, G. and Manganiello, P. (1987) A double blind,
placebo controlled study of the use of
methylprednisolone in infertile men with sperm
associated immunoglobulins. Fertil. Steril, 47, 295301.
Hargreave, T.B. (1995) Debate on the pros and cons of
varicocele treatment. Hum. Reprod., 10 (Suppl. 1),
151-157.
Hennekens, C.H. and Buring, J.E. (1993) Observational
evidence. Ann. N Y. Acad. Scl, 703, 18-24.
Howards, S.S. (1995) Treatment of male infertility. N.
Engl. J. Med, 332, 312-316.
International Conference on Harmonisation of Technical
Requirements for Registration of Pharmaceuticals for
Human Use (ICH) (1996) Good Clinical Practice:
Consolidated Guideline, Geneva, E6GCPD12, pp.
1-71.
Izzo, P., Canale, D., Bianchi, B. et al. (1984) The
treatment of male subfertility with kallikrein.
Andrologia, 16, 156-161.
Jaschke, R., Guyatt, G. and Sackett, D.L. (1994) User's
guides to the medical literature, III: How to use an
article about a diagnostic test. Are the results of the
study valid? J. Am. Med. Assoc, 271, 389-391.
Johnson, A.G. (1994) Surgery as a placebo. Lancet, 344,
1140-1142.
Keck, C , Behre, H.M., Jockenhovel, F. and Nieschlag,
E. (1994) Ineffectiveness of kallikrein in treatment of
idiopathic male infertility: a double-blind, randomized,
placebo-controlled trial. Hum. Reprod., 9, 325-329.
Kleijnen, J., de Craen, A.J.M., van Everdingen, J. and
Krol, L. (1994) Placebo effect in double-blind clinical
trials: a review of interactions with medications.
Lancet, 344, 1347-1351.
Kliesch, S., Behre, H.M. and Nieschlag, E. (1994) High
efficacy of gonadotropin or pulsatile GnRH treatment
in hypogonadotropic hypogonadal men. Eur. J.
Endocrinol., 131, 347-354.
Kliesch, S., Behre, H.M. and Nieschlag, E. (1995)
Recombinant human follicle stimulating hormone and
73
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Male Reproductive Health and Dysfunction. Springer,
Heidelberg, pp. 87-111.
Black, N. (1996) Why we need observational studies to
evaluate the effectiveness of health care. Br. Med. J.,
312, 1215-1218.
Blendy, J.A., Kaestner, K.H., Weinbauer, G.F. et al.
(1996) Severe impairment of spermatogenesis in mice
lacking the CREM gene. Nature, 380, 162-165.
Campbell, E.J.M. (1987) The diagnosing mind. Lancet,
329, 849-851.
Collins, J.A., Burrows, E.A., Yeo, J. and YoungLai, E.V.
(1993) Frequency and predictive value of antisperm
antibodies among infertile couples. Hum. Reprod., 8,
592-598.
Comhaire, F.H. (1990) Treatment of idiopathic testicular
failure with high-dose testosterone undecanoate: a
double blind pilot study. Fertil. Steril, 54, 689-693.
Comhaire, F.H., Rowe, P.J. and Farley, T.M. (1986) The
effect of doxycycline in infertile couples with male
accessory gland infection: a double blind prospective
study. Int. J. Androl., 9, 91-98.
Cooper, T.G. (1996) Implementation of quality control
in the andrology laboratory. Int. J. Androl., 19, 67-68.
Cooper, T.G., Neuwinger, J., Bahrs, S. and Nieschlag,
E. (1992) Internal quality control of semen analysis.
Fertil. Steril., 58, 172-178.
Davidoff, F , Haynes, B., Sackett, D.L. and Smith, R.
(1995) Evidence-based medicine, a new approach to
teaching the practice of medicine. Br. Med. J., 310,
1085-1086.
Dawson, E.B., Harris, W.A., Teter, M.C. and Powell,
L.C. (1992) Effect of ascorbic acid supplementation
on the sperm quality of smokers. Fertil. Steril., 58,
1034-1039.
Detsky, A.S., Naylor, CD., O'Rourke, K. et al. (1992)
Incorporating variations in the quality of individual
randomized trials into meta-analysis. /. Clin.
Epidemiol, 45, 255-265.
Deutsche Gesellschaft fur Endokrinologie (DGE) (1991)
Zur Therapie des Hodenhochstandes. EndokrinologieInformation, 15, 20-22.
Eggert-Kruse, W., Hofsass, A., Haury, E. et al. (1991)
Relationship between local anti-sperm antibodies and
sperm mucus interaction in vitro and in vivo. Hum.
Reprod, 6, 267-276.
Eggert-Kruse, W., Rohr, G., Bockem-Hellwig, S. et al.
(1995) Immunological aspects of subfertility. Int. J.
Androl., 18 (Suppl. 2), 43-52.
ESHRE Andrology Special Interest Group (1996)
Consensus workshop on advanced diagnostic
andrology techniques. Hum. Reprod., 11, 1463-1479.
ESHRE Capri Workshop (1996) Guidelines to the
prevalence, diagnosis, treatment and management of
infertility, 1996. Hum. Reprod., 11, 1775-1807.
Evidence-Based Medicine Working Group (1992)
Evidence-based medicine: a new approach to teaching
A.Kamischke and E.Nieschlag
74
A.C. (1992) Pregnancies after intracytoplasmatic
sperm injection of single spermatozoa. Lancet, 430,
17-18.
Pescosolido, D., Curatola, A. and Gentili, S. (1985) A
new gonadotropin treatment for idiopathic moderate
oligoasthenospermia. Acta Eur. Fertil, 16, 129-132.
Pusch, H. (1989) Oral treatment of oligozoospermia
with testosterone-undecanoate: results of a double
blind study. Arch. Androl, 2, 479-486.
Rolf, C , Behre, H.M. and Nieschlag, E. (1995)
Tamoxifen bei mannlicher Infertilitat: Analyse einer
fragwiirdigen Therapie. Dtsch. Med. Wochenschr.,
121, 33-39.
Ronnberg, L. (1980) The effect of clomiphene citrate
on different sperm parameters and serum hormone
levels in preselected infertile men: a controlled double
blind study. Int. J. Androl, 3, 479^86.
Sackett, D.L. and Cook, D. (1993) Can we learn anything
from small trials? Ann. N.Y. Acad. Sci., 703, 25-31.
Schill, W.B. (1978) Kallikrein im Doppelblindversuch
bei idiopathischer Infertilitat. Hautarzt, 28, 308-312.
Schill, W.B. (1986) Medical treatment of male infertility.
In Insler, V. and Lunenfeld, B. (eds), Infertility: Male
and Female. Churchill Livingstone, Edinburgh, pp.
533-573.
Schopohl, J., Mehltretter, G., von Zumbusch, R. et al.
(1991) Comparison of gonadotropin-releasing
hormone and gonadotropin therapy in male patients
with idiopathic hypothalamic hypogonadism. Fertil.
Steril, 56, 1143-1150.
Scottish Infertility Group (1984) Randomized trial of
mesterolone versus vitamin C for male infertility. Br.
J. Urol, 56, 740-744.
Simoni, M. and Nieschlag, E. (1995) FSH in therapy:
physiological basis, new preparations and clinical use.
Reprod. Med. Rev., 4, 163-177.
Sokol, R., Steiner, B., Bustillo, M. et al. (1988) A
controlled comparison of the efficacy of clomiphene
citrate in male infertility. Fertil. Steril, 49, 865-870.
Spiteri-Grech, J. and Nieschlag, E. (1993) Paracrine
factors relevant to the regulation of spermatogenesis
review. J. Reprod. Fertil, 98, 1-14.
Takihara, H., Sakatoku, J. and Cockett, A.T.K. (1991)
The pathophysiology of varicocele in male infertility.
Fertil. Steril, 55, 861-868.
Tesarik, J., Mendoza, C. and Testart, J. (1995) Viable
embryos from injection of round spermatids into
oocytes. N. Engl. J. Med., 333, 525.
Torok, L. (1985) Treatment of oligozoospermia with
tamoxifen (open and controlled studies). Andrologia,
17, 497-501.
van Alphen, M.M.A., van de Kant, H.J.G. and de Rooij,
D.G. (1988) Follicle-stimulating hormone stimulates
spermatogenesis in the adult monkey. Endocrinology,
123, 1449-1455.
Vandekerckhove, P.O., Donavan, P.A., Lilford, R. and
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
human chorionic gonadotropin for induction of
spermatogenesis in a hypogonadal male. Fertil. Steril,
63, 1326-1328.
Knuth, U.A., Hdnigl, W., Bals-Pratsch, M. et al (1987)
Treatment of severe oligozoospermia with hCG/hMG.
A placebo-controlled double blind trial. J. Clin.
Endocrinol. Metab., 65, 1081-1087.
Leifke, E. and Nieschlag, E. (1996) Evidence-based
andrology: the importance of controlled clinical trials.
In Hansson, V., Levy, F.O.L. and Tasken, K. (eds),
Signal Transduction in Testicular Cells: Basic and
Clinical Aspects. Springer, Heidelberg, pp. 287-306.
Madgar, I., Weissenberg, R., Lunenfeld, B. et al. (1995)
Controlled trial of high spermatic vein ligation for
varicocele in infertile men. Fertil. Steril., 63,120-124.
Mahmoud, A.M., Tuyttens, C.L. and Comhaire, F.H.
(1996) Clinical and biological aspects of male immune
infertility: a case-controlled study of 86 cases.
Andrologia, 28, 191-196.
Mauss, J. (1974) Ergebnisse der Behandlung von
Fertilitatsstorungen des Mannes mit Mesterolon oder
einem Placebo. Arzneim.-Forsch., 2A, 1338-1341.
Micic, S., Hadzi-Djokic, J. and Dotlic, R. (1988)
Pentoxifyllin treatment of oligoasthenospermic men.
Ada Eur. Fertil, 19, 135-137.
Moilanen, J. and Hovatta, O. (1995) Excretion of alphatocopherol into human seminal plasma after oral
administration. Andrologia, 27, 133-136.
Morris, J.A. and Gardner, M.J. (1988) Calculating
confidence intervals for relative risks (odds ratios)
and standardised ratios and rates. Br. Med. J., 296,
1313-1316.
Nieschlag, E. (1993) Current therapy: care for the
infertile male. Clin. Endocrinol., 38, 123-133.
Nieschlag, E. (1997) Classification of andrological
disorders. In Nieschlag, E. and Behre, H.M. (eds),
Andrology: Male Reproductive Health and
Dysfunction. Springer, Heidelberg, pp. 81-83.
Nieschlag, E. and Behre, H.M. (1997) Andrology: Male
Reproductive Health and Dysfunction. Springer,
Heidelberg.
Nieschlag, E., Hertle, L., Fischedick, A. and Behre,
H.M. (1995) Treatment of varicocele: counselling as
effective as occlusion of the vena spermatica. Hum.
Reprod., 10, 347-353.
O'Donovan, P.A., Vandekerckhove, P., Lilford, R.J. and
Hughes, E. (1993) Treatment of male infertility: is it
effective? Review and meta-analysis of published
randomized controlled trials. Hum. Reprod., 8,
1209-1222.
Olive, D.L. (1986) Analysis of clinical fertility trials: a
methodologic review. Fertil. Steril, 45, 157-171.
Organisation for Economic Co-Operation and
Development (OECD), Paris (1992) The OECD
Principles of Good Laboratory Practice. OECD, Paris,
Environment monograph no. 45.
Palermo, G., Joris, H., Devroey, P. and Van Steirteghem,
Conventional treatments of male infertility
response into ordered categories. Statistics Med., 13,
2503-2515.
Wilcox, A.J., Weinberg, C.R. and Baird, D.D. (1995)
Timing of sexual intercourse in relation to ovulation.
Effects on the probability of conception, survival of
the pregnancy, and sex of the baby. N. Engl. J. Med.,
333, 1517-1521.
WHO Task Force on the Diagnosis and Treatment of
Infertility (1989) Mesterolone and idiopathic male
infertility: a double-blind study. Int. J. AndroL, 12,
254-264.
World Health Organization (1992a) WHO Laboratory
Manual for the Examination of Human Semen and
Sperm—Cervical Mucus Interaction, 3rd edn.
Cambridge University Press, Cambridge.
World Health Organization (WHO) Special Programme
of Research, Development and Research Training in
Human Reproduction, Task Force on the Prevention
and Management of Infertility (1992b) A doubleblind trial of clomiphene citrate for the treatment of
idiopathic male infertility. Int. J. AndroL, 15, 299-307.
World Health Organization (1992c) The influence of
varicocele on parameters of fertility in a large group
of men presenting to infertility clinics. Fertil. Steril.,
57, 1289-1293.
World Health Organization (1993a) Internal Quality
Control for Assays of Reproductive Hormones:
Guidelines for
Laboratories. World Health
Organization, Geneva.
World Health Organization (1993b) Guidelines for Good
Clinical Practice (GCP)for Trials on Pharmaceutical
Products. World Health Organization, Geneva.
Yanushpolski, E.T., Politch, J.A., Hill, J.A. and
Anderson, D.J. (1995) Antibiotic therapy and
leukocytospermia:
a
prospective
randomized
controlled study. Fertil. Steril., 63, 142-147.
75
Downloaded from http://humrep.oxfordjournals.org/ by guest on September 9, 2014
Harada, T.W. (1993) Infertility treatment: from
cookery to science. The epidemiology of randomized
controlled trials. Br. J. Obstet. Gynaecol, 100,
1005-1036.
Vandekerckhove, P., Lilford, R. and Hughes, E. (1995)
Kinin enhancing drugs for male infertility. In Lilford,
R., Hughes, E. and Vandekerckhove, P. (eds),
Subfertility Module of the Cochrane Database of
Systematic Reviews. In The Cochrane Library, The
Cochrane Collaboration, Issue 2. Oxford: Update
Software.
Vandekerckhove, P., Lilford, R. and Hughes, E. (1996)
The medical treatment of idiopathic oligo/
asthenospermia:
androgens
(mesterolone
or
testosterone) versus placebo or no treatment. In
Lilford, R., Hughes, E. and Vandekerckhove, P. (eds),
Subfertility Module of the Cochrane Database of
Systematic Reviews. In The Cochrane Library, The
Cochrane Collaboration, Issue 2. Oxford: Update
Software.
Wagner, R.O.F. and Warsch, F. (1984) Pulsatile LHRH
therapy of 'slow pulsing oligospermia': indirect
evidence for a hypothalamic origin of the disorder.
Acta Endocrinol, 105 (Suppl. 264), 142-145.
Walker, A.M., Martin-Moreno, J.M. and Artalejo, F.R.
(1988) Odd man out: a graphical approach to metaanalysis. Am. J. Public Health, 78, 961-966.
Weinbauer, G.F., Gromoll, J., Simoni, M. and Nieschlag,
E. (1997) Physiology of testicular function In
Nieschlag, E. and Behre, H.M. (eds), Andrology:
Male Reproductive Health and Dysfunction. Springer,
Heidelberg, pp. 25-51.
Whitehead, A. and Jones, N.M.B. (1994) A metaanalysis
of clinical trials involving different classifications of