REVIEWS

Varicocele and male infertility

Christian Fuglesang S. Jensen1,2, Peter Østergren1,2, James M. Dupree3, Dana A. Ohl3,
Jens Sønksen1,2 and Mikkel Fode1,4
Abstract | The link between varicoceles and male infertility has been a matter of debate for
more than half a century. Varicocele is considered the most common correctable cause of
male infertility, but some men with varicoceles are able to father children, even without
intervention. In addition, improvements in semen quality after varicocelectomy do not always
result in spontaneous pregnancy. Studies regarding possible pathophysiological mechanisms
behind varicocele-induced infertility have tried to address these controversies. Oxidative
stress seems to be a central mechanism; however, no single theory is able to explain the
differential effect of varicoceles on infertility. As a consequence, careful patient selection for
treatment based on couple fertility status, varicocele grade, and semen quality is critical for
achieving a chance of a subsequent pregnancy. A substantial amount of data on the effects of
varicocelectomy has been gathered, but inadequate study design and considerable
heterogeneity of available studies mean that these data are rarely conclusive. Current
evidence suggests a beneficial effect of varicocelectomy on semen quality and pregnancy
outcomes in couples with documented infertility only if the male partner has a clinically
palpable varicocele and affected semen parameters.

A clinical varicocele is defined as abnormally dilated Epidemiology of varicoceles and infertility

and tortuous veins in the pampiniform plexus of the The first evidence supporting a link between varicoceles
spermatic cord and is graded according to findings on and infertility was published in 1952 by Tulloch11. In this
palpation (FIG. 1; TABLE 1). Varicoceles are widely con- case report, Tulloch described how varicocele repair in
sidered the most common correctable cause of male a man with azoospermia resulted in natural conception.
infertility1 and studies conducted over more than half Since then, the epidemiological evidence supporting
a century have associated varicoceles with negative a link between varicoceles and infertility has grown
effects on semen quality2,3, sperm function4,5, testicular substantially. Today, the incidence of varicocele among
1
Department of Urology, histology6–8, and reproductive hormones2,9. However, men with primary infertility is estimated at 35–44%,
Herlev and Gentofte Hospital, most men with a varicocele are able to father children whereas the incidence in men with secondary infertility
Herlev Ringvej 75, and the exact connection between varicoceles and is ~45–81%12,13. Incidence of varicocele is <1% in pre­
DK‑2730 Herlev, Denmark.
infertility is, therefore, controversial. Oxidative stress adolescent boys aged 2–10 years, but increases to ~14%
2
Faculty of Health and
Medical Sciences, University seems to have an increasingly important role10, but the in adolescents aged 15–19 years14. In the general popu-
of Copenhagen, Blegdamsvej potential pathophysiology of varicocele-induced infer- lation of adult men the incidence is reported at ~15%15.
3B, DK‑2200 Copenhagen, tility is subject to many theories. Likewise, the effects These data are supported a cross-sectional study of 7,035
Denmark of varicocelectomy and recommendations regarding men with a median age of 19 years, 15.7% of whom had
3
Department of Urology,
University of Michigan,
patient selection for surgery have been extensively a clinical varicocele diagnosed on physical examination2.
1500 E Medical Center Dr, discussed and studied without the generation of Participants were recruited when attending a compul-
Ann Arbor, conclusive answers. sory medical examination during evaluation for mili-
Michigan 48109, USA. This Review summarizes the epidemiological evi- tary service and were, therefore, not selected on the basis
4
Department of Urology,
dence behind the connection between varicocele and of their fertility status, making them comparable to
Zealand University Hospital,
Sygehusvej 10, infertility and describes the proposed pathophysiological similarly aged men from the general population.
DK‑4000 Roskilde, Denmark. aetiologies of the disorder. With a focus on high-quality Crude epidemiological findings support a link
Correspondence to M.F. studies regarding varicocele surgery and by scrutiniz- between varicoceles and infertility. However, not all
[email protected] ing the results, patient selection, and surgical methods, men with varicoceles are infertile — for example, a study
doi:10.1038/nrurol.2017.98 we provide concrete clinical recommendations for of 598 men with proven fertility seeking a vasectomy
Published online 4 Jul 2017 management of varicocele-associated infertility. showed that 16% had a varicocele16. In addition, 45–65%

NATURE REVIEWS | UROLOGY VOLUME 14 | SEPTEMBER 2017 | 523

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Key points might be part of the explanation how a left-sided clinical

varicocele can have bilateral effects on testicular function.
• Varicoceles have detrimental effects on semen quality, sperm function, The precise mechanism by which varicocele might
and pregnancy outcomes in some men cause infertility is still unknown. No single factor is
• Oxidative stress seems to have a central role in the pathogenesis of believed to be responsible for the negative testicular
varicocele-induced infertility effects; instead the pathogenesis is believed to be complex
• Current evidence supports beneficial effects of varicocelectomy on semen and multifactorial, with several proposed mechanisms
parameters, sperm function, and pregnancy outcomes in men from couples with acting together. In this complex pathophysiological net-
documented infertility with clinical varicoceles and affected semen parameters work, oxidative stress seems to have a central role (FIG. 2).
• The optimal technique for performing varicocelectomy is a microsurgical approach, A meta-analysis from 2006 that evaluated four studies
owing to reduced complication rates and increased pregnancy rates compared with reporting on similar reactive oxygen species (ROS) in
other techniques 118 infertile men with varicoceles and 76 healthy sperm
• Current evidence supports varicocele repair in adolescents with clinical varicoceles donors showed significantly higher ROS concentrations
when progressive testicular growth retardation and/or impaired semen quality in men with varicoceles compared with sperm donors
is observed (weighted mean difference 0.73; 95% CI 0.40–1.06;
• Varicocelectomy might have beneficial effects on pregnancy outcomes following P <0.0001) and significantly lower levels of total anti-
assisted reproduction oxidant capacity (TAC) (386 fewer trolox equivalents
(measurement unit for antioxidant strength) 95% CI
-556.56–-216.96; P <0.00001)22. Increased oxidative stress
of men with grade 1–3 varicoceles (n = 1,102) have can be a result of the combination of increased ROS and
normal semen parameters2. Thus, a causal relation- decreased TAC. Increased pressure on venous walls
ship between varicocele and infertility cannot be might result in release of ROS as demonstrated in a study
established through epidemiological studies. In fact, this comparing surgically removed great saphenous veins that
relationship is still a matter of debate and has sparked were sufficient, insufficient, or varicose23. Antioxidant
intense research into the pathophysiology behind power was significantly higher in sufficient veins com-
varicocele-induced infertility. pared with insufficient or varicose veins (P < 0.001)
and, likewise, levels of markers of oxidative stress were
How varicoceles cause infertility significantly higher in varicose and in­­sufficient veins
Varicoceles occur when blood flows backwards into the compared with sufficient veins (P < 0.001). In addition,
internal spermatic vein resulting in vascular dilation of higher grades of varicoceles have been shown to correlate
the veins in the pampiniform plexus (FIG. 1). This patho- with higher levels of oxidative stress24, and levels of bio­
logical reflux of blood is thought to be caused by con­ markers of oxidative stress have been found to decrease
genital insufficient or absent venous valves17 and is most after varicocele repair, with a corresponding improve-
often found on the left side as the hydrostatic pressure is ment in semen parameters25. Oxidative stress can harm
higher on this side owing to the perpendicular drainage of germ cells directly or indirectly through influencing
the left internal spermatic vein into the left renal vein17–19. non­spermatogenic cells and the basal lamina of the semi­
The hydrostatic pressure on the left side can increase even niferous tubules resulting in induction of apoptosis26. The
further as the superior mesenteric artery crosses the left ROS and TAC balance shift leads to oxidation of fatty
renal vein and exerts additional pressure often termed ‘the acids in spermatozoa membranes causing changes in
nutcracker effect’ (REFS 17,18). As a consequence, up to sperm morphology, motility, and fertilizing capabilities26.
90% of men with varicocele are diagnosed with a unilat- Other possible pathophysiological mechanisms
eral, left-sided varicocele17. However, venographic studies involved in varicocele-induced male infertility include
have shown bilateral abnormal venous reflux in ~80% of scrotal hyperthermia, hypoxia, reflux of renal and adre-
men with varicoceles, a fact illustrated in a study including nal metabolites, hormonal imbalances, and the forma-
255 men diagnosed with clinical or subclinical varicoce- tion of antisperm antibodies. Venous reflux increases
les based on either physical examination, thermo­graphy, scrotal temperature and impairs normal spermato­
Doppler ultrasonography or venography20. On physi- genesis, which usually takes place at a temperature 2°C
cal examination, a left-sided or right-sided vari­cocele below core body temperature18. In addition heat stress
was detected in 89.4% and 0.8% of men, respectively can cause production of reactive oxygen species (ROS)
(no data are presented on the number of bilateral cases increasing the levels of oxidative stress17,27. Markers of
on physical examination). By contrast, when using veno­ hypoxia are increased in dilated veins of patients with
graphy, a right-sided varicocele was detected in 82.4% varicocele, and testicular tissue ischaemia is thought
of patients and a left varicocele in 98.4%, with 80.8% of to occur as the venous pressure in the internal sper-
the men presenting with a bilateral varicocele20. Another matic vein exceeds testicular arteriolar pressure, with
study of 840 men clinically suspected of having a vari- resulting detrimental effects on testicular function
cocele demonstrated abnormal venous reflux on the through inflammatory reactions and the hypoxia induc-
right side in 86% of participants (no data are available ible factor (HIF) pathway27,28. Varicoceles are thought
for the left side)21. Both studies lack information on the to have a negative effect on Sertoli cell function and
number of clinically diagnosed men and on varicocele Leydig cell function, resulting in an imbalance in the
grade. However, detection of abnormal venous reflux hypo­thalamic–pituitary–gonadal axis with reduced tes-
on the right side in a substantial number of patients tosterone levels ultimately affecting spermatogenesis29.

524 | SEPTEMBER 2017 | VOLUME 14 www.nature.com/nrurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

reactions17,27. Reflux of renal and adrenal metabolites

has been shown to increase oxidative stress; increased
hydrostatic pressure causes reflux of adrenal and renal
Left renal vein metabolites into the internal spermatic vein and, sub-
sequently, into the testes causing vasoconstriction of
testicular arterioles leading to hypoxia and impaired
spermatogenesis17,27.
Left internal One question that remains unanswered is why only
spermatic vein
a fraction of men with varicoceles develop infertility.
The link between varicocele size and fertility might
be one explanation, with the possibility that only
large vari­coceles have an effect on fertility, as increas-
ing varicocele grade is associated with a corresponding
decrease in semen quality2. Another possibility is that
No valves some men have a higher TAC, making them less sus-
ceptible to the damages of increasing ROS. Moreover,
considering infertility as a clinical sequela of varicocele
is meaningless unless the female partner is taken into
account, and optimal female fertility status might be able
to compensate for varicocele-induced negative effects on
the male reproductive system. This differential effect of
Plexus pampiniformis
vari­coceles on a couple’s fertility highlights the need for
Varicocele careful patient selection when considering performing
varicocele repair.

Effects of varicocele repair

Figure 1 | Anatomy of the venous drainage from the testes. Varicoceles arise when Numerous studies have examined the effects of vari­
blood flows backwards into the internal spermatic vein resultingNature Reviews
in vascular | Urology
dilation of cocele repair on semen parameters, sperm function,
the veins in the pampiniform plexus. Reflux of blood is thought to be caused by
pregnancy outcomes, and reproductive hormones.
congenital insufficient or absent venous valves and is most often found on the left side as
the hydrostatic pressure is higher on this side owing to the perpendicular drainage of the
Although optimal semen parameters increase the prob-
left internal spermatic vein into the left renal vein. The hydrostatic pressure on the left ability of conception30, semen parameters are only surro-
side can increase even further as the superior mesenteric artery crosses the left renal gate markers of fertility, meaning that studies that report
vein and exerts additional pressure often termed ‘the nutcracker effect’. them as end points might not reflect clinically relevant
outcomes.

They might also be associated with a breach in the Effects on semen quality
blood–testis barrier resulting in the production of In 2011, a meta-analysis showed significant increases
antisperm antibodies. Levels of antisperm antibodies in sperm concentration after varicocelectomy (random
are higher in infertile men compared with fertile men; effect model combined improvement 12.32 × 106 sperm/
however, the levels among infertile men with or with- ml; 95% CI 9.45–15.19, P <0.0001), in total motility
out varicoceles are equal. Thus, the association between (improvement 10.86%; 95% CI 7.07–14.65, P <0.0001)
varicocele and antisperm antibodies is not entirely clear29. and in sperm progressive motility (improvement 9.69%;
Whereas some of these mechanisms act on their own 95% CI 4.86–14.52, P = 0.003), based on 22, 17, and
to cause negative testicular effects, others also influence five prospective studies, respectively31. The included
the level of oxidative stress highlighting the central role random­ized prospective trials, and nonrandomized
of this pathophysiological mechanism. Consequently, prospective trials showed considerable heterogeneity
heat stress has been shown to increase oxidative stress between the studies in regards to patient characteris-
by production of ROS from mitochondrial membranes, tics, diagnostic criteria, and treatment methods, mak-
cytoplasm, and peroxisomes, and likewise hypoxia ing comparison between the studies difficult. An earlier
might increase oxidative stress through inflammatory meta-analysis from 2007 (REF. 32) included ~1,200 infer-
tile men with a unilateral or bilateral clinical varicocele
and ≥1 affected semen parameter. This meta-­analysis
Table 1 | Grading of varicoceles on men in an upright position
evaluated the effects of varicocele repair on semen
Varicocele grade Clinical features parameters and showed a statistically significant increase
Subclinical Examined with Doppler ultrasonography and defined as in sperm concentration (9.71 × 106 sperm/ml; 95% 
reverse venous blood flow during Valsalva manoeuvre or CI 7.34–12.08, P <0.00001) and motility (9.92%; 95% CI
ectasia of the spermatic vein (>3 mm) 4.90–14.95, P = 0.0001) after inguinal microsurgical vari-
1 Only palpable during Valsalva manoeuvre cocelectomy. Similar effects were seen after high ligation
2 Palpable without Valsalva manoeuvre varicocelectomy, with statistically significant increases
in sperm concentration (12.03 × 106 sperm/ml; 95% 
3 Visible and palpable without Valsalva manoeuvre CI 5.71–18.35, P = 0.0002) and motility (11.72%; 95% CI

NATURE REVIEWS | UROLOGY VOLUME 14 | SEPTEMBER 2017 | 525

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Varicocele surgical varicocele repair without testicular-artery spar-

ing, which might have limited any potential beneficial
effects of the varicocelectomy. In a 2012 randomized
controlled trial by Mansour and colleagues34, male part-
↑ Hydostatic pressure and venous reflux
ners with varicoceles from 136 infertile couples were
randomly assigned to either varicocelectomy or obser-
vation. After 6 months, men who had undergone vari-
↑ Scrotal Hypoxia Reflux of toxic cocelectomy showed a significantly greater improvement
temperature metabolites
in sperm concentration than those who were observed
(75% versus 12.5%; P = 0.001). However, all men from
both groups included in the study had normal semen
Oxidative stress parameters at baseline, making the study less relevant
↑ Reactive oxygen species
↓ Total antioxidant capacity
to clinical practice. An earlier randomized controlled
trial published by Abdel-Meguid et al.35 in 2011 is more
relevant, as it included 145 infertile men with palpable
varicocele and ≥1 affected semen parameter. They were
Impaired testicular function randomized to receive either subinguinal micro­surgical
varicocelectomy (n = 73) or were observed without
treatment (n = 72). After 12 months, the means of all
Figure 2 | The central role of oxidative stress in the pathophysiology of varicocele
Nature Reviews | Urology
induced male infertility. Although varicocele pathogenesis is multifactorial, oxidative semen parameters improved significantly compared
stress seems to have a central role. Increased oxidative stress can be a result of the with baseline in the treatment group (18.1 ± 5.8 versus
combination of increased ROS and decreased TAC. Increased pressure on venous walls 32.2 ± 10.6 × 106 sperm/ml, 25.3 ± 12.8 versus 41.0 ± 10%,
might result in release of ROS and higher grades of varicoceles have been shown to and 31.2 ± 4.1 versus 39. ± 4.5% for sperm concentration,
correlate with higher levels of oxidative stress. Oxidative stress can harm germ cells motility and morphology, respectively,all P <0.0001)
directly or indirectly through influencing nonspermatogenic cells and the basal lamina of whereas no difference was observed in the control group.
the seminiferous tubules resulting in induction of apoptosis. In addition, spermatozoa Overall, good-quality randomized controlled trials
DNA and lipid membranes are attacked directly by ROS, resulting in sperm DNA damage, are lacking in this field, a concern that is reflected by
increased DNA fragmentation, and lipid peroxidation. Other potential mechanisms the heterogeneity of included studies in the available
involved in varicocele-induced male infertility include scrotal hyperthermia, hypoxia,
meta-analyses. However, the randomized controlled trial
reflux of renal and adrenal metabolites, hormonal imbalances, and the formation of
antisperm antibodies; these mechanisms might act on their own to cause negative by Abdel-Meguid et al.35 does support the findings of the
testicular effects, or they might also influence the level of oxidative stress, highlighting meta-analyses, strongly suggesting that varicocelectomy
the central role of this pathophysiological mechanism. Heat stress has been shown to has beneficial effects on semen parameters.
increase oxidative stress by production of ROS from mitochondrial membranes,
cytoplasm, and peroxisomes, and likewise hypoxia might increase oxidative stress Effects on levels of oxidative stress
through inflammatory reactions. Hurtado et al.25 evaluated the levels of oxidative stress
biomarkers in spermatozoa, seminal plasma, and
peripheral blood in 36 men treated for a left-sided clin-
4.33 to 19.12, P = 0.002). Morphology improved sig- ical varicocele. They observed that the oxidative stress
nificantly after both microsurgery and high ligation levels were higher in men with varicocele compared with
(3.16%; 95% CI 0.72–5.60, P = 0.01). A limitation of the 33 men in a control group with proven fertility, and that
meta-analysis is the inclusion of both randomized and these levels normalized at different times after vari­
observational studies, but the authors argued the validity cocelectomy depending on the specific biomarker inves-
of this approach on the basis that they performed ana­ tigated. For example, malondialdehyde levels (measured
lyses on effect change over time after vari­cocelectomy, using a thiobarbituric acid reactive substances (TBARS)
with the patient acting as his own control rather than assay) normalized after 3 months and TAC normalized
comparison with no‑treatment groups. However, inclu- after 3 months25. In a retrospective study from 2008,
sion of patients with affected semen quality at baseline Sakamoto and colleagues36 investigated the effects of
and lack of a control group mean that whether the subinguinal microsurgical varicocelectomy on oxida-
observed effect in either study is a result of the varicocele tive stress markers and antioxidant capacity in 15 nor-
repair or whether it is a result of regression towards the mozoospermic men and 15 oligozoospermic men with
mean remains unclear. clinical varicoceles. Six months after vari­cocelectomy,
Assessing the results of randomized controlled trials the oxidative stress markers 8‑hydroxy‑2ʹde­
is likely to provide a better understanding of the effect oxyguanosine (8‑OHdG) and hexanoyl-lysine were
of varicocelectomy on semen parameters. A randomized statistically significantly reduced (10.3 ± 4.7 umol/l ver-
study by Breznik et al.33 found no difference in post­ sus 6.2 ± 2.5 umol/l, P <0.001 and 137.3 ± 67.9 umol/l
operative semen parameters when comparing 41 infer- versus 90.9 ± 28.5 umol/l, P = 0.005) and similarly the
tile men with varicoceles in the no‑treatment group with antioxidant superoxide dismutase showed signifi-
48 infertile men with varicoceles in the varicocele-repair cantly lower activity (85.8 ± 5.8% versus 78.1 ± 8.1%,
group. However, one limitation of this study is that 12 of P = 0.01). A prospective study by Chen et al.37 on 30
the men in the treatment group were treated using sclero­ infertile men with clinical varicoceles observed a
therapy or embolization, whereas the rest underwent statistically significant decrease in oxidative damage,

526 | SEPTEMBER 2017 | VOLUME 14 www.nature.com/nrurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

measured as a reduction in spermatozoa DNA content Another major criticism of the trial is that only 125 of
of 8‑OHdG (10.27 ± 2.24 /105 dG versus 5.95 ± 1.46 /105 226 couples fulfilling the inclusion criteria completed
dG, P <0.001) 6 months after subinguinal micro­surgical the study. Among the couples who did not complete
varicocelectomy37. Similarly they observed a signif- the study 23 couples opted for assisted reproduction,
icant increase in antioxidant capacity measured as an 43 couples were lost after randomization, 14 couples
increase in seminal plasma protein thiols and ascorbic were lost during follow‑up, 11 female partners devel-
acid (0.77 ± 0.75 nmol/ml versus 3.00 ± 1.17 nmol/ml and oped endometriosis or tubal blockage, two men
1.87 ± 0.40 mg/dl versus 3.12 ± 0.94 mg/dl respectively, developed epididymitis, four couples chose the other
P <0.001). Thus, current evidence supports a beneficial treatment than the one they were randomized to, one
effect of varicocele repair on oxidative stress. man was suspected of having an intestinal tumour, two
couples separated, and one couple used condoms. In the
Sperm DNA integrity randomized trial by Breznik and colleagues33, no sig-
Beneficial effects of varicocelectomy are also observed nificant differences were found in pregnancy outcomes
when evaluating sperm DNA integrity. A meta-­analysis between the group undergoing varicocele repair and
by Wang et al.4 showed that 240 men with clinical vari­ the control group (13 of 38 versus 22 of 41, respec-
coceles (included from seven prospective and retro­ tively; P >0.05). However, this study is also subject to
spective studies) had significantly higher levels of limitations, including the use of different methods for
sperm DNA damage compared with 176 healthy fertile varicocele repair that did not spare the testicular artery.
volunteers, and that varicocele repair on 177 men with A 2001 Cochrane review reported no significant
clinical varicoceles (included from six prospective and increase in spontaneous pregnancy rates in couples fol-
retrospective studies) significantly decreased sperm lowing varicocele repair compared with those of couples
DNA damage compared with presurgery levels, meas- in whom the male partner’s varicocele was untreated
ured by either SCSA, TUNEL or Comet assay (mean (OR 1.15; 95% CI 0.73–1.83)41. Updates to this review in
difference -3.37%; 95% CI -4.09–-2,65, P <0.0001). 2004 (REF. 42) and 2008 (REF. 43) showed similar results.
Similarly, Smit et al.38 showed decreased sperm DNA In 2011, a meta-analysis by Baazeem et al.31 also failed to
fragmentation after varicocelectomy in their prospective show a significant benefit on natural pregnancy outcome
study on 49 men with clinical varicocele (35.2 ± 13.1% following varicocelectomy with an OR of 2.23 (95% CI
versus 30.2 ± 14.7%, P = 0.019). The included men all 0.86–5.78). However, the latest version of the Cochrane
had a history of >1 year infertility, and 2 years after vari­ review, published in 2012, included two new randomized
cocelectomy 37% of the couples achieved a spontaneous controlled trials making a total number of 10 trials and
pregnancy and 24% achieved pregnancy with assisted 894 men with clinical or subclinical varicocele, and this
reproductive techniques. The mean DNA fragmentation update did report an improvement in spontaneous preg-
index after varicocelectomy was significantly higher in nancy rates after varicocelectomy (OR 1.47; 95% CI 1.05–
couples who did not achieve a pregnancy (P = 0.033) 2.05)44, concluding that varicocelectomy might improve
highlighting the clinical relevance of sperm DNA test- a couple’s chance of pregnancy with a number needed
ing. However, assessment of sperm DNA damage is to treat of seven. Collectively, the Cochrane reviews
not currently recommended in clinical practice, as the have been criticized for their inclusion of men with sub­
data generated using currently available methods for clinical varicocele and normal semen parameters and,
assessing sperm DNA damage do not correlate with as a consequence, a subgroup analysis excluding these
reproductive outcomes39. men was performed in the latest version44. The subgroup
analysis clarified the beneficial effect of vari­cocelectomy
Effects on pregnancy outcomes on pregnancy rates (OR 2.39; 95% CI 1.56–3.66); how-
The most clinically relevant question is whether or ever, the authors note that the results are suggestive
not varicocelectomy improves pregnancy outcomes. rather than conclusive owing to the low quality of evi-
This end point is difficult to assess as follow‑up pro- dence. Another meta-analysis45 reported a pregnancy
tocols differ and because female fertility factors are an rate of 33% (31 of 96) in surgically treated men compared
important potential confounder. This obstacle is illus- with 15.5% (27 of 174) in untreated men, corresponding
trated in a randomized controlled trial by Nieschlag to an OR of 2.87 (95% CI 1.33–6.20). The analysis was
et al.40 from 1998 in which male partners with clinical based on two randomized trials and three observational
varicoceles from 125 infertile couples were randomly studies including infertile men with an abnormal semen
assigned to treatment (ligation or embolization) or analyses and a palpable varicocele. These data are simi-
observation. After 12 months follow‑up duration, the lar to those obtained in the randomized controlled trial
pregnancy rates in each group were similar (29% in by Abdel-Meguid and colleagues35, in which a similar
the vari­cocelectomy group and 25.4% in the observa- odds ratio for achieving a spontaneous pregnancy
tion group). Female partners from couples achieving after varicocelectomy was reported (OR 3.04; 95% CI
a pregnancy were significantly younger than those 1.33–6.95).
from couples not achieving a pregnancy (28.8 years Overall, the available evidence supports a beneficial
versus 31.2 years; P <0.05), highlighting the impor- effect of varicocelectomy on pregnancy outcomes, but
tance of female factors on reproductive outcomes in large sufficiently powered randomized controlled trials
varicocele studies. Data regarding any differences with homogenous patient populations are lacking to
in female age between groups were not reported. substantiate the partly conflicting results.

NATURE REVIEWS | UROLOGY VOLUME 14 | SEPTEMBER 2017 | 527

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Effects on reproductive hormones effect of varicocelectomy on reproductive hormones

The two major functions of the testicles are the produc- are difficult to draw. Interestingly, testicular tissue from
tion of spermatozoa in the seminiferous tubules stimu- infertile men with varicoceles has been shown to contain
lated by the gonadotrophin follicle-stimulating hormone fewer Leydig cells than those of healthy men and signs
(FSH) and production of the male androgen testos­ of Leydig cell apoptosis49. Further studies examining
terone, which is produced by Leydig cells in response the effects of varicocelectomy on Leydig cell function
to the gonadotrophin luteinizing hormone (LH). Thus and including pathology assessment are needed; such
in addition to having an effect on spermatozoa a vari­ studies will also help to further elucidate the association
cocelectomy might also affect the production of testos- between varicocele and reproductive hormone levels.
terone and subsequently the level of LH. Data regarding
the effect of varicocelectomy on reproductive hormones Overall effects of varicocelectomy on fertility
are conflicting. Cayan et al.9 reported that varicocelec- The overall results of clinical varicocele repair to treat
tomy resulted in a statistically significant decrease in subfertility seem to be positive. However, the hetero­
FSH from 15.21 mIU/ml to 10.82 mIU/ml (P = 0.01) and geneity of available studies with regards to patient
an increase in testosterone from 5.63 ng/ml to 8.37 ng/ selection, varicocelectomy method, and reported out-
ml (P = 0.01). However, none of the men included in comes makes comparisons difficult. In addition, many
this study were hypogondal at any time. Hurtado et al.25 studies are inadequately designed and of poor quality.
compared testosterone, LH, and FSH levels in men As a result, further studies are needed to substantiate the
with varicoceles to those in healthy sperm donors and possible beneficial effects of varicocelectomy on semen
showed that serum total testosterone was significantly parameters, sperm function, pregnancy outcomes, and
lower (424 ± 18 ng/dl (mean ± SEM) versus 298 ± 17 ng/ reproductive hormones.
dl, P <0.001) and gonadotrophin levels were signifi-
cantly higher (FSH 10.7 ± 0.3 IU/L versus 17.5 ± 0.2 Who and when to treat
IU/L, P <0.001; LH 5.3 ± 0.2 ng/dl versus 8.1 ± 0.1 ng/dl, Possible progressive nature of varicoceles
P <0.001) in men with varicoceles compared with healthy Epidemiological evidence suggests that the negative
controls. Testosterone levels normalized following var- effect of varicoceles on fertility might be progressive,
icocelectomy, reaching comparable levels to those of as varicoceles have a higher incidence among men with
healthy controls 1–8 months after the operation. Levels secondary infertility compared to men with primary
of FSH and LH initially decreased after varicocelectomy infertility12,13. Likewise, the negative effect of vari­coceles
and were comparable to the levels of the healthy controls on semen parameters has been reported to progress
1–3 months after the procedure, but eventually increased over time50. Such effects are partly supported by a cross-­
and were once again significantly higher than those of sectional study of 7,035 men among whom 15.7% had
healthy donors 6–8 months after varicocelectomy25. a varicocele2. As grade of varicocele increased, a corre-
A retrospective study of 96 men showed no change in sponding decline in semen quality and inhibin B lev-
testosterone levels after varicocelectomy; however, the els, and an increase in FSH and LH levels were noted,
indication for varicocelectomy in these men was infer- whereas testosterone levels did not vary with varicocele
tility, palpable swelling or scrotal pain, making the study grade2. The observed changes might be explained by a
population heterogeneous46. Another retro­spective study varicocele-induced reduction in Sertoli cell function
that evaluated 272 men undergoing vari­cocelectomy for and spermatogenesis resulting in a diminished produc-
infertility or hypogonadism and stratified them accord- tion of inhibin B from the Sertoli cells and an increase
ing to age into three groups (≤30 years, 30–39 years, in FSH from the pituitary gland as a response to the
and ≥40 years) demonstrated significant increases in affected spermatozoa production and lower inhibin
serum testosterone in all groups after vari­cocelectomy B levels. Furthermore, the increase in LH might indicate
(73 ± 32 ng/dl, 59 ± 25 ng/dl, and 93 ± 25 ng/dl, respec- a subtle Leydig cell dysfunction although not affecting
tively; P = 0.03, P = 0.02, P = 0.001, respectively). testosterone levels. Although the study does not provide
Stratifying men according to preoperative testosterone definitive evidence for the progressive nature of vari­
levels revealed that only patients with a testosterone level coceles, the observed changes are likely to be attributa-
<400 ng/dl at baseline showed significant improvements ble to the varicocele and not to other factors separating
in testosterone levels after varicocelectomy. A study by fertile and infertile men, as participants in the study
Tanrikut et al.47 compared presurgical and postsurgical came from a general population not selected on the basis
testosterone levels in a group of 200 men with clinical of fertility status. Thus, the deleterious effects of vari­
varicoceles that were primarily corrected on the basis of coceles do seem to increase with increasing vari­cocele
infertility. Postoperative serum testosterone was signifi- grade, which corresponds to an association between
cantly increased compared with baseline levels (454 ng/ high varicocele grade and high levels of ROS24. However,
dl versus 358 ng/dl; P <0.001). A recent review by Dabaja not all studies support this hypothesis of progressive
and Goldstein48 concluded that microsurgical varicoce- detrimental effects. In fact, one small (n = 57) study
lectomy has beneficial effects on testosterone levels and showed that semen parameters declined more in men
that the effect was more pronounced in patients with without varicocele than in those with varicocele over
baseline testosterone <300 ng/dl. However, the some- an 8‑year period, suggesting that possible detrimental
what conflicting results and retrospective design of most effects happen early in life51. Nevertheless, prophylactic
studies mean that definitive conclusions regarding the treatment of varicoceles has been suggested to counteract

528 | SEPTEMBER 2017 | VOLUME 14 www.nature.com/nrurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

possible progressive effects. However, this approach With regards to semen quality, only men with affected
does not seem reasonable in the majority of men with semen parameters should be offered vari­cocelectomy.
undocumented infertility. This consideration reflects that the majority of studies
of varicocele and infertility include men with abnormal
Patient selection and timing of treatment semen parameters, providing evidence to support sur-
In the absence of other symptoms (for example, pain) gery for this group of patients. In addition, men with
only men from couples with documented infertility normal semen parameters tend to glean little benefit
should be considered for varicocele repair, as the major- from varicocele repair in terms of pregnancy outcome,
ity of men with a varicocele are fertile, highlighting the as demonstrated by the Cochrane reviews44. In men
fact that prophylactic varicocelectomy is inappropriate with azoospermia, a thorough evaluation of the patient
in most cases. The criteria for selecting men to undergo including a full medical history, physical examination,
varicocelectomy remain a matter of debate. hormonal evaluation, and genetic testing (karyo­type and
With regards to patient selection based on varicocele azoospermia factor (AZF) microdeletions) is mandatory.
grade, the benefits of repairing clinically palpable (grade These investigations help to distinguish nonobstructive
1–3) varicoceles are well documented, as demonstrated azoospermia (NOA) from obstructive azoospermia
by the previously discussed studies evaluating the effect (OA), the latter of which could benefit from reconstruc-
of varicocele repair on pregnancy outcomes. By contrast, tive surgery. This work‑up also helps to exclude the pos-
no conclusive evidence supports repair of subclinical sibility that the varicocele is an incidental finding in a
varicoceles. Dhabuwala et al.52 reported improvement patient with azoospermia due to another cause in which
in semen quality after repair of subclinical varicoceles retrieval of spermatozoa is close to impossible, for exam-
resulting in pregnancies of the partner in eight of 16 ple AZFa/b microdeletions56. If such causes are excluded,
treated men. The number of achieved pregnancies was a varicocelectomy can be performed in men with NOA,
comparable to that of men treated for clinical varicocele resulting in beneficial effects on sperm retrieval rates
(18 of 38, P = 0.86). In a randomized controlled trial by (SRR), as illustrated in a meta-analysis showing an
Yamamoto and colleagues53, 85 men with subclinical var- increased SRR in men with NOA who underwent vari-
icocele received either high ligation or no treatment. No cocelectomy compared with men with NOA who did not
difference was seen in pregnancy rates (6.7% versus 10%, undergo varicocelectomy (OR 2.65; 95% CI 1.69–4.14;
P = 0.578), although those who underwent high ligation P <0.001)57. However, final conclusions regarding the
demonstrated significant increases in sperm density use of varicocelectomy in men with NOA require results
and total motile sperm (P <0.006 and P <0.008, respec- from future randomized trials.
tively). Grasso et al.54 randomized 68 men with a left- Age becomes a matter for discussion when var-
sided subclinical varicocele to either high ligation or no icocelectomy is considered in adolescent males.
treatment and showed no improvement in semen quality An optimal age range for varicocele repair has not been
or pregnancy outcomes in either group. In the study by suggested, nor is there an advised lower or upper age
Yamamoto and co-workers53, the men were diagnosed limit. Although prophylactic varicocelectomy is gen-
using scrotal thermography, whereas Dhabuwala et al.52 erally not recommended in adolescent patients, owing
and Grasso et al.54 used Doppler ultrasonography, mak- to the unpredictable nature of varicocele-induced
ing direct comparisons between the studies difficult. infertility, the procedure might be of value in carefully
In a 2016 meta-analysis, Kim et al.55 compared vari­ selected adolescents with progressive deleterious testic-
cocelectomy to no treatment in 548 men with sub­clinical ular effects. Mori et al.58 evaluated 360 adolescent males
varicocele. A statistically significant improvement in (aged 14–18 years) and diagnosed a grade 2 or grade 3
forward-­progressive motility was seen in sperm from the varicocele in 27.8% of patients, associated with consider-
treatment group (weighted mean difference 3.94; 95% CI able testicular asymmetry compared with boys who had
1.24–6.65), but no differences were noted in sperm con- a grade 1 varicocele or no varicocele. This observation
centration (0.92; 95% CI -0.36–2.19), total motility (3.83; demonstrates testicular growth retardation, presuma-
95% CI -4.07–11.73), sperm morphology (0.61; 95%  bly caused by the presence of an adolescent varicocele.
CI -1.29–2.50), or pregnancy outcomes (OR 1.29; 95% CI Although sperm progressive motility and concentration
0.99–1.67) between groups. This meta-analysis is biased were lower in the boys with a grade 2 or 3 varicocele, lev-
by the heterogeneity of studies, which included men with els were still within the normal reference limits defined
unilateral subclinical varicocele, those with combined by the WHO. Other studies have also demonstrated var-
clinical left-sided varicocele, and men with right-sided icocele-induced negative effects on semen quality and
subclinical varicocele. In addition, the subclinical vari- testicular volume in adolescents17,59. In addition, vari-
coceles are diagnosed using different methods, different cocelectomy has been shown to improve semen quality
surgical techniques are used to treat, and the control and enable testicular growth catch‑up in this patient
groups consist of men having no treatment at all and men group60,61. Varicocelectomy is, therefore, considered an
being treated with clomiphene citrate. These confound- option for adolescents presenting with a clinical vari-
ers highlight the lack of standardization, which makes cocele and ipsilateral impaired testicular growth and/
drawing comparisons difficult. Repair of subclinical vari­ or affected semen quality. However, testicular growth
coceles seems to have little effect on pregnancy outcomes; can happen in a differential manner: Kolon et al.62
as a consequence, varicocelectomy should only be offered studied 161 boys with unilateral left-sided clinical vari­
to men presenting with a clinically palpable varicocele. cocele, 71 of whom had undergone ≥3 ultrasonography

NATURE REVIEWS | UROLOGY VOLUME 14 | SEPTEMBER 2017 | 529

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

examinations during follow‑up monitoring; 54% of 33.20% 30.07% for high ligation, inguinal vari­
these boys initially had a testicular volume differential cocelectomy, embolization, and laparascopic approaches,
of >15%. After 2 years, this proportion dropped to 15%. respectively; P = 0.001). Ding et al.65 restricted their
The observed growth retardation in adolescent boys meta-analysis to include only randomized controlled
with varicoceles might, therefore, not be caused by the trials in infertile men with a clinical varicocele,
varicocele itself. In addition, some adolescents with affected semen parameters, and with a partner with-
varicoceles experience an improvement in their semen out female-factor infertility. Based on analyses of 1,015
quality without treatment63. Finally, only 20% of adoles- men undergoing either high ligation, laparoscopic, or
cents with a varicocele will experience fertility problems microsurgical varicocelectomy, microsurgery was supe-
later in life3; as a consequence, the risk of overtreatment rior to high ligation and similar to laparoscopy with
is considerable in this patient group. Thus, progression regards to pregnancy rate (OR microsurgery versus
of the deleterious testicular effects should be observed high ligation 1.63; 95% CI 1.19–2.23 and OR micro-
before performing varicocelectomy. However, before surgery versus laparo­scopic 1.37; 95% CI 0.84–2.24).
concrete recommendations on this matter can be made, In addition, microsurgery was found to be superior to
randomized studies are required on varicocele repair in both high ligation and laparoscopy with regards to vari­
adolescents examining sperm function and/or long-term cocele recurrence (OR microsurgery versus high liga-
effects on fertility. tion 0.13; 95% CI 0.07–0.25 and OR microsurgery versus
Finally, patient selection could include consid- laparoscopic 0.12; 95% CI 0.06–0.32) and incidence of
eration of reported negative predictive factors for post­operative hydrocele (OR microsurgery versus high
vari­cocelectomy success, including low preoperative tes- ligation 0.09; 95% CI 0.03–0.30 and OR microsurgery
tosterone, low-volume testes, and elevated FSH levels17. versus laparoscopic 0.05; 95% CI 0.01–0.36). Based on
However, owing to a lack of good quality evidence, clear the observed differences in the meta-analyses by Cayan
recommendations for patient selection based on these and colleagues64 and Ding and co-workers65, micro­
parameters are not possible. surgical varicocelectomy seems to be the method of
choice if microsurgical expertise is available.
Techniques for varicocelectomy
Several methods for varicocele repair have been sug- Surgical techniques
gested and studied including open surgical, lapara­ Various different surgical techniques have been used for
scopic, microsurgical, and percutaneous radiological varicocele repair over the past decades. An open retro-
approaches, all of which are associated with different peritoneal high ligation is performed through an inci-
risks of recurrence and possible complications (TABLE 2). sion medial to the anterior superior iliac spine where
Recurrence-free and complication rates reported in the internal spermatic vein is identified as it exits the
Table 2 are largely from a meta-analysis by Cayan et al.64 inguinal canal. Ligation of the vein is performed while
evaluating the different approaches in 4,473 infertile the internal spermatic artery is preserved66. The exter-
men with a clinical unilateral or bilateral varicocele nal spermatic and cremasteric veins are not accessible
and abnormal semen parameters, excluding men with through this approach. A laparoscopic varicocelectomy
azoospermia. The meta-analysis also examined is performed as a retroperitoneal approach using lap-
post­operative natural pregnancy rates and found a aroscopic magnification making identification of the
significantly higher pregnancy rate following micro­ internal spermatic artery and lymphatics easier17. When
surgical vari­cocelectomy (41.97%) compared with all performing an inguinal varicocelectomy the spermatic
other vari­cocelectomy techniques (37.69%, 36.00%, cord is accessed through an incision over the inguinal

Table 2 | Methods of varicocele repair: techniques, effects, and complications

Technique Risk of recurrence Risk of Rarely reported complications
hydrocele
Open retroperitoneal high 14.97% 8.24% None reported in available literature
ligation
Laparoscopic surgery 4.30% 2.84% Literature reports injury to the testicular artery and lymphatic vessels,
scrotal subcutaneous emphysema, inferior epigastric artery injury, bleeding,
epididymitis, severe scrotal pain, wound infection, pulmonary embolism,
intestinal injury, nerve damage, peritonitis64,77,78
Inguinal surgery 2.63% 7.30% Literature reports wound hematoma, wound separation and postoperative
pain owing to incision of the external oblique fascia17,79
Microsurgical inguinal or 1.05% 0.44% Literature reports wound infection, haematoma, epididymal pain,
subinguinal surgery ecchymosis and injury to the testicular artery and lymphatics64,80,81
Embolization Risk of recurrence Technical failure Literature reports thrombophlebitis, testicular atrophy, haematoma,
infection, allergy to contrast media, venous perforation, coil migration,
12.7% 13.05% flank pain64,69,82
*Recurrence and hydrocele risk data from Cayan, S., Shavakhabov, S. & Kadioglu, A. Treatment of palpable varicocele in infertile men: a meta-analysis to define the
best technique. J. Androl. 30, 33–40 (2009)64. Rare complications as reported in the published literature.

530 | SEPTEMBER 2017 | VOLUME 14 www.nature.com/nrurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

canal. Both the internal and external spermatic veins are not reported, these data suggest the benefit of bi­­
can be identified and ligated, while sparing the testic- lateral repair whenever a bilateral clinical vari­cocele is
ular artery and lymphatics67. Finally, for inguinal or present. However, evidence regarding bilateral or
sub­inguinal microsurgery an incision is made over the unilateral repair of bilateral clinical varicoceles is limited.
inguinal canal or just below the external inguinal ring, Conflicting results have been reported regarding
respectively. In the latter, the external oblique fascia is bilateral repair in patients with a clinical left-sided var-
spared for incision resulting in less postoperative pain. icocele in combination with a subclinical right-sided
In both approaches the spermatic cord is mobilized. varicocele. In a prospective randomized trial, Elbendary
Using an operating microscope the branches of the et al.73 found significantly better improvement of sperm
internal spermatic vein can be identified and ligated in concentration (15 ± 4.3 to 23 ± 4.9 × 106 sperm/ml versus
addition to external cremasteric veins while preserving 15.1 ± 4.1 to 21 ± 4.2 × 106 sperm/ml, P = 0.008), sperm
the vas deferens, testicular artery, and lymphatics17,28. motility (36.7 ± 8.8 to 50.5 ± 10.1% versus 37.8 ± 8.7 to
40.5 ± 9.2%, P <0.001), progressive motility (17.6 ± 4.4 to
Radiological techniques 26.1 ± 5.2% versus 17.9 ± 4.5 to 19.6 ± 4.3%, P <0.001),
Radiological varicocelectomy implies occlusion by and pregnancy rates (61.6% versus 31.9%, P = 0.04)
image-guided placement of a coil (embolization) or after bilateral repair (n = 73) than after unilateral left-
injection of a sclerosing agent (sclerotherapy)68,69.These sided repair (n = 72) using an inguinal open approach.
approaches have been criticized for their risk of a failed However, the researchers used a 2 mm cut-off thresh-
ligation, which ranges from 4 to 27%65,68. However, sub- old for diagnosing a subclinical varicocele as opposed
sequent studies have shown the radiological approach to to the more commonly used 3 mm cut-off. To account
be efficient and minimally invasive with low recurrence for this discrepancy, a subgroup analysis was performed,
rates (5.9%)69. Comparative studies have reported con- which included 37 of the participants who complied
flicting results, both in favour of radiological interven- with the 3 mm cut-off. This analysis revealed similar
tion70 and in favour of surgical intervention71, but the results in favour of bilateral repair. Conversely, another
meta-analysis by Cayan et al.64 suggested that surgical prospective randomized study that included 51 men in
intervention was better than embolization with regards the bilateral repair group and 53 men in the left-only
to spontaneous pregnancy rates (41.97% for micro- repair group showed no differences in postoperative
surgery versus 33.20% for embolization, P = 0.001) semen parameters or pregnancy outcomes between the
and recurrence rates (1.05% for microsurgery versus groups74. Based on the available evidence, final conclu-
12.70% for embolization, P = 0.001)64. However, a lack of sions regarding bilateral varicocele repair in men with
high-quality studies comparing radiological intervention a clinical left-sided varicocele in combination with a
to surgical intervention means that drawing final con- subclinical right-sided varicocele cannot be made and,
clusions as to whether radiological intervention should consequently, bilateral repair in this situation is not
be used as a first-line treatment is difficult. Owing supported.
to the good results of microsurgical varicocelectomy in
the existing literature, this technique seems preferable if Varicocelectomy and assisted reproduction
microsurgical expertise is available. Improvements in assisted reproductive techniques (ART)
have led to questions regarding whether vari­cocelectomy
Bilateral repair should be performed, especially in couples in whom a
The benefit of a bilateral varicocele repair is the subject female factor limits the chance of post­operative suc-
of much debate. Cases of bilateral clinical varicocele and cessful natural pregnancy or intra­uterine insemination.
cases of a left-sided clinical varicocele in combination In these cases, in vitro fertilisation (IVF) with or with-
with a subclinical right-sided varicocele are of particular out intracytoplasmic sperm injection (ICSI) could be
clinical relevance as these conditions are the most fre- more relevant than varicocele repair to achieve a natural
quently occurring bilateral cases of varicoceles in which pregnancy. However, most cost-effectiveness analyses
at least one side is a clinical varicocele. favour varicocelectomy over ART17, notwithstanding
Scherr and Goldstein prospectively compared the the possible psychological benefit that some couples
effect of unilateral or bilateral microsurgical vari­ find in achieving a natural pregnancy. Furthermore,
cocelectomy in 91 men with a grade 2–3 left-sided varicocelectomy has been shown to improve both SRR
varicocele and a simultaneous grade 1 right-sided vari­ and pregnancy outcomes of ART in men with varicocele
cocele72. Bilateral repair was performed in 65 men, and azoospermia or oligozoospermia75. These data are
whereas the rest had unilateral left repair. Bilateral supported by a meta-ana­lysis evaluating potential ben-
repair was associated with a significantly greater effect efits of varicocelectomy on ART outcomes, which was
on postoperative sperm concentration (bilateral repair: based on four retrospective studies including nonazoo-
23.8 ± 29.5 to 48.6 ± 61.3 × 106 sperm/ml (157.6% change) spermic men with a clinical vari­cocele undergoing ICSI76.
versus unilateral repair: 41.1 ± 40.9 to 59.5 ± 66.7 × 106 Significantly higher pregnancy rates (OR 1.59; 95% CI
sperm/ml (44.8% change), P <0.05) and motile 1.19–2.12) and live birth rates (OR 2.17; 95% CI 1.55–
sperm concentration (bilateral repair: 12.1 ± 17.7 to 3.06) were achieved in the group of men who had vari-
23.7 ± 31.8 × 106 sperm/ml (95.8% change) versus uni- cocelectomy before ICSI compared with those who did
lateral repair: 19.5 ± 21.4 to 27.8 ± 34.8 × 106 sperm/ml not have a varicocelectomy before ICSI76. Although larger
(42.6% change), P <0.05). Although pregnancy outcomes trials with a prospective randomized design evaluating

NATURE REVIEWS | UROLOGY VOLUME 14 | SEPTEMBER 2017 | 531

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Clinical varicocele the effect of vari­cocelectomy on ART outcomes are

(grade I–III)? needed, the available data suggest that vari­cocelectomy
will continue to be important in the treatment of
Yes No varicocele-induced infertility.

Clinical recommendations
Documented couples
infertility (>1 year)? No Surgery not indicated Treatment of clinical varicoceles in adult men should
only be used to improve fertility in cases of documented
infertility of the couple and affected semen parameters
Yes No (FIG. 3). In cases of clinical varicoceles in adolescent men,
a microsurgical varicocelectomy should be offered for
Affected semen fertility preservation if progressive testicular growth
parameters?
retardation and/or impaired semen quality is observed.

Yes Bilateral clinical

Conclusions
No Varicoceles can have deleterious effects on testicular
function, impairing semen quality, sperm function and
Unilateral? Bilateral microsurgical
varicocelectomy pregnancy outcomes for some men. Current evidence
No
supports varicocele repair in men from couples with
Yes documented infertility presenting with a clinical vari-
Clinical and subclinical cocele and affected semen parameters. Evidence does
not support the correction of subclinical vari­coceles.
Unilateral microsurgical
varicocelectomy Consequently, bilateral varicocele repair should only be
Microsurgical varicocelectomy
performed if the varicocele is palpable on both sides.
of clinical side If microsurgical expertise is available, the optimal var-
icocelectomy method is a microsurgical procedure.
Figure 3 | Treatment algorithm for adult infertile men with varicoceles. Evidence Adolescents presenting with a clinical varicocele should
Nature
does not support the correction of subclinical varicoceles. Bilateral Reviewsrepair
varicocele | Urology
be offered varicocelectomy for fertility preservation
should only be performed if the varicocele is palpable on both sides. If microsurgical only if a progressive testicular growth retardation and/
expertise is available, microsurgery is the optimal varicocelectomy method. or impaired semen quality is observed.

1. Dubin, L. & Amelar, R. D. Etiologic factors in 1294 13. Jarow, J. P., Coburn, M. & Sigman, M. Incidence of stress biomarkers and hormonal profile in human
consecutive cases of male infertility. Fertil. Steril. 22, varicoceles in men with primary and secondary patients undergoing varicocelectomy. Int. J. Androl.
469–474 (1971). infertility. Urology 47, 73–76 (1996). 30, 519–530 (2007).
2. Damsgaard, J. et al. Varicocele is associated with 14. Akbay, E., Cayan, S., Doruk, E., Duce, M. N. & 26. Romeo, C. & Santoro, G. Free radicals in adolescent
impaired semen quality and reproductive hormone Bozlu, M. The prevalence of varicocele and varicocele- varicocele testis. Oxid. Med. Cell. Longev. 2014,
levels: a study of 7035 healthy young men from six related testicular atrophy in Turkish children and 912878 (2014).
European countries. Eur. Urol. 70, 1019–1029 adolescents. BJU Int. 86, 490–493 (2000). 27. Cho, C. L., Esteves, S. C. & Agarwal, A. Novel insights
(2016). 15. Saypol, D. C. Varicocele. J. Androl. 2, 61–71 (1981). into the pathophysiology of varicocele and its
3. [No authors listed.] The influence of varicocele on 16. de Castro, M. P. & Mastrorocco, D. A. Reproductive association with reactive oxygen species and sperm
parameters of fertility in a large group of men history and semen analysis in prevasectomy fertile DNA fragmentation. Asian J. Androl. 18, 186–193
presenting to infertility clinics. World Health men with and without varicocele. J. Androl 5, 17–20 (2016).
Organization. Fertil. Steril. 57, 1289–1293 (1992). (1984). 28. Masson, P. & Brannigan, R. E. The varicocele. Urol.
4. Wang, Y. J., Zhang, R. Q., Lin, Y. J., Zhang, R. G. & 17. Hamada, A., Esteves, S. C. & Agarwal, A. Varicocele Clin. North Am. 41, 129–144 (2014).
Zhang, W. L. Relationship between varicocele and and Male Infertility: Current Concepts, Controversies 29. Pastuszak, A. W. & Wang, R. Varicocele and testicular
sperm DNA damage and the effect of varicocele and Consensus (Springer International Publishing, function. Asian J. Androl. 17, 659–667 (2015).
repair: a meta-analysis. Reprod. Biomed. Online 25, 2016). 30. Bonde, J. P. et al. Relation between semen quality and
307–314 (2012). 18. Naughton, C. K., Nangia, A. K. & Agarwal, A. fertility: a population-based study of 430 first-
5. Vigil, P., Wohler, C., Bustos-Obregon, E., Comhaire, F. Pathophysiology of varicoceles in male infertility. Hum. pregnancy planners. Lancet 352, 1172–1177 (1998).
& Morales, P. Assessment of sperm function in fertile Reprod. Update 7, 473–481 (2001). 31. Baazeem, A. et al. Varicocele and male factor infertility
and infertile men. Andrologia 26, 55–60 (1994). 19. McClure, R. D. & Hricak, H. Scrotal ultrasound in the treatment: a new meta-analysis and review of the role
6. Abdelrahim, F. et al. Testicular morphology and infertile man: detection of subclinical unilateral and of varicocele repair. Eur. Urol. 60, 796–808 (2011).
function in varicocele patients: pre-operative and post- bilateral varicoceles. J. Urol. 135, 711–715 (1986). 32. Agarwal, A. et al. Efficacy of varicocelectomy in
operative histopathology. Br. J. Urol. 72, 643–647 20. Gat, Y., Bachar, G. N., Zukerman, Z., Belenky, A. & improving semen parameters: new meta-analytical
(1993). Gornish, M. Varicocele: a bilateral disease. Fertil. approach. Urology 70, 532–538 (2007).
7. Dubin, L. & Hotchkiss, R. S. Testis biopsy in Steril. 81, 424–429 (2004). 33. Breznik, R., Vlaisavljevic, V. & Borko, E. Treatment of
subfertile men with varicocele. Fertil. Steril. 20, 21. Gat, Y. et al. Right varicocele and hypoxia, crucial varicocele and male fertility. Arch. Androl. 30,
51–57 (1969). factors in male infertility: fluid mechanics analysis of 157–160 (1993).
8. Etriby, A., Girgis, S. M., Hefnawy, H. & Ibrahim, A. A. the impaired testicular drainage system. Reprod. 34. Mansour Ghanaie, M. et al. Effects of varicocele repair
Testicular changes in subfertile males with varicocele. Biomed. Online 13, 510–515 (2006). on spontaneous first trimester miscarriage: a
Fertil. Steril. 18, 666–671 (1967). 22. Agarwal, A., Prabakaran, S. & Allamaneni, S. S. randomized clinical trial. Urol. J. 9, 505–513 (2012).
9. Cayan, S. et al. The effect of microsurgical Relationship between oxidative stress, varicocele and 35. Abdel-Meguid, T. A., Al‑Sayyad, A., Tayib, A. &
varicocelectomy on serum follicle stimulating infertility: a meta-analysis. Reprod. Biomed. Online Farsi, H. M. Does varicocele repair improve male
hormone, testosterone and free testosterone levels in 12, 630–633 (2006). infertility? An evidence-based perspective from a
infertile men with varicocele. BJU Int. 84, 1046–1049 23. Krzysciak, W. & Kozka, M. Generation of reactive randomized, controlled trial. Eur. Urol. 59, 455–461
(1999). oxygen species by a sufficient, insufficient and varicose (2011).
10. Agarwal, A., Hamada, A. & Esteves, S. C. Insight into vein wall. Acta Biochim. Pol. 58, 89–94 (2011). 36. Sakamoto, Y., Ishikawa, T., Kondo, Y., Yamaguchi, K. &
oxidative stress in varicocele-associated male 24. Allamaneni, S. S., Naughton, C. K., Sharma, R. K., Fujisawa, M. The assessment of oxidative stress in
infertility: part 1. Nat. Rev. Urol. 9, 678–690 Thomas, A. J. Jr & Agarwal, A. Increased seminal infertile patients with varicocele. BJU Int. 101,
(2012). reactive oxygen species levels in patients with 1547–1552 (2008).
11. Tulloch, W. S. Consideration of sterility; subfertility in varicoceles correlate with varicocele grade but not 37. Chen, S. S., Huang, W. J., Chang, L. S. & Wei, Y. H.
the male. Edinb. Med. J. 59, 29–34 (1952). with testis size. Fertil. Steril. 82, 1684–1686 (2004). Attenuation of oxidative stress after varicocelectomy
12. Gorelick, J. I. & Goldstein, M. Loss of fertility in men 25. Hurtado de Catalfo, G. E., Ranieri-Casilla, A., in subfertile patients with varicocele. J. Urol. 179,
with varicocele. Fertil. Steril. 59, 613–616 (1993). Marra, F. A., de Alaniz, M. J. & Marra, C. A. Oxidative 639–642 (2008).

532 | SEPTEMBER 2017 | VOLUME 14 www.nature.com/nrurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

38. Smit, M. et al. Decreased sperm DNA fragmentation 54. Grasso, M. et al. Low-grade left varicocele in patients 72. Scherr, D. & Goldstein, M. Comparison of bilateral
after surgical varicocelectomy is associated with over 30 years old:the effect of spermatic vein ligation versus unilateral varicocelectomy in men with
increased pregnancy rate. J. Urol. 189, S146–S150 on fertility. BJU Int. 85, 305–307 (2000). palpable bilateral varicoceles. J. Urol. 162, 85–88
(2013). 55. Kim, H. J. et al. Clinical significance of subclinical (1999).
39. Practice Committee of the American Society for varicocelectomy in male infertility: systematic review 73. Elbendary, M. A. & Elbadry, A. M. Right subclinical
Reproductive Medicine. The clinical utility of sperm and meta-analysis. Andrologia 48, 654–661 (2016). varicocele: how to manage in infertile patients with
DNA integrity testing: a guideline. Fertil. Steril. 99, 56. Jungwirth, A. et al. EAU Guidelines on Male Infertility clinical left varicocele? Fertil. Steril. 92, 2050–2053
673–677 (2013). (European Association of Urology, 2016). (2009).
40. Nieschlag, E., Hertle, L., Fischedick, A., Abshagen, K. 57. Esteves, S. C., Miyaoka, R., Roque, M. & Agarwal, A. 74. Zheng, Y. Q. et al. Efficacy of bilateral and left
& Behre, H. M. Update on treatment of varicocele: Outcome of varicocele repair in men with varicocelectomy in infertile men with left clinical and
counselling as effective as occlusion of the vena nonobstructive azoospermia: systematic review and right subclinical varicoceles: a comparative study.
spermatica. Hum. Reprod. 13, 2147–2150 (1998). meta-analysis. Asian J. Androl. 18, 246–253 (2016). Urology 73, 1236–1240 (2009).
41. Evers, J. L., Collins, J. A. & Vandekerckhove, P. Surgery 58. Mori, M. M., Bertolla, R. P., Fraietta, R., Ortiz, V. & 75. Kirby, E. W., Wiener, L. E., Rajanahally, S., Crowell, K.
or embolisation for varicocele in subfertile men. Cedenho, A. P. Does varicocele grade determine & Coward, R. M. Undergoing varicocele repair before
Cochrane Database Syst. Rev. 1, CD000479 (2001). extent of alteration to spermatogenesis in assisted reproduction improves pregnancy rate and
42. Evers, J. L. & Collins, J. A. Surgery or embolisation for adolescents? Fertil. Steril. 90, 1769–1773 (2008). live birth rate in azoospermic and oligospermic men
varicocele in subfertile men. Cochrane Database Syst. 59. Bong, G. W. & Koo, H. P. The adolescent varicocele: to with a varicocele: a systematic review and meta-
Rev. 3, CD000479 (2004). treat or not to treat. Urol. Clin. North Am. 31, analysis. Fertil. Steril. 106, 1338–1343 (2016).
43. Evers, J. H., Collins, J. & Clarke, J. Surgery or 509–515 (2004). 76. Esteves, S. C., Roque, M. & Agarwal, A. Outcome of
embolisation for varicoceles in subfertile men. 60. Laven, J. S. et al. Effects of varicocele treatment in assisted reproductive technology in men with treated
Cochrane Database Syst. Rev. 3, CD000479 (2008). adolescents: a randomized study. Fertil. Steril. 58, and untreated varicocele: systematic review and
44. Kroese, A. C., de Lange, N. M., Collins, J. & Evers, J. L. 756–762 (1992). meta-analysis. Asian J. Androl. 18, 254–258
Surgery or embolization for varicoceles in subfertile 61. Li, F. et al. Effect of varicocelectomy on testicular (2016).
men. Cochrane Database Syst. Rev. 10, CD000479 volume in children and adolescents: a meta-analysis. 77. Miersch, W. D., Schoeneich, G., Winter, P. &
(2012). Urology 79, 1340–1345 (2012). Buszello, H. Laparoscopic varicocelectomy: indication,
45. Marmar, J. L. et al. Reassessing the value of 62. Kolon, T. F. et al. Transient asynchronous testicular technique and surgical results. Br. J. Urol. 76,
varicocelectomy as a treatment for male subfertility growth in adolescent males with a varicocele. J. Urol. 636–638 (1995).
with a new meta-analysis. Fertil. Steril. 88, 639–648 180, 1111–1114 (2008). 78. Tan, S. M., Ng, F. C., Ravintharan, T., Lim, P. H. &
(2007). 63. Chu, D. I. et al. The natural history of semen Chng, H. C. Laparoscopic varicocelectomy: technique
46. Resorlu, B., Kara, C., Sahin, E. & Unsal, A. The parameters in untreated asymptomatic adolescent and results. Br. J. Urol. 75, 523–528 (1995).
significance of age on success of surgery for patients varicocele patients: a retrospective cohort study. 79. Ross, L. S. & Ruppman, N. Varicocele vein ligation in
with varicocele. Int. Urol. Nephrol. 42, 351–356 J. Pediatr. Urol. 13, 77.e1–77.e5 (2017). 565 patients under local anesthesia: a long-term
(2010). 64. Cayan, S., Shavakhabov, S. & Kadioglu, A. Treatment review of technique, results and complications in light
47. Tanrikut, C. et al. Varicocele as a risk factor for of palpable varicocele in infertile men: a meta-analysis of proposed management by laparoscopy. J. Urol.
androgen deficiency and effect of repair. BJU Int. 108, to define the best technique. J. Androl. 30, 33–40 149, 1361–1363 (1993).
1480–1484 (2011). (2009). 80. Jungwirth, A. et al. Clinical outcome of microsurgical
48. Dabaja, A. A. & Goldstein, M. When is a varicocele 65. Ding, H. et al. Open non-microsurgical, laparoscopic subinguinal varicocelectomy in infertile men.
repair indicated: the dilemma of hypogonadism and or open microsurgical varicocelectomy for male Andrologia 33, 71–74 (2001).
erectile dysfunction? Asian J. Androl. 18, 213–216 infertility: a meta-analysis of randomized controlled 81. Kumar, R. & Gupta, N. P. Subinguinal microsurgical
(2016). trials. BJU Int. 110, 1536–1542 (2012). varicocelectomy: evaluation of the results. Urol. Int.
49. Benoff, S., Marmar, J. L. & Hurley, I. R. Molecular and 66. Palomo, A. Radical cure of varicocele by a new 71, 368–372 (2003).
other predictors for infertility in patients with technique; preliminary report. J. Urol. 61, 604–607 82. Lenk, S., Fahlenkamp, D., Gliech, V. & Lindeke, A.
varicoceles. Front. Biosci. (Landmark Ed.) 14, (1949). Comparison of different methods of treating
3641–3672 (2009). 67. Ivanissevich, O. Left varicocele due to reflux; varicocele. J. Androl. 15 (Suppl.), 34S–37S (1994).
50. Chehval, M. J. & Purcell, M. H. Deterioration of semen experience with 4,470 operative cases in forty-two
parameters over time in men with untreated years. J. Int. Coll. Surg. 34, 742–755 (1960). Author contributions
varicocele: evidence of progressive testicular damage. 68. Will, M. A. et al. The great debate: varicocele C. F. S. J. researched data for the article and wrote the man-
Fertil. Steril. 57, 174–177 (1992). treatment and impact on fertility. Fertil. Steril. 95, uscript. M. F. and C. F. S. J. made substantial contributions to
51. Lund, L. & Larsen, S. B. A follow‑up study of semen 841–852 (2011). discussion of content. All authors reviewed and edited the
quality and fertility in men with varicocele testis and 69. Crestani, A. et al. Antegrade scrotal sclerotherapy of manuscript before submission.
in control subjects. Br. J. Urol. 82, 682–686 internal spermatic veins for varicocele treatment:
(1998). technique, complications, and results. Asian J. Androl. Competing interests statement
52. Dhabuwala, C. B., Hamid, S. & Moghissi, K. S. Clinical 18, 292–295 (2016). J.M.D. receives grant funding from Blue Cross Blue Shield of
versus subclinical varicocele: improvement in fertility 70. Sautter, T., Sulser, T., Suter, S., Gretener, H. & Hauri, D. Michigan for his role in the Michigan Value Collaborative and
after varicocelectomy. Fertil. Steril. 57, 854–857 Treatment of varicocele: a prospective randomized Michigan Urological Surgery Improvement Collaborative and
(1992). comparison of laparoscopy versus antegrade owns common stock in Lipocine. D.A.O is a consultant for
53. Yamamoto, M., Hibi, H., Hirata, Y., Miyake, K. & sclerotherapy. Eur. Urol. 41, 398–400 (2002). Coloplast and Pfizer and receives research funding from
Ishigaki, T. Effect of varicocelectomy on sperm 71. Yavetz, H. et al. Efficacy of varicocele embolization Endo. J.S. is a consultant for Astellas, Neotract, and Pfizer,
parameters and pregnancy rate in patients with versus ligation of the left internal spermatic vein for and is a shareholder of Multicept. M.F. is a consultant and
subclinical varicocele: a randomized prospective improvement of sperm quality. Int. J. Androl. 15, speaker for Astellas. C.F.S.J and P.B.Ø. declare no competing
controlled study. J. Urol. 155, 1636–1638 (1996). 338–344 (1992). interests.

NATURE REVIEWS | UROLOGY VOLUME 14 | SEPTEMBER 2017 | 533

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.