World J Urol (2011) 29:645–650
DOI 10.1007/s00345-011-0701-4
TOPIC PAPER
Color Doppler ultrasound imaging in varicoceles: is the venous
diameter sufficient for predicting clinical and subclinical
varicocele?
A. Pilatz • B. Altinkilic • E. Köhler •
M. Marconi • W. Weidner
Received: 27 February 2011 / Accepted: 10 May 2011 / Published online: 24 May 2011
Ó Springer-Verlag 2011
Abstract
Purpose Investigating the diagnostic value of color
Doppler ultrasound for defining the varicocele grade
according to WHO criteria.
Methods A total of 217 men (129 with clinical varicocele
and 88 without clinical varicocele) were investigated by
physical examination and color Doppler ultrasound and
categorized according to WHO varicocele criteria (0,
subclinical, I, II, and III). Diameter and reflux of the largest
vein in the pampiniform plexus were measured bilaterally
with the patient in the supine position in rest and during the
Valsalva maneuver. To assess the possibility of differen-
tiating varicocele grade by venous diameter, optimal cut-point
values were determined by receiver-operator characteristic
(ROC) analysis.
Results With increased varicocele grade, a larger vein
diameter was more significant in rest and during Valsalva
(in all cases P \ 0.05), except between grade I and grade
II. Retrograde peak flow velocities were similar in every
group (in all cases P [ 0.1). Only grade III varicoceles
demonstrated significantly increased peak flow values
A. Pilatz, B. Altinkilic contributed in equal parts.
M. Marconi was a fellow in ‘‘Clinical Andrology’’ at the Department
of Urology in Giessen (2007–2008), Scholarship MIDEPLAN Chile.
A. Pilatz (&)
B. Altinkilic
E. Köhler
M. Marconi

W. Weidner
Department of Urology, Pediatric Urology and Andrology,
University Hospital Giessen and Marburg GmbH—Giessen,
Justus-Liebig University Giessen, Rudolf-Buchheim-Strasse 7,
35385 Giessen, Germany
e-mail: adrian.pilatz@chiru.med.uni-giessen.de
M. Marconi
Urology Unit, Complejo Hospitalario San Jose, Santiago, Chile
compared with all other grades (P \ 0.001). There were no
side-related differences when comparing identical varico-
cele grades (in all cases P [ 0.1). Venous diameters above
2.45 mm in rest (sensitivity 84%, specificity 81%) or
2.95 mm during Valsalva (sensitivity 84%, specificity
84%) predicted the presence of a clinical varicocele.
Conclusions Our findings support the hypothesis that
clinical varicoceles can be predicted with high accuracy
based only on the diameter of testicular veins using cut-
point values of [2.45 mm in rest or [2.95 mm during
Valsalva maneuver in the supine position.
Keywords Varicocele
Color Doppler ultrasound

Infertility
Venous diameter
Abbreviations
CDU Color Doppler ultrasound
ROC Receiver-operator characteristic
Introduction
Varicoceles are abnormal dilations of the pampiniform
plexus. This condition was already known by A. Celsus
(first century AD), who gave a detailed report of a vari-
cocele and its operative therapy [1].
Varicoceles are found in 20–40% of infertile men, but in
only 15% of the general population [2, 3]. Although it is
widely accepted that idiopathic varicoceles are mainly
evident on the left side [3], some authors propose a bilat-
eral disease [4, 5]. The direct association between varico-
cele and male infertility is unknown, but a negative impact
on semen and testicular volume is well documented [2].
123
646
However, a substantial number of men with a clinical
varicocele have no fertility problems [2]. Although vari-
cocele repair is discussed controversially [6], current data
support the idea that therapy of clinical varicoceles does
have a beneficial effect on improving seminal parameters
as well as assisting infertile couples to achieve pregnancy
spontaneously [7, 8].
The clinical diagnosis of varicocele is made by physical
examination in a warm room [9], but the specificity of this
assessment is only about 70% [3, 10]. In contrast, color
Doppler ultrasound has a sensitivity of 97% and a speci-
ficity of 94% [10]. Due to its noninvasiveness, color
Doppler ultrasound has replaced the former gold standard
of venography [11]. Although men with impaired fertility
are often investigated for the presence of a varicocele by
scrotal ultrasound, no generally accepted criteria are
available. This is because several modalities have been
used for ultrasound-based diagnosis. In this connection,
various possibilities have been investigated: measuring
venous diameters in an upright [12–14] or supine position
[15, 16], measuring only venous diameters [14, 17], and
measuring venous diameter and retrograde flow [3, 18–20].
Other studies have involved heterogeneous populations
comprising healthy controls and infertile patients with
clinical, subclinical or no varicoceles, resulting in different
cut-points for venous diameters and retrograde peak flow
[3, 12–20].
The purpose of this study was to determine whether
specific scrotal color Doppler ultrasound (CDU) parame-
ters (basal venous diameter, maximum venous diameter
during Valsalva maneuver, continuous spontaneous reflux,
peak retrograde flow) can be used to predict varicocele
grade according to WHO criteria in patients with and
without clinical varicocele.
Patients and methods
After approval of the institutional review board, we con-
ducted a prospective study from 2004 to 2010 on color
Doppler ultrasound in 129 patients with clinical varicocele
(median age 30 years, range 12–62 years) and 88 healthy
controls without clinical varicocele (median age 35 years,
range 19–61 years). The study excluded men who had
undergone previous varicocele repair, scrotal surgery, or
hernia repair, men who had suffered a symptomatic vari-
cocele (e.g., renal cell cancer), men with a history of uro-
genital infection, and men with single testes. Controls were
enrolled during routine prevention care in our outpatient
department.
The presence of a clinical varicocele was determined by
palpation and observation in the standing position before
and during the Valsalva maneuver. Classification was in
123
World J Urol (2011) 29:645–650
accordance with the WHO: 0 (no varicocele), grade I
(palpable during Valsalva maneuver), grade II (palpable
without Valsalva maneuver), and grade III (visible through
the scrotal skin). Subclinical varicocele was classified as
non-palpable, but with retrograde reflux under Valsalva
maneuver as visualized using CDU [9].
The study was carried out by two experienced investi-
gators (A.P., E.K.) to avoid bias in the diagnosis of vari-
cocele and mistakes in interpreting the CDU findings.
High-frequency ultrasound with color Doppler imaging
was performed using a 7.5 MHz transducer (SonoAce
8800, Sonoace Deutschland GmbH, Marl, Germany).
During the investigation, diameter and reflux of the largest
vein in the pampiniform plexus were measured bilaterally
with the patient in the supine position. Basal vein diameter
and continuous spontaneous reflux were obtained when the
patient was relaxed. Maximum vein diameter and peak
retrograde flow were evaluated during the Valsalva
maneuver (Fig. 1 a–d). At least three separate measure-
ments of each variable were made and subsequently
averaged.
Vein diameters and reflux in the different groups were
compared with non-parametric tests for unpaired samples.
Differences in vein diameter before and during Valsalva
maneuver were compared with non-parametric test for
paired samples. The cutoff points, sensitivity, and speci-
ficity for vein diameters were determined by receiver-
operator characteristic (ROC) analysis. The ‘‘optimal’’
operating point of the ROC curve was chosen as the value
that produced the point closest to 100% sensitivity and
100% specificity. The results were expressed as medians. A
value of P \ 0.05 was considered statistically significant.
Statistical analyses were performed using PASW Statistics
18 for Windows (SPSS GmbH Software, Munich,
Germany).
Results
A total of 434 testicles in 217 men were examined and
scanned by color Doppler ultrasound. In 258 testicular
units from 129 patients (with clinical varicocele), 146
clinical varicoceles (grade I: 11 right, 52 left; grade II: 7
right, 42 left; grade III: 1 right, 33 left) and 26 subclinical
varicoceles (25 right, 1 left) were recorded. A subclinical
varicocele was identified in 18 out of 176 testicular units (8
right and 10 left) from 88 controls (without clinical vari-
cocele). In all subclinical and clinical varicoceles, a ret-
rograde flow was observed during the Valsalva maneuver.
The median basal vein diameter in a relaxed supine
position was 1.6 mm (grade 0), 2.2 mm (subclinical),
3.0 mm (grade I), 2.8 mm (grade II), and 2.9 mm (grade
III) on the right and 1.4 mm (grade 0), 1.8 mm
World J Urol (2011) 29:645–650
Fig. 1 a B-scan of a grade III
varicocele in relaxed supine
position. b B-scan of the same
varicocele during Valsalva
maneuver. c Color Doppler
ultrasound of a grade III
varicocele depicting continuous
spontaneous reflux. d Color
Doppler ultrasound of a grade
III varicocele during Valsalva
maneuver
(subclinical), 3.0 mm (grade I), 3.2 mm (grade II), and
3.4 mm (grade III) on the left side, respectively. There
were no side-related differences regarding identical grades
(in each case P [ 0.1). With increasing varicocele grade, a
larger basal vein diameter became significant (in each case
P \ 0.05), except between grade I and grade II (Fig. 2a).
During Valsalva maneuver, a median increase of
0.3 mm in all vein diameters compared to baseline was
recorded (P \ 0.001). Median maximum vein diameter
was 1.8 mm (grade 0), 2.7 mm (subclinical), 3.7 mm
(grade I), 4.1 mm (grade II), and 3.4 mm (grade III) on the
right and 1.7 mm (grade 0), 2.2 mm (subclinical), 3.65 mm
(grade I), 3.75 mm (grade II), and 4.7 mm (grade III) on
the left side, respectively. Again, there were no side-related
differences when comparing identical grades (in each case
P [ 0.3), but with increasing varicocele grade, a larger
maximum vein diameter was significant (in each case
P \ 0.05), except between grade I and grade II (Fig. 2b).
We assessed the ability of the basal venous diameter, the
maximum venous diameter during Valsalva maneuver, and
the arithmetic difference of both in order to distinguish
testicular units with no clinical varicocele from those with
clinical varicocele. The area under the ROC curve was
0.883 for basal venous diameter and 0.904 for maximum
venous diameter, indicating this value to be a good clas-
sifier. On the other hand, the area under the ROC curve was
only 0.779 for the diameter difference before and during
Valsalva maneuver (Fig. 3a). The optimal cutoff point for
647
discriminating between testicular units without clinical
varicocele and with clinical varicocele in the relaxed
supine position was 2.45 mm (sensitivity 84%, specificity
81%). During Valsalva maneuver, the optimal cutoff point
was 2.95 mm (sensitivity 84%, specificity 84%) in the
same testicular units. Discriminating between testicular
units without varicocele and with subclinical varicocele on
the basis of venous diameters resulted in less favorable
results. The area under the ROC curve was 0.655 for basal
venous diameter, 0.673 for maximum venous diameter, and
0.602 for the diameter difference of both (Fig. 3b). All
possible cutoff points showed sensitivity and specificity
both below 70%. When calculating side-specific cutoff
diameters, no advantage was found.
In 146 clinical grade I-III varicoceles, 17 cases with a
continuous spontaneous reflux were found (11.6%). Due to
the low number of cases, an association between varicocele
grade and reflux altitude could not be demonstrated.
Median retrograde peak flow was 8 cm/s (subclinical),
7 cm/s (grade I), 12 cm/s (grade II), and 13.95 cm/s (grade
III) on the right and 7 cm/s (subclinical), 11 cm/s (grade I),
11 cm/s (grade II), and 19.85 cm/s (grade III) on the left
side, respectively (Fig. 4). Identical grades showed similar
retrograde peak flow values when comparing both sides (in
each case P [ 0.1). While subclinical, grade I, and grade II
had similar peak flow velocities (in each case P [ 0.1),
grade III varicoceles demonstrated significantly increased
peak flow values (P \ 0.001).
123
648
Fig. 2 Vein diameters subject to varicocele grade and based on the
side investigated. a is before and b is during Valsalva maneuver. No
side-related difference in vein diameter (in each case P [ 0.1).
Significant increase in venous diameter to the next higher WHO
group (in each case P \ 0.05), except between grade I and grade II.
Box and whisker plots display median values, quartiles, and
(o) outliers
Discussion
Varicocele is the most common cause of infertility in men
[11]. It is accompanied by reflux, causing dilation in the
pampiniform plexus. This condition is routinely used as a
diagnostic tool in detecting varicocele by physical exami-
nation [9]. Detection of a varicocele by palpation is closely
associated with increased vein diameter. In this context,
Hoekstra and coworkers noted that veins smaller than
2.5 mm are never palpable, while scrotal veins with
diameters larger than 3.5 mm are always palpable [12].
Although clinical diagnosis of varicocele can be made
directly by physical examination, the specificity of this
assessment is only about 70% [3, 10]. Due to the high
sensitivity (97%) and specificity (94%) of color Doppler
123
World J Urol (2011) 29:645–650
Fig. 3 ROC curve analysis for predicting a clinical varicocele and
b subclinical varicocele by venous diameters in rest, during Valsalva
maneuver and using the arithmetic difference of both
ultrasound, this method is recommended in cases of doubt
and for confirming diagnosis [10, 11, 21]. Color Doppler
ultrasound detects a varicocele either by measuring
augmented scrotal vein size or by demonstrating blood
refluxing through the pampiniform plexus during the
Valsalva maneuver. The measurement of retrograde peak
flow creates a possibility to detect further subclinical
varicoceles [9].
Our prospective study sought to determine whether
color Doppler ultrasound can predict the grade of sub-
clinical and clinical varicocele. In contrast to others [5, 14,
15, 17], we evaluated a large group of 217 men: 129 suf-
fering from clinical varicocele and 88 without clinical
varicocele, who served as controls. In order to obtain
complete evaluation of the status in the spermatic veins,
physical examination was performed in the standing posi-
tion, and sonographic findings were noted in the supine
position in rest and during Valsalva maneuver. Further-
more, all scanned testicular units were analyzed separately
World J Urol (2011) 29:645–650
Fig. 4 Retrograde peak flow subject to varicocele grade and based on
the side investigated. No side-related difference in peak flow (in each
case P [ 0.1). Only grade III demonstrated significantly increased
peak flow compared to other groups (#P \ 0.001). Box and whisker
plots display median values, quartiles, and (o) outliers
on both sides, including varicocele grade before and during
Valsalva maneuver. Unfortunately, some other authors
mixed up different grades [12, 14, 16] or did not specify if
the results were found in an upright or a supine position, in
rest or during Valsalva [15, 18, 20]. This complicates any
proper interpretation. Furthermore, the use of primarily
pooled data for venous diameters of different testicular
sides [13, 16, 20] is questionable because of the varying
venous drainage. Nevertheless, our current results clearly
show that venous diameters and retrograde peak flow
velocities do not differ between sides.
The main goal of our study was to consider the impor-
tance of venous diameter as a predictor of varicocele grade.
The investigations showed that venous diameters in a
relaxed supine position and during Valsalva maneuver
increased significantly from a lower to the next higher
grade on both sides. Only grades I and II did not differ
significantly. This increase in venous diameter is in
accordance with the data of other authors [3, 20]. However,
in contrast to our findings, some studies noted a pro-
nounced overlap in venous diameters, making it impossible
to differentiate properly between subclinical, grade I, and
no varicoceles [3, 13, 18, 20].
Interestingly, the caliber increase of venous diameter in
the relaxed position in comparison with the Valsalva
maneuver is only slight, with a median of 0.3 mm, and not
useful in differentiating varicocele grade. Here, our results
are in accordance with previous studies [14, 17].
Using ROC analysis, we identified the optimal cutoff
point for discriminating between testicular units without
and with clinical varicocele in a relaxed supine position
to be 2.45 mm (sensitivity 84%, specificity 81%).
649
Interestingly, Hussein discovered that in veins larger than
2.5 mm, reflux can be found in 85.6% of the cases [19].
These data are absolutely in accordance with our results
and reflect the importance of the 2.45-mm vein diameter
cutoff in rest for predicting clinical varicocele. During
Valsalva maneuver, the optimal cutoff point climbed to
2.95 mm (sensitivity 84%, specificity 84%). Several cutoff
values for venous diameters (2.0, 2.5, and 3.0 mm) have
been proposed [15, 19, 20], but to our knowledge, they
have never been determined by ROC analysis. Our results
clearly indicate that, in many cases, Doppler evaluation can
be omitted, because a B-scan of vein diameters in rest or
during Valsalva maneuver predicts clinical varicocele with
high accuracy.
However, the diagnosis of subclinical varicocele based
only on the diameter of the veins is characterized by a high
number of false positives and negatives (sensitivity and
specificity each below 70%). Thus, to detect subclinical
varicocele, color Doppler ultrasound is mandatory. Here, our
findings are in accordance with the results of other authors [3,
13, 18, 20]. Moreover, it remains debatable whether sub-
clinical varicocele is clinically relevant at all [6, 18].
Although the peak retrograde flow increased with higher
varicocele grades, we detected a significant difference only
in patients with grade III. In all other groups, peak flow
velocities were comparable at about 9 cm/s, as found by
others [13]. This makes the peak flow velocities useless for
differentiating between grades. Nevertheless, retrograde
peak flow velocities might be of importance for predicting
progression or a new onset of testicular asymmetry in
adolescents [22].
Conclusions
Our data support the hypothesis that by means of color
Doppler ultrasound, clinical varicoceles can be predicted
with high accuracy (sensitivity [80%, specificity [80%)
based only on the diameter of testicular veins using cut-
point values of 2.45 mm in rest or 2.95 mm during the
Valsalva maneuver in the supine position. However, sub-
clinical varicoceles cannot be reliably differentiated from
testicular units without varicocele. The venous diameter
difference before and during Valsalva as well as peak
flow velocities are unsuitable for distinguishing between
varicocele grades.
Conflict of interest The present authors have no conflict of interest.
References
1. Noske HD, Weidner W (1999) Varicocele–a historical perspec-
tive. World J Urol 17:151–157
123
650
2. World Health Organization (1992) The influence of varicocele on
parameters of fertility in a large group of men presenting to
infertility clinics. Fertil Steril 57:1289–1293
3. Sakamoto H, Saito K, Shichizyo T et al (2006) Color Doppler
ultrasonography as a routine clinical examination in male infer-
tility. Int J Urol 13:1073–1078
4. Gat Y, Zukerman Z, Chakraborty J, Gornish M (2005) Varico-
cele, hypoxia and male infertility. Fluid Mechanics analysis of
the impaired testicular venous drainage system. Hum Reprod
20:2614–2619
5. Trussell JC, Haas GP, Wojtowycz A, Landas S, Blank W (2003)
High prevalence of bilateral varicoceles confirmed with ultraso-
nography. Int Urol Nephrol 35:115–118
6. Evers JH, Collins J, Clarke J (2009) Surgery or embolisation
for varicoceles in subfertile men. Cochrane Database Syst
Rev:CD000479
7. Marmar JL, Agarwal A, Prabakaran S et al (2007) Reassessing
the value of varicocelectomy as a treatment for male subfertility
with a new meta-analysis. Fertil Steril 88:639–648
8. Abdel-Meguid TA, Al-Sayyad A, Tayib A, Farsi HM (2010)
Does varicocele repair improve male infertility? An evidence-
based perspective from a randomized, controlled trial. Eur Urol
9. WHO (2000) Manual for the standardized investigation, diag-
nosis and management of the infertile male. Cambridge Univer-
sity Press, Cambridge
10. Trum JW, Gubler FM, Laan R, van der Veen F (1996) The value
of palpation, varicoscreen contact thermography and colour
Doppler ultrasound in the diagnosis of varicocele. Hum Reprod
11:1232–1235
11. EAU (2010) EAU Guidelines edition presented at the 25th EAU
Annual Congress: EAU Guidelines Office, Arnhem, The
Netherlands
12. Hoekstra T, Witt MA (1995) The correlation of internal spermatic
vein palpability with ultrasonographic diameter and reversal of
venous flow. J Urol 153:82–84
123
World J Urol (2011) 29:645–650
13. Cina A, Minnetti M, Pirronti T et al (2006) Sonographic quan-
titative evaluation of scrotal veins in healthy subjects: normative
values and implications for the diagnosis of varicocele. Eur Urol
50:345–350
14. Orda R, Sayfan J, Manor H, Witz E, Sofer Y (1987) Diagnosis of
varicocele and postoperative evaluation using inguinal ultraso-
nography. Ann Surg 206:99–101
15. Schiff JD, Li PS, Goldstein M (2006) Correlation of ultrasound-
measured venous size and reversal of flow with Valsalva with
improvement in semen-analysis parameters after varicocelec-
tomy. Fertil Steril 86:250–252
16. Kocakoc E, Serhatlioglu S, Kiris A et al (2003) Color Doppler
sonographic evaluation of inter-relations between diameter,
reflux and flow volume of testicular veins in varicocele. Eur J
Radiol 47:251–256
17. Wolverson MK, Houttuin E, Heiberg E, Sundaram M, Gregory J
(1983) High-resolution real-time sonography of scrotal varico-
cele. AJR Am J Roentgenol 141:775–779
18. Caskurlu T, Tasci AI, Resim S, Sahinkanat T, Ekerbicer H (2003)
Reliability of venous diameter in the diagnosis of subclinical
varicocele. Urol Int 71:83–86
19. Hussein AF (2006) The role of color Doppler ultrasound in
prediction of the outcome of microsurgical subinguinal varico-
celectomy. J Urol 176:2141–2145
20. Jarow JP, Ogle SR, Eskew LA (1996) Seminal improvement
following repair of ultrasound detected subclinical varicoceles.
J Urol 155:1287–1290
21. Liguori G, Trombetta C, Garaffa G et al (2004) Color Doppler
ultrasound investigation of varicocele. World J Urol 22:378–381
22. Kozakowski KA, Gjertson CK, Decastro GJ, et al (2009) Peak
retrograde flow: a novel predictor of persistent, progressive and
new onset asymmetry in adolescent varicocele. J Urol
181:2717–2722; discussion 2723