Selection of Saphenous Vein Conduit in Varicose
Vein Disease
Joseph D. Cohn, MD, FACS, and Keith F. Korver, MD, FACS
Department of Surgery, Sutter Medical Center of Santa Rosa, Santa Rosa, California
Background. Limbs with varicose veins are difficult to
assess as a source of saphenous vein conduit. Anatomic,
histologic, and ultrasound studies demonstrate two types
of longitudinal veins in the lower extremities. The great
saphenous vein is deep to the saphenous fascia. Acces-
sory saphenous veins are superficial to this layer and
have thin walls with diminished muscle cells and elastic
fiber. Accessory saphenous veins dilate and form varicos-
ities. Segments of great saphenous veins are often suit-
able as coronary conduits. No studies have assessed the
suitability of saphenous veins as coronary artery con-
duits in patients with varicose vein disease.
Methods. Intraoperative high-resolution ultrasound
studies were performed in coronary artery bypass graft
procedures to assess lower extremity venous morphology
in limbs of 77 patients without known venous disease, in
19 limbs with venous telangiectases, and in 23 limbs with
varicose veins.
I t has been presumed that saphenous veins in limbs
with varicose veins cannot routinely be used as coro-
nary artery conduits. This assertion is based on the view
that venous dilatation and tortuosity affect all superficial
veins in limbs with varicose vein disease. Alternative
conduit sites and options for use in patients with venous
varicosities have been reported [1, 2]. MacFarlane and
colleagues [3] documented that great saphenous vein
(GSV) segments below the knee do not become varicose
even in patients with extensive below-knee varicosities.
Fligelstone and coworkers [4] and Hammarsten and
associates [5] performed ultrasound studies and venog-
raphy in limbs with varicosities after high saphenous
vein ligation and multiple stab phlebectomy. They re-
ported that GSV length and lumen dimensions were
suitable for vascular conduit in more than 60% of limbs.
Mellire [6] performed preoperative ultrasound assess-
ment in patients with varicose vein disease and demon-
strated that more than 80% of limbs had normal or
slightly dilated GSV segments.
Ultrasound studies in limbs with venous telangiectases
demonstrate reflux into reticular veins, and this is often
associated with incompetence of larger epifascial veins.
In some cases, there is reflux into the deep venous system
[7, 8]. Thibault and associates [9] performed ultrasound
Accepted for publication Nov 4, 2005.
Address correspondence to Dr Cohn, 5773 Shiloh Ridge, Santa Rosa, CA
95403; e-mail: jcohn@alum.mit.edu.
2006 by The Society of Thoracic Surgeons
Published by Elsevier Inc
CARDIOVASCULAR
Results. Dilated great saphenous vein segments were
identified in 6% of normal limb venous segments com-
pared with 21% of segments in limbs with telangiectases
(p 0.027) and 22% of segments in limbs with varicosi-
ties (p 0.012). The incidence of absent or hypoplastic
great saphenous vein segments is increased in limbs with
varicosities (35%) compared with normal limbs (21%; p
0.032). In the calf, at least one great saphenous vein
segment suitable for coronary artery bypass grafting is
present in 70% of limbs with varicosities and in 89% of
limbs with telangiectases.
Conclusions. Ultrasound studies document that vari-
cose veins are limited to accessory saphenous veins.
Great saphenous vein conduits, identified by ultrasonog-
raphy, are available in limbs with varicose vein disease.
(Ann Thorac Surg 2006;81:1269 74)
2006 by The Society of Thoracic Surgeons
imaging in patients with symptomatic telangiectases and
demonstrated that 23% have major sources of venous
reflux within the GSV. No studies have systematically
assessed the suitability of saphenous veins as coronary
artery conduits in patients with venous telangiectases or
varicose veins.
Correlation of anatomic and ultrasound studies dem-
onstrates two types of longitudinal veins in the lower
extremities [10 13]. The GSV lies deep to the saphenous
fascia and above the muscular fascia. Accessory saphe-
nous veins (ASVs) are superficial to the saphenous facial
layer. The muscular and saphenous fasciae join after
arching over the GSV, delimiting the saphenous com-
partment. This compartment contains the GSV, facial
septa, adipose tissue, saphenous artery, and saphenous
nerve [1117]. All longitudinal veins devoid of a fascial
envelope are ASVs. Accessory saphenous veins may
dilate and develop into varicose veins. In varicose vein
disease, dilatation and reflux may occur within GSVs, but
development of varicosities does not occur owing to the
firm investing saphenous and muscular fascia [15, 18].
Accessory saphenous veins may be anterior, posterior,
or lateral to the GSV. Segments of ASVs and GSVs may
be absent, atretic, or hypoplastic. The transition of an
ASV as a branch of the GSV may be seamless, indistinct
even during anatomic dissection. Each of the two veins
has differing anatomic, histologic, and physiologic char-
acteristics [11, 16]. On the basis of these findings and
0003-4975/06/$32.00
doi:10.1016/j.athoracsur.2005.11.013
 1270 COHN AND KORVER
SAPHENOUS CONDUIT IN VARICOSE VEIN DISEASE
additional studies, it has been postulated that ASVs are
prone to varicose changes [12, 14, 18].
The purpose of this study is to assess GSVs in limbs
with telangiectases and varicosities and to document the
suitability of GSVs as coronary artery conduits in limbs
CARDIOVASCULAR
with venous varicosities. Great saphenous vein site se-
lection criteria and issues related to venous varicosities in
patients undergoing coronary artery bypass graft surgery
are examined.
Material and Methods
This study was reviewed by the Sutter Health Central
Area Institutional Review Committee and received ex-
empt status approval. The study involves record review
of intraoperative ultrasound reports and data on patients
from September 22, 2004, through October 26, 2005.
This study was performed to define the incidence of
GSVs suitable for use as coronary artery conduits in
limbs with venous telangiectases and varicose veins.
Consecutive, intraoperative ultrasound studies were per-
formed in 154 limbs without venous disease (77 patients),
in 19 limbs (12 patients) with telangiectases, and in 23
limbs (14 patients) with varicosities. Studies were per-
formed in patients undergoing elective and emergent
coronary artery bypass graft surgery. For the purpose of
this study, venous telangiectases represent intradermal
telangiectases present in a fan-shaped or cluster pattern,
located in the thighs or calves, behind the knees, or over
the ankles. The individual veins are less than 2 mm in
diameter. Varicose veins are recognized as irregular
dilated veins in a subcutaneous location anywhere in the
lower extremity.
Intraoperative venous duplex ultrasound studies are
performed routinely in all coronary artery bypass graft
patients using a high-resolution duplex scanner and
model 15L8 linear array probe (Acuson Sequoia 512
Ultrasound System, Acuson Corp, Mountain View, CA).
After endotracheal intubation, a high thigh tourniquet is
placed around the proximal thigh using a 1-inch rubber
drain secured tightly with a clamp. This provides com-
pression of the saphenous vein distal to the saphe-
nofemoral junction. Venous distention approximates the
distention achieved during vein conduit preparation.
Real-time transverse axial scanning provides sufficient
imaging to assess venous anatomy. Scans are performed
from the ankle to the proximal thigh, with the leg
externally rotated and the knee flexed. Initial ultrasound
depth adjustment is 20 mm with 13.0 MHz probe scan
frequency. Frequent tissue pressure with the scan head is
used to compress the saphenous vein to identify areas of
venous thrombosis. Venous abnormalities and internal
lumen diameters of the distended vein segments are
recorded. Scan time for each leg is usually 3 to 4 minutes.
Additional scan time is necessary if the GSV is unsatis-
factory and ASV segments require identification as con-
duit options. Ultrasound studies are performed during
intraoperative preparatory procedures, avoiding unnec-
essary operative delays. Details and methodology of the
procedure have been previously reported [19].
Ann Thorac Surg
2006;81:1269 74
Fig 1. Intraoperative ultrasound recording of great saphenous vein
(G) and accessory saphenous vein (A). Transaxial image of the mid
thigh. The great saphenous vein is located beneath the saphenous
fascia and measures 2.7 mm in internal lumen diameter. The acces-
sory saphenous vein measures 2.6 mm in diameter. White arrow-
heads indicate saphenous fascia. Black arrowheads identify muscular
fascia. Vertical axis, full scale, is 20 mm.
Patient data include age, sex, type of surgery, and ankle
to brachial artery pressure index. Accessory saphenous
veins and GSVs are evaluated in the proximal and distal
calf and thigh. For the purpose of this study, a vein
segment is defined as representing the length of vein at
each of these four sites. Characteristics of the four ASV
and four GSV segments in each limb, two in the thigh
and two in the calf, are recorded.
Accessory saphenous vein and GSV segments are
described as normal if they measure 2 to 5 mm in internal
lumen diameter and extend half of the length of the calf
or thigh, a length suitable for at least one coronary artery
conduit. A small segment is defined as a narrow ASV or
GSV segment, less than 2 mm in diameter, or not
extending half of the length of the calf or thigh. Nar-
rowed segments may represent absent, atretic, or hypo-
plastic veins [10]. Vein segments less than 2 mm in
diameter are not considered for conduit use. Dilated ASV
and GSV segments are identified as veins greater than 5
mm internal lumen diameter and extending approxi-
mately half of the length of the calf or thigh. Thrombosed
veins are defined as incompletely compressible vein
segments. On a few occasions, more then one abnormal-
ity is present within a limb segment. A vein segment is
assigned a single category based on the major portion of
the vein segment fulfilling the defined criteria.
An example of an ultrasound recording of a GSV and
an ASV is shown in Figure 1. An ultrasound image of
varicosities associated with a GSV is shown in Figure 2.
Statistical analysis of categorical data is performed by
the Mann-Whitney rank sum test statistic. All probability
value statistics are two-tailed. Percentage values are
rounded to whole numbers. In some instances, percent-
age values do not add to 100% as a result of whole
number rounding. Vein diameter measurements, identi-
fied in Figures 1 and 2, are obtained from scanned
images.
Ann Thorac Surg COHN AND KORVER 1271
2006;81:1269 74 SAPHENOUS CONDUIT IN VARICOSE VEIN DISEASE

(6%). All dilated segments were located in the thigh.

Absent or hypoplastic venous segments were docu-
mented in 21% of venous segments. The remaining 74%
of venous segments were in the range 2 to 5 mm internal
lumen diameter and would be suitable for use as coro-

Fig 2. Intraoperative ultrasound recording of varicose veins (V) and
great saphenous vein (large white arrowhead). Transaxial image of
the mid calf. Varicose veins are located above the saphenous fascia.
The small white arrowheads identify the saphenous fascia and adja-
cent border of varicose veins. The great saphenous vein is located
beneath the saphenous fascia and above the muscular fascia and
measures 1.6 mm in internal lumen diameter. Vertical axis, full
scale, is 30 mm.
Results
Patient demographics include coronary artery bypass
graft procedures in 76 men and 26 women. Mean age for
all patients is 67.6 years. Average saphenous vein con-
duits per patient are 2.2. Additional procedures include
internal mammary artery anastomoses (n 50), aortic
valve replacement (n 18), and mitral valve repair or
replacement (n 9). Venous telangiectases were identi-
fied in 19 (10%) limbs, and varicosities were present in 23
(12%) limbs. Vein harvesting was performed endoscopi-
cally with the Guidant Vasoview Uniport Plus kit
(Guidant Corp, Santa Clara, CA) in 34 instances. Surgical
excision was performed at 126 limb sites. In some pa-
tients, endoscopic vein harvesting and surgical excision
were both performed. More than one conduit was usually
obtained at each vein excision site.
CARDIOVASCULAR
nary artery conduits. No thrombosed venous segments
were identified. The number of dilated, small, and nor-
mal-caliber GSV segments and their percentage limb
distributions are shown in Table 1.
On the basis of venous segments measuring 2 to 5 mm
internal lumen diameter, at least one GSV segment
suitable for use as a coronary artery conduit was present
in every limb. At least two segments could be harvested
in 91% of limbs. Three segments were available in 61% of
limbs, and four venous segments could be obtained in
42% of limbs.
Venous Telangiectases
Dilated GSV segments in limbs with venous telangiecta-
ses were identified in 16 of 76 venous segments (21%). All
dilated GSV segments were located in the thigh. Small
GSV segments were recorded in 13 venous segments
(17%). Venous segments suitable for use as coronary
conduits were identified in 62% of limb segments. No
thrombosed venous segments were identified. Distribu-
tions of dilated, small, and normal venous segment
dimensions in limbs with telangiectases are recorded in
Table 2.
In all limbs with venous telangiectases, at least one
GSV segment was identified as suitable for use as a
coronary artery conduit. In 84% of limbs, at least two GSV
segments were available for conduit use. At least three
GSV segments could be used in 37% of limbs, and four
venous segments, the entire GSV, could be used for
bypass grafting in 26% of limbs with telangiectases.
The distribution of the size of limb segments between
normal limbs and limbs with telangiectases was com-
pared using the Mann-Whitney rank sum test. There was
a statistical difference in size distribution of limb seg-
ments in the proximal thigh (p 0.007) and distal thigh (p
0.036) but not in the two calf segments (both p 0.6).

Normal Limbs Varicose Veins

Dilated vein segments, in limbs without known venous A similar pattern of increased dilated vein segments was
disease, were identified in 34 of 616 venous segments found in limbs with varicose veins. Dilated GSV seg-

Table 1. Distribution of Great Saphenous Vein Segments in Limbs Without Venous Diseasea
Normal Small Dilated Thrombus

Site n % n % n % n %

Thigh proximal 129 84 4 3 21 14 0 0

Thigh distal 101 66 40 26 13 8 0 0
Calf proximal 95 62 59 38 0 0 0 0
Calf distal 128 83 26 17 0 0 0 0
a
Data are from great saphenous vein segments of 154 limbs.
Great saphenous vein lumen diameter distributions in vein segments in the proximal and distal thigh and calf in limbs without venous disease. Dilated
segments represent vein segments with internal lumen caliber greater than 5 mm. Small vein segments measure less than 2 mm lumen diameter or are
less than half of the calf or thigh in length. Normal vein segments are between 2 and 5 mm internal lumen diameter. Four great saphenous vein segments
of approximately equal length are identified in each limb. Percentage values may not total 100% owing to whole number rounding.
 1272 COHN AND KORVER Ann Thorac Surg
SAPHENOUS CONDUIT IN VARICOSE VEIN DISEASE 2006;81:1269 74

Table 2. Distribution of Great Saphenous Vein Segments in Limbs With Telangiectasesa

Normal Small Dilated Thrombus

Site n % n % n % n %
CARDIOVASCULAR

Thigh proximal 9 47 0 0 10 53 0 0
Thigh distal 8 42 5 26 6 32 0 0
Calf proximal 13 68 6 32 0 0 0 0
Calf distal 17 89 2 11 0 0 0 0
a
Data are obtained from great saphenous vein segments of 19 limbs.
Great saphenous vein internal lumen diameter distributions in vein segments in the proximal and distal thigh and calf in limbs with telangiectases. Vein
segment categories are defined in the text and in Table 1. Great saphenous vein category distributions are significantly different in limbs without venous
disease compared with limbs with telangiectases in the proximal thigh (p 0.007) and distal thigh (p 0.036) but not in calf segments.

ments were detected in 20 of 92 venous segments (22%).
Small venous segments were found in 35% of venous
segments, and normal-caliber venous segments were
documented in 40%. Thrombosed venous segments, all
in the distal calf, were detected in 3% of venous seg-
ments. Frequency distributions of GSV segment catego-
ries are shown in Table 3.
Varicosities imaged by ultrasound were always located
superficial to the saphenous fascia, within the subcuta-
neous adipose tissue. In the event a GSV was not able to
be ultrasonically detected between the muscular and
saphenous fascia, ASVs were readily detected in a sub-
cutaneous location. Normal diameter and dilated GSVs
were easily identified. No GSV segment was documented
to be tortuous.
At least one GSV segment, measuring between 2 and 5
mm internal lumen diameter, could be harvested in 78%
of limbs with varicosities. Two GSV segments were
suitable as conduits in 57% of limbs. The availability of
GSV segments within a limb for use as coronary conduits
in normal limbs and in limbs with telangiectases and
varicose veins is illustrated in Figure 3.
Data analysis illustrates a predominance of normal-
sized GSV segments within the thigh and calf in normal
limbs and in the calf in limbs with venous telangiectases.
Venous dilatation within the thigh in limbs with telangi-
ectases and varicosities results in a reduction in suitable
GSV segments for conduit use. In varicose vein disease,
calf GSV segments have the highest likelihood for pro-
viding at least one suitable GSV segment for use as a
coronary conduit. In 5 of 23 limbs with varicosities (22%),
a satisfactory GSV segment for conduit use could not be
identified in the entire limb.
Great saphenous vein size distribution was compared
at the various vessel segment levels in normal limbs and
in limbs with varicose veins using the Mann-Whitney
rank sum test. There were statistically significant differ-
ences in GSV segment size distribution in the proximal
thigh (p 0.002), distal thigh (p 0.006), and proximal
calf (p 0.001). Differences in distribution categories in
the distal calf were of borderline significance (p 0.052).
Although demographic findings were similar in nor-
mal limbs and in limbs with telangiectases and varicos-
ities, the high frequency of dilated combined with absent
or hypoplastic GSV segments in the thigh resulted in a
reduction in the use of endoscopic vein harvesting in
these patients. In patients with normal limbs, endoscopic
vein harvesting was performed in 43% of cases. Endo-
scopic vein harvesting was performed in one limb with
telangiectases and in no limbs with varicose veins.
Accessory Saphenous Veins
Accessory saphenous veins and collateral veins are
readily distinguished from GSVs by their location above
the saphenous fascia. Longitudinal ASV segments, in the
range greater than 2 mm and less than 5 mm internal
lumen diameter and extending half of the calf or thigh,
were observed in 144 of 616 normal limb venous seg-
ments (23%). In limbs with venous telangiectases, ASV
segments in this diameter and length range were present

Table 3. Distribution of Great Saphenous Vein Segments in Limbs With Varicose Veinsa
Normal Small Dilated Thrombus

Site n % n % n % n %

Thigh proximal 10 43 3 13 10 43 0 0
Thigh distal 8 35 8 35 7 30 0 0
Calf proximal 5 22 16 70 2 9 0 0
Calf distal 14 61 5 22 1 4 3 13
a
Data are obtained from great saphenous vein segments of 23 limbs.
Great saphenous vein internal lumen diameter distributions in vein segments in the proximal and distal thigh and calf in limbs with varicose veins. Vein
segment categories are defined in the text and in Table 1. Thrombosed veins indicate noncompressible vein segments. Great saphenous vein category
distributions in the proximal thigh, distal thigh, and proximal calf are significantly different compared with their distributions in normal limbs (all p
0.006). Category differences in the distal calf segments are of borderline significance (p 0.052).
Ann Thorac Surg
2006;81:1269 74
Fig 3. Percent great saphenous vein (GSV) segment distribution for
normal-caliber vein segments, internal lumen diameter greater than
2 mm and less than 5 mm, in normal limbs, limbs with venous tel-
angiectases, and limbs with varicose veins. Four great saphenous
vein segments are identified in each limb. The abscissa notations
1, 2, and 3 indicate that at least one, two, and three nor- on ultrasound examination [12, 13]. In the diameter range
mal-size great saphenous vein segments are identified by ultrasound
in the limb. The abscissa notation 4 indicates that four great sa-
phenous vein segments are identified. There is a decreased incidence
of suitable great saphenous vein segments for conduit use in limbs
with varicose veins compared with normal limbs (p 0.001). The
distribution difference between normal limbs and limbs with telangi-
ectases is of borderline significance (p 0.054).
in 30% of limb segments. In limbs with varicosities, ASV
segments were documented in 28% of the limb segments.
A few dilated segments and a single thrombosed ASV
segment in a limb with varicosities were observed. There
were no statistical differences in distribution of small,
normal-caliber, dilated, and thrombosed vein segments
in normal limbs compared with limbs with telangiectases
and varicosities (all p 0.32). Varicosities were always
observed above the saphenous fascia. Great saphenous
vein segments adjacent to either varicosities or ASVs
may be of normal caliber (Fig 1), absent, hypoplastic (Fig
2), dilated, or thrombosed.
Comment
Normal-appearing limbs, without evidence of venous
disease, have a low incidence of dilated GSV segments
(6%), all present in the thigh. The incidence of normal-
sized GSV segments in the thigh is 75% and in the calf is
72%. The remaining GSV segments are small, measuring
less than 2 mm in diameter. These small GSV segments
cannot be used for saphenous conduits. Similar results
have been reported by others [14, 20 24].
Limbs with telangiectases and varicosities have a high
incidence of dilated and small GSV segments, compared
with normal limbs. The finding of GSV dilatation in limbs
with telangiectases supports the concept that reflux oc-
curs in venous tributaries draining into the GSV, result-
ing in proximal dilatation of GSV segments in the thigh
COHN AND KORVER 1273
SAPHENOUS CONDUIT IN VARICOSE VEIN DISEASE
[9]. Findings in our study indicate that two suitable GSV
segments could be harvested in 84% of limbs with telan-
giectases. In limbs with varicosities, the possibility of
harvesting two normal-sized segments is reduced to 57%.
In limbs with telangiectases or varicosities, thigh seg-
CARDIOVASCULAR
ments have a higher incidence of small and dilated GSV
segments compared with calf segments. Similar findings
have been reported by MacFarlane and coworkers [3].
Ultrasound studies can identify GSV lumen and length
characteristics to optimize surgical site selection [19].
Both dilated and small venous segments contribute to
a reduction in available sites for harvesting saphenous
vein conduits. The finding of an increased incidence of
small GSV segments in limbs with varicose veins is
consistent with the investigations of Caggiati and Ricci
[14] and Caggiati and Mendoza [15], who described an
increased incidence of dilated ASV segments in limbs
with proximal GSV atresia. They postulated that atretic
GSV segments were causative in the development of
varicose veins by diverting blood flow through adjacent
ASVs, which then became dilated and varicose.
Accessory saphenous veins have characteristic features
2 mm to 5 mm, they are common in normal limbs,
occurring in 23% of venous segments. A slightly higher
frequency is recorded in limbs with telangiectases and
varicose veins. Because of their thin wall structure, ASVs
may not be suitable for use as coronary artery conduits.
Definitive studies are needed to resolve this issue. Ultra-
sonic identification of ASVs is important to consider in
saphenous site selection to select optimal saphenous
conduits for coronary artery bypass graft.
In this study, varicosities within GSV segments were
not identified on ultrasound studies. Surgical excision of
GSV segments, in limbs with varicose veins, also did not
identify varicose GSV segments. In some of these GSV
segments, venous valves were incompetent but vein
diameter was of normal caliber, and these segments were
suitable for use as coronary conduits.
Intraoperative ultrasound provides a rapid and effec-
tive means to assess GSV segments suitable for use as
coronary artery conduits. Ultrasound guidance, particu-
larly in limbs with telangiectases and varicosities, allows
identification of suitable saphenous vein conduits, opti-
mizing surgical site selection. The use of ASV segments
for bypass conduits may not be comparable with GSV
conduits because of their differing histologic and physi-
ologic characteristics. Long-term assessment, comparing
the use of ASVs and GSVs as coronary artery conduits,
requires further investigation.
References
1. Moritz A, Grabenwger F, Wolner E. Mesh tube-calibrated
varicose veins for coronary artery bypass grafting. Ann
Thorac Surg 1994;57:240 2.
2. Canver CC. Conduit options in coronary artery bypass
surgery. Chest 1995;108:1150 55.
3. MacFarlane R, Godwin RJ, Barabas AP. Are varicose veins
and coronary artery bypass surgery compatible? Lancet
1985;2:859.
 1274 COHN AND KORVER
SAPHENOUS CONDUIT IN VARICOSE VEIN DISEASE
4. Fligelstone L, Carolan G, Pugh N, Shandall A, Lane I. An
assessment of the long saphenous vein for potential use as a
vascular conduit after varicose vein surgery. J Vasc Surg
1993;18:836 40.
5. Hammarsten J, Pedersen P, Cederlund C-G, Campanello M.
CARDIOVASCULAR
Long saphenous vein saving surgery for varicose veins. Eur
J Vasc Surg 1990;4:361 4.
6. Mellire D. Why and when to preserve the saphenous veins
of varicose patients to serve as an arterial conduit. J Mal Vasc
1994;19:216 21.
7. Weiss RA, Weiss MA. Doppler ultrasound findings in retic-
ular veins of the thigh subdermic lateral venous system and
implication for sclerotherapy. J Dermatol Surg Oncol 1993;
19:94751.
8. Somjen GM, Ziegenbein R, Johnston AH, Royle JP. Anatom-
ical examination of leg telangiectases with duplex scanning.
J Dermatol Surg Oncol 1993;19:940 5.
9. Thibault P, Bray A, Wlodarczyk J, Lewis W. Cosmetic leg
veins: evaluation using duplex venous imaging. J Dermatol
Surg Oncol 1990;16:612 8.
10. Caggiati, A, Bergan JJ, Gloviczki P, Eklof B, Allegra C,
Partsch H. Nomenclature of the veins of the lower limb:
extensions, refinements, and clinical application. J Vasc Surg
2005;41:719 24.
11. Papadopoulos NJ, Sherif MF, Albert EN. A fascial canal for
the great saphenous vein: gross and microanatomical obser-
vations. J Anat 1981;132:3219.
12. Caggiati A. The saphenous venous compartments. Surg
Radiol Anat 1999;21:29 34.
13. Lova RM, Cappelli M, Macchi C, et al. Macroscopic anatomy
of the superficial veins of the lower limb: a systematic review
based upon echography. It J Anat Embryol 2001;106:21520.
Ann Thorac Surg
2006;81:1269 74
14. Caggiati A, Ricci S. The caliber of the human long saphenous
vein and its congenital variations. Ann Anat 2000;182:195201.
15. Caggiati A, Mendoza E. Segmental hypoplasia of the great
saphenous vein and varicose disease. Eur J Vasc Endovasc
Surg 2004;28:257 61.
16. Caggiati A. Facial relationships of the long saphenous vein.
Circulation 1999;100:25479.
17. Lescali F, Germouty I, Chevalier JM, Enon B, Moreau P,
Pillet J. Extrinsic arterial supply of the great saphenous vein:
an anatomic study. Ann Vasc Surg 1986;1:2737.
18. Caggiati A. Fascial relationships and structure of the tribu-
taries of the saphenous veins. Surg Radiol Anat 2000;22:
191 6.
19. Cohn JD, Korver KF. Optimizing saphenous vein site selec-
tion using intraoperative venous duplex ultrasound scan-
ning. Ann Thorac Surg 2005;79:20137.
20. Bagi P, Schroeder T, Sillesen H, Lorentzen JE. Real time
B-mode mapping of the greater saphenous vein. Eur J Vasc
Surg 1989;3:1035.
21. Head HD, Brown MF. Preoperative vein mapping for coro-
nary artery bypass operations. Ann Thorac Surg 1995;59:
144 8.
22. Kupinski AM, Evans SM, Khan AM, et al. Ultrasonic char-
acterization of the saphenous vein. Cardiovasc Surg 1993;1:
5137.
23. Ruoff BA, Cranley JJ, Hannan LA, et al. Real-time duplex
ultrasound mapping of the greater saphenous vein before in
situ infrainguinal revascularization. J Vasc Surg 1987;6:107
13.
24. Van Dijk LC, Wittens CH, Pieterman H, van Urk H. The
value of pre-operative ultrasound mapping of the greater
saphenous vein prior to closed in situ bypass operations.
Eur J Radiol 1996;23:2357.