Composite Arterial Conduits for a Wider Arterial
Myocardial Revascularization
Antonio M. Calafiore, MO, Gabriele Di Giammarco, MO, Nicola Luciani, MO,
Nicola Maddestra, MO, Ernesto Di Nardo, MO, and Romeo Angelini, MO
Cattedra di Cardiochirurgia, Universita di Chieti, Chieti, Italy
From October 1991 to May 1993, 130 patients were sub-
mitted to myocardial revascularization using complex
preformed arterial conduits. The age ranged from 29 to 75
years (mean age, 60.1 years); 121 patients were male. One
hundred twenty-six patients had double- or triple-vessel
disease. The mean ejection fraction was 0.53 (range, 0.22
to 0.79); only 6 patients had an ejection fraction less than
0.35. In 6 cases the procedure was a reoperation. We used
360 arterial conduits, 163 of which as free grafts (3 left
internal mammary arteries, 16 right internal mammary
arteries, 86 inferior epigastric arteries, 57 radial arteries,
and 1 right gastroepiploic artery). One hundred fifty-four
free grafts were anastomosed to one or both internal
mammary arteries and one to a radial artery. We con-
T he advantages in using the internal mammary artery
(IMA) in myocardial revascularization have been
accepted broadly, both for its durability and for better
survival compared with saphenous vein [1-3]. In the
attempt to limit the use of vein and considering the
growing demand for multiple coronary revascularization,
either in primary or in secondary operations, which are
not always feasible with both IMAs, new arterial conduits
are being tested. Some of them, such as the right gastro-
epiploic artery (RGEA) or the radial artery (RA), had
already been used in coronary operations without signif-
icant success or with discouraging results [4]. A third
arterial graft, the inferior epigastric artery (IEA), has been
used recently [5-7]. Although the RGEA can be used as
either a pedicled or a free graft, RA and IEA can only be
used as free grafts.
Arterial conduits used as free grafts show the same late
patency as saphenous vein [8], the common point being
the aortic anastomosis. For this reason we decided to
perform as a routine proximal anastomoses of arterial free
grafts on the IMA. This technique already was used in
particular situations by other surgeons [5]; we extended it
to arrange complex arterial conduits before establishing
extracorporeal circulation, tailoring them according to the
topography of coronary lesions. In this report we describe
the technique used and the results.
Accepted for publication Nov 24, 1993.
Address reprint requests to Dr Di Giammarco, Clinica Cardiochirurgica,
Ospedale US. Camillo de Lellis," via Forlanini, 50, 66100 Chieti, Italy.
1994 by The Society of Thoracic Surgeons
structed 136 complex arterial conduits (branched, length-
ened, or both). In 6 cases a double arterial system had to
be used in a single patient. There was no operative
mortality, and no inotropic or mechanical supports were
used. The overall mortality rate was 1.5%. Earlyangio-
graphic controls (between the 7th and 15th postoperative
days) demonstrated 100% patency; late angiographic con-
trols (at a mean interval of 9.5 months after operation)
documented a mean patency rate ranging from 94.1% of
the radial arteries to 100% of the left internal mammary
arteries and right gastroepiploic arteries. At a mean
follow-up of 7.2 months (range, 1 to 15 months) all
patients are alive without recurrence of symptoms.
(Ann Thorae Surg 1994;58:185-90)
Material and Methods
Patients
From October 1991 to May 1993, 130 patients were sub-
mitted to myocardial revascularization using complex
preformed arterial conduits; in 6 cases the procedure was
a reoperation. The clinical characteristics of the patients
are summarized in Table 1.
Three-hundred ninety-five conduits were employed
altogether: 128 LIMAs, 69 RIMAs, 86 IEAs, 20 RGEAs, 57
RAs, and 35 saphenous veins. Of 360 arterial conduits,
163 were used as free grafts: 3 LIMAs, 16 RIMAs, 86 lEAs,
57 RAs, and 1 RGEA. Five RAs and 1 RIMA had to be
anastomosed on ascending aorta; lEA and RGEA had to
be anastomosed to a saphenous vein in 1 patient each. Of
the remaining 155 free grafts, 154 were anastomosed to
one or both IMAs and one to an RA, according to the
coronary lesions, constructing 136 complex arterial con-
duits; these systems, together with the donor vessel,
included 2 conduits in 117 patients, 3 in 18 patients, and
4 in 1 patient. In 6 cases a double complex arterial system
had to be used. Table 2 shows the different arrangements
of the grafts in the composite conduits.
Surgical Technique
The IMA, harvested as a pedicle, is divided distally after
systemic heparinization. The artery is cannulated by a
I-mm Teflon needle and is gently injected with 10 mL of
a solution containing papaverine (1 mg/1 mL of normal
saline solution). A hemostatic clip then is applied distally
to allow the artery to dilate for at least 10 minutes.
The lEA is harvested from a paramedian incision run-
0003-4975/94/$7.00
186 CALAFIORE ET AL Ann Thorac Surg
CORONARY REVASCULARIZATIONWITH ARTERIAL GRAFTS 1994;58:185--90

Table 1. Clinical Characteristics of the Patients When the LIMA had to be used to reach the left anterior
descending artery, different arrangements of the free
Variable Value
grafts on the donor conduit were adopted to achieve a
Mean age (y) (range) 60.1 (29-75) more complete revascularization of the left territories. The
Sex (MlF) 121/9 more commonly used arrangement was the lEA on a
l-vessel disease 4 diagonal branch, whereas the RA was employed to reach
2-vessel disease 46 more distant branches.
3-vessel disease 80 When the left anterior descending artery had to be
LMT 11 revascularized with the RIMA, the RA was used on the
Mean EF (range) 0.53 (0.22-0.79) right coronary territories. This conduit had to be used as
>0.50 70 a donor vessel in a patient who had a LIMA damaged
0.50-0.36 54 proximally during the harvesting.
<0.35 6
Urgent/elective 34/96 END-TO-END ANASTOMOSIS. A free graft was used 19 times
Redo 6 to lengthen an RIMA or an LIMA directed to an otherwise
not reachable vessel or to execute a sequential anastomo-
EF = ejection fraction; LMT = left main trunk. sis. This technique allowed us to employ an IMA acciden-
tally cut or partially dissected. In this case the extremity of
ning from a few centimeters below the transverse umbil- the free graft and that of the donor conduit were obliquely
icalline to the groin. After its fascia is opened, the rectus mouthed to assure a wider anastomosis. The conduit then
muscle is displaced medially to expose the first part of the was treated in the same manner as above.
lEA. The artery is then followed proximally, up to about
2 ern from its origin, and distally up to its first muscular COMBINED END-TO-END AND END-TO-SIDE ANASTOMOSIS. In
collaterals, for about 6 cm of length. The vessel is flushed 2 patients both types of anastomoses were used. In these
with the same solution as above to wash out the blood cases the RIMA, too short to be employed as a donor
from the intimal layer; papaverine solution also is used to
immerse the artery before its use. Table 2. Patterns of Arrangement of the Arterial Grafts in
The RA is used only in the patients with normal Allen the Composite Conduits
test. The artery is harvested from a few centimeters down Pattern of Arrangement
its origin to the wrist, together with its satellite veins and
Inflow Branches or No. of No. of Distal
the surrounding adipose tissue. It is treated as the lEA. Conduit Lengthening Conduits Anastomoses
The free grafts always are anastomosed to the donor
conduit before extracorporeal circulation is established, Branches
laying them down on a folded pad put between the LIMA RIMA 9 19
branches of the sternal retractor beside the donor conduit LIMA lEA 48 97
to assure a relatively stable operating field. LIMA RA 29 62
The anastomosis between the donor conduit and the LIMA RIMA/lEA 3 9
free graft is performed as end-to-side, end-to-end, or LIMA lEA/lEA 3 9
both. LIMA IEAIRA 9 28
LIMA IEA/RIMA/IEA 1 4
END-TO-SIDE ANASTOMOSIS. The posterior surface of IMA RIMA LIMA 1 2
is always preferred as the site of the anastomosis of the RIMA lEA 5 10
free grafts. After the underlying fascia is incised and the RIMA RA 6 12
IMA carefully dissected for about 15 mm, the artery is RA LIMA 1 2
clamped proximally to the established point. Then the Total 115 254
donor conduit is incised for about 8 mm. The conduit is Lengthening
laid down on a folded pad put between the branches of LIMA IEA 3 5
the sternal retractor to make the operating field relatively LIMA RA 2 2
stable. Then the free graft, mouthed obliquely at its tip, is RIMA LIMA 1 1
anastomosed with 7-0 or 8-0 suture (Ethicon Inc, Somer- LIMA lEA 8 8
ville, NJ; or Sharpoint, Reading, PA). The donor conduit LIMA RA 5 8
is declamped proximally after the free graft has been Total 19 24
clamped temporarily and the site of the anastomosis has Lengthening/
been inspected carefully. The free graft is irrigated with branch
5 mL of the solution mentioned above, and its tip is finally LIMA RIMA/lEA 1 2
ligated with a hemostatic clip. LIMA RIMAIRA 1 2
To avoid bleeding from the little collaterals that are not Total 2 4
visible before pharmacologic dilation, hemostasis of the Grand total 136 282
free graft is checked again. In the case of the RA this lEA = inferior epigastric artery; LIMA = left internal mammary
procedure is particularly important. artery; RA = radial artery; RIMA = right internal mammary artery.
Ann Thorac Surg CALAFIORE ET AL 187
1994;58: 185-90 CORONARY REVASCULARIZATION WITH ARTERIAL GRAFTS

vessel, had to be cut down. Then it was anastomosed

end-to-side to the LIMA (donor conduit) and was length-
ened with the IEA (1 case) or with the RA (1 case).
All the anastomoses between the conduits were per-
formed under total heparinization and before establish-
ment of extracorporeal circulation. Cardiopulmonary by-
pass was instituted between the right atrium and the
ascending aorta in all patients. Myocardial protection was
achieved with intermittent blood cardioplegia delivered in
an antegrade fashion through the aortic root; cold cardio-
plegia was used in 41 patients and warm cardioplegia in
the remaining 89.
In all cases the donor conduit was anastomosed first to
tailor better the side branches when necessary. Only the
IMA was anchored to the epicardium with two stitches
passed through the muscular remnants of the distal part
of its pedicle.
The coronary anastomoses were performed with 7-0 or
8-0 sutures (Ethicon or Sharpoint).
Of 389 distal anastomoses, with an average of three
anastomoses per patient, 35 were venous and 354 were
arterial; 282 of the latter were performed using complex
preformed arterial conduits. There were six sequential
Fig 1. Left internal mammary artery (LIMA) (to diagonal branch)
anastomoses using 1 RIMA, 3 RAs, and two lengthened Y-branched with radial artery (RA) (to obtuse marginal side-to-side
conduits (LIMA to lEA and RIMA to RA). In 100 patients and to posterolateral branch end-to-side) at 7 days.
only arterial grafts were employed with a number of
anastomoses ranging from two to five.
The mean cardiopulmonary bypass time was 64 min- myocardial necrosis due to an acutely thrombosed IMA; 1
utes (range, 31 to 93 minutes) and the mean aortic patient died on the fourth day postoperatively of pancre-
cross-clamping time was 42 minutes (range, 20 to 69 atic shock. The overall in-hospital mortality rate was
minutes). 1.5%.
After aortic declamping a continuous intravenous infu- No patient received pharmacologic or mechanical sup-
sion of diltiazem was administered up to the first postop- port in the intensive care unit. Enzymatic or electrocar-
erative day. It then was administered orally at a dose of diographic signs were observed in 3 patients during the
60 mg three times a day (up to 6 months if the RA is stay in the intensive care unit; all conduits were angie-
present; up to 1 month in the other cases), and aspirin graphically patent and the only change was an apical
was given at a dose of 325 milligrams daily up to 2 years akinesia in 1 patient, compared with the preoperative
after operation. control.
The bleeding up to 24 hours after the operation was low
(mean standard deviation, 743 471 mL) and only 35
Results patients needed a blood transfusion (mean, 0.7 blood
There was no need of inotropic or mechanical support in units/patient).
the operating theater. One patient had to be reoperated on for bleeding and 2
One patient died on the first day postoperatively of for sternal dehiscence. The last 2 patients were diabetics
who had to be submitted to bilateral IMA harvesting.

Table 3. Angiographic Controls Angiographic Controls

Early angiographic controls have been performed be-
Early Control Late Control
(7-15 days) (6-14 mol tween the seventh and 15th days postoperatively; late
controls have been performed at a mean interval of 9.5
Arterial No. Controlledl Patency No. Controlledl Patency months after operation (range, 6 to 14 months). The
Conduit No. Patent (%) No. Patent (%)
results are reported in Table 3. We did not observed any
LIMA 43/43 100 31/31 100 graft disease or spasm, the patterns being full patency or
RIMA 35/35 100 23/22 95.6 complete occlusion. The more common angiographic fea-
IEA 34/34 100 21/20 95.2 tures of these preformed conduits are shown in Figures 1
RA 26/26 100 17/16 94.1 through 7.
RGEA 5/5 2/2
Clinical Outcome
lEA = inferior epigastric artery; LIMA left internal mammary
artery; RA = radial artery; RGEA = right gastroepiploic artery; At a mean follow-up of 7.2 months (range, 1 to 15 months)
RIMA = right internal mammary artery. all patients are alive without recurrence of symptoms.
188 CALAFIORE ET AL Ann Thorac Surg
CORONARY REVASCULARIZATION WITH ARTERIAL GRAFTS 1994;58:185-90

Fig 2. Left internal mammaryartery (LIMA) (to left anterior de-

scending) Y-branched with inferior epigastric artery (lEA) (to diago-
nal branch) at 13 months.

Fig 4. Right internal mammary artery (RIMA) (to diagonal branch)

Comment Y-branched with inferior epigastric artery (lEA) (to left anterior de-
scending) at 14 months.
The long-term results of myocardial revascularization
have improved steadily with the use of one or both IMAs
[1, 9, 10). This fact has led to increasing use of this artery
in primary or in repeated operations. Other arterial con-

Fig 3. Left internal mammaryartery (LIMA) (to obtuse marginal) Fig 5. Left internal mammary artery (LIMA) (to left anterior de-
Y-branched with inferior epigastric artery (lEA) (to diagonal branch) scending) Y-branched with radial artery (RA) (to obtuse marginal) at
at 12 months. 10 months.
Ann Thorae Surg
1994;58:185-90
duits recently have been introduced into clinical practice
in the belief that arteries could be the conduits of choice
for myocardial revascularization [5-7, 11-13]. However,
their use has not become popular for various reasons: the
lack of late angiographic controls, prolonged cardiopul-
monary bypass and cross-clamping times, and some tech-
nical difficulties. Moreover all these conduits (except
RGEA) can be used only as free grafts, so they need aortic
anastomosis either direct or with the interposition of a
pericardial patch [5].
It is generally accepted that the behaviour of the arterial
free grafts (IMA or different arteries) proximally anasto-
mosed to the aorta is similar to that of saphenous vein for
the first year after the operation. After this period they
show the same fate as the pedicled arterial conduits [8]. In
our opinion there could be two reasons. One is the
mismatch between the aorta and the conduit wall in terms
of thickness and diameter; this fact implies technical
difficulties in executing the anastomosis and could lead to
an impairment of flow. Moreover-and in our opinion
this is the main reason-these arteries are normally third-
or fourth-order branches of the aorta, so they usually are
submitted to a pattern of flow that is quite different and
cannot tolerate the flow of the ascending aorta. When
these conduits are anastomosed to the aorta the abrupt
increase in pressure wave can result in wall stretching
with intimal tearing and subsequent development of
premature hyperplasia.
Furthermore, the RA offers a constant diameter and a
length sufficient to reach almost all coronary branches.
The lEA, on the contrary, has a constant diameter only in
the first 5 to 6 em of its course; then it shows a progressive
and sensible caliber reduction. When it has to be anasto-
Fig 6. Right internal mammaryartery (RIMA) lengthened with ra-
dialartery (RA) to posterior descending artery at 10 months. The ar-
row indicates the end-to-end anastomosis between RIMA and RA.
CALAFIORE ET AL 189
CORONARY REVASCULARIZATION WITH ARTERIAL GRAFTS
Fig 7. Left internal mammaryartery (LIMA) (to left anterior de-
scending) V-branched with radial artery (RA) (to posterolateral
branch), V-branched with inferior epigastric artery (lEA) (to diagonal
branch) at 11 months.
mosed to the ascending aorta, the lEA must be harvested
in all its length, from the groin to the umbilicus, where the
diameter of the artery is very small; moreover, many
muscular collaterals must be divided, thereby exposing
the rectus muscle to ischemia. For these reasons we tend
always to use the IMA and the RGEA as pedicled grafts
and to branch or lengthen the IMA with the lEA or the
RA.
The technique described above is feasible and the
anastomoses so performed are completely safe; further-
more, they can be checked easily before the conduits are
put in site. The immediate and the mid-term results
reported in this series are encouraging. However, long-
term follow-up is needed to assess its value compared
with IMA.
References
1. Loop FO, Lytle BW, Cosgrove OM, et al. Influence of the
internal mammary artery graft on lO-year survival and other
cardiac events. N Engl J Med 1986;314:1-6.
2. Zeff RH, Kongatahworn C, Iannone LA, et al. Internal
mammary artery versus saphenous vein graft to the left
anterior descending coronary artery: prospective random-
ized study with lO-year follow-up. Ann Thorac Surg 1988;45:
451-4.
3. Barner HB, Standeven JW, Reese J. Twelve-year experience
with internal mammary artery for coronary artery bypass. J
Thorac Cardiovasc Surg 1985;90:668-75.
4. Mills NL, Everson CT. Right gastroepiploic artery: a third
conduit for coronary artery bypass. Ann Thorac Surg 1989;
47:706-11.
5. Puig LB, CiongoIIi W, Cividanes GL, et al. Inferior epigastric
artery as a free graft for myocardial revascularization. J
Thorac Cardiovasc Surg 1990;99:251-5.
6. Barner HB, Naunheim KS, Fiore AC, Fischer VW, Harris HH.
190 CALAFIORE ET AL
CORONARY REVASCULARIZATION WITH ARTERIAL GRAFTS
Use of the inferior epigastric artery as a free graft for
myocardial revascularization. Ann Thorac Surg 1991;52:
429-37.
7. Buche M, Schoevaerdts JC, Lavagie Y, et al. Use of the
inferior epigastric artery for coronary bypass. J Thorac Car-
diovasc Surg 1992;103:665--70.
8. Loop FD, Lytle BW, Cosgrove OM, Golding LAR, Taylor pc,
Stewart RW. Free (aorto-coronary) internal mammary artery
graft: late results. J Thorac Cardiovasc Surg 1986;92:827-31.
9. Galbut DL, Traad EA, Dorman M], et al. Seventeen year
experience with bilateral internal mammary artery grafts.
Ann Thorac Surg 1990;49:195--201.
10. Fiore AC, Naunheim KS, Dean P, et al. Results of internal
A Special Topical Meeting
Preparing Your Practice for Change:
Thoracic Surgery Into the Next Decade
Hyatt Regency Hotel
Atlanta, Georgia, September 24-25, 1994
"Preparing Your Practice for Change" will be a unique
topical meeting focusing on how the practice of thoracic
surgery is being affected as the healthcare reform process
sweeps the country. All those interested in keeping their
thoracic surgery practices well positioned in the future
will benefit from this meeting. Thoracic surgeons from all
practice arrangements and all areas of subspecialization as
well as surgical practice administrators are encouraged to
attend this meeting.
The meeting is sponsored by The Society of Thoracic
Surgeons under the direction of its Major Issues Commit-
tee.
The program will feature a variety of nationally re-
nowned experts including many members of The Society
who will discuss the major concerns and issues affecting
the decision of how to adapt a thoracic surgery practice for
the coming change.
The program will begin on Saturday, September 24, at
8:30 AM and will conclude on Sunday, September 25, at
1:00 PM. Session topics on Saturday morning include
discussion of the political and environmental forces that
affect thoracic surgery practices. Various practice patterns
Ann Thorae Surg
1994;58: 18S-90
thoracic artery grafting over fifteen years: single versus
double grafts. Ann Thorac Surg 1990;49:202-9.
11. Pym J, Brown PM, Charrette ED, Parker JO, West RO.
Gastroepiploic--coronary anastomosis. J Thorac Cardiovasc
Surg 1987;94:256-9.
12. Suma H, Wanibuchi Y, Terada Y, Fukuda S, Takayama T,
Furuta S. The right gastroepiploic artery graft. Clinical and
angiographic midterm results in 200 patients. J Thorac Car-
diovasc Surg 1993;105:615--23.
13. Acar C, [ebara V, Portoghese M, et al. Revival of the radial
artery for coronary bypass grafting. Ann Thorac Surg 1992;
54:652-60.
will be highlighted to identify critical issues affecting
successful adaptation to change. A special luncheon
speaker, H. Ross Perot, will address the Physician's Role
in Healthcare Reform. Saturday afternoon will focus on
getting the most from your healthcare dollar. Discussion
of contract negotiations for thoracic surgery services will
be undertaken by representatives of managed care orga-
nizations as well as hospital administrators. The feasibility
of capitation for thoracic surgery practice will be re-
viewed. The afternoon will conclude with an exploration
of quality assessment by database studies as well as
outcome analysis, both of which may be mandated in the
future. On Sunday morning the meeting will address the
role of industry in the new scheme of practice, and there
will be a discussion of thoracic surgical training and
distribution of specialty services as they relate to future
healthcare needs.
Further details on this meeting will be mailed to all
members of The Society in the near future and may be
requested from The Society of Thoracic Surgeons, 401 N
Michigan Ave, Chicago, IL 60611; (312) 644-6610.