Total Revascularization With T Grafts
Alfred J. Tector, MD, Susan Amundsen, PA-C, Terence M. Schmahl, MD,
David C. Kress, MD, and Mohan Peter, MD
Midwest Heart Surgery Institute and St. Lukes Medical Center, Milwaukee, Wisconsin
The T graft is constructed by anastomosing the proximal
end of the free right internal thoracic artery to the side of
the attached left internal thoracic artery. Besides adding
considerable reach to the right internal thoracic artery,
this technique allows the left anterior descending coro-
nary artery and its branches to be bypassed with the
attached left internal thoracic artery. Two hundred
eighty-seven patients, aged 34 to 86 years (mean age, 64.6
years) received an average of 4.4 internal thoracic artery
to coronary artery anastomoses. Sixty-nine patients had
left main disease, 33 were undergoing first-time reoper-
ations, and two were reoperated on for the second time.
C oronary artery bypass grafting is the most effective
treatment for patients with symptomatic three-
vessel coronary artery disease, as long as the graft and
any native coronary artery distal to the bypass graft
remain free of obstruction. The ideal coronary artery
bypass graft supplies ample blood to the ischemic areas of
myocardium during rest and exercise, and remains free of
obstruction from spasm or intraluminal deposits for the
remainder of the patients life. It should be accessible from
a median sternotomy incision and not be in the way if
reentry of the sternum is required at a later time. The
For editorial comment, see page 8.
internal thoracic artery (ITA) has remained patent for
more than 20 years, and is the bypass graft that is most
apt to meet all of these requisites.
After Demikhov [l]first successfully bypassed the left
anterior descending coronary artery (LAD) with the left
ITA (LITA) in dogs with good long-term results, Vineberg
[2] used the LITA and right ITA (RITA) to indirectly
revascularize the entire myocardium in humans. In 1960
Goetz and associates [3] performed the first ITA coronary
artery bypass in a human, grafting the RITA to the right
coronary artery. The LITA was first anastomosed to the
LAD in humans in 1965 by Kolessov [4]. Green and
colleagues [5] popularized the technique and introduced
the concept of using magnification to improve the preci-
sion of constructing these anastomoses. The use of bilat-
eral ITA grafts was first reported by Suzuki and associates
in 1973 (61.
Presented at the Twenty-ninth Annual Meeting of The Society of Thoracic
Surgeons, San Antonio, TX, Jan 25-27, 1993.
Address reprint requests to Dr Tector, 2901 W Kinnickinnic River Pkwy,
Milwaukee, WI 53215.
0 1994 by The Society of Thoracic Surgeons
Ejection fraction ranged from 0.20 to 0.70. Operative
mortality was 1.7%.Twenty-six patients had postopera-
tive graft visualization, and 94.7% of the grafts were
open. All 45 bypass grafts from the left internal thoracic
artery were patent, and 91% of those from the right
internal thoracic artery were unobstructed. This proce-
dure requires technical skill with internal thoracic arter-
ies, but it has the potential of significantly improving
long-term event-free survival and reducing the need for
reoperation in patients undergoing coronary artery by-
pass grafting.
(Ann Thorac Surg 1994;57:33-9)
As coronary artery bypass grafting has progressed from
an experimental procedure to an accepted clinical treat-
ment of coronary atherosclerosis, it has been clearly
demonstrated that bypassing the LAD with an attached
LITA significantly improves long-term survival [7-121.
This information strongly encouraged us [12] and others
[13] to extend the use of the ITA to bypass as many
coronary arteries as possible with a combination of bilat-
eral attached and free ITAs, sequential bypass grafts, and
Y grafts. It is easy to reach most coronary arteries and
their branches with ITA grafts. Inadequate length of the
free RITA, however, has often limited our ability to
bypass the posterior marginal branches of the circumflex
and right coronary arteries with ITA grafts [12].
The concept of anastomosing another bypass graft to an
attached 1TA was introduced by Mills [14] to avoid placing
vein grafts into a severely atherosclerotic ascending aorta.
Sauvage [15] has employed this technique to bypass all
coronary arteries supplying the left ventricle with ITA
grafts. To overcome the deficiency in length of the free
RITA and to always bypass the LAD with the attached
LITA, we have anastomosed the RITA to the side of the
attached LITA forming a T graft. This can place the RITA
10 cm or more closer to the distal circumflex and right
coronary artery branches, allowing it to reach these ves-
sels. Throughout the years many surgeons have cast
doubts concerning the ability of the ITA to deliver ade-
quate flow even to single coronary arteries [16, 171. As we
have extended the number of coronary arteries and their
branches bypassed with ITA grafts, we have found like
others [13] that the meticulously prepared and anasto-
mosed ITA has sufficient flow to supply the entire myo-
cardium. The T-graft technique is described in detail and
the results of its use in 287 patients operated on from
September 1990 through August 1992 are reported.
0003-4975/94/$7.00
 TECTORETAL
TOTAL ITA REVASCULAREATION
Material and Methods
Piztients
TWOhundred eighty-seven patients (62 women and 225
men) whose ages ranged from 34 to 86 years (mean age,
64.6 years) were totally revascularized with ITA T grafts
between September 1990 and August 1992. Indications for
the operation were severe angina, a positive stress test, or
a prior coronary event. The patients had significant ob-
structions in all three coronary arteries and received an
average of 4.4 ITA coronary grafts per patient. In addi-
tion, 69 of the patients had left main stenosis greater than
60%, and 33 patients in the group were operated on for
the second time (two for the third time). Ejection fractions
ranged from 0.20 to 0.70. Preoperative stress tests were
performed in 110 patients, and all had positive results.
Two to four coronary anastomoses were made with the
RITA, whereas one to four anastomoses were constructed
with each LITA graft. Some of our patients with valvular
disease as well as three-vessel coronary artery disease
received ITA T grafts, but these grafts were performed on
a more selected basis and were not included in the series.
A few patients had left subclavian artery stenosis and the
LITA was detached and anastomosed to the attached
RITA, but these patients were excluded from the series.
Operative Technique
The heart is exposed through a median sternotomy inci-
sion, and an inverted T incision is made in the pericar-
dium. The right and left pleural space;s are entered and
the ITAs are examined for adequacy of pulse. High power
6x magnification is used for dissection, preparation, and
anastomosis of the ITA grafts.
DISSECTION OF THE LEFT INTERNAL THORACIC ARTERY. With
a Morris retractor the left hemithorax is sprung upward
and to the left to facilitate mobilization of the LITA. The
cautery blade is covered with a piece of gum rubber or
plastic tubing, leaving only the tip exposed to prevent
damage to the ITA. After heparin is administered, dissec-
bon of the ITA begins at its bifurcation into the superior
epigastric and musculophrenic arteries. Although some
surgeons have had good success with skeletonized ITA
grafts [13], our preference is to dissect the ITA with a
pedicle of its surrounding muscle, fat, and fascia. This
minimizes disruption of the internal elastic lamina, which
can lead to intimal hyperplasia as was discovered in dogs
when skeletonized ITAs were compared histologically
with ITA pedicle grafts [18]. When the surrounding tissue
is left attached to the ITA, the chance of twisting the ITA
is reduced and the graft can be easily sutured to the
epicardium. Careful dissection proceeds proximally to the
thoracic inlet, and caution is taken not to injure the
subclavian vein. Much of the tissue between the ITA and
chest wall is separated bluntly with the cautery blade, and
the branches are cauterized. The LITA is cut distally
including a few centimeters of the superior epigastric and
musculophrenic branches.
Ann Thorac Surg
1994;573>9
PREPARATION OF THE LEFT INTERNAL THORACIC ARTERY.
Arterial and venous branches are secured with hemostatic
clips. The ITA is examined for adequacy of pulse, size,
injury, dissection, and the presence of atherosclerosis.
Some areas of injury can be repaired without compromis-
ing the ITA flow. The investing fascia is dissected to
expose the adventitia on the posterior surface of the
artery. One of the most critical steps in preparation of the
LITA, as it is responsible for the entire flow to the bypass
grafts, is dilating it with papaverine. We carefully inject a
solution of 1 mL papaverine in 30 mL saline solution into
the distal end of the LITA after the LITA is occluded with
a bulldog clamp near its origin. We agree with Mills and
Bringaze [19] that gentle injection of the papaverine
solution directly into the lumen of the ITA allows for the
greatest contact between the receptors in the ITA wall and
the solution. It is probably the vasodilatory effect of the
papaverine on the receptors in the ITA wall, and not
hydrostatic pressure, that gives maximum distention and
flow to the LITA. Forceful hydrostatic dilation can injure
the ITA. Since we have adopted this technique, almost
every uninjured ITA is capable of producing a sufficient
amount of "critical flow" (the amount of flow the ITA is
immediately capable of delivering to the ischemic myo-
cardium after anastomosis) [lo].
DISSECTION OF THE RIGHT INTERNAL THORACIC ARTERY. The
right hemithorax is sprung upward and to the right with
the Morris retractor. The RITA is mobilized in a similar
manner as the LITA. Near the thoracic inlet the RITA is
dissected free from the thymus gland to increase the
available length. The phrenic artery and vein may be cut,
but care must be taken not to injure the phrenic nerve.
After the RITA is detached from its origin the investing
fascia is dissected on the posterior surface down to the
adventitia. This exposes the artery for performing the
side-to-side anastomoses, increases length by 5% to 15%
[20], and identifies constricting bands. Papaverine solu-
tion is gently injected into the proximal end of the free
RITA after the artery is clamped at its most distal end with
a bulldog clamp.
CONSTRUCTION OF THE T GRAFT. To bypass all three coro-
nary arteries with ITA grafts it is essential to obtain
maximum graft length. Some measures that help to ac-
complish this include (1)dissection of the ITA from below
its bifurcation into the superior epigastric and musculo-
phrenic branches up to the thoracic inlet, (2) freeing the
investing fascia from the posterior wall of the ITA, and,
most importantly, (3) anastomosis of the RITA to the
attached LITA, instead of the ascending aorta, as a T graft.
Attaching the RITA in the form of a T not only enhances
its reach but prevents kinking at the anastomosis.
A slit is made in the ITA pedicle perpendicular to the
artery, on the side opposite the surgeon, where the T graft
is to be constructed. Then a vertical 1-cm incision is made
in the posterior wall of the LITA at the level of the left
atrial appendage (if the LITA is lifted from the chest, it is
the same site where it crosses the anterior table of the
sternum). Care is taken not to injure the intima on the
Ann Thorac Surg
1994:573>9
A B
Fig 1 . Technique of T-graft anastomosis. A 1 cm parallel incision is made into the attached left internal thoracic artery (LITA). The proximal end
of the detached right internal thoracic artery (RITA) is anastomosed to the side of the LITA with a continuous suture. (A) The anterior surface is
sutured first. The RITA is then turned over and lifted out of the chest, and (B) the posterior row is completed. (C) Completed amstornosis.
anterior wall of the ITA when making the arteriotomy.
The anastomosis is constructed with a continuous 8-0
monofilament suture (Fig 1).The edges of the pedicles are
sutured together to prevent twisting at the anastomotic
site. The clamp is removed, and the LITA and RITA are
checked for adequacy of flow and pulse. If the flow is
questionable or the pulse is diminished in the RITA, the
anastomosis should be taken down and redone because it
is likely narrowed. The LITA pedicle is attached to the
epicardium near the LAD. This makes it easy to measure
the appropriate length of graft needed to reach the first
coronary artery or branch to be bypassed with the RITA
and LITA.
CONSTRUCTION OF INTERNAL THORACIC ARTERY TO CORO-
NARY ARTERY ANASTOMOSES. The ascending aorta or aortic
arch is cannulated. A two-staged venous cannula is intro-
duced into the right atrium, and a coronary sinus catheter
is placed for retrograde perfusion of cardioplegia. A
13-gauge needle is inserted into the ascending aorta for
venting and infusion of antegrade blood cardioplegia.
Cardiopulmonary bypass and ischemic cardiac arrest are
used. The myocardium is protected with antegrade and
retrograde cardioplegia as has been described by Buck-
bergs group [21]. Cold blood induction cardioplegia is
administered antegrade and retrograde, then cold main-
tenance blood cardioplegia is infused every 15 to 20
minutes, or if cardiac activity occurs. Warm blood sub-
strate-enriched cardioplegia is given antegrade and retro-
grade after completion of the last anastomosis.
The ITA coronary anastomoses are constructed with 8-0
monofilament suture while the patient is on cardiopulmo-
nary bypass during ischemic arrest after infusion of cold
blood cardioplegia. Details of the technique are described
elsewhere [lo]. The side-to-side anastomoses are only
c o n s t i ~ ~ t eparallel
d to the coronary artery. We no longer
TECTOR ET AL 35
TOTAL ITA REVASCULARIZATION
C
use perpendicular side-to-side anastomoses because they
are at risk of narrowing. Because native arteries do not
run in a perfectly straight course, gentle curves are made
in the ITAs between each anastomosis. Terminal end-to-
side anastomoses are constructed either parallel or per-
pendicular to the coronary artery. The ITA pedicle is
attached to the epicardium with sutures to prevent twist-
ing or kinking.
The causes of each operative death and major compli-
cations are reported below. Postoperative infarctions were
diagnosed by electrocardiographic changes or myocardial
enzyme level elevation. One hundred seven patients had
postoperative stress tests, and 26 patients underwent
postoperative catheterization with ITA visualization on
angiography. Data are expressed as the mean standard *
deviation where applicable.
Results
Five patients (1.7%)died within the first 30 days after the
procedure. Three patients had cerebrovascular accidents:
1 patient had basilar artery insufficiency; 1 patient had
embolization to the brain, kidneys, and lower extremity;
and 1 patient fully awakened and had a massive stroke 12
hours later.
Two patients died of cardiac-related causes. One pa-
tient had low cardiac output and cardiac arrest in the
intensive care unit. He was returned to the operating
room, and a saphenous vein graft (SVG) was anasto-
mosed to the acute marginal and the posterior descending
branches of the right coronary artery. Right ventricular
failure developed, and the patient could not be weaned
from cardiopulmonary bypass. The other patient was
stable postoperatively for 8 hours and was not receiving
any inotropic support. He had a sudden cardiac arrest and
was returned immediately to the operating room and
36 TECTOR ET AL A n n Thorac Surg
TOTAL ITA REVASCULARIZATION 1994;5733-9

placed on cardiopulmonary bypass. There was a dissec-

t:ion of the LITA caudad to the T graft attachment in the
region where the bulldog clamp was placed, and no flow
was coming through the LITA. An SVG was anastomosed
t'o his LAD, diagonal, and circumflex coronary arteries but
the patient could not be weaned from cardiopulmonary
bypass.
The average cross-clamp time for the group was 77.8 &
15 minutes. Sixty-four patients (22%) required small to
moderate amounts of dopamine or do'butamine for sev-
eral hours postoperatively. Fifteen patients (5%) needed
em intraaortic balloon pump for a few days after the
operation. One of these patients had the intraaortic bal-
loon pump placed preoperatively, and 2 patients were
urgent cases. Five patients who received intraaortic bal-
loon pumps had reoperations. The average left ventricular
ejection fraction of the 15 patients was 0.34 (range, 0.20 to A
0.57). None of these patients had postoperative myocar-
dial infarctions. This suggests that the balloons were
required for myocardial stunning in an already compro-
mised ventricle rather than inadequate ITA flow.
We immediately added SVGs in 4 patients who mani-
fested decreased wall motion on echocardiography before
coming off cardiopulmonary bypass. These patients had
uneventful postoperative courses and were not included
in the study. Four other patients were returned to the
operating room from the intensive care unit. Three of
these patients had cardiac arrests, and 1 had electrocar-
diographic changes of an anterior myocardial infarction.
Saphenous vein grafts were placed in all of these patients.
Two of these patients, described earlier in detail, died.
'The other 2 survived without further difficulty.
Three patients (1%)were diagnosed. as having periop-
Ierative myocardial infarcts. Sixteen patients (5.6%) re-
Iquired reexploration for bleeding. Ten patients (2.9%)had
sternal wound infections and required muscle flaps for
closure. Six of these patients were diabetic: 2 were insulin-
dependent and 4 were non-insulin-dependent.
Twenty-six patients have been restudied by angiogra- B
phy, and 107 of 113 coronary anastomoses evaluated Fig 2 . Postoperative studies of two diflerent Tgrafts.
(94.7%)were patent (Fig 2). None of the 45 grafts from the
attached LITA to the LAD and diagonal branches were
closed. Sixty-two of the 68 grafts (91%) from the RITA
were open. Five of the six occluded grafts had been positive preoperative stress test result that was un-
anastomosed to the right coronary artery branches, and changed postoperatively. This patient was asymptomatic.
the other was placed to the second marginal branch of the In 3 patients the stress test was stopped because of patient
circumflex artery. In 1 patient the portion of the RITA fatigue and there were no electrocardiographic changes.
between its attachment to the LITA and the first anasto- A final patient had undergone a reoperation and did not
moses to the most anterior marginal branch of the circum- have postoperative chest pain.
flex artery was occluded. Minimal obstruction in the first
circumflex marginal branch was present in this patient,
and this branch perfuses the remainder of the RITA graft,
Comment
which is patent. The ultimate achievement in coronary artery bypass graft-
One hundred seven patients have undergone postop- ing is to have all patients undergoing the operation
erative stress tests, of which 96 (89.7%) had negative survive free from coronary events, angina, and reopera-
results and 11 had positive results. Five of the patients tion for as long as possible. Knowing that the greatest
with positive postoperative stress test results have under- cause of failure of the operation is obstruction of the
gone coronary angiography and had one or more grafts bypass graft, selection of an ideal bypass graft that re-
occluded. One patient had all grafts patent on restudy. mains free of obstruction appears to be one of the most
Another patient had extensive ventricular scarring and a important factors in achieving this goal. Kroncke and
Ann Thorac Surg
1994;573?-9
Fig 3. Postoperative study of left internal thoracic artery graft placed
21 years ago by us.
associates [22] reported that only 14% of patients studied
5 years after revascularization demonstrated progression
of atherosclerosis or new atherosclerosis in their coronary
arteries distal to the anastomoses. Furthermore, bypass-
ing just the LAD with the attached LITA and using SVGs
for the circumflex and right coronary arteries has been
shown to significantly enhance survival 10 or more years
after operation [7, 10, 111. The ITA remains free from
obstruction because of its low risk of intimal thickening
due to a more elastic media [23] and its freedom from
significant atherosclerosis [24]. We have seen ITAs at
recatheterization that are widely patent and entirely free
from obstruction 21 years after grafting (Fig 3).
Anastomosing the free RITA to the attached L E A can
bring the RITA as much as 10 cm or more closer to the
distal circumflex and right coronary artery branches. This
makes it possible for us to totally bypass nearly every
patient with three-vessel disease with ITA grafts. With
this procedure, the principle of bypassing the LAD (the
most important coronary artery) with the attached LITA,
the graft with the greatest patency, is preserved [25]. Leg
incisions and their complications are eliminated. The use
of other arteries that are more prone to spasm and intimal
hyperplasia is avoided [23]. No graft is anastomosed to
the ascending aorta, reducing the chance of embolization
and the hindrance of a graft on the aorta in the event that
an aortic valve operation is necessary at a later date.
Sternal reentry is safer if required in the future because
the ITAs are well out of the way, and the chance of
atherosclerotic emboli from dissecting near the bypass
grafts as seen in the SVG is eliminated.
Reliance upon the attached LITA to supply sufficient
flow to meet the demands of all of the revascularized
areas, and the increased incidence of sternal infection
TECTOR ET AL 37
TOTAL ITA REVASCULARIZATION
when both ITAs are dissected from the chest wall, are the
two main drawbacks of the procedure. Proximal injury to
the LITA can impair flow and be life threatening, as
occurred in 1 of the patients who died. If the left subcla-
vian artery is known to be obstructed at the time of
operation, the LITA can be detached and anastomosed to
the attached RITA as a T graft. If later atherosclerosis
causes obstruction of the left subclavian artery, a left
carotid-subclavian artery bypass, axillary-axillary artery
bypass, endarterectomy, or subclavian angioplasty can be
performed. In one instance, we detached the LITA and
anastomosed it to the ascending aorta. A bypass graft can
also be interposed between the ascending aorta and the
LITA.
The incidence of sternal infection was somewhat higher
in this group of patients, particularly in the patients with
diabetes. Mobilization of the ITA devascularizes the ster-
nal side from which it is dissected [26]. Strict adherence to
sterile technique, careful opening of the sternum in the
middle, tight secure approximation of the sternum with
many sutures, and elimination of dead space when clos-
ing the fascia are measures that will most likely assist in
reducing infections in the diabetic patients.
We have noticed inadequate perfusion by ITA grafts on
rare occasions through the years. Often this can be
detected, before or just after coming off cardiopulmonary
bypass, by electrocardiographic changes or new wall
motion abnormalities on the transesophageal echocardio-
gram. Immediate placement of an SVG to the area sus-
pected of insufficient perfusion is usually the only mea-
sure that corrects the problem. The cause of this
hypoperfusion has been attributed to spasm [27] and flow
capacity [28] of the graft. We believe, however, that
reduction in ITA flow is most often the result of even
slight narrowing of ITA coronary anastomoses or kinking
or stretching of the ITA. On occasion, when flow through
the ITAs after completion of the T graft was inadequate,
redoing the anastomoses resulted in excellent flow in both
limbs of the graft. Another time, inadequate flow to the right
coronary branch was caused by kinking of the artery, from
the way the pedicle was sutured to the epicardium. Chang-
ing the stitch released the kink in the ITA, and the
electrocardiogram and echocardiogram immediately re-
turned to normal.
If there is any doubt about the adequacy of flow, the
safest approach is to place an SVG to the suspected area
before removing the patient from cardiopulmonary by-
pass. Initially we were extremely cautious of hypoperfu-
sion, but as we attain experience with T grafts we rarely
see this problem.
It would be ideal to obtain postoperative visualization
of the grafts in all 287 patients. Postoperative studies are
becoming more and more difficult to obtain because of
refusal by the patient, insurance companies, and Medi-
care. The graft patency in the 26 patients who underwent
postoperative graft visualization is acceptable. It was
encouraging to see that the high patency of the LITA
attached to LAD, as attained in the past, was achieved in
this group. The one T-graft failure may have been a result
of competition of flow with the first circumflex marginal
38 TECTOR ET AL
TOTAL ITA REVASCULARIZATION
branch that was bypassed and had minimal disease. The
greatest incidence of graft failure occurred at the distal
end of t h e RITA. Stretching or kinking of t h e graft, o r
possibly the presence of diffusely diseased right coronary
artery branches, m a y be responsible for these closures.
Another possible reason for the closure of the RITA
end-to-side anastomosis is that the superior epigastric o r
musculophrenic branch was used. Although s o m e au-
thors [29] report increased graft closure i n these vessels
when used as a Y graft, we have found satisfactory
patency i n the Y grafts we have placed.
Total revascularization with the ITA, the most ideal
bypass graft, has the potential to significantly increase
long-term event-free survival and dramatically reduce t h e
need for reoperation i n patients with three-vessel disease.
The procedure can be performed w i t h acceptable mortal-
ity i n patients of all ages, including those with poor
ventricles o r left m a i n disease, and even i n some patients
requiring reoperation. The average cost per case i n pa-
tients operated on t h e first time w i t h reasonable ventric-
ular function was $18,000, and t h e average cost per
multiple-vessel a n g i o p l a s t y i s more than half t h a t
a m o u n t . In patients requiring a second angioplasty, total
revascularization w i t h T grafts could be the more cost-
effective form of treatment. Dedication to impeccable
precision of technique i n the preparation and anastomosis
of ITA grafts is p a r a m o u n t to the success of this proce-
dure. We can recommend it to the cardiovascular surgeon
with extensive experience and interest i n ITA grafting.
We are indebted to all of the cardiologists for performing post-
operative angiography and graft visualization, to Mr Alfred
Anderson for statistical advice, and to Kelly Tector for preparing
the manuscript.
Supported by a grant from the St. Luke's Medical Center Foun-
dation.
References
1. Demikhov VP. Experimental transplantation of vital organs.
(Authorized translation by Basil Haig, MA, MB, and B.
Chin.) New York: Consultants Bureau, 1962:221.
2. Vineberg AM. Development of an anastomosis between the
coronary vessels and a transplanted internal mammary ar-
tery. Can Med Assoc J 1946;55:117-9.
3. Goetz RH, Rohman M, Haller JD, Dee R, Rosenak SS.
Internal mammary-coronary artery anastomosis. A nonsu-
ture method employing tantalum rings. J Thorac Cardiovasc
Surg 1961;41:378-86.
4. Kolessov VI. Mammary artery-coronary artery anastomosis
as a method of treatment for angina pectoris. J Thorac
Cardiovasc Surg 1967;54:53544.
5. Green GE, Stertzer SH, Reppert EH. Coronary arterial by-
pass grafts. Ann Thorac Surg 1968;5:443-50.
6. Suzuki A, Kay EB, Hardy JD. Direct anastomosis of the
bilateral internal mammary artery to the distal coronary
artery, without a magnifier, for severe diffuse coronary
atherosclerosis. Circulation 1973;48(Suppl 3):190-7.
7. Loop FD, Lytle BW, Cosgrove DM, et al. Influence of the
internal mammary artery graft on 10 year survival and other
cardiac events. N Engl J Med 1986;314:1-6.
8. Okies JE, Page US, Bigelow JC, et al. The left internal
mammary artery: the graft of choice. Circulation 1984;
7O(Suppl 1):213-21.
Ann Thorac Surg
1994;5733-9
9. Ochsner JL. Institutional variations influencing results: arte-
rial versus venous grafts. Ciirculation 1982;65(Suppl 2):81-5.
10. Tector AJ, Schmahl TM, Crouch JD, Canino VR, Heckel RC.
Sequential, free and Y intemal thoracic artery grafts. Eur
Heart J 1989;1O(Suppl H):7--7.
11. Acinapura AJ, Jacobowitz IJ. Kramer MD, Zisbrod Z, Cun-
ningham JN. Internal mammary artery bypass: thirteen years
of experience. Influence of angina and survival in 5125
patients. J Cardiovasc Surg 1.992;33:554-9.
12. Tector AJ, Schmahl TM, Canino VR. Expanding the use of
the internal mammary artery to improve patency in coronary
artery bypass grafting. J Thorac Cardiovasc Surg 1986;
91~9-16.
13. Sauvage LR, Wu HD, Koivalsky TE, et al. Healing basis and
surgical techniques for complete revascularization of the left
ventricle using only the internal mammary arteries. Ann
Thorac Surg 1986;42:449-65.
14. Mills NL. Physiologic and technical aspects of internal mam-
mary artery coronary artery bypass grafts. In: Cohn LH, ed.
Modern techniques in surgery. Cardio-thoracic Surgery. Mt.
Kisco, NY: Futura, 1982;48:1-19.
15. Sauvage LR. Extensive myocardial revascularization using
only internal thoracic arteries for grafting the anterior de-
scending, circumflex and right systems. In: Meyens WO, ed.
CABG update. Part 11. Cardiac surgery state of the art
reviews. Philadelphia: Hanky and Belfus, 1992;6:397-419.
16. Kouchoukos NT. Coronary artery bypass grafts: how many?
What kind? Ann Thorac !hrg 1976;22:60&2.
17. Flemma RJ, Singh HM, Tector AJ, Lepley D Jr, Frazier BL.
Comparative hemodynamic properties of vein and mammary
artery in coronary bypaiss operations. Ann Thorac Surg
1975;20:619-27.
18. Daly RC, McCarthy PM, Orszulak TA, Schaff HV, Edwards
WD. Histologic comparison of experimental coronary artery
bypass grafts. J Thorac Cardiovasc Surg 1988;9619-29.
19. Mills NL, Bringaze WL. Preparation of the internal mammary
artery graft. Which is the best method? J Thorac Cardiovasc
Surg 1989;98:73-9.
20. Fogarty TJ, Mollenauer KH. Preparation of the internal
mammary artery for coronary artery bypass grafting. J Car-
diac Surg 1991;6:322-5.
21. Buckberg GD. Antegradeiretrograde blood cardioplegia to
ensure cardioplegic distribution: operative techniques and
objectives. J Cardiac Surg 1'989;4:21638.
22. Kroncke GM, Kosolcharoen P, Clayman JA, et al. Five-year
changes in coronary arteries of medical and surgical patients
of the Veterans Administration Randomized Study of Bypass
Surgery. Circulation 1988;7S(Suppl 1):1&50.
23. Van Son JAM, Smedts F, de Wilde PCM, et al. Histological
study of the internal mammary artery with emphasis on its
suitability as a coronary artery bypassgraft. AnnThorac Surg
1993;55:10&13.
24. Kay HR, Korns ME, Flemma RJ, Tector AJ, Lepley D.
Atherosclerosis of the internal mammary artery. Ann Thorac
Surg 1976;21:504-7.
25. Tector AJ, Schmahl TM, Cainino VR. The internal mammary
artery graft: the best choice for bypass of the diseased left
anterior descending coronary artery. Circulation 1983;
68(Suppl 2):214-7.
26. Seyfer AE, Shriver CD, Miller TR, Graeber GM. Sternal blood
flow after median sternotomy and mobilization of the inter-
nal mammary arteries. Surgery 1988;104:899-904.
27. Sarabu MR, McClung JA, Fass A, Reed GE. Early postoper-
ative spasm in left internal mammary artery bypass grafts.
Ann Thorac Surg 1987;44:199-200.
28. Jones EL, Lattouf OM, Weintraub WS. Catastrophic conse-
quences of internal mammary artery hypoperfusion. J Thorac
Cardiovasc Surg 1989;98:902-7.
29. Morin JE, Hedderich G. Poirier NL, et al. Coronary artery
bypass using internal mammary artery branches. Ann Tho-
rac Surg 1992;54:9114.
Ann Thorac Surg
1994;5733-9
DISCUSSION
DR LEO CUELLO (San Antonio, TX): I rise to congratulate Dr
Tector for an excellent presentation and for his excellent results
with the T graft. Second, I want to make the members of The
Society aware that this report is important as a piece of evidence
on medicolegal cases.
As you are aware, there are a few members of this Society who
are testifying for the plaintiff lawyers. Last year a very prominent
member of this Society under oath told jurors that it is a
physiologic impossibility for this procedure to supply the left
ventricle. He also stated that it would be absolutely impossible
for a patient to survive with such a graft when the right coronary
artery is occluded even if you had added a side branch to it.
DR TECTOR This group of patients and previous work in the
past by us and other investigators around the world have to most
everyones satisfaction demonstrated that the internal thoracic
artery (that is properly prepared and constructed) is capable of
supplying a sufficient amount of blood to the myocardium in
almost every patient.
DR DELOS M. COSGROVE I11 (Cleveland, OH): Doctor Tector,
I congratulate you on excellent results. I realize that your refer-
ence to the T graft may be the shape of the anastomosis but it also
might refer to Tector. This is not an easy anastomosis to
perform. Obviously this is not to be taken lightly as one starts
ones career as a revascularization surgeon.
I have three questions. First, you are putting all of your eggs in
the basket of the left internal thoracic artery; are you concerned
about the development of atherosclerosis in the subclavian artery
or the potential for dissection or injury to the left internal thoracic
artery? Second, how often have you had to abandon this proce-
dure for reasons of moderate flow? Finally, why do you find that
it is an advantage for you to anastomose to a small internal
thoracic artery as opposed to doing a free graft of the right
internal thoracic artery to the aorta?
TECTOR ET AL 39
TOTAL ITA REVASCULARIZATION
DR TECTOR. Thank you for your comments. In answer to the
first question, I have thought about what would happen if
obstruction to the left subclavian artery developed years after this
procedure, and there are a number of ways that this can be
remedied.
If the obstruction happens late, left carotid subclavian bypass,
axillary artery-to-axillary artery bypass, endarterectomy of the
subclavian artery, or angioplasty could be performed. Also, in 1
patient we detached the left internal thoracic artery and anasto-
mosed it directly to the aorta. You could also place a graft from
the ascending aorta to the left internal thoracic artery. In a few
patients in whom we have noticed subclavian obstruction at the
time of operation we have left the right internal thoracic artery
attached and anastomosed the left internal thoracic artery to the
right internal thoracic artery as a T graft.
I do not recall having to completely abandon the procedure,
but there have been several instances where we have added
saphenous vein grafts, particularly when we first started per-
forming this operation. We rarely add vein grafts any more.
As far as the T graft anastomosis is concerned, I think that the
patency of the T graft is considerably better than it is in the free
graft that is anastomosed to the ascending aorta. This is probably
because the internal thoracic artery is considerably thinner than
the ascending aorta. Also, as I mentioned before, it does add
significant length and allows the right internal thoracic artery to
reach all the posterior branches of the myocardium.
DR JOHN W. HAMMON, JR (Winston-Salem, NC): Doctor
Tector, I know it is difficult to recatheterize patients after these
procedures, but do you have any other objective data such as
thallium exercise testing in these patients to document that they
are not having ischemia during exercise postoperatively?
DR TECTOR: One hundred seven of the patients have had stress
tests postoperatively, and 90% of these stress tests have had
negative results; we are in the process of trying to find out how
many patients complain of angina in the postoperative period.