JOURNAL

OF SURGICAL

Collateral

RESEARCH

53,578-587

(1992)

Blood Flow from Skeletal Muscle to Normal Myocardium’

JOHN D. MANNION, M.D.,2 PETER D. BUCKMAN, M.D., MICHAEL G. MAGNO, Department

of Surgery, Division

of Cardiothoracic

Surgery, Jefferson Submitted

Medical

for publication

PH.D., ANDFRED DIMEO,

College, 1025 Walnut Street, Philadelphia,

B.S.

Pennsylvania

19107

June 24. 1991

eta1 muscle, skin, omentum, lung, and jejunum were studied both clinically and experimentally [l-7]. Although the concept of indirect revascularization led to the development of the Vineberg procedure, in which the internal mammary pedicle was implanted into the myocardium [8-lo], indirect revascularization was ultimately abandoned. Clinical studies lacked objective evidence of benefit. Experimental studies were hampered by an inability to quantitate blood flow to ischemic myocardium and to distinguish between blood flow derived from intramyocardial or extramyocardial collaterals. There are both clinical and experimental reasons why indirect myocardial revascularization merits reinvestigation. First, there is a clinical need. There is a large cohort of patients with extensive distal vessel coronary disease who would benefit from an effective indirect revascularization procedure. Also, there is extensive clinical evidence that a well-developed collateral network can be beneficial. [ll-191 An experimental reason to restudy indirect revascularization lies in the development of new methodologies [20, 211 to measure blood flow. The use of microspheres permits a quantification and differentiation of the sources of collateral blood flow, allowing an accurate assessment of the utility of a muscle graft. Finally, it is known that a chronically ischemic area of human myocardium receives its collateral blood supply through a plexus of vessels located in the subendocardium [ 22,231. Recent research has been directed toward understanding this system [24, 251. The establishment of a complementary epicardial source of collaterals is also feasible. For these reasons, we attempted to induce collateral formation from skeletal muscle to the epicardial surface of the heart and quantitate skeletal muscle to myocardial collateral blood flow in a goat model.

Collateral blood vessels from skeletal muscle to myocardium might supplement intramyocardial collaterals during periods of acute myocardial ischemia. This study was conducted to verify the existence of such collaterals and to measure their contribution to collateral flow. In 12 male goats, the innate coronary collateral system to a moderate size myocardial risk area was defined with colored microspheres, and a latissimus dorsi pedicle flap was then apposed to the heart. After 3 weeks, skeletal muscle to myocardial collaterals were characterized by (a) creation of vascular casts (three animals); (b) estimation of skeletal muscle to myocardial collateral blood flow (three animals); and, (c) measurement of total collateral blood flow to the risk area (innate plus skeletal muscle to myocardial collateral flow). Under a dissecting microscope the vascular casts revealed direct communications from the skeletal muscle which penetrated deeply into the myocardium. With the coronary artery to the risk area open, the estimated myocardial collateral blood flow derived from the muscle flap was 0.01,0.02, and 0.04 ml/min. With the coronary artery to the risk area closed, there was no significant increase in total coronary collateral blood flow. Although the quantity of blood flow delivered by skeletal muscle collaterals was small, this study demonstrates that clearly identified collateral blood vessels form between skeletal muscle and myocardium in a cardiomyoplasty model. This raises the possibility that, under conditions more favorable to their development, extramyocardial collaterals from skeletal muscle might be exploited to augment the intramyocardial col0 1992 Academic Press, Inc. lateral system.

INTRODUCTION Methods to indirectly revascularize ischemic myocardium were examined extensively before the advent of coronary bypass surgery. Among other procedures attempted, pedicle flaps composed of mediastinal fat, skel-

Experimental

1 Supported by Grant HL-41918 from the National and Blood Institute, Bethesda, MD. 2 To whom correspondence should be addressed.

Twelve castrate male goats weighing 28-48 kg were studied. The animals were cared for following the “Principles of Laboratory Animal Care” prepared by the National Society for Medical Research and the “Guide for

0022.4804/92 $4.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form

Heart,

METHODS AND MATERIALS

Lung,

578 Inc. reserved.

Design

MANNION

ET AL.: COLLATERAL

the Care and Use of Laboratory Animals” prepared by the National Academy of Science (NIH Publication No. 80-23, revised 1978). All animals underwent a left thoracotomy, and a moderate size branch of the circumflex coronary artery was dissected out. The innate coronary collateral circulation of the risk area was defined, as previously described [26], by left atria1 injection of microspheres, with the artery to the risk area alternately open, then occluded. The animals subsequently underwent a latissimus dorsi cardiomyoplasty, in which the muscle was mobilized on its neurovascular pedicle and used to cover the risk area. After a 3-week period, during which collaterals would be expected to form, total myocardial collateral blood flow to the risk area (intramyocardial collateral flow plus skeletal muscle to myocardial collateral flow) was measured in nine animals by left atria1 injection of microspheres. Risk area blood flow was measured with the artery to the risk area both open and occluded. In five of these animals, the latissimus flap was then removed from the heart, and the residual, innate collateral flow to the risk area was again measured. In four animals, the flap was not removed from the heart, and an injection of microspheres was made directly into the thoracodorsal artery. From these data, we estimated the skeletal muscle to myocardial collateral blood flow. Three animals were used to prepare vascular casts.

BLOOD

of Innate Myocardial

Collateral Blood Flow

For the initial surgical procedure, all animals were placed under general endotracheal anesthesia employing 1.0% Forane, following induction with ketamine (500 mg im) and thiopental(25 mg/kg iv). One gram of cefazolin (iv) was given pre-operatively. Employing sterile surgical technique, the left latissimus dorsi muscle was mobilized on its neurovascular pedicle, as previously described [27]. A left anterolateral thoracotomy was then performed through the fifth intercostal space, and the heart was exposed. The circumflex coronary artery was identified, and a moderate size obtuse marginal branch was dissected for a length of approximately 1.5 cm. A reversible vessel occluder (In Vivo Metrics, Heldsberg, CA) was placed around the artery, and tacked to the myocardium to prevent kinking. An indwelling left atria1 silastic catheter was inserted into the left atria1 appendage. A temporary plastic catheter was introduced into the descending aorta through a pursestring suture. With the coronary artery unoccluded, 15 million colored microspheres (E-Z Trac, Los Angeles, CA), 15 pm in diameter, were injected through the left atria1 catheter (Control Injection l), and reference blood was simultaneously withdrawn from the aorta at 12.7 ml/min. The dissected coronary blood vessel was temporarily occluded. During the last 1.5 min of a 3.5min occlusion, a second injection of microspheres

579

was given (Ischemic Injection 1). The selection of microsphere colors for injection was randomized from six that were available for use. The blood flow to the risk area after the first acute occlusion is the innate collateral blood flow. Latissimus

Dorsi Cardiomyoplasty

After measurement of innate collateral blood flow to the risk area, the latissimus pedicle flap was delivered into the chest through the second intercostal space. The surface to be applied to the heart was abraded with sterile velcro; the epicardium was similarly abraded. The temporary plastic aortic cannula was removed, and the latissimus graft was used to cover the risk area and adjacent myocardium. The reversible vessel occluder was left in place. Several sutures were employed to tack the skeletal muscle to the pericardium. The incision was closed, and the left pleural space evacuated of residual air. The left atria1 catheter and the end of the reversible occluder were either implanted subcutaneously or brought out through the skin on the back of the neck. The animals were allowed to recover for 3 weeks. Postoperative pain was controlled with morphine (0.2 mg/kg, im), and the animals were maintained on oral penicillin as a single daily dose for 5 days. Daily local care was taken of the silastic catheters where they emerged from the skin. Evaluation

Measurement

FLOW

of Collateral Blood Vessels

Measurement of total myocardial collateral blood flow to the risk area (innate collateral blood flow plus skeletal muscle to myocardial collateral blood flow). After a 3week recovery period, total collateral blood flow to the risk area was measured. The animals were anesthetized with pentobarbital. The femoral artery was cannulated by means of a cut-down, for withdrawal of reference blood samples, and for continuous blood pressure monitoring. In two animals, due to difficulties with clotting of the left atria1 cannulae, microsphere injections were made via a catheter passed into the left ventricle through a left carotid artery cutdown. With the thorax unopened, and consequently with the latissimus dorsi undisturbed, blood flow measurements were obtained by left atria1 injection of microspheres, with the marginal branch of the circumflex coronary artery open (Control Injection 2), and during the last 1.5 min of a 3.5-minute occlusion (Ischemic Injection 2). Measurement of residual intramyocardial collaterals. The thorax was then opened and the latissimus dorsi was then dissected off the heart. A third series of injections was then made, with the circumflex artery open (Control Injection 3), and with the artery occluded (Ischemit Injection 3). After the last injection, with the coronary artery still occluded, brilliant green dye was injected into the left atrium to delineate the potential area at risk for myocardial ischemia. The animals were there-

580

JOURNAL

OF SURGICAL

RESEARCH:

after terminated by means of an overdose of pentobarbital and KCl, and the hearts and latissimus dorsi grafts were excised from the thorax. Samples were taken of other organs for assessment of blood flow. Estimation of skeletal muscle to myocardial collateral blood flow to the risk area. In four animals, after Control 2 and Ischemic 2 injections, the thorax was not opened initially. Rather, the thoracodorsal artery was dissected out and 1.5 to 2 million microspheres were directly injected into the artery with a 25-gauge needle. The microspheres were injected slowly, so that they were distributed by blood in the thoracodorsal artery under normal perfusion pressures. During and at the completion of the injection process, withdrawal of blood from the thoracodorsal artery was used to confirm placement of the needle tip in the lumen of the artery. In one animal, technical problems with thoracodorsal injection precluded its use.

53, NO. 6, DECEMBER

1992

by dicing with a scalpel and boiling for 60 min in 2 N NaOH, and then centrifuged. The resulting pellet was washed with surfactant and recentrifuged, and the pellet resuspended in 30 to 50 ~1 of surfactant. The final sample volume was measured with an Eppendorf micropipet. The microspheres were counted on a hemocytometer under the microscope. Reference blood samples were hemolyzed, centrifuged, and digested with 2.0 N NaOH and processed as above. Adequate numbers of microspheres were used for blood flow determinations: during the control periods, in the areas at risk and in the normally perfused areas of myocardium, microsphere counts per sample averaged 5000; during the occlusion periods, in the areas at risk, the counts per sample ranged from 0 to 800. Blood Flow Calculations Blood flows are expressed as ml/min/g were calculated according to the following

Vascular Casts The three animals that developed respiratory or renal infections were used for anatomical study of vascular connections between the skeletal muscle and myocardium. Vascular casts were made by transsection of the latissimus pedicle, cannulation of the thoracodorsal artery, and injection of Microfil (Canton Biomedical Products, Boulder CO) silicone rubber. The heart, with attached graft, was then removed and sectioned transversely along its long axis. Gross examination and subgross examination with a dissecting microscope of these specimens were performed following serial dehydration with graded ethanol solution and clearing with methyl salicylate. Histochemical

VOL.

Analysis

Cubic blocks of tissue containing myocardium, skeletal muscle to myocardial interface, and latissimus dorsi were then fixed in a 1% formaldehyde solution. Transverse sections ranging from 5 to 20 pm in thickness were then cut. Histological evaluation was performed with Hematoxylin-Eosin stained sections. Colored Microsphere

Processing

The excised hearts were cut in l.O-cm transverse sections along the long axis of the heart. The ischemic area was well defined by the brilliant green dye. The heart muscle was divided into ischemic, ischemic border, and normal zones based on the brilliant green staining. Maps of the sectioned hearts were traced onto clear plastic sheets. The hearts were then divided into approximately 2-g specimens from each zone, and subdivided into epicardial, mid, and endocardial segments. The techniques of tissue and blood preparation and microsphere counting have been described in detail previously [21, 261. Briefly, tissue specimens were prepared

Flow =

of tissue and formula 1:

(microspheres/gram) X reference flow 9 (1) reference microspheres

where reference flow is the withdrawal rate of the reference sample, and reference microspheres is the number of microspheres in the reference sample. Cardiac output was calculated from the same formula using the total number of microspheres injected in place of the tissue count. Estimates of skeletal muscle to myocardial collateral flow could then be made by formula 2: Skeletal to myocardial flow counts per gram myocardium X latissimus flow , Flow = counts per gram latissimus

(2)

where latissimus flow is that measured during the previous control period. Calculations were performed on an IBM PS/2, using a laboratory programmed Lotus spreadsheet. Statistical analysis consisted of a t test. Statistical significance was declared when P < 0.05. RESULTS

Gross Surgical Findings No animals developed evidence of wound infections, and the transfer of the latissimus dorsi did not appear to interfere with the animals’ posture or locomotion. At about 2 f weeks, three animals died: two animals developed upper respiratory infections and one died of renal complications. In all the animals, there was evidence of dense connections between the latissimus and all surrounding

MANNION

ET AL.:

COLLATERAL

structures. The lung was firmly adherent to the latissimus, and the latissimus was firmly adherent to the heart. Dissection of the latissimus from the lung and from the heart produced a moderate amount of fine, diffuse bleeding. There appeared to be a thin layer of scar between the skeletal muscle and the myocardium. The distal aspect of the latissimus dorsi was edematous. The area at risk for myocardial ischemia was clearly delineated by the vital stain. The area at risk for infarction represented 27 f 5.9% (n = 8) of the weight of the left ventricle. In 11 animals, the latissimus dorsi flap covered the risk area, and extended on to normal myocardial areas. In one animal, the latissimus partially covered the risk area. Vascular Casts and Histological

Analysis

The major veins on the epicardial surface of the heart were filled with the Microfil; the large epicardial arteries were not observed to contain the silicone rubber. The fact that of the large epicardial vessels only the veins were filled is not inconsistent with the existence of precapillary collaterals in which the path of least resistance is through the capillaries and into the veins. Examination of cut sections of the interface between the latissimus and the myocardium, under the dissecting microscope at approximately 50x, demonstrated direct vascular connections between the skeletal muscle and the heart. In several sections, deep myocardial penetration of the Microfil-highlighted blood vessels could be appreciated as shown by the representative micrograph in Fig. 1. Histological

Examination

Hematoxylin and Eosin staining of the formalin fixed sections revealed in general a preservation of the latissimus dorsi architecture; there were some areas of focal degeneration. A scar several millimeters in thickness was interposed between the latissimus muscle and the heart. The latissimus dorsi muscle was densely filled with the silicone rubber and there were multiple vascular channels containing the silicone rubber in the interposed scar. Hemodynamic

Data

Listed in Table 1 are the hemodynamic parameters associated with the six different injections of microspheres. During the last four time periods, there were no significant differences in mean blood pressure, heart rate, or tension time index. Lead 2 of the EKG demonstrated isoelectric ST segments during all control periods, and ST segment elevation during ischemic periods. There was no apparent benefit from the muscle flap in preventing electrocardiographic evidence of acute transmural ischemia.

BLOOD

581

FLOW

Blood Flow to Peripheral

Organs

Peripheral blood flow measurements were obtained by microsphere analysis in eight animals. Blood flows to peripheral organs are listed in Table 2, and suggest that the microspheres were well mixed. The higher renal and liver blood flows during Control 1 and Ischemic 1 injections probably are secondary to the use of Forane as an anesthetic, with its vasodilator properties [28]. Blood flows to the latissimus dorsi muscles are listed in Table 3. During Control Injection 1, the left latissimus dorsi blood flow is higher than the right. The elevated blood flow on the left side is probably due to inadvertent stimulation of the muscle which occurred when mobilizing it. This suggests that after mobilization the distal muscle was not ischemic, since it had a slightly higher blood flow than the nonoperated side. During Control Injection 3, the left proximal latissimus flow is again higher than the right. This observation can again be attributed to the muscle stimulation associated with stripping the muscle off of the ventricle. There are no values for the left distal latissimus during the third Control and Ischemic Injections 3 because the muscle was stripped off the myocardium and the graft was transected. Blood Flow in Normal Myocardium Blood flow measurements in normally perfused areas of the left ventricle are contained in Table 4. Blood flows to normal myocardium were divided into epicardial, mid-myocardial, and endocardial, The epicardial blood flow was lower than the endocardial blood flow (P < 0.05, paired t test). Blood flow in the mid-myocardium was of an intermediate value. The goat species, therefore, has a transmural distribution of myocardial flow similar to that reported for the dog [29] and pig [30]. Collateral Blood Flow Innate coronary collateral blood fIow. Regional myocardial blood flows to the ischemic, ischemic border, and normal areas of the myocardium are presented in Table 5. Before the graft, acute occlusion of the coronary artery significantly lowered blood flow to the ischemic zone (P < 0.05). Within the ischemic border zone, blood flow fell to 54% of the control value (P < 0.05). These data show that the goat has a poorly developed system of innate coronary collaterals. Total coronary collateral blood flow to the risk area (innate, intramyocardial collateral flow plus skeletal muscle to myocardial collateral j-low). After 3 weeks with the latissimus dorsi flap in place, the Control Injection 2 blood flow to the risk area was normal, suggesting that there was no permanent ischemic damage or fibrosis around the reversible coronary occluder. When the artery to the area at risk was occluded, the blood flow to the risk area was not higher than that measured during

582

JOURNAL

OF SURGICAL

RESEARCH:

VOL.

53, NO. 6, DECEMBER

1992

FIG. 1. Micrograph of vascular cast of skeletal muscle-myocardial interface. Bar: 3.0 mm. ENDO: Endocardium. White arrow: Vessel deep within the myocardium filled with microfil injected into the thoracodorsal artery. S: Scar. Black arrow: A vessel coursing from the latissimus dorsi into the myocardium. L: Latissimus dorsi muscle.

the Ischemic Injection 1 period. Statistical analysis of the responses to occlusions 1 and 2 were made using percentages because of the differences in Control 1 and Control 2 flows and because of the variability between animals. Blood flow during the occlusion was expressed as a percentage of its respective control value. Within the Ischemic zone, the reduction in blood flow during the occlusions was significant. However, the presence of the pedicle graft during occlusion 2 did not significantly change the fall in blood flow. To determine if the skeletal muscle to myocardial collaterals could be contributing to the total collateral blood flow, the flap was removed surgically, and the innate collateral system again measured. There was no significant change in the intramyocardial collateral blood

flow before and after removal of the flap. Thus, with the flap in position, it appears that the total coronary collateral blood flow had not increased from the low values characteristic of a species with a poorly developed innate collateral system. Estimation of skeletal muscle to myocardial blood flow. The vascular casts suggested that there were vascular connections between the muscle and the heart, but the casts could not distinguish between precapillary connections and venous connections. Microspheres were directly injected into the thoracodorsal artery, reasoning that if the connections were venous, the microspheres would be largely trapped by the lung; those microspheres that entered the left atrium would be evenly distributed throughout the heart. If the vascular connec-

MANNION

ET AL.: COLLATERAL

TABLE

ISCH 1

BP (mm Hd SE FIR (bpm) SE GTI (I-k X bpm) SE E.0. (ml/min) SE n Note. BP, Mean arterial

Parameters CON 2

ISCH 2

CON 3

ISCH 3

77.3 11.6 3

63.9 6.5 3

117.3 6.8 8

106.5 6.2 8

104.2 9.1 5

99.2 9.6 5

93.7 1.9 3

84.3 1.2 3

93.4 7.9 8

96.5 7.5 8

103.6 4.9 5

105.6 5.4 5

9896 1025 3

7810 891 3

13360 1556 8

12410 1241 8

13094 1222 5

12725 1280 5

4025 597 4

4338 686 7

3695 551 7

3144 352 8

3417 808 4

3471 645 3

pressure; HR, heart rate; TTI, time-tension

index (systolic

DISCUSSION

This experiment demonstrates tions exist between a latissimus

TABLE Blood

Flow

to Peripheral

pressure X heart rate); C.O., cardiac output.

mus did. The percentage trapped is indicated by the height of the bar and the number at the top of the bar. Most of the microspheres were trapped by regions under the wrap and in the epicardial portion of the ventricle. However, it should be noted that the midmyocardium and endocardium also trapped microspheres injected into the thoracodorsal artery. The estimated collateral blood flows from skeletal muscle to myocardium in three animals were 0.01, 0.02, and 0.04 ml/min.

tions were precapillary, the microspheres would be concentrated in myocardium underneath the flap. The distribution of the microspheres injected into the thoracodorsal artery is shown in Fig. 2. This schematic diagram shows a representative slice of the left ventricle cut longitudinally through the septum and “unrolled.” The area of the large rectangle is proportional to the area of the ventricle. The area of the small rectangle labeled Latissimus Dorsi is proportional to that of the Latissimus wrap and its position shows that it is placed over most of the freewall of the ventricle. The small boxes within the ventricle are drawn to scale and show the proportion and location of regions within the ventricle that trapped 2% or more microspheres than the latissi-

Left kidney Mean SE n Right kidney Mean SE n Liver Mean SE n Right ventricle Mean SE n

583

FLOW

1

Hemodynamic CON 1

BLOOD

that vascular connecdorsi pedicle flap and

2 Tissues

(ml/min/g)

CON 1

ISCH 1

CON 2

ISCH 2

CON 3

ISCH 3

5.06 1.28 5

5.44 0.88 6

2.78 0.43 7

2.93 0.67 7

2.59 0.66 5

2.63 0.97 4

4.32 1.30 5

5.22 1.35 6

2.73 0.50 7

3.10 0.82 7

2.33 0.76 5

3.18 1.22 4

0.81 0.48 3

0.94 0.25 3

0.25 0.12 4

0.35 0.19 4

0.38 0.16 4

0.43 0.14 3

0.77 0.45 3

1.48 0.61 4

1.28 0.36 5

1.24 0.29 5

0.83 0.14 5

0.77 0.16 5

584

JOURNAL

OF SURGICAL

RESEARCH:

VOL.

TABLE Blood

Left distal latissimus dorsi Mean SE n Left proximal lattissimus dorsi Mean SE n Right latissimus dorsi Mean SE n

Flow

3

to Latissimus

Dorsi

(ml/min/g)

ISCH 1

CON 2

ISCH 2

CON 3

ISCH 3

0.09 0.02 6

0.10 0.02 7

0.07 0.02 8

0.05 0.01 8

X X X

X” X X

0.06 0.02 6

0.08 0.02 6

0.02 0.00 7

0.03 0.01 7

0.21 0.09 5

0.14 0.07 3

0.02 0.02 3

0.08 0.03 3

0.03 0.01 4

0.03 0.01 4

0.02 0.01 4

0.03 0.02 3

flow because the muscle was transected

myocardium in a cardiomyoplasty model. After a 3-week time period, with no prolonged ischemic stimulus from the myocardium, the vascular connections appear to perfuse myocardial capillaries, but with only a fraction of the blood provided to the latissimus dorsi muscle. This collateral flow represents a small fraction of the blood flow needed by viable myocardium. Although not an important source of collaterals in quantitative terms in this experiment, the fact that these precapillary connections exist raises the possibility that they might be exploited to treat ischemic heart disease. There are several potential reasons why the collateral connections demonstrated between skeletal muscle and myocardium were not abundant enough to ameliorate acute myocardial ischemia. First, it is likely that the principal stimulus to collateral development in our preparation was wound healing. The stimulus for angiogenesis in wound healing is hypoxia. This stimulus is transient, as it disappears soon after the oxygen tension in the wound normalizes [31]. A more prolonged stimulus might be required. Second, the most potent stimulus for angiogenesis in the heart is ischemia-either repeated episodes of acute

Transmural

Blood

Flow

and stripped

from the heart.

ischemia or a chronic ischemic state [32, 331. Neither type of ischemia was present in this experiment. Third, it is possible that the type of collagen in the extracellular matrix may have inhibited more extensive collateral development. Specifically, on type 1 or type 3 collagen, endothelial cells are reported to form monolayers [34, 351, whereas on type 4 collagen, endothelial cells form tubelike structures. It is possible that the type of collagen scar between the latissimus dorsi and the myocardium inhibited neovascularization. Another possibility is that insufficient time was allowed for collateral development. The skeletal muscle was in contact with the myocardium for a period of 3 weeks. It might be argued that the heart was providing collaterals to the latissimus and that the stimulus for the collateral development between the skeletal muscle and heart was distal latissimus ischemia. This possibility is unlikely because resting blood flow to the distal latissimus dorsi after mobilization was sufficient to maintain cell viability with the muscle at rest, as it was with the canine model [36]. Furthermore, the failure of the distal latissimus blood flow to fall during ischemic injec-

TABLE

* Significantly

1992

CON 1

a There are no values for distal left latissimus

Endocardium Mean SE n Epicardium Mean SE n

53, NO. 6, DECEMBER

4

Gradient

in Normal

Myocardium

CON 1

ISCH 1

CON 2

ISCH 2

CON 3

ISCH 3

2.01* 0.59 6

2.41 0.55 7

1.96* 0.41 8

1.78 0.30 8

1.23* 0.15 6

1.23 0.19 5

1.70 0.52 6

1.98 0.49 7

1.39 0.27 8

1.31 0.19 8

0.98 0.14 6

1.02 0.23 5

greater then epicardial

flow by a paired t test (P < 0.05).

MANNION

ET AL.:

COLLATERAL

TABLE Regional CON 1 Ischemic zone Mean SE

n Percentage

1.85 0.61 4

ISCH 1

0.08 0.02 5

585

FLOW

5

Blood

Flows

CON 2

1.22 0.35 6

(ml/min/g) ISCH 2

0.10 0.02 6

CON 3

0.96 0.19 5

ISCH 3

0.27 0.19 5

change

Mean SE n Ischemic border zone Mean SE n Percentage change Mean

-93.7* 1.7 4 2.17 0.89 4

0.69 0.21 5

-9o.o* 2.3 6 1.34 0.23 6

-54.1*

SE

SE

4 2.06 0.64 4

1.70 0.30 5

0.57 0.11 6

-90.6* 5.8 4 1.07 0.10

5

-51.9* 12.6 6

10.9

72 Normal myocardium Mean n Percentage Mean

Myocardial

BLOOD

1.58 0.30 6

1.47 0.20 6

0.53 0.16 5 -44.5 23.6 4

1.07 0.15 5

1.08 0.19 5

change 1.7

SE

16.4

n

4

6.5 16.7 6

5.6 22.5 4

Note. The percentage change during ischemia 1, 2, and 3 is calculated from control 1, 2, and 3, respectively. *P < 0.05. Comparison of the changes in flow during ischemia 1, ischemia 2, and ischemia 3 showed that there were no significant differences among these responses.

tion 2 argues against the heart providing blood to the latissimus. Regardless of the mechanisms for the development of the new vascular connections, it is clear that blood flow was directed from the latissimus to the heart. Microspheres injected into the thoracodorsal artery were dis-

FIG. 2. Schematic diagram of an opened myocardial slice showing the distribution of microspheres injected into the thoracodorsal artery and trapped by myocardial capillaries. The columns refer to the percentage of microspheres recovered in the heart in relation to the average latissimus dorsi count. Note that the myocardial collateral blood flow from the latissimus dorsi muscle concentrated in the epicardium underneath the muscle flap. The height of the bars and the numbers at the top represent the microsphere counts as a percentage of the weighted average of the number of counts found in the latissimus dorsi.

tributed by the blood flow under normal perfusion pressure. The fact that the microspheres were trapped by myocardial capillaries located principally underneath the muscle graft demonstrates that the skeletal muscle to myocardial collaterals was precapillary. Our data extend earlier observations of others by providing histological and quantitative blood flow data showing that pedicle flaps provide capillary connections between skeletal muscle and myocardium [37, 381. The concept of using skeletal muscle to augment cardiovascular function is not new. In recent years it has been investigated with increasing interest, since Macoviak and Stephenson demonstrated its potential [37,38]. Among others, Magovern [39] and Chacques and Carpentier [40] have performed latissimus dorsi cardiomyoplasty in selected patients. Approximately 100 such operations have now been performed worldwide. An early clinical observation is that the patients appear to improve clinically, although there is often not a dramatic increase in ejection fraction. It is interesting to speculate that the symptomatic improvement might be related, through an unknown mechanism, to latissimus collaterals. Among the limitations of this study is the identification of latissimus to myocardial collaterals by injection of microspheres into the thoracodorsal artery and the subsequent estimation of skeletal muscle to myocardial collateral blood flow. Injection of microspheres at this downstream location eliminates blood flow from intra-

586

JOURNAL

OF SURGICAL

RESEARCH:

myocardial collaterals and blood flow arising from possible collaterals from the internal mammary artery, bronchial arteries, and directly from the lung. However, microsphere mixing at a downstream location is not optimal and could result in streamlining. Nonetheless, the microsphere count in the distal latissimus was relatively uniform, suggesting that streamlining was not a significant factor. The estimation of collateral flow from the latissimus was based on the latissimus flow during a previous control injection, and was not a measured value. Since the muscle was at rest, the blood flow should not have changed significantly over that time period. In summary, vascular channels were found to connect skeletal muscle and normal myocardium in a cardiomyoplasty model. Under physiologic conditions, the vascular channels provided blood flow to myocardial capillaries. The flow was insufficient to ameliorate acute myocardial ischemia. It is possible that the collateral blood flow from skeletal muscle could be augmented by establishing conditions more favorable to angiogenesis.

REFERENCES 1.

Beck, C. S. A new blood supply to the heart by operation. Surg. Gynecol. Obst. 61: 407-410,1935. 2. Beck, C. S. The development of a new blood supply to the heart by operation. Ann. Surg. 102: 801-813,1935. 3. Beck, C. S., Tichy, V. L., and Moritz, A. R. The production of a collateral circulation to the heart. Am. Heart J. 10: 849-873,

1935.

VOL. 53, NO. 6, DECEMBER portance

of coronary

collateral

1992 circulation.

14. Carroll,

R. J., Varani, M. S., and Falsetti, H. L. The effect of collateral circulation on segmental left ventricular contraction. Circulation 60: 709-713, 1974.

15’ Williams,

D. O., Amsterdam, E. A., Miller, R. R., and Mason, D. T. Functional significance of coronary collateral vessels in patients with acute myocardial infarction: Relation to pump performance, cardiogenic shock and survival. Am. J. Cardiol. 37:

345-351,1976. 16.

Rogers, W. J., Hood, W. P., Mantle, J. A., Baxley, W. A., Kirklin, J. K., Zorn, G. L., and Nath, H. P. Return of left ventricular function after reperfusion in patients with myocardial infarction: Importance of subtotal stenoses or intact collaterals. Circulation 69: 338-349, 1984.

17. Nitzberg, W. D., Nath, H. P., Rogers, W. J., Hood, W. P., Whit-

low, P. L., Reeves, R., and Baxley, W. A. Collateral flow in patients with acute myocardial infarction. Am. J. Cardiol. 66: 729736, 1985.

18. Saito, Y., Yasuno, M., Ishida, M., Suzuki, K., Matoba,

Y., Emura, M., and Takahashi, M. Importance of coronary collaterals for restoration of left ventricular function after intracoronary thrombolysis. Am. J. Cardiol. 55: 1259-1263, 1985.

19. Nohara,

R., Kambara, H., Murakami, T., Kadota, K., Tamaki, S., and Kawai, C. Collateral function in early acute myocardial infarction. Am. J Cardiol. 52: 955-959, 1983.

20. Makowski,

E. L., Meschia, G., Droegemueller, W., and Battaglia, F. C. Measurement of umbilical artery blood flow to the sheep placenta and fetus in utero. Circ. Res. 23: 623-631, 1968.

21. Hale, S. L., Alker, K. J., and Kloner, radioactive, myocardial

R. A. Evaluation of noncolored microspheres for measurement of regional blood flow in dogs. Circulation 78: 428-434, 1988.

22. Schwarz, F., Schaper, J., Becker, V., Kubler, W., and Flameng, W. Coronary collateral vessels: Their significance for left ventricular histologic structure. Am. J. Cardiol. 49: 291-295, 1982.

4. Carter, B. N., Gall, E. A., and Wadsworth,

23. Schaper, W., Schaper, J., Xhonneux,

5.

24. Schaper, W., Gorge, G., Winkler,

6.

7.

8.

9.

10.

11. 12.

13.

C. L. An experimental study of collateral coronary circulation produced by cardiopneumopexy. Surgery 25: 489-509, 1949. O’Shaughnessy, L. An experimental method ofproviding a collateral circulation to the heart. &it. J. Surg. 23: 665-670, 1936. Key, J. A., Kergin, F. G., Martineau, Y., and Leckey, R. G. A method of supplementing the coronary circulation by a jejunal pedicle graft. J. Thoracic Surg. 28: 320-330, 1954. Neumann, C. G., Moran, R. E., VonWedel, J., Lord, J. W., and Hinton, J. W. Reduction of coronary artery blood flow preparatory to revascularization of the heart via pedicled flap of skin. Pi&t. Reconstruct. Surg. 13: 85-94, 1954. Vineberg, A. M. Development of anastomosis between the coronary vessels and a transplanted internal mammary artery. J. Thorac. Sug. 18: 839-850, 1949. Vineberg, A. Treatment of coronary insufficiency by implantation of the internal mammary artery into the left ventricular myocardium. J. Thoracic Surg. 23: 42-54, 1952. Vineberg, A., and Buller, W. Technical factors which favor mammary-coronary anastomosis. J. Thoracic Surg. 30: 411-435, 1955. Baroldi, G. Coronary heart disease: significance of morphologic lesions. Am. Heart J. 85: l-5, 1973. Frick, M. H., Valle, M., Korhola, O., Rukimaki, E., and Wiljasalo, M. Analysis of coronary collaterals in ischemic heart disease by angiography during pacing induced ischemia. Br. Heart J. 38: 186-196,1976. Frye, R. L., Gura, G. M., Chesebro, J. H., and Ritman, E. L. Complete occlusion of the left main coronary artery and the im-

Mayo Clin. Proc. 52:

742-745,1977.

R., and Vandesteene, R. The morphology of intercoronary anastomoses in chronic coronary artery occlusion. Cardiouasc. Res. 3: 315-323, 1969.

eral circulation 1988.

B., and Schaper, J. The collatof the heart. Prog. Cardiouasc. Dis. 31: 57-77,

25. Schaper, W., Sharma, H., Quinkler,

W., Marker& T., Wunsch, M., and Schaper, J. Molecular biologic concepts of coronary anastomoses. J. Am. Coil. Cardiol. 15: 513-518, 1990.

26. Brown, W. E., Magno, M. G., Buckman,

P. D., Dimeo, F., Gale, D. R., and Mannion, J. D. The coronary collateral circulation in normal goats. J. Surg. Res. 51: 54-59, 1991.

27. Mannion,

J. D., Acker, M. A., Hammond, R. L., Faltemeyer, W., Duckett, S., and Stephenson, L. W. Power output of skeletal muscle ventricles in circulation: short-term studies. Circulation

76:155-162,1987. 28. Stevens, W. C., Cromwell, T. H., Halsey, M. J., Eger, E. I., Shakespeare, T. F., and Bahlman, S. H. The cardiovascular effects of a new inhalation anesthetic, Forane, in human volunteers at a constant arterial carbon dioxide tension. Anesthesiology 35:

8-16,197l. 29. Fedor, J. M., McIntosh,

D. M., Rembert, J. C., and Greenfield, J. C. Coronary and transmural myocardial blood flow responses in awake domestic pigs. Am. J. Physiol. 235: H435-H444, 1978. 30. Roth, D. M., Maruoka, Y., Rogers, J., White, F. C., Longhurst, J. C., and Bloor, C. M. Development of coronary collateral circulation in left circumflex Ameroid-occluded swine myocardium. Am. J. Physiol. 253: H1279-88, 1987. 31. Knighton, D. R., Hunt, T. K., Scheuenstuhl, H., and Halliday,

MANNION

ET AL.:

COLLATERAL

B. J. Oxygen tension regulates the expression of angiogenesis factor by macrophages. Science 221: 1283%1285,1983. 32.

Goldsmith, H. S. Salvage of end stage ischemic extremities intact omentum. Surgery 88: 732-736,198O.

33.

Schaper, W. The Collateral North-Holland, 1971.

34.

Madri, J. A., and Pratt, B. M. Endothelial cell-matrix interactions: In vitro models of angiogenesis. J. Histochem. Cytochem. 34(l): 85-91, 1986.

35.

Schor, A. M., and Schor, S. L. Tumour 141: 385-413, 1983.

36.

Mannion, J. D., Velchik, M., Hammond, R., Alavi, A., Mackler, T., Duckett, S., Staum, M., Hurwitz, S., Brown, W., and Stephenson, L. W. Effects of collateral blood vessel ligation and electrical conditioning on blood flow in dog latissimus dorsi muscle. J. Surg. Res. 47: 332-40, 1989.

Circulation

by

of the Heart. Amsterdam:

angiogenesis.

37.

38.

39.

J. Pathol.

40.

BLOOD

FLOW

587

Macoviak, J. A., Stephenson, L. W., Kelly, A., Likoff, M., Reichek, N., and Edmunds, L. H. Partial replacement of the right ventricle with a synchronously contracting diaphragmatic skeletal muscle autograft. Proceedings of the Third Meeting of the Znternational Society for Artificial Organs. 5: 550-555, 1981. Macoviak, J. A., Stephenson, L. W., Alavi, A., Kelly, A. M., and Edmunds, L. H. Effects of electrical stimulation on diaphragmatic muscle used to enlarge right ventricle. Surgery 90: 271277, 1981. Magovern, G. J., Heckler, F. R., Park, S. B., Christlieb, I. Y., Liebler, G. A., Burkholder, J. A., Maher, T. D., Benckart, D. H., Magovern, G. J., Jr., and Burkholder, J. A. Skeletal muscle for dynamic cardiomyoplasty. Ann. Thorac. Surg. 45: 614-619, 1988. Chacques, J. C., Grandjean, P. A., and Carpentier, A. Latissimus dorsi dynamic cardiomyoplasty. Ann. Thorac. Surg. 47: 600604, 1989.