Catheterization and Cardiovascular Diagnosis 25:8!5-90 (1992)
Original Studies Internal Mammary Artery Angiography Should Be a Routine Component of Diagnostic Coronary Angiography -
Alan Feit, MD, C.V. Reddy,
MD,
Carlos Cowley, MD, Bassiema Ibrahim, MD, and Zvi Zisbrod, MD
Left internal mammary artery (LIMA) angiography was performed with diagnostic coronary angiography In 130 cases for which the coronary findings made use of the LIMA as a bypass graft a consideration. In 98% of the cases the approach to LIMA angiography was femoral with a 5F LIMA catheter first directed into the proximal subclavian and then advanced over a guldewire placed into the distal subclavian well beyond the origin of the LIMA. After withdrawing the wire the catheter was brought proximally to selectively cannulate and visualize the LIMA wlth nonlonlc contrast media. The only complication was a single transient occipital visual field loss. LIMA caliber too narrow to permit use as a graft was found twice, LIMA occlusion unrelated to prior surgery was found once, and LIMA occlusion related to prior surgery was found twice. Subclavian and/or vertebral stenosis was present five times. Large proximal branches of the LIMA best identified prior to surgery were present 12 times. Based on this experience, LIMA angiography 1) can be performed safely with a high degree of success, 2) demonstrates significant findlngs in 15% of cases, and 3) should therefore be performed whenever coronary angiographic findings make It appropriate to consider LIMA to coronary artery bypass grafting. Key words: coronary bypass, coronary angiography, LIMA angiography
INTRODUCTION
angioplasty or bypass surgery. In recommending a revascularization approach the possibility of initial success In view of its durability and longevity compared to and restenosis with angioplasty must be balanced against saphenous vein grafts (SVGs) [I] and its low incidence the possibility of initial success and long-term patency of atherosclerosis [2-61 the internal mammary artery with surgery, with the long-term benefit of an IMA graft (IMA) has become the graft of choice for coronary artery being superior to that of a SVG graft. We have long been bypass. Although the need for pre-operative angiography of the opinion that revascularization decisions such as of the IMA is well recognized in specialized circumthese should most appropriately be made with the knowlstances, such as previous saphenous vein graft bypass edge that the surgeon will or will not be able to use the surgery [7] or previous mastectomy and radiation therapy IMA as a graft. This requires angiographic demonstra[8], recommendations regarding the role for IMA angition of the IMA at the time of diagnostic catheterization ography in the more routine setting are less clear. Singh rather than the expectation that the surgeon will probably has suggested that initial diagnostic coronary arteriografind the IMA suitable. phy should routinely include IMA angiography [5-61, In those patients for whom bypass surgery is clearly and Krijne has recommended that routine pre-operative the most attractive option we have long felt that demonangiography of the IMA is “indeed unnecessary” [9], stration of the IMA angiographically allows the surgeon while a recent review by Kuntz and Baim made no specific recommendation but did observe that “pre-operative evaluation of the IMA at the time of initial diagnostic From the Division of Cardiology, Department of Medicine, Slate Uniangiography may also become more widespread if safe versity of New York Health Science Center at Brooklyn, Brooklyn, and rapid IMA catheterization techniques can be devel- New York. oped” [lo]. Others have recommended routine pre-operative IMA angiography on the basis of isolated case Received July 6, 1991; revision accepted September 17, 1991. reports [11,12]. Address reprint requests to Alan Feit, MD, Director, Interventional Patients having coronary artery disease requiring re- Cardiology, State University of New York Health Science Center at vascularization often have the option of either coronary Brooklyn, Box 1199, 450 Clarkson Avenue, Brooklyn, N Y 11203. 0 1992 Wiley-Liss, Inc.
86
Feit et al.
to plan what operation will be performed and to identify particular elements of the IMA anatomy which may require special attention at surgery, for example proximal branches which may be used as secondary bypass conduits and/or which may require intentional ligation to improve IMA flow, therefore saving time and preparation in the operating room, as well as providing for a more controlled operation. For these reasons we decided to perform IMA angiography as a routine part of coronary angiography whenever the coronary angiographic findings in our judgement made use of the IMA as a bypass graft a clinical consideration.
TABLE 1. Patient Characteristics. Age (yr)
30-39 40 -49 50-59 60-69 70-79 >80 Totals
Male
Female
0 14
2
27 28 9 0
25 9 I
78
52
I 14
*Total = 130.
brought proximally to selectively cannulate and visualize the IMA with nonionic contrast media. Our technique for IMA angiography was arrived at MATERIALS AND METHODS independently, but is quite similar to the technique deBetween October 1989 and January 1991 we per- scribed by Kuntz and Bairn [lo], with the exception that formed 734 diagnostic coronary angiography proce- we often found it necessary to use a Wholey (ACS) dures. IMA angiography was performed as a part of the guidewire rather than a standard J tip guidewire. The following parameters were recorded for each case: diagnostic coronary angiographic procedure in 130 cases hypertension; presence or absence of equality in blood (18%). in both arms; presence or absence of peripheral pressures Initially IMA angiography was usually limited to the vascular disease, including TIAs, bruits, claudication, or LIMA and was performed in the setting of bypassable peripheral bypass surgery; diabetes with or without inlesions in the LAD. As we became more proficient at the sulin required; renal failure; radiation therapy to the cannulation of the IMA, bilateral IMA angiography was chest; and previous bypass surgery using saphenous vein performed. Bilateral IMA angiography was only pergrafts. The following angiographic findings were reformed if in the judgement of the operator more than one IMA study: size at the fourth intercostal corded for each bypass graft was required, and the lesions were located space; presence or absence of a significant proximal in positions which made bypass with the IMAs possible. branch; site of origin in proximal, mid, or distal third of The decision whether or not to perform IMA angiograsubclavian; presence or absence of a subclavian and/or phy was an individual decision made by one of the three vertebral stenosis, and presence or absence of an IMA primary operators in this study based on that operator’s lesion; and the time required for IMA cannulation. In all interpretations of the coronary anatomy and the lesion cases non-ionic contrast media was employed. distribution. We were not attempting to define IMA angiography in all settings. Rather we were interested in the IMA angiRESULTS ographic findings only in the subset of patients who had coronary artery lesions for which IMA bypass grafting A total of 130 patients were studied. The patient charwould be an appropriate consideration. These results acteristics by age and sex are shown in Table I. Nineteen have been obtained in a selected group of patients with percent of the females were age 70 or greater while 12% coronary artery disease, and may not be applicable to of the males were age 70 or greater. Males outnumbered females by 1.5 to I . Of the patient population 8.5% had other populations. Patient age ranged from 39 to 82. Those who had peripheral vascular disease as evidenced by either TIAs, undergone previous bypass surgery were entered into the claudication, or previous peripheral vascular surgery, study only if they had SVGs used as bypass grafts at 8.5% had diabetes mellitus, and I .5% had either signifsurgery, and if the native IMAs had never been dissected icant renal failure as evidenced by creatinine greater than or visualized previously. 2.5%or previous renal transplantation. In those cases employing the femoral approach IMA As can be seen from Table I1 the LIMA was successangiography was performed by first directing a 5F IMA fully cannulated 93% of the time and the RIMA 83% of catheter into the proximal subclavian or innominate ar- the time. In our experience there is a short but very real teries and then advancing a guidewire into the distal sub- learning curve for cannulating the IMA. In six of the first clavian artery well beyond the origin of the IMA. The ten attempts to cannulate the LIMA the operator required catheter was then advanced over the wire beyond the greater than five min for successful selective cannulaIMA and, after withdrawing the wire, the catheter was tion. However, only ten of the next 120 cases required
IMA Angiography Should Be Routine TABLE II. Results of Attempted IMA Cannulations LIMA Number attempts Successful selective cannulation, 5 min Complications Total successful selective cannulations
TABLE 111. Angiographic Findings RIMA
I30
18
105
I1
16" Ib(0.8%)
121 (93%)
87
4 0
15 (83%)
'Six of these occurred in the first ten cases. Vransient occipital visual field loss after a non-selective injection.
more than 5 min for LIMA cannulation. We expect that with continued RIMA cannulations our success rate will improve as well. Only one complication related to IMA angiography occurred. This was an episode of transient occiptal visual field loss which occurred in a 75-yr-old hypertensive female. In this case the LIMA could not be selectively cannulated and was non-selectively visualized on an injection into the subclavian. In the nine cases of failure to selectively cannulate the LIMA and the three failures to selectively cannulate the RIMA, failure appeared to result from tortuosity in the proximal subclavian or proximal innominate arteries. Although wires could be positioned into the appropriate artery, advancing the preformed catheters caused the wire to twist out of the target artery. In two other cases upsizing from a 5F IMA catheter to a 7F catheter provided greater control and made selective cannulation possible. Only two of our cases were performed by the brachial approach. Both of these were done from the left arm to permit visualization of the LIMA. Table I11 lists the significant findings. A large proximal branch defined as a branch equalling at least 50% of the diameter of the IMA and occurring within two rib interspaces from the origin of the IMA was found in 9.2%. IMA caliber too small (judged relative to the target artery) to be utilized at bypass surgery was found in I .5%, and a totally occluded IMA (verified at coronary bypass surgery) was found once. Subclavian stenoses were found in 2.3%, vertebral stenosis was found twice (1.5%), and IMAs cut off at the time of previous bypass surgery and snared in the wire sutures were noted on two occasions. Figures 1 to 5 illustrate some of these findings. Excluding the two cases in which the IMAs were snared by wire sutures at the time of previous bypass surgery, 15% of the IMAs studied exhibited findings which would be of importance to the surgeon at the time of coronary bypass surgery. These two cases have been excluded since the need to perform IMA angiography in the setting of prior saphenous vein graft coronary artery bypass surgery is well established [7].
Significant findings
LIMA' (8)
RIMAh (B)
Large proximal branch Caliber too small Occluded Subclavian stenosis Vertebral stenosis Cut off at prior CABG
12 (9.2) 2 (1.5) 1 (0.8) 3 (2.3) 2 (1.5) 2 (1.5)
0
Total
22 (16)
Total excluding cut off at prior CABG
20 (15)
0 0 0 0 1 (.OSS)
'N = 130. bN = 18.
Fig. 1. This proximal branch of more than one-half the dlame ter of the LIMA arose within two rib Interspaces of the LIMA'S origin from the subclavian artery.
DISCUSSION
Conflicting recommendations regarding the need for selective visualization of the IMAs prior to coronary bypass surgery have persisted. Based on our experience it is clear that selective IMA angiography is required at the time of diagnostic coronary angiography if the coronary angiographic findings suggest that bypass surgery utilizing the IMA will be an appropriate option. The complication rate attributable to selective IMA angiography is small. We attribute this to the technique employed which required that a guidewire first be placed down the subclavian artery and then the catheter advanced over the guidewire. The low complication rate demonstrates that selective IMA cannulation can be performed safely. However. and just as importantly, the findings indicate that the knowledge gained from selective cannulation of the IMA is very relevant to any planned surgery. Ignoring the two patients whose IMAs were trapped in the wire sutures at
88
Feit et al.
Fig. 2. a: This LIMA was of too narrow caliber to serve as a graft to any of the target branches of the left coronary artery, b. Note that the magnification is increased from a to b, but compare the arterial caliber with the 5F catheter in a and the 6F catheter in b.
Fig. 3. a: This LIMA could not be visualized, and was verified to be totally occluded at surgery. b: The RlMA was normal.
previous bypass surgery we identified six out of 130 patients whose IMAs could not be used as coronary bypass conduits. Of these, two were of too small caliber, one was totally occluded (this was verified at the time of coronary bypass surgery), and three had significant proximal subclavian stenoses. Thus, between 4% and 5% of the native IMAs were unsuitable for bypass surgery. Significant proximal IMA branches occurred in 12 patients for a 9.2% occurrence rate. It is important that these large proximal branches be identified prior to surgery so that the surgeon knows how far back to dissect the artery in order to ligate them. For the extremely large lateral costal branches which were identified on two oc-
casions, I .5%, the entire surgical approach may be modified as these branches may be brought down to be used as bypass grafts with the IMA. We found no correlation between the IMA angiographic anatomy and the presence or absence of peripheral vascular disease, diabetes mellitus, or renal failure. Similarly we found no correlation between the appearance of the IMA and the patient’s age or sex. During the course of this study we have been impressed by the variable appearance of internal mammary arteries. The size range at the fourth intercostal space has been from less than 1 mm to greater than 4 mm. In other words, some are large, some are small, and some have
IMA Angiography Should Be Routine
89
results are different from those of Krijne [9], who observed such a low incidence of findings which would be significant prior to bypass surgery. It is our opinion that non-selective visualization of the IMA is often inadequate and should be discouraged in favor of becoming proficient at selective cannulation, a technique of importance not only because of the issues raised here, but also because of the number of IMAs which require post bypass assessment. We realize that the data presented were obtained in a selected population of patients with coronary artery disease. However, these were all patients with obstructive coronary lesions which suggested to the operator that the internal mammary arteries would be appropriate as coronary bypass grafts. Overwhelmingly this meant that proximal left anterior descending disease was present. Fig. 4. A large ulcerated left subclavian stenosis proximal to Although our results were obtained in a selective population, it is our belief that these results should apply to the LIMA. any other population with coronary artery disease coming for diagnostic coronary angiography. significant proximal branches. Clearly defining anatomOur experience indicates that selective IMA angiograical abnormalities such as those listed in Table 111 is of phy can be performed safely with a high degree of sucvalue prior to surgery. However, defining normal IMA cess, demonstrates significant findings in 15% of cases, anatomy prior to surgery is important in that it allows the and should therefore be performed whenever diagnostic physicians and patients to know that a bypass with longer coronary angiography makes it appropriate to consider patency is expected, and it allows the surgeon to plan IMA to coronary grafting. pre-operatively which vessel will serve as which bypass graft, and better assess the need for vein harvesting. In the early stages of this study we frequently perACKNOWLEDGMENTS formed non-selective IMA injections prior to seating the Grateful appreciation is extended to the following ancatheter selectively, and we have been and remain impressed that non-selective injections often fail to ade- giographers and fellows whose skills and efforts helped quately visualize the internal mammary artery; consider- make this work possible: Robert Kipperman, Arnold able and unpredictable streaming artifact varied from Einhorn, Tak Kwan, Robert Vaccarino, Madu Darawhat, case to case. We believe that this is the reason why our Joseph Idicula, Dino Paiusco, Donald Kruez, Chan-
Fig. 5. a, and b: Two views of a LIMA with a large lateral costal branch, arising very proximally.
90
Feit el al.
dranath Das, Ralph Morales, David Schifter, Marcel Pintea, K.S. Chandrasekhar, Nieca Goldberg, and Gerard Hanley.
REFERENCES I . Lewis M, Dehmer G: Coronary bypass using the internal mammary artery (Editorial). Am J Cardiol 56:480-482, 1985. 2. Finci L, Meier B, Steffenio G, Doriot PA, Rutishauser W: Nonselective preoperative digital subtraction of internal mammary arteries. Cathet Cardiovas Diagn 19:13-16, 1990. 3. Sisto T, Isola J: Incidence of atherosclerosis in the internal mammary artery. Ann Thorac Surg 472384-886, 1989. 4. Kay H, Korns M, Flemma R. Tector A, Lepley D: Atherosclerosis of the internal mammary artery. Ann Thorac Sure 21504501. 1976. 5 . Singh R: Radiographic anatomy of the internal mammary arteries. Cathet Cardiovas Diagn 7:373-386, 1981.
6. Singh R: Atherosclerosis of the internal mammary arteries. Cardiovas Intervent Radio1 6:72-77, 1983. 7. Davies SW: Pre-operative imaging of the internal thoracic artery (letter to the editor). Am J Cardiol 67:334, 1991. 8. Hanet C. Marschand E, Kukeyux A: Left internal mammary artery occlusion after mastectomy and radiotherapy therapy. Am J Cardiol 65:1044-1045. 1990. 9. Krijne R, Deng M, Heinich K, Sons H, Knan A: Semi-selective angiography of the internal mammary arteries as a preparation for coronary bypass surgery. Am J Cardiol 66:377-378, 1990. 10. Kuntz R , Baim D: Internal mammary angiography: A review of technical issues and newer methods. Cathet Cardiovas Diagn 20: 10-16. 1990. I I . Stullman W, Hilliard G: Unrecognized internal mammary artery stenosis treated by percutaneous angioplasty after coronary bypass surgery. Am Heart J 113:393-395, 1987. 12. Raner W, Sadler T, Liggett M: Internal mammary arteriography prior to coronary artery bypass surgery. Chest 641523-524, 1973.
Original Studies Internal Mammary Artery Angiography Should Be a Routine Component of Diagnostic Coronary Angiography -
Alan Feit, MD, C.V. Reddy,
MD,
Carlos Cowley, MD, Bassiema Ibrahim, MD, and Zvi Zisbrod, MD
Left internal mammary artery (LIMA) angiography was performed with diagnostic coronary angiography In 130 cases for which the coronary findings made use of the LIMA as a bypass graft a consideration. In 98% of the cases the approach to LIMA angiography was femoral with a 5F LIMA catheter first directed into the proximal subclavian and then advanced over a guldewire placed into the distal subclavian well beyond the origin of the LIMA. After withdrawing the wire the catheter was brought proximally to selectively cannulate and visualize the LIMA wlth nonlonlc contrast media. The only complication was a single transient occipital visual field loss. LIMA caliber too narrow to permit use as a graft was found twice, LIMA occlusion unrelated to prior surgery was found once, and LIMA occlusion related to prior surgery was found twice. Subclavian and/or vertebral stenosis was present five times. Large proximal branches of the LIMA best identified prior to surgery were present 12 times. Based on this experience, LIMA angiography 1) can be performed safely with a high degree of success, 2) demonstrates significant findlngs in 15% of cases, and 3) should therefore be performed whenever coronary angiographic findings make It appropriate to consider LIMA to coronary artery bypass grafting. Key words: coronary bypass, coronary angiography, LIMA angiography
INTRODUCTION
angioplasty or bypass surgery. In recommending a revascularization approach the possibility of initial success In view of its durability and longevity compared to and restenosis with angioplasty must be balanced against saphenous vein grafts (SVGs) [I] and its low incidence the possibility of initial success and long-term patency of atherosclerosis [2-61 the internal mammary artery with surgery, with the long-term benefit of an IMA graft (IMA) has become the graft of choice for coronary artery being superior to that of a SVG graft. We have long been bypass. Although the need for pre-operative angiography of the opinion that revascularization decisions such as of the IMA is well recognized in specialized circumthese should most appropriately be made with the knowlstances, such as previous saphenous vein graft bypass edge that the surgeon will or will not be able to use the surgery [7] or previous mastectomy and radiation therapy IMA as a graft. This requires angiographic demonstra[8], recommendations regarding the role for IMA angition of the IMA at the time of diagnostic catheterization ography in the more routine setting are less clear. Singh rather than the expectation that the surgeon will probably has suggested that initial diagnostic coronary arteriografind the IMA suitable. phy should routinely include IMA angiography [5-61, In those patients for whom bypass surgery is clearly and Krijne has recommended that routine pre-operative the most attractive option we have long felt that demonangiography of the IMA is “indeed unnecessary” [9], stration of the IMA angiographically allows the surgeon while a recent review by Kuntz and Baim made no specific recommendation but did observe that “pre-operative evaluation of the IMA at the time of initial diagnostic From the Division of Cardiology, Department of Medicine, Slate Uniangiography may also become more widespread if safe versity of New York Health Science Center at Brooklyn, Brooklyn, and rapid IMA catheterization techniques can be devel- New York. oped” [lo]. Others have recommended routine pre-operative IMA angiography on the basis of isolated case Received July 6, 1991; revision accepted September 17, 1991. reports [11,12]. Address reprint requests to Alan Feit, MD, Director, Interventional Patients having coronary artery disease requiring re- Cardiology, State University of New York Health Science Center at vascularization often have the option of either coronary Brooklyn, Box 1199, 450 Clarkson Avenue, Brooklyn, N Y 11203. 0 1992 Wiley-Liss, Inc.
86
Feit et al.
to plan what operation will be performed and to identify particular elements of the IMA anatomy which may require special attention at surgery, for example proximal branches which may be used as secondary bypass conduits and/or which may require intentional ligation to improve IMA flow, therefore saving time and preparation in the operating room, as well as providing for a more controlled operation. For these reasons we decided to perform IMA angiography as a routine part of coronary angiography whenever the coronary angiographic findings in our judgement made use of the IMA as a bypass graft a clinical consideration.
TABLE 1. Patient Characteristics. Age (yr)
30-39 40 -49 50-59 60-69 70-79 >80 Totals
Male
Female
0 14
2
27 28 9 0
25 9 I
78
52
I 14
*Total = 130.
brought proximally to selectively cannulate and visualize the IMA with nonionic contrast media. Our technique for IMA angiography was arrived at MATERIALS AND METHODS independently, but is quite similar to the technique deBetween October 1989 and January 1991 we per- scribed by Kuntz and Bairn [lo], with the exception that formed 734 diagnostic coronary angiography proce- we often found it necessary to use a Wholey (ACS) dures. IMA angiography was performed as a part of the guidewire rather than a standard J tip guidewire. The following parameters were recorded for each case: diagnostic coronary angiographic procedure in 130 cases hypertension; presence or absence of equality in blood (18%). in both arms; presence or absence of peripheral pressures Initially IMA angiography was usually limited to the vascular disease, including TIAs, bruits, claudication, or LIMA and was performed in the setting of bypassable peripheral bypass surgery; diabetes with or without inlesions in the LAD. As we became more proficient at the sulin required; renal failure; radiation therapy to the cannulation of the IMA, bilateral IMA angiography was chest; and previous bypass surgery using saphenous vein performed. Bilateral IMA angiography was only pergrafts. The following angiographic findings were reformed if in the judgement of the operator more than one IMA study: size at the fourth intercostal corded for each bypass graft was required, and the lesions were located space; presence or absence of a significant proximal in positions which made bypass with the IMAs possible. branch; site of origin in proximal, mid, or distal third of The decision whether or not to perform IMA angiograsubclavian; presence or absence of a subclavian and/or phy was an individual decision made by one of the three vertebral stenosis, and presence or absence of an IMA primary operators in this study based on that operator’s lesion; and the time required for IMA cannulation. In all interpretations of the coronary anatomy and the lesion cases non-ionic contrast media was employed. distribution. We were not attempting to define IMA angiography in all settings. Rather we were interested in the IMA angiRESULTS ographic findings only in the subset of patients who had coronary artery lesions for which IMA bypass grafting A total of 130 patients were studied. The patient charwould be an appropriate consideration. These results acteristics by age and sex are shown in Table I. Nineteen have been obtained in a selected group of patients with percent of the females were age 70 or greater while 12% coronary artery disease, and may not be applicable to of the males were age 70 or greater. Males outnumbered females by 1.5 to I . Of the patient population 8.5% had other populations. Patient age ranged from 39 to 82. Those who had peripheral vascular disease as evidenced by either TIAs, undergone previous bypass surgery were entered into the claudication, or previous peripheral vascular surgery, study only if they had SVGs used as bypass grafts at 8.5% had diabetes mellitus, and I .5% had either signifsurgery, and if the native IMAs had never been dissected icant renal failure as evidenced by creatinine greater than or visualized previously. 2.5%or previous renal transplantation. In those cases employing the femoral approach IMA As can be seen from Table I1 the LIMA was successangiography was performed by first directing a 5F IMA fully cannulated 93% of the time and the RIMA 83% of catheter into the proximal subclavian or innominate ar- the time. In our experience there is a short but very real teries and then advancing a guidewire into the distal sub- learning curve for cannulating the IMA. In six of the first clavian artery well beyond the origin of the IMA. The ten attempts to cannulate the LIMA the operator required catheter was then advanced over the wire beyond the greater than five min for successful selective cannulaIMA and, after withdrawing the wire, the catheter was tion. However, only ten of the next 120 cases required
IMA Angiography Should Be Routine TABLE II. Results of Attempted IMA Cannulations LIMA Number attempts Successful selective cannulation, 5 min Complications Total successful selective cannulations
TABLE 111. Angiographic Findings RIMA
I30
18
105
I1
16" Ib(0.8%)
121 (93%)
87
4 0
15 (83%)
'Six of these occurred in the first ten cases. Vransient occipital visual field loss after a non-selective injection.
more than 5 min for LIMA cannulation. We expect that with continued RIMA cannulations our success rate will improve as well. Only one complication related to IMA angiography occurred. This was an episode of transient occiptal visual field loss which occurred in a 75-yr-old hypertensive female. In this case the LIMA could not be selectively cannulated and was non-selectively visualized on an injection into the subclavian. In the nine cases of failure to selectively cannulate the LIMA and the three failures to selectively cannulate the RIMA, failure appeared to result from tortuosity in the proximal subclavian or proximal innominate arteries. Although wires could be positioned into the appropriate artery, advancing the preformed catheters caused the wire to twist out of the target artery. In two other cases upsizing from a 5F IMA catheter to a 7F catheter provided greater control and made selective cannulation possible. Only two of our cases were performed by the brachial approach. Both of these were done from the left arm to permit visualization of the LIMA. Table I11 lists the significant findings. A large proximal branch defined as a branch equalling at least 50% of the diameter of the IMA and occurring within two rib interspaces from the origin of the IMA was found in 9.2%. IMA caliber too small (judged relative to the target artery) to be utilized at bypass surgery was found in I .5%, and a totally occluded IMA (verified at coronary bypass surgery) was found once. Subclavian stenoses were found in 2.3%, vertebral stenosis was found twice (1.5%), and IMAs cut off at the time of previous bypass surgery and snared in the wire sutures were noted on two occasions. Figures 1 to 5 illustrate some of these findings. Excluding the two cases in which the IMAs were snared by wire sutures at the time of previous bypass surgery, 15% of the IMAs studied exhibited findings which would be of importance to the surgeon at the time of coronary bypass surgery. These two cases have been excluded since the need to perform IMA angiography in the setting of prior saphenous vein graft coronary artery bypass surgery is well established [7].
Significant findings
LIMA' (8)
RIMAh (B)
Large proximal branch Caliber too small Occluded Subclavian stenosis Vertebral stenosis Cut off at prior CABG
12 (9.2) 2 (1.5) 1 (0.8) 3 (2.3) 2 (1.5) 2 (1.5)
0
Total
22 (16)
Total excluding cut off at prior CABG
20 (15)
0 0 0 0 1 (.OSS)
'N = 130. bN = 18.
Fig. 1. This proximal branch of more than one-half the dlame ter of the LIMA arose within two rib Interspaces of the LIMA'S origin from the subclavian artery.
DISCUSSION
Conflicting recommendations regarding the need for selective visualization of the IMAs prior to coronary bypass surgery have persisted. Based on our experience it is clear that selective IMA angiography is required at the time of diagnostic coronary angiography if the coronary angiographic findings suggest that bypass surgery utilizing the IMA will be an appropriate option. The complication rate attributable to selective IMA angiography is small. We attribute this to the technique employed which required that a guidewire first be placed down the subclavian artery and then the catheter advanced over the guidewire. The low complication rate demonstrates that selective IMA cannulation can be performed safely. However. and just as importantly, the findings indicate that the knowledge gained from selective cannulation of the IMA is very relevant to any planned surgery. Ignoring the two patients whose IMAs were trapped in the wire sutures at
88
Feit et al.
Fig. 2. a: This LIMA was of too narrow caliber to serve as a graft to any of the target branches of the left coronary artery, b. Note that the magnification is increased from a to b, but compare the arterial caliber with the 5F catheter in a and the 6F catheter in b.
Fig. 3. a: This LIMA could not be visualized, and was verified to be totally occluded at surgery. b: The RlMA was normal.
previous bypass surgery we identified six out of 130 patients whose IMAs could not be used as coronary bypass conduits. Of these, two were of too small caliber, one was totally occluded (this was verified at the time of coronary bypass surgery), and three had significant proximal subclavian stenoses. Thus, between 4% and 5% of the native IMAs were unsuitable for bypass surgery. Significant proximal IMA branches occurred in 12 patients for a 9.2% occurrence rate. It is important that these large proximal branches be identified prior to surgery so that the surgeon knows how far back to dissect the artery in order to ligate them. For the extremely large lateral costal branches which were identified on two oc-
casions, I .5%, the entire surgical approach may be modified as these branches may be brought down to be used as bypass grafts with the IMA. We found no correlation between the IMA angiographic anatomy and the presence or absence of peripheral vascular disease, diabetes mellitus, or renal failure. Similarly we found no correlation between the appearance of the IMA and the patient’s age or sex. During the course of this study we have been impressed by the variable appearance of internal mammary arteries. The size range at the fourth intercostal space has been from less than 1 mm to greater than 4 mm. In other words, some are large, some are small, and some have
IMA Angiography Should Be Routine
89
results are different from those of Krijne [9], who observed such a low incidence of findings which would be significant prior to bypass surgery. It is our opinion that non-selective visualization of the IMA is often inadequate and should be discouraged in favor of becoming proficient at selective cannulation, a technique of importance not only because of the issues raised here, but also because of the number of IMAs which require post bypass assessment. We realize that the data presented were obtained in a selected population of patients with coronary artery disease. However, these were all patients with obstructive coronary lesions which suggested to the operator that the internal mammary arteries would be appropriate as coronary bypass grafts. Overwhelmingly this meant that proximal left anterior descending disease was present. Fig. 4. A large ulcerated left subclavian stenosis proximal to Although our results were obtained in a selective population, it is our belief that these results should apply to the LIMA. any other population with coronary artery disease coming for diagnostic coronary angiography. significant proximal branches. Clearly defining anatomOur experience indicates that selective IMA angiograical abnormalities such as those listed in Table 111 is of phy can be performed safely with a high degree of sucvalue prior to surgery. However, defining normal IMA cess, demonstrates significant findings in 15% of cases, anatomy prior to surgery is important in that it allows the and should therefore be performed whenever diagnostic physicians and patients to know that a bypass with longer coronary angiography makes it appropriate to consider patency is expected, and it allows the surgeon to plan IMA to coronary grafting. pre-operatively which vessel will serve as which bypass graft, and better assess the need for vein harvesting. In the early stages of this study we frequently perACKNOWLEDGMENTS formed non-selective IMA injections prior to seating the Grateful appreciation is extended to the following ancatheter selectively, and we have been and remain impressed that non-selective injections often fail to ade- giographers and fellows whose skills and efforts helped quately visualize the internal mammary artery; consider- make this work possible: Robert Kipperman, Arnold able and unpredictable streaming artifact varied from Einhorn, Tak Kwan, Robert Vaccarino, Madu Darawhat, case to case. We believe that this is the reason why our Joseph Idicula, Dino Paiusco, Donald Kruez, Chan-
Fig. 5. a, and b: Two views of a LIMA with a large lateral costal branch, arising very proximally.
90
Feit el al.
dranath Das, Ralph Morales, David Schifter, Marcel Pintea, K.S. Chandrasekhar, Nieca Goldberg, and Gerard Hanley.
REFERENCES I . Lewis M, Dehmer G: Coronary bypass using the internal mammary artery (Editorial). Am J Cardiol 56:480-482, 1985. 2. Finci L, Meier B, Steffenio G, Doriot PA, Rutishauser W: Nonselective preoperative digital subtraction of internal mammary arteries. Cathet Cardiovas Diagn 19:13-16, 1990. 3. Sisto T, Isola J: Incidence of atherosclerosis in the internal mammary artery. Ann Thorac Surg 472384-886, 1989. 4. Kay H, Korns M, Flemma R. Tector A, Lepley D: Atherosclerosis of the internal mammary artery. Ann Thorac Sure 21504501. 1976. 5 . Singh R: Radiographic anatomy of the internal mammary arteries. Cathet Cardiovas Diagn 7:373-386, 1981.
6. Singh R: Atherosclerosis of the internal mammary arteries. Cardiovas Intervent Radio1 6:72-77, 1983. 7. Davies SW: Pre-operative imaging of the internal thoracic artery (letter to the editor). Am J Cardiol 67:334, 1991. 8. Hanet C. Marschand E, Kukeyux A: Left internal mammary artery occlusion after mastectomy and radiotherapy therapy. Am J Cardiol 65:1044-1045. 1990. 9. Krijne R, Deng M, Heinich K, Sons H, Knan A: Semi-selective angiography of the internal mammary arteries as a preparation for coronary bypass surgery. Am J Cardiol 66:377-378, 1990. 10. Kuntz R , Baim D: Internal mammary angiography: A review of technical issues and newer methods. Cathet Cardiovas Diagn 20: 10-16. 1990. I I . Stullman W, Hilliard G: Unrecognized internal mammary artery stenosis treated by percutaneous angioplasty after coronary bypass surgery. Am Heart J 113:393-395, 1987. 12. Raner W, Sadler T, Liggett M: Internal mammary arteriography prior to coronary artery bypass surgery. Chest 641523-524, 1973.
