In Vivo Coronary Angioscopy

J AM COLL CARDIOL 1311 1983.1(5) 1311-4

J. RICHARD SPEARS, MD, FACC , H. JOHN MARAIS, MD, JUAN SERUR, MD, OLEG POMERANTZEFF, ° Eng, ROBERT P. GEYER , PhD, ROBERT S. SIPZENER, MD, RONALD WEINTRAUB, MD, FACC, ROBERT THURER, MD, FACC, SVEN PAULIN, MD, RICHARD GERSTIN, BS, WILLIAM GROSSMAN, MD, FACC Boston. Massachusetts

The feasibility of in vivocoronary angioscopywas tested catheterization technique and in patients during open utilizing a 1.8 mm angioscope in vesselswhere blood had heart surgery. The results demonstrate the feasibility been replaced by optically clear liquids, including a new and potential clinical usefulness of direct visualization perfluorocarbon emulsion. After trials in postmortem of intravascular anatomy and disease, analogous to en• canine and human coronary arteries, in vivo intralum• doscopy of other organ systems. inal visualization was accomplished in the dog with a

At present, the primary diagnostic method for evaluating technique , progressing from in vitro studies of postmortem coronary anatomy in humans beings is selective coronary hearts to in vivo studies of canine and human coronary angiography. Although this method delineates the coronary arteries. lumen with reasonable accuracy , it allows recognition of Methods pathologic changes principally by identifying filling defects. Description of angioscope. The prototype Olympus ul• Direct visualization of the internal surface of coronary ar• trathin fiberscope used in this study has an external diameter teries would permit not only precise determination of coro• of 1.8 mm and consists of two separate fiberoptic bundles . nary cross-sectional area, but also identification of different One bundle transmits light from an external source through types of vascular disease (for example, thrombus versus the fiberscope into the vessel being examined and the second atheroma) . The major barriers to in vivo visualization of bundle, the image guide, transmits the reflected image back coronary arteries in the past include the opacity of blood to to an adjustable eyepiece. To reduce the outer diameter of visible light and the large size of commercially available this prototype unit to 1.8 mm, the angulation system and fiberoptic scopes relative to coronary artery dimensions (l,2). channels usually provided in endoscopes were not included . The recent development of a high resolution 1.8 mm fiber• Illumination was provided by an Olympus ILK-3 150 watt optic scope, as well as a translucent perfluorocarbon emul• halogen light source. Photography was accomplished by sion blood substitute, has made prolonged in vivo visual• using a special endoscopic camera adapter that was attached ization of the coronary arteries possible. In this report, we to the eyepiece of the endoscope and then fitted to the OM• describe our initial studies during the development of this 2 camera body. The standard ground glass focusing screen was replaced by a special Fresnel lens focusing screen de• signed exclusively for endoscopy . With this photographi c From the Charles A. Dana Research Institute and Harvard-Thorndike system, we were able to perform the endoscopic procedure Laboratory of Beth Israel Hospital. Department of Medicine. Cardiovas• cular Division, and the Radiology Research Center of the Department of through the viewfinder of the 35 mm camera, taking pictures Radiology. Beth Israel Hospital and Harvard Medical School. Boston. whenever appropriate. Massachusetts. This work was supported in part by Young Investigatorship Description of perfluorocarbon emulsion. A new Research Award HL2738-02 from the National Heart. Lung, and Blood Institute. Nauonal Institutes of Health, Bethesda. Maryland. Manuscript translucent perfluorocarbon emulsion blood substitute has received September 21. 1982; revised manuscnpt received December 7. recently been produced by one of us (R.P.G.). Perfluoro• 1982; accepted December 10. 1982 carbon-containing blood replacement preparations can pro• Address for reprints: J. Richard Spears , MD. Department of Medicine, Cardiovascu lar Division . Beth Israel Hospital. 330 Brookline Avenue. vide both oxygen transport and maintenance of oncotic pres• Boston, Massachusetts 022 I5. sure (3-5). These compounds must be emulsified, but the

resulting emulsions are usually fairly turbid. This turbidity graft surgery, angioscopy was performed after completion would interfere with intravascular visualization by means of the distal anastomosis of the saphenous vein graft to the of fiberoptics; therefore, it is necessary to utilize prepara• left anterior descending coronary artery. Through a side tions that are as low in turbidity as possible. By means of branch of the vein graft, the angioscope was advanced across sonication using suitable emulsifying agents and perfluo• the distal anastomosis and into the native coronary artery. rocarbons, the translucent perfluorocarbon-containing blood In three additional patients, the angioscope was advanced replacement preparation was made. As in previous studies, anterograde into the left anterior descending coronary artery electrolytes and an oncotic agent such as hydroxyethyl starch directly through the arteriotomy. Intraluminal visualization were added. The completed preparation was sterilized by was achieved in each patient with the aid of the translucent filtration and equilibrated with a mixture of 95% oxygen cardioplegic solution normally used to perfuse the vein grafts and 5% carbon dioxide, and the pH was adjusted to 7.45. and coronary arteries. Postmortem studies. Studies were initially conducted on postmortem canine and human hearts from which blood Results had been washed out with normal saline solution. Structures Postmortem canine studies. Initial development of the within the left ventricle were viewed with the fiberscope angioscopic technique involved studies in postmortem ca• which had been passed retrograde across the aortic valve. nine hearts in which details of light source and intensity, An 8 French Cordis sheath was used to gain access to the photographic technique and catheter manipulation could be left main coronary artery. The vessel was perfused at phys• investigated in a totally controlled setting. Figures 1 and 2 iologic pressures with either saline solution or the perfluo• illustrate the detail and photographic quality of images of rocarbon emulsion by way of the sidearm of the sheath. The the mitral valve commissure (Fig. 1) and papillary muscles angioscope was then passed through the rubber diaphragm with chordae (Fig. 2) obtained in the postmortem saline• of the Cordis sheath into the left coronary artery orifice. filled heart. The black spots in these and the other figures Selective angioscopy of the proximal segments of the cir• represent broken optical fibers that were not transmitting cumflex and anterior descending coronary arteries was then the reflected image back to the observer and external camera. performed. Figure 3 shows the lumen of the left anterior descending In vivo canine studies. Adult mongrel dogs weighing artery (viewed from the aortic orifice of the left coronary 20 to 25 kg were anesthetized with morphine sulfate and artery) in a postmortem canine heart using similar techniques. chloralose, intubated and then ventilated with a Harvard In vivo canine studies. After successfully defining the respiratory pump. Cutdown procedures were performed over technical factors necessary to obtain photographic images both carotid arteries. A thoracotomy was performed over under ideal static conditions (Fig. lA, B and C), we pro• the left lateral chest and the heart was suspended in a peri• ceeded to in vivo studies of the normal canine coronary cardial cradle. A 7 French NIH cathether was placed in the vasculature using oxygenated Ringer's lactate solution and descending aorta for measurement of arterial pressure. A perfluorocarbon solution. Figures ID and IE show the in• 21 gauge angiocatheter was inserted into a distal branch of ternal (lumen) surface of the left coronary artery in anes• the left anterior descending coronary artery. Arterial and thetized dogs; these photographs were taken through the intracoronary pressures were measured with Statham P23D6 ultrathin angioscope that had been advanced by a guiding strain gauges and recorded on an Electronics for Medicine catheter into the ostium of the left coronary artery. The ostia multichannel photographic recorder. of arterial branches of the left coronary artery are clearly A 10 French guiding catheter was constructed to facilitate visible. entrance of the fiberscope into the left main coronary artery. Clinical studies. After completion of the in vivo canine After insertion into the carotid artery, the tip of the guiding studies, we studied four patients during the course of coro• catheter was advanced to the left main coronary artery under nary artery bypass graft surgery. Figure 1F shows an ath• fluoroscopy. Ifdistal coronary artery pressure decreased as erosclerotic plaque within the lumen of the left anterior a result of ostial obstruction by the guiding catheter, per• descending artery in one of these patients, after the ultrathin fusion of the left coronary artery could be achieved by angioscope was passed into the artery by way of a side autotransfusion through a connection to the proximal end branch of the saphenous vein graft. The plaque appears of the NIH catheter. After insertion of the fiberscope in all white and is raised from the wall, partially occluding the dogs, visualization was achieved by sheath sidearm infusion vessel lumen. of either oxygenated Krebs-Ringer's solution or the per• fluorocarbon emulsion, both of which were maintained at 37°C before injection. Discussion In vivo human studies. Informed consent was obtained The development of an ultrathin fiberoptic instrument from all patients in whom coronary angioscopy was at• (6-8) has made possible the intraluminal visualization of tempted. In one patient undergoing coronary artery bypass coronary arteries in vivo. Replacement of intraarterial blood IN VIVO CORONARY ANGlOSCOPY J AM Call CARDIOl 1313 1983.1(5).1311-4

Figure 1. A, Mitral valve commissure as viewed from left ventricle of postmortem canine heart filled with normal saline solution. B, Canine anterior-superior papillary muscle and chordae ten• dineae in postmortem left ventricle filled with normal saline solution. C, Left anterior descending coronary artery, filled with normal saline solution in postmortem canine heart. D, In vivo view of trifurcation of left coronary artery into (from left to right) circumflex, intermediate and anterior descending branches in the dog during perfusion with oxygenated Krebs-Ringer's solution. E, In vivo view of bifurcation of left mainstem coronary artery into circumflex (left) and anterior de• scending (right) branches during injection of perfluorocarbon emulsion into canine coronary artery. F, In vivo view of human left anterior descending coronary artery lumen showing encroachment by atherosclerotic plaque. Blood replacement with cardioplegic solution permitted intraluminal visualization.

with clear liquids such as simple crystalloid solutions or a the coronary vessels but recognizes pathologic changes prin• recently developed translucent perfluorocarbon emulsion is cipally by identifying filling defects. The nature of these necessary. Perfluorocarbon emulsions have been used for defects (for example, atheroma, thrombus, plaque, hem• transfusion in patients with a remarkable lack of significant orrhage) frequently cannot be determined by the coronary side effects (9) and are currently being tested in clinical angiogram; moreover, the cardiac surgeon performing by• trials. Moreover, they have been shown experimentally to pass surgery is rarely able to provide information concerning have a protective effect on ischemic myocardium (10) and the nature of the coronary obstructions that prompted the to be efficacious when used for total cardiopulmonary by• operative procedure. Coronary angioscopy at the time of pass (11). However, a rare idiosyncratic adverse pulmonary diagnostic cardiac catheterization or during cardiac surgery, reaction, which appeared to be immunologic in origin and or both, should provide important information regarding the was successfully treated with corticosteriods and diphen• nature of the underlying disease process. hydramine, was recently reported (12). During cardiac catheterization, it may also be possible Clinical applications. There are numerous potential ap• to differentiate atheroma from residual thrombus in patients plications of coronary angioscopy. The principal current undergoing fibrinolytic treatment for acute myocardial in• method for evaluation of human coronary anatomy is coro• farction (13). Insight may be gained into the mechanism of nary angiography, which defines accurately the lumen of abrupt reclosure in patients undergoing percutaneous trans- 1314 J AM COLL CARDIOL SPEARS ET AL 1983.1(5) 1311-4

luminal coronary angioplasty (14,15). Cardiac surgeons References may find the angioscope useful as a means for evaluating I. Tanabe T, Tokota A, Sugie S. Cardiovascular fiberoptic endoscopy. the patency of vein graft anastomoses to coronary arteries development and clinical apphcanon. Surgery 1980;87:375-9. during open heart surgery. In addition, preliminary studies 2. Moser KM, Shure D, Harrell JA, Tulumello J. Angioscopic Visual• with lasers on artherosclerotic plaques (16) raise the pos• ization of pulmonary emboli. Chest 1980;77: 198-20I. sibility of laser-mediated "atheroplasty" under direct vis• 3. Clark LC Jr, Kaplan S, Becattirn F. The physiology of synthenc blood. J Thorac Cardiovasc Surg 1970;60:757-73. ualization, if this modality can be coupled to a miniature 4. Geyer RP. Whole animal perfusion with fluorocarbon suspensions. fiberscope. Fed Proc 1970;29:1758-63. Riskof adverse effects. Even though no untoward effect 5. Geyer RP. Fluorocarbon-polyol artificial blood substItutes. N Engl J from coronary angioscopy was observed in any of the human Med 1973;289:1077-82. or dog studies, instrumentation of coronary arteries with the 6. Wood RE, Fink RJ. Applications of flexible fiberoptic bronchoscopes angioscope is likely to carry a finite risk, although small, in infants and children. Chest 1978;73(suppl 1):1-737-40. 7. Rucker RW, Silva WJ, Worcester Cc. Fiberopnc bronchoscopic of an adverse effect, such as provocation of coronary spasm, nasotracheal intubation in children. Chest 1979;76.56-8. dissection or perforation. Until the incidence and severity 8. Vauthy PA, Weinfeld I, Katzman G, Knshnau V. Fiberoptic bron• of potential risks are better defined, coronary angioscopy choscopy in the neonate (abstr). Pediatr Res 1979.13:508. should be performed only under a research protocol by ex• 9. Mitsuno T, Ohyanagi H, Naito R. Chrucal studies of a perfluoro• perienced angiographers skilled in techniques of catheter chemical whole blood substitute (Fluosol-DA). Ann Surg 1982;195:60• 9. manipulation and at centers where potential complications 10. Glogar DH, Kloner RA, Muller J, DeBoer LWV, Braunwald E. fluo• can be treated immediately. In this regard, because the po• rocarbons reduce myocardial ischemic damage after coronary occlu• tential risks are not yet defined, it is unknown whether an sion. Science 1980;211:1439-41. attempt to obtain additional diagnostic information with an• II. Kanter KR, Jaffin JH, Ehrlichman RJ, Flaherty JT, Golt VL, Gardner gioscopy during surgery or cardiac catheterization is war• TJ. Superiority of perfluorocarbon cardioplegia over blood or crys• talloid cardioplegia. Circulation 1981;64(suppl 11):11-75-83. ranted clinically. Moreover, it should be emphasized that 12. Vercellotti GM, Hammerschmidt DE, Craddock PR, Jacob HS. Ac• visualization of the lumen is currently limited to proximal tivation of plasma complement by perfluorocarbon artificial blood: coronary artery segments because of the lack of an angu• probable mechanism of adverse pulmonary reactions in treated patients and rationale for corticosteriod prophylaxis, Blood 1982;59:1299• lation mechanism in the fiberscope. 1304. In this study, we reported on the development of meth• 13. Markis JE, Malagold M, Parker JA, et al. Myocardial salvage after odology for coronary angioscopy and documented its prac• intracoronary thrombolysis with streptokmase in acute myocardial in• tical application in patients undergoing cardiac surgery. The farction. Assessment by intracoronary thallium-201 N Engl J Med in vivo canine studies have encouraged us to believe that 1981;305:777-82. similar studies of human coronary anatomy may be possible 14. Gruntzig AR, Jennmg A, Siegenthaler WE. Non-operative dilatation at the time of diagnostic cardiac catheterization, and efforts of coronary artery stenosis. Percutaneous transluminal angioplasty. N Engl J Med 1979;301:61-8. are currently underway to test appropriate guiding catheters to facilitate coronary angioscopy in patients. IS. Kent KM, Bentivoglio LG, Block PC, et al. Percutaneous transluminal coronary angioplasty: report from the Registry of the National Heart, Lung, and Blood InstItute. Am J Cardiol 1982;49:2011-20. We gratefully acknowledge the expert technical assistance of Robert Lowe and Deborah Shropshire in the performance of the canine studies. 16. Lee G, Ikede RM, Kozina J, Mason DT. Laser dissolution of coronary atherosclerotic obstrucnons. Am Heart J 1981;102:1074-5