In vivo human comparison of intravascular ultrasonography and angiography

In vivo human comparison of intravascular ultrasonography and angiography M a r w a n Tabbara, M D , R o d n e y White, M D , Douglas Cavaye, F R A C S , and George Kopchok, BS, Torrance, Calif. This study evaluates the in vivo correlation' ofintravascular ultrasonography and uniplanar angiography in determining the luminal dimensions of normal and moderately atherosclerotic human arteries. Five French and 8F rotating A scan intravascular ultrasound catheters were used to obtain 48 images in four superficial femoral arteries, five iliac arteries, and one aorta in eight patients undergoing vascular surgery. Cross-sectional areas measured by intravascular ultrasonography were compared to cross-sectional areas calculated by uniplanar angiography of the same location in the vessel. Maximum and minimum luminal diameters were also measured from intravascular ultrasound images. An ellipticity index was defined as the maximum/minimum diameter ratio (max/min) and ranged from 1.0 to 1.8 (mean, 1.2). Comparison of the cross-sectional areas measured from intravascular ultrasound images and those calculated from uniplanar angiography showed no significant difference at any level of ellipticity studied. However, when the values of cross-sectional areas were analyzed in groups corresponding to the diameter of the vessel, that is, aortic, iliac, and femoral, the values for the iliac arteries calculated from uniplanar angiography were significantly greater by 9.8% -+ 0.7% (n = 29, p = 0.03) when compared to those measured by intravascular ultrasonography. In addition to providing accurate luminal determinations, intravascular ultrasound images displayed transmural morphology, the location and character of the atherosclerotic lesions, and the thickness of the vessel wall. We conclude that intravascular ultrasound imaging provides accurate, novel information regarding human vessels and that this technology may play a significant role in future diagnostic and interventional therapies. (J Vase SURG 1991;14: 496-504.)

Intravascular or intraluminal ultrasonography (IUS) is an imaging modality that enables transmural visualization of blood vessels by an ultrasound transducer placed at the tip of a catheter that is introduced into the vascular system either percutaneously or through an arteriotomy or venotomy. Imaging of the vessel lumen and wall is performed in a tomographic fashion by advancing the tip of the catheter to the areas of interest. >3 A computerized processing system displays real time images of the vessels. Luminal dimensions obtained with use of IUS catheters have been shown to correlate significantly with values measured from corresponding histologic specimens and vessel angiograms both in vitro and in vivo2 12 From the Department of Surgery, Harbor-UCLA Medical Center, Torrance. Experimentalwork performedin this study supported in part by C-VIS, Inc., Sunnyvale,Calif. Presented at the Sixth AnnualMeeting of the Western Vascular Society, Palm Springs, Calif.,Jan. 13-16, 1991. Reprint requests: Rodney A. White, MD, Harbor-UCLA Medical Center, 1000 West Carson St., Torrance, CA 90509. 24/6/30559 496

Uniplanar angiography can be quite accurate in defining vessel luminal cross-sectional area if the vessel is circular, as it is in most normal and mildly diseased arteries. Clinically significant atherosclerotic occlusive disease is usually eccentrically positioned in the arterial lumen, and the lumen may be either circular or elliptic in shape, although most are circular. 13-1s In instances where the lumen is elliptic, biplanar angiograms more accurately define luminal cross-sectional areas and calculation of percent area stenosis. 16 This study compares IUS and uniplanar angiography for determining the luminal dimensions of normal and atherosderotic human arteries in patients undergoing vascular surgery. The ability of IUS to delineate the morphology of the arterial wall and distribution of lesions was also addressed. MATERIAL AND METHODS U l t r a s o u n d imaging Intraluminal ultrasound images of arterial segments were obtained by use of 5F (30 MHz) and 8F (20 MHz) catheters with the ultrasound transducers

Volume 14 Number 4 October 1991

located at the tip of the device (Cardiovascular Imaging Systems, C-VIS, Inc., Sunnyvale Calif.). The ultrasound beam is projected at a 90-degree angle to the long axis of the catheter by a mirror rotating at 15 cycles per second (cps) at the end of a coaxial shaft inside the catheter. The imaging elements are housed in a plastic chamber at the tip of the catheter that is filled with saline solution through a separate port at the proximal end. The system provides 360-degree cross-sectional slices of the vessel in a plane perpendicular to the catheter. Gray scale, real-time images are displayed on a monitor and recorded by an on-line, high-resolution video recorder. Photographs of the images were obtained from an in-line camera. Measurement of vessel dimensions, luminal diameters, and cross-sectional areas were performed by an on-line processing unit that calculates the area of the lumen outlined on still images on the monitor. Alternatively, the areas can be calculated by use of a digitized pad (Microplan II Laboratory Computer Systems, Inc., Cambridge, Mass.) from calibrated photographs of the images. All IUS images were obtained during operation. At the time of a concomitant vascular procedure, the catheters were introduced through an arteriotomy. For each image obtained, a distance between the rotating mirror and the entry site marked with a metal clip was recorded to identify the location of the image for comparison to angiograms of the vessel. Anteroposterior orientation of the vessel and catheter was not correlated in most images. Although this would be needed in many studies to acquire accurate data, it does not affect the comparison of crosssectional area calculated from anteroposterior anglograms to those measured from IUS images. Luminal cross-sectional areas and maximum and minimum luminal diameters were measured for each IUS image. Before performing the ultrasound examinations, all patients were entered in a clinical protocol with a consent approved by our institution's human subjects and research committees.

Angiography Preoperative uniplanar angiograms corresponding to the segments of vessels that were imaged by IUS were used to determine luminal dimensions. A surgical clip placed at the proximal end of the arteriotomy allowed identification of the site of the introduction of the IUS catheter by comparing the angiograms to a plain radiograph taken after the procedure to identify the clip. The angiogram and the postprocedure film were overlapped, and the catheter introduction site was marked on the angiograms after

Intravascular ultrasonography compared to angiography 497

correction for differences in radiograph magnification. Luminal diameters were then measured from the angiogram at the site of the IUS images by use of a digitized pad and a computerized image-analysis system. The magnification of arteriograms was determined in each study by use of radiopaque markers of known length. The measured diameters were corrected by the corresponding value in each instance. A cross-sectional area was calculated with the diameter measured from the angiograms by use of the following formula: area -- "rr(~)~ that is, assuming that the lumen of the vessel imaged was circular. Statistical methods Intravascular ultrasonography enabled measurement of maximum and minimum luminal diameters of arterial segments. This permitted calculation of an ellipticity index for each IUS image, which was defined as the ratio of maximum to minimum (max/min). As this index increases, the lumen becomes more elliptic. Luminal cross-sectional areas measured by IUS were compared to those calculated by uniplanar angiography in groups defined by the variables constituting each eUipticity index, beginning with an ellipticity index = 1 and increasing by 0.1 for each subsequent test. In addition, the values were analyzed in groups corresponding to the diameter of the vessels studied, that is, aortic, iliac, and femoral. The comparisons were done by use of a paired Student t test and the correlations were done with use of linear regression analysis.

RESULTS The images obtained by IUS clearly delineated vessel wall morphology in most examinations. Both normal artery wall and soft plaque were easily identified (Fig. 1). Interpretation of images is enhanced during continuous real-time studies or by viewing video recordings. This is particularly true in diseased vessels. Calcified arterial wall plaque attenuated the ultrasound signal beyond the lesions, identifying the location and distribution of calcification (Fig. 2). Compared to the angiograms, IUS images provided more detailed information regarding the location, eccentricity, and cross-sectional area of the lumen (Fig. 3). A total of 48 ultrasound images were analyzed. Ellipticity index ranged from 1.0 to 1.8 (mean, 1.2). The plot of ellipticity index values versus the number of IUS images at a particular ellipticity index indicated that most of the images had ellipticity indexes of 1.0 to 1.3 that is, circular or nearly circular

498

Journal of VASCULAR SURGERY

Tabbara et al.

Fig. 1. Intravascular ultrasound image of human iliac artery by use of an 8F catheter. U, Ultrasound probe; single arrow, normal vessel wall; double arrfr~v, soft plaque. Shadow a, is an artifact produced by a thin metalic wire along the long axis of the catheter. This can be used to orient the catheter in the vessel by noting its location when the catheter is inserted and carefully maintaining its position throughout the examination.

Fig. 2. Intravascular ultrasound image of human iliac artery with use of an 8F catheter in iliac artery. Note calcific plaque (arrow) with complete attenuation of the signal beyond the lesion. lumens in most arteries (Fig. 4). Comparison o f luminal cross-sectional areas measured by IUS to those calculated by angiography for each ellipticity index value at 0.1 increments from 1.0 to 1.8 showed no significant difference. Moreover, the correlation between the areas for indexes o f 1.0 (n = 16, r = 0.98) and 1.1 (n = 13, r = 0.96) was significant (p < 0.05), above 1.1 values were inadequate to

determine a significant correlation. Vv~nen the mean luminal cross-sectional area for the arteries was examined in groups corresponding to aortic, iliac or femoral locations, the iliac values calculated by angiography (44.0 _ 17.2 m m 2) were greater than those measured by IUS (39.7 +__ 1.9 mm2)p = 0.03, n = 29, Table I. The correlation o f values was also good (r = 0.8,p < 0.05) (Fig. 5).

Volume 14 Number 4 October 1991

Intravascular ultrasonography compared to angiography

499

Fig. 3. Comparison of angiography and IUS in the common and external iliac arteries. A, normal lumen; B, severe stenosis in the external iliac artery; and C, normal vessel distal to the lesion. Note the three-layer appearance of the muscular artery wall in the normal arterial segment. Single arraws delineate a 77% luminal compromise in ultrasound image B. Double arrows identify the echolucent media in each figure.

DISCUSSION

Arteriography is the "gold standard" for imaging the peripheral vasculature. The area of the vessel lumen is estimated by calculating this dimension with use of vessel diameters measured from uniplanar or biplanar angiograms. Unless the lumen is circular, the accuracy of this method can be limited. Transcutaneous ultrasonography can image arteries in a tomographic (perpendicular to the long axis) or coronal projection (parallel to the long axis). 17 However, it is limited by the intervening skin and soft tissue, which will attenuate the signal and decrease signal penetration. Use of lower frequencies ( < 10

MHz) allows definition of deeper arteries but sacrifices resolution. Intravascular ultrasonography is able to image vessels in a tomographic projection at higher frequencies (20 to 30 MHz) because blood is the only intervening substance between the probe and the arterial wall, and because less depth of penetration is required. Recently, coronal reconstructions of arteries has become possible by constructing threedimensional maps of serial images. 18 The images obtained by IUS clearly delineate vessel wall morphology) 9-21 The luminal circumference is the best seen layer in an artery because blood

500

Journal of VASCULAR SURGERY

Tabbara et al.

20

,

0

15 NO.

1,1

I0

IUS 1,2

IMAGES 5

1,]

1.4 • •

0

' 1.0

1,7

1,6

•

•

1,8

I

!.5

2.0

Ellipticity index

Fig. 4. Comparison of ellipticity index and the number of images obtained. Note that 73% of the images have an ellipticity index value of 1.0 to 1.3.

Table I. Lumen cross-sectional areas (mm 2) Aorticsegmen~

IUS

Angiography

No. of cases 5 5 Minimum 153.9 141.0 Maximum 191.1 219.0 Mean ± SD 173.2 + 15.3 163.9± 31.4 % mean difference+ SD 5.7 ± 9.8

and intima have very different acoustic impedances. On the contrary, the outer adventitial interface can be difficult to visualize because the adventitia has the same acoustic impedance as the surrounding 6ssue. In muscular arteries the media represents a hypoechoic layer, defining the outer limits o f any noncalcified plaque with the luminal circumference defining the inner limits. This allows accurate localization and measurement o f the thickness o f the plaque. 22 Calcification impedes ultrasound transmission and makes reliable measurement o f thickness difficult. 23 Our results have shown that luminal crosssectional areas obtained by IUS correlate very well with those o f angiography in normal and minimally diseased arteries. These results are consistent with other studies imaging primarily the coronary circulation. 24-26 In this study most o f the lumens imaged were only mildly elliptic, with 73% o f the images having eUipticity indexes o f 1.3 or less. Therefore the sample o f vessel segments imaged was

Iliac segments IUS

Angiography

29 29 16.6 15.9 78.5 84.9 39.7 + 16.9 44.0 ± 17.2 9.8 ± 0.7

Superficialfemoral artery segments IUS

Angiography

14 14 5.3 5.7 24.6 23.8 11.7 + 6.4 12.3 _+ 7.2 4.9 ± 6.5

not random and was dictated by the ability to pass the IUS catheter (diameter, 1.7 and 2.7 mm) through thc lumen. In diseased arteries with elliptic lesions, the calculation o f the luminal cross-scctional area by angiography is probably inaccurate. This assumption is supported by the finding that there was a significant difference in cross-sectional calculated from angiography and those measured by IUS, with the values determined from angiography being greater when the iliac vessels were cxamined alone. The aortic and femoral segments were probably not significant because the aortic images were o f large circular vessels, and because highly stenotic femoral lesions are eliminated from evaluation as they would not permit passage o f the IUS catheter. In addition to the limitations o f angiography in defining luminal dimensions o f elliptic vessels, it gives no information on vessel wall morphology. In this regard, other imaging modalities are being used

Volume 14 Number 4 October 1991

Intravascular ultrasonography compared to angiography 501

80 B

60

_~

•

40 °°o o ~ O

OO

20 O

0

I

I

20

40

__

l

60

80

100

ANGLO Fig. 5. Linear regression of luminal cross-sectional area (mm 2) measured by IUS and luminal cross-sectional area calculated from angiography for the iliac artery segments. Y = 4.1 + 0.8x. (n = 29, r = 0.80,p < 0.05).

to help define the extent and morphology of the atherosclerotic plaque. Angioscopy can clearly visualize the lumen of a blood vessel but has limited ability in defining vessel wall morphology and the distribution of the atherosclerotic plaque. Intravascular ultrasonography enhances the intraluminal perspective of angioscopy by defining luminal dimensions and vessel wall morphology. 27 Despite the theoretic considerations and the emerging data comparing arteriography to IUS, the former imaging modality is still regarded as the gold standard for lesion evaluation before and after procedures are performed and for guiding the application of interventional devices. 28'29 As an invasive catheter-based technology, IUS is restricted to diagnostic applications during invasive procedures, and to enhancing the precision of endovascular surgical therapy. Recent studies suggest that angiography is deficient in assessing the outcome of endovascular procedures. 3° In this respect, IUS promises to provide a control for evaluation of the indications for, and determining the outcome of intravascular methods. In the future, IUS may become the principal guidance method for ablative endovascular techniques such as atherectomy or pulsed-laser angioplasty, 31'~2enabling precise debulking of lesions and avoiding ablation of normal wall. Theoretically, this will decrease the incidence of vessel wall perforation and help limit recurrance of lesions related to inadequate removal. CONCLUSION In normal and mildly diseased arteries, uniplanar angiography is relatively accurate in measuring luminal diameters and calculating a luminal cross-

sectional area. In significantly diseased vessels, IUS may provide a more accurate evaluation of vessel wall and luminal morphology and plaque distribution. Intravascular ultrasonography also has the potential of guiding endovascular ablative techniques and assessing accurately the end result of therapy. REFERENCES 1. White RA. Intravascular ultrasound. In Kadir S, ed. Current practice of interventional radiology. Philadelphia: BC Decker Inc, 1991:36-41. 2. Pandian NG, Kreis A, Brockway B, et al. Ultrasound angioscopy: real time, two dimensional intraluminal ultrasound imaging of blood vessels. Am I Cardiol 1988;62: 493 -4. 3. Kopchok GE, White RA, Guthrie C, Hsiang Y, Rosenbaum D, White GH. Intraluminal vascularultrasound: preliminary report of dimensional and morphologic accuracy. Ann Vasc Surg 1990;4:291-6. 4. Kopchock G, White R, White G. Intravascular ultrasound: a new potential modality for angioplasty guidance. Angiology 1990;41:785-92. 5. Nissen SE, Gurley JC, Grines CL, Booth DC, Fischer C, DeMaria AN. Comparison of intravascular ultrasound and angiography in quantitation of coronary dimensions and stenoses in man: impact of lumen eccentricity [Abstract]. Cir~lation 1990;82(supp1111):1II-440. 6. Tabbara M, Kopchok G, White R. In vitro and in vivo evaluation of intraluminal ultrasound in normal and atherosclerotic arteries. Am J Surg 1990;160:556-60. 7. Meyer CR, Chiang EH, Fechner KP, et al. Feasibilityof high resolution intravascularultrasonic imaging catheters. Radiology 1988;168:113-6. 8. Yock PG, Johnson EL, Linker DT. Intravascular ultrasound development and clinical potential. Am J Cardiac Imag 1988;2:185-93. 9. Hodgson J, Eberle MI, Sarakus AD. Validation of a new real time percutaneous intravascular ultrasound imaging catheter [Abstract]. Circulation 1988;78(supp111):11-21. 10. Nissen SE, Grines CL, Gurely JC, et al. Application of new

502

11.

12.

13.

14. 15.

16.

17.

18. 19. 20.

21.

22.

Journal of VASCULAR SURGERY

Tabbara et al.

phased-array ultrasound imaging catheter in the assessment of vascular dimensions. Circulation 1990;81:660-6. Neville RF, Bartorelli AL, Leon MB, Potldn B, Gessert J, Sidawy AN. Validation and feasibility of in vivo intravascular ultrasound imaging with a new flexible catheter. Surg Forum ACS 1989;75:314-6. Mallery IA, Tobis JM, Gessert J, et al. Evaluation of an intravascular ultrasound imaging catheter in porcine peripheral and coronary arteries in vivo [Abstract]. Circulation 1988;78(suppl II):II-21. Zarins C, Zamra MA, Glagov S, Correlation of postmortem angiography with pathologic anatomy: quantitation of atherosclerotic lesions. In: Bond MG, Insull W, Glagov S, et al., eds. Clinical diagnosis of atherosclerosis, New York: Springer Verlag, 1983:283-306. Roberts KW. Coronary arteries in coronary heart disease: morphologic observations. Pathobiol Annu 1975;5:249. Waller BF. The eccentric coronary atherosclerotic plaque: morphologic observations and clinical relevance. Clin Cardiol 1989; 12:14-20. Sumner DS, Russell JB, Miles RD. Pulsed Doppler arteriography and computer assisted imaging of carotid bifurcation. In: Bergan IJ, Yao IST, eds. Cerebrovascular insufficiency. New York: Grune and Stratton, 1983:115-35. Pignoli P, Tremoli E, Poll A, et al. Intimal plus medial thickness of the arterial wall: a direct measurement with ultrasound imaging. Circulation 1986;6:1399-406. Cavaye DM, Tabbara MR, Kopchok GE, et al. Three dimensional vascular ultrasound imaging. Am Surg (In press). Fitzgerald PJ, Yock PG, Ring ER. Catheter tipped ultrasonic intravascular imaging. Perspect Vasc Surg 1989;2:113-25. Yock PG, Linker DT, Angelsen BAJ. Two dimensional intravascular ultrasound: technical development and initial clinical experience. J Am Soc Echocardiogr 1989;2:296-304. Pandian NG. Intravascular and intracardiac ultrasound imaging. An old concept, now on the road to reality. Ciraalation 1989;80:1091-4. Mallery ]A, Tobis JM, Griffith l, et al: Assessment of normal

23.

24.

25.

26.

27. 28.

29.

30.

31.

32.

and atherosclerotic arterial wall thickness with an intravascular ultrasound imaging catheter. Am Heart J 1990;119:1392400. Mallery JA, Tobis JM, Gessert l, et aL Identification of tissue components in human atheroma by an intravascular ultrasound catheter [Abstract]. Circulation 1988;78(suppl II):II22. Waller BF. Anatomy, histology, and pathology of the major epicardial coronary arteries relevant to echocardiographic imaging techniques. 1 Am Soc Echocardiogr 1989;2:232-52. Arnett EN, Isner IM, Redwood DR, et al. Coronary narrowing in coronary heart disease: comparison of cineangiographic and necropsy findings. Ann Int Med 1979;9I: 350-6. Vlodaver Z, Frech R, Van Tassel RA, Edwards IE. Correlation of the antemortem coronary arteriogram and post mortem specimen. Circulation 1973;47:162-9. White R. Indications for fiberoptic angioscopy and intraluminal ukrasound. Compr Ther 1990;16:23-30. White R. Alternative techniques for treating discrete atherosclerotic occlusive lesions: current perspective. Semin Vascul Surg 1989;2:113-6. White GH, White RA, Kopchok GE. Ancillary modalities for endovascular surgery: guidance systems, vascular stents, and methods to prevent restenosis. Semin Vasc Surg 1989;2:17987. Tobis I, MaUery JA, Gessert ], et al. Intravascular ultrasound cross-sectional arterial imaging before and after balloon angioplasty in vitro. Circulation 1989;80: 873-82. Yock PG, Linker DT, White NW, et al. Clinical applications ofintravascular ultrasound imaging in atherectomy. Int J Card Imaging 1989;4:117-25. Artez HT. Intraluminal ultrasound guidance of transverse laser coronary atherectomy. Proceeding of optical fibers in medicine V SPIE 1990;1202:68-78.

Submitted Feb. 7, 1991; accepted Apr. 26, 1991.

DISCUSSION Dr. J. Dennis Baker (Los Angeles, Calif.). This report represents another step in the ongoing program at Harbor/UCLA Medical Center to evaluate IUS imaging in peripheral arteries. Their in vivo studies have demonstrated the accuracy that can be achieved in measurement of diameters as well as cross-sectional areas of both normal and diseased arteries. As shown today, the commercially available system provides good quality images of arterial wall morphologic features. This study focuses on comparing estimates of severity of stenosis obtained from (1) area measurements by ultrasound and (2) area estimates from a single transverse diameter measurement on an angiogram. As would be expected from the asymmetry of atherosclerotic plaques, the derived estimate was higher than the measured area. A reasonable correlation was due to the large proportion of

symmetrical stenosis found in early lesions. One would expect a greater discrepancy between the two techniques when including more advanced disease. In such cases uniplanar angiography has long been recognized as inadequate. Errors can certainly be reduced with multiple projections. To obtain a really fair comparison of the two techniques, it is necessary to use each at its maximum. Do you have any cases comparing biplanar or multiplanar angiographic measurements with IUS imaging? It is clear that ultrasound images can provide a detailed view of plaque morphology, unobtainable with angiograplay or even angioscopy. Have you made any clinical use of morphology information? Several of the imaging devices have received Food and Drug Administration approval and are being marketed. Therefore, it is very important to determine the clinical

Volume 14 Number 4 October 1991

relevance, if any, of this new technology. Is IUS simply an eloquent clinical research tool or should we be considering introducing it into our practices? It is important to keep in mind the costs that are going to be involved. At least one of the units is in the $100,000 price range. Beyond the initial equipment purchase, each catheter costs $600. Therefore one must estimate a cost per patient examination of $800 to $1,000. Do you consider that this technology has a clearly indicated clinical use in any of the following situations: (1) Diagnostic workup before surgical reconstruction; (2) diagnostic workup before a percutaneous procedure; (3) as an adjunct to balloon dilation; (4) as an adjunct to other endovascular procedures; (5) follow-up evaluation? Dr. Marwan Tabbara. Thank you Dr. Baker, for your insightful comments. In this paper, the luminal crosssectional areas of arterial segments were calculated by use of only uniplanar arteriograms. Biplanar arteriograms were not used because they are not routinely done in our institution unless specifically requested by the surgeon. I agree that using biplanar angiography in calculating luminal cross-sectional area would be a more accurate reflection of the measured area that can be obtained by IUS. Intraluminal ultrasonography has not been routinely used for diagnostic workup before vascular surgical reconstructions. In some instances, it can be used as an adjuvant test to arteriography, for example, a case of an infrainguinal revascularization where there was a question about the hemodynamic significance of an lilac lesion. During operation, a 5F catheter was passed through the common femoral arteriotomy in a retrograde manner, and the lesion was imaged. Exact percent stenosis was measured on line and was found to be 77%. An inflow procedure was done in conjunction with the infrainguinal bypass. We have also used IUS as an adjunct to arteriography in diagnosing an acute dissection of the thoracic and abdominal aorta. It showed precisely the proximal and distal extent of the dissection and clearly defined the true and false lumens. The most significant impact of IUS will be in the field of endovascular surgery. It will allow accurate selection of the appropriate sized balloon for transluminal dilation. Moreover, continuous monitoring of the procedure can be done. The end result of the intervention can be objectively and accurately assessed by measuring the resultant luminal cross-sectional area and the extent of intimal dissections. IUS has been coupled to the Simpson atherectomy device tO define the exact location of the atherosclerotic plaque and possibly prevent injury of the non diseased vessel wall. The amount of residual plaque can be accurately assessed and the end point of the procedure precisely defined. The quality of the IUS image depends on the power and resolution of the ultrasound beam. Vessels with relatively large luminal diameters ( > 5 to 8 mm) should be imaged with the 8 F, 20 M H z catheter. It allows deeper penetration of the ultrasound beam and full imaging of the vessel wall. Smaller sized vessels (diameters <5 mm) should be imaged with the 5F, 30 M H z catheter because

Intravascular ultrasonography compared to angiography 503

depth of penetration is not important and resolution is better with the higher frequency probes. The luminal shape of a vessel can be accurately imaged with the appropriate catheter even if calcification is present in the atherosclerotic plaque. The blood and infima (normal or diseased) have very different acoustic impedances, therefore the interface will appear as a hyperechoic line. However, calcification will attenuate the passage of the ultrasound beam through the full thickness of the vessel wall making the assessment of the size and morphology of the calcified plaque very difficult. We have not seen any iatrogenic injuries as a result of the passage of the IUS catheters. The tips are covered with a soft plastic material, and they are passed over guide wires that keep them relatively concentric in the vessel lumen. The information obtained during an endovascular procedure from IUS, in our opinion, far outweighs the risk of injury to the vessel wall. Intimal dissections or inadequate plaque removal at the end of an endovascular procedure can be easily imaged. Appropriate action to correct any of these problems would probably lead to a better overall result. Dr. Cornelius Olcott IV, (Palo Alto, Calif.). As you know, we share our institution with John Simpson and his cardiology colleagues. One of the things that we have learned a little bit about is IUS, and I have been impressed with it. We have started a study comparing angioscopy, arteriography, and IUS. It appears that when you use these three devices for any type of endovascular procedure, and that might well include endarterectomy as well, that IUS is much better. You can actually see the planes better. When you compare it to angioscopy, what you are actually able to tell about the wall of the artery and the depth of atherectomy is much better with IUS. Dr. George Andros (Burbank, Calif.). I am surprised to learn that intravascular ultrasonography is so costly. If this procedure helps to reduce restenosis after endovascular procedures it is easier to rationalize its expense and use. Percutaneous transluminal angioplasty is the one form of infrarenal revascularization that, according to our hospital administration, makes money under DRGs. The D R G reimbursement is more than twice the costs - t h e costs, not charges. With that sort of profit margin, additional steps to reduce restenosis may be justified. Unfortunately, it is still too early to know whether endovascular ultrasonography will give us the information we need to prevent restenosisassuming we knew what to do with that information. I have two questions regarding choice of catheter size. Will you get the same sort of information if you put a 5F device into a 6 or 7 mm diameter vessel that will accommodate the larger 8F probe? Would you need one size probe for the suprainfrageniculate femoral and popliteal and another for the infrapopliteal vessels? Dr. Tabbara. No, a definite difference exists in the quality of the pictures or the images that we get depending on the catheter. Usually the bigger the catheter, the more powerful, you can get more ukrasound energy out of it and

504

Tabbara et aL

you can image it further so the depth of the imaging is higher. So, you would better use a bigger catheter as you go for bigger arteries; obviously, the frequency is also another factor that will affect the resolution or the quality of the images. Usually the bigger catheters have lower resolution so it is advisable if you want to go for big arteries like the aorta or the iliacs that you use an 8F catheter that has a 20 MHz frequency. If you go for the smaller ones, it is better offto use the smaller catheters just for the size, and a 30 MHz catheter, will give you better resolution. Dr. Linda M. Reilly (San Francisco, Calif.). Since one of the principal advantages of this technique is its ability to accurately assess the size of the vessel, in what percentage of these studies are you unable to assess size of the vessel because of extensive calcification in the lesion, and what in percentage do you end up really estimating the size by mentally drawing a line as to where the outside wall of the vessel really is? Dr. Tabbara. When it comes to luminal dimensions, you can do that every single time because the calcification would actually enhance your interface between the blood and the first layer. So you always see that interface first. However, it will completely eliminate any kind of evaluation of the plaque itself because of elimination of the signals. But in every case, if you can cross the lesion you will be able to image it with an adequate catheter. Dr. Sam S. Ahn (Los Angeles, Calif.). We have had a limited experience at UCLA with ultrasonography, and one concern that I have with all these different devices now is the potential injury to the intima by the passage of these multiple devices. I do not kuow how significant that is going to be. Right now when an atherectomy catheter or a balloon causes some injury and then you put an angioscope down the vessel and then you put an ultrasound catheter down, you multiply the potential for injury. The question to you is have you seen any injury created by the ultrasonic catheter in your experience? Dr. Tabbara. We have not really seen any injuries, but I think it is like any other endovascular device. Actually, the injury done will be much less than any interventional device because it is a diagnostic device-in comparison to laser angioplasty or atherectomy- or the injury has already been produced by that device, and the ultrasonography probably will be more of benefit because it shows you the injury, intimal flaps, or whatever you left there, or what you have

Journal of VASCULAR SURGERY

not done. I think that outweighs the disadvantage of the possible injury, which is definitely there, but we have not really seen it. Dr. Eugene F. Bernstein (La Jolla, Calif.). I think this is a very powerful technology, and we will all learn to use it within the next couple of years. There are at least 10 groups developing IUS equipment in the United States, the United Kingdom, and Europe at the moment, and it is clearly on the horizon. In comparing this method and angioscopy, it seems to me there are two advantages for IUS. One is that you do not have to deal with blood washout, so it is amenable for percutaneous procedures. The second is that it gives you information about the depth of the lesion, which you do not get from angioscopy. If you are using these devices to monitor how much you are removing in an atherectomy procedure, IUS would be the better device. In addition, several groups are developing computer technology for a 3-D reconstruction of the vessel, using not only lateral transducers, but also forward-looking transducers. This technique may become the most sensitive of all the methods we have to diagnose intimal and placque ulceration. Dr. Rodney A. White. The cost issue is an extremely pertinent consideration. In my opinion, future interventional catheters will be based on lidS, with other modalities such as angioscopy fiberoptics and therapeutic devices (i.e., laser fibers, atherectomy instruments, or balloons) added as needed, At present, each device alone costs a significant amount ($100,000 or more), but combined systems could be developed more economically. For example, current intravascular ultrasound systems could be modified to also accommodate angioscopy by the addition of a light source without requiring duplication of the entire imaging system. Additional modalities could be added at limited expense if the delivery catheters are developed to handle several devices simultaneously in a one-pass sole-therapy catheter. Although the cost of hardware for such a device would be significant, combining technologies to produce more effective therapy would justify the investment. Intravascular ultrasound technology also favorably affects the cost issue by providing a control for angioplasty data, enabling, for the first time, accurate assessment of the indications for and results of interventions. The technology promises to enhance the efficacy of current devices by improving the cost-effectiveness of developing endovascular techniques.