Intravascular ultrasonography: Validation studies and preliminary intraoperativ¢ observations Richard F. Neville, M D , R o b e r t W. H o b s o n , II, M D , Zafar Jamil, M D , Gary B. Breitbart, M D , R o b e r t J. Anderson, M D , A n t o n i o L. Bartorelli, M D , Martin B. Leon, M D , Newark, N.J. and Bethesda, A/Id. Intravascular ultrasonography is emerging as an important imaging modality to assess the presence, distribution, and extent of atherosclerotic vascular disease. To determine the accuracy and clinical utility of intravascular ultrasonography, a flexible catheter-based system was used to generate two-dimensional, cross-sectional vas~lar images. In 23 arteries of 11 sheep 206 in vivo images demonstrated an echo-free lumen surrounded by three distinct concentric acoustic transitions corresponding to intima, media, and adventitia. Ultrasound measurements of lumen diameter and area correlated significantly with those of corresponding arteriographic measurements obtained by use of digital calipers (r = 0.91, r = 0.86). To evaluate clinical feasibility, intraoperative images (n = 160) were obtained in 10 patients undergoing vascular bypass or hemodialysis access procedures. The images depicted luminal configuration and arterial wall morphologic characteristics. Measurements of lumen diameter and lumen area correlated closely with corresponding intraoperative arteriography (r = 0.81, r = 0.79). The ultrasound images demonstrated arterial stenoses, intimal hyperplasia, intraluminal thrombus, polytetrafluoroethylene graft material, and anastomotic sites. We conclude that flexible catheterbased ultrasonography produces images that accurately demonstrate arterial wall architecture, lumen diameter, and area. Intraoperative application can produce images that provide unique information thus expanding the clinical potential of ultrasonography as a guidance system for vascular procedures. (J VASC SURG 1991;13:274-83.)
Intravascular ultrasonography is an imaging modality able to delineate arterial wall architecture, characterize atherosclerotic plaque composition, and precisely measure lumen diameter and area. The ability to generate this unique information provides the potential to assess complex vascular bypass procedures and endarterectomy surfaces. As a real-time guidance system, intravascular ultrasonography may also prove useful in the deployment of endovascular therapeutic devices such as intraarterial stents, atherectomy catheters, and laser angioplasty. In this capacity intra-
From the Section of Vascular Surgery, (Drs. Neville, Hobson, Jamil, Breitbart, and Anderson), Department of SurgeqT,Universityof Medicineand Dentistry- New JerseyMedicalSchool, Newark, N.J., and the National Heart, Lung, and Blood Institute, NationalInstitutesof Health (Drs. Bartorelliand Leon), Bethesda, Md. Presented at the Thirty-eighth ScientificMeeting of the North American Chapter, International Sociew for Cardiovascular Surgery, Los Angeles, Calif.,June 4-6, 1990. Reprint requests: RichardF. Neville,MD, Universiwof Medicine and Dentistry - New JerseyMedicalSchool,MSB G-532, 185 South Orange Ave., Newark, NJ 07103. 24/6/26369 274
vascular ultrasonography could guide appropriate management by defining plaque composition and morphology as well as provide a preprocedure and postprocedure assessment of the arterial wall and lumen. Preliminary studies using prototype probes have found ultrasound images accuratebr determine plaque composition, arterial wall architecture, lumen diameter, and lumen area. >6 The evolution of ultrasound technology from rigid, metallic probes to a flexible catheter-based system has improved the safety and expanded potential applications of this intraluminal imaging modality. Our study was designed to evaluate the accuracy o f catheter-based intravascular ultrasonography and d e t e r ~ e q t s ability to generate information concerning vascular architecture and luminal configuration during a clinical trial in the operating room. MATERIAL AND METHODS The system. Intravascular ultrasound images were obtained using a 6F, flexible catheter 20 cm in length (Intertherapy, Inc., Costa Mesa, Calif.). An,
Volume t3 Number 2 February 1991
Intravascular u#rasonography 275
ultrasoun~ Beam
,%.
~
Transducer
/
/-TORQUE LUMEN FLUSH PORT
/
/
/r
/ /
Image Plane
TORQUE KNOB
/ /
I
\
\ \
,.
\
~TOUHY-BORST " -ACOUSTIC
SUBASSEMBLY
Fig. 1. Schematic diagram of the intravascular ultrasound catheter. The transducer at the catheter's tip (see inset) emitted the ultrasound beam, which was redirected by the distal mirror and reflected at a 90 degree angle to the catheter's longitudinal axis to insonate the arterial wall.
acoustic subassembly at the catheter tip contained the ultrasound transducer, a piezoelectric crystal with a 1 mm outside diameter and 20 M H z center freq u e n t . By means of pulsed-echo techniques, the ultrasound beam emitted from the transducer was reflected perpendicular to the catheter's longitudinal axis by a distal mirror set at a 45 degree angle (Fig. 1). The signal returning from the tissue was reflected back to the transducer and transmitted by a flexible torque shaft to an 8 bit analog to digital (A/D) converter. The A / D converter transformed the radial ultrasound data into rectangular format for computer processing and viewing on a 640 x 480 pixel grey scale video monitor. Circumferential rotation of the torque shaft inside the catheter produced cross-sectional images of the vessel wall. Positional orientation of the catheter was defined by two potentiometers attached to a torque knob at the proximal end of the torque shaft. Heparinizcd saline (5000 units per liter of normal saline) was infused through the catheter lumen to provide lubrication for the torque shaft, purge air from the system, and prevent blood from entering the catheter lumen where it could impair the movement of the torque shaft. Images were displayed on the monitor and recorded in digital format on floppy disks for subsequent analysis. Animal in vivo studies. Catheter-based intravascular ultrasonography was used to study 23 normal arteries (10 femoral, 13 carotid) in 11 sheep (58 + 4 kg). The animals were cared for in accordance with the "Principles of Laboratory Animal
Care" and the "Guide for the Care and Use of Laboratory Animals" (NIH Publication 80-23). After administration of intramuscular thiopental sodium, each animal was intubated and anesthetized with halothane. Each artery was exposed surgically, and the ultrasound catheter was inserted through a 10F introducer sheath. Manual rotation of the catheter produced images at various sites with fluoroscopic localization of the catheter tip. In 14 arteries (11 carotid, 3 femoral) arteriograms were obtained before ultrasound imaging. Radiopaque markers were used to indicate the specific imaging location for the comparison of arteriographic and ultrasound images from the same arterial segment. The catheter was then fluoroscopically guided to the marked arterial segment for ultrasound imaging. A postprocedure arteriogram was obtained in each case. Intraoperative studies. Ten patients (6 men, 4 women) were studied in the operating room during vascular surgical procedures. Eleven procedures were performed in the 10 patients including aortobifemoral bypass (2), iliac artery angioplasty (1), superficial femoral artery laser-assisted angioplasty (1), Cimino shtmt revision (2), brachial artery-axillary vein shunt construction (4), and excision of a brachial artery pseudoaneurysm (1). In each case appropriate vascular control was obtained, and the ultrasound catheter was introduced through an arteriotomy necessary for the vascular procedure being performed. Vessel loops were used to maintain hemostasis while the ultrasound catheter was in the vessel. During angio-
276
Journal of VASCULAR SURGERY
Neville et al.
Fig. 2. An in vivo ultrasound image of a sheep carotid artery generated by the flexible catheter. The cross-sectional image demonstrates three concentric layers around an echo-free lumen; intima (I), media (M), and adventitia (A).
Fig. 3. This intraoperative image of a radial artery obtained during construction of a Cimino shunt delineates the echolucent lumen, intima (I), media (M), and adventitia (A).
plasty the catheter was inserted through the standard guide sheath. Manual rotation of the torque shaft in the catheter produced intraoperative ultrasound images. Images were obtained from the radial artery, brachial artery, axillary artery, iliac artery, superficial femoral artery, axillary vein, antecubital veins, polytetrafluoroethylene (PTFE) graft, and Dacron graft material. Lo-
calization of the catheter tip was possible by direct palpation in most cases because of the surgical exposure necessary for the operative procedure. When palpation was not possible, the length of catheter inserted into the vessel was measured to localize the tip. In the laser-assisted angioplasty, an intraoperative radiograph was obtained to delineate catheter position. In five procedures preoperative arteriography or intraoperative completion arteriography was performed. Quantitative measurements. Lumen diameter and cross-sectional area were measured for those ul;.. trasound images with corresponding arteriogramsl The ultrasound measurements were performed by use of computer video planimetry. Arteriographic lumen diameter of each corresponding arterial segment was measured with standard caliper techniques based on comparison with a radiopaque object of known size in the field. Arteriographic cross-sectional area was then calculated from the measured diameter (area = *rr2). Arteriographic and ultrasound measurements of corresponding arterial segments were compared by means of Student's t test analysis of paired data. Analysis of variance (ANOVA) was used to test for significant differences among sample means and variances. A p value less than 0.05 was considered a statistically significant difference. Correlations of paired data were determined for two variables by use of Pearson's correlation coefficient. The regression lines were compared to the line of identity (slope =.
Volume 13 Number 2 February 1991
Intravascular ultrasonography 277
Fig. 4. Intraoperative images of an axillary vein (A) and PTFE hemoaccess graft (B) both depict luminal configuration and the lumen-intima/graft interface.
Fig. 5. Intraoperative images of a superficial femoral artery, (A) and iliac arteU (B) demonstrate the luminal irregularity of circumference (arrows) consistent with these stenotic arteries. Intraluminal echo-reflections indicative of calcium are noted in the superficial femoral artery. 1, y intercept = 0) for each correlation to test for the level of significance (p < 0.05). RESULTS Animal in vivo images. Using the flexible ultrasound catheter, 206 images were generated from 10 normal femoral arteries (85 images) and 13 normal carotid arteries (121 images). These in vivo images demonstrated three distinct concentric layers around a central echo-free lumen; a highly reflective inner layer, an echo-free middle layer, and an echodense outer layer. Corresponding to the intima, media, and adventitia of the vessel wall, these layers clearly delineated the acoustic transitions at the lumcn-intima, intima-media, and media-adventitia interfaces (Fig. 2). Intraoperative images. One hundred sixty images were obtained in the operating room from arterial segments (n = 97), venous segments (n = 18), graft material (n = 34), and anastomotic sites (n = 11). Images were generated and recorded as
two-dimensional, 360 degree cross-sectional representations of the arterial wall, venous wall, or graft. One hundred percent of the images demonstrated the vascular luminal configuration by clearly defining the lumen-intima interface. Sixty-seven percent (65 / 97) of arterial images allowed delineation of the arterial wall architecture as echo-bright intima, echolucent media, and echo-dense adventitia (Fig. 3). The remaining arterial images did not provide resolution capable of differentiating media from adventitia but did demonstrate luminal configuration. The venous and PTFE images also depicted the lumcn-intima interface (Fig. 4). The intraoperative study provided an analysis of images obtained in a variety of clinical settings. Arterial stenoses examined in diseased superficial femoral and external iliac arteries displayed a reduction in lumen area and an alteration of the circular crosssectional appearance of the normal arterial image (Fig. 5). Intimal hyperplasia was noted in radial artery segments at the anastomotic sites ofhemodialysis
Journal of VASCULAR
278 Neville et al.
SURGERY
Fig. 6. A hyperreflective lumen-intima interface consistent with a hyperplastic intimal response is noted in the venous portion of the anastomotic site of a failed Cimino shunt.
Fig. 7. Intraoperative image of superficial femoral artery after laser-assisted balloon angioplasty. A central lumen (12) the size of the laser probe (2.5 mrn) is surrounded by a thickened layer of echo-reflective material before the arterial media is seen (M).
12"-~
-
o ANIMAL STUDY
m
E lo lad t~-,I
8
r = o.91 ---
/ o ~ n o /
p
Intravascular ultrasonography is an imaging modality able to delineate arterial wall architecture, characterize atherosclerotic plaque composition, and precisely measure lumen diameter and area. The ability to generate this unique information provides the potential to assess complex vascular bypass procedures and endarterectomy surfaces. As a real-time guidance system, intravascular ultrasonography may also prove useful in the deployment of endovascular therapeutic devices such as intraarterial stents, atherectomy catheters, and laser angioplasty. In this capacity intra-
From the Section of Vascular Surgery, (Drs. Neville, Hobson, Jamil, Breitbart, and Anderson), Department of SurgeqT,Universityof Medicineand Dentistry- New JerseyMedicalSchool, Newark, N.J., and the National Heart, Lung, and Blood Institute, NationalInstitutesof Health (Drs. Bartorelliand Leon), Bethesda, Md. Presented at the Thirty-eighth ScientificMeeting of the North American Chapter, International Sociew for Cardiovascular Surgery, Los Angeles, Calif.,June 4-6, 1990. Reprint requests: RichardF. Neville,MD, Universiwof Medicine and Dentistry - New JerseyMedicalSchool,MSB G-532, 185 South Orange Ave., Newark, NJ 07103. 24/6/26369 274
vascular ultrasonography could guide appropriate management by defining plaque composition and morphology as well as provide a preprocedure and postprocedure assessment of the arterial wall and lumen. Preliminary studies using prototype probes have found ultrasound images accuratebr determine plaque composition, arterial wall architecture, lumen diameter, and lumen area. >6 The evolution of ultrasound technology from rigid, metallic probes to a flexible catheter-based system has improved the safety and expanded potential applications of this intraluminal imaging modality. Our study was designed to evaluate the accuracy o f catheter-based intravascular ultrasonography and d e t e r ~ e q t s ability to generate information concerning vascular architecture and luminal configuration during a clinical trial in the operating room. MATERIAL AND METHODS The system. Intravascular ultrasound images were obtained using a 6F, flexible catheter 20 cm in length (Intertherapy, Inc., Costa Mesa, Calif.). An,
Volume t3 Number 2 February 1991
Intravascular u#rasonography 275
ultrasoun~ Beam
,%.
~
Transducer
/
/-TORQUE LUMEN FLUSH PORT
/
/
/r
/ /
Image Plane
TORQUE KNOB
/ /
I
\
\ \
,.
\
~TOUHY-BORST " -ACOUSTIC
SUBASSEMBLY
Fig. 1. Schematic diagram of the intravascular ultrasound catheter. The transducer at the catheter's tip (see inset) emitted the ultrasound beam, which was redirected by the distal mirror and reflected at a 90 degree angle to the catheter's longitudinal axis to insonate the arterial wall.
acoustic subassembly at the catheter tip contained the ultrasound transducer, a piezoelectric crystal with a 1 mm outside diameter and 20 M H z center freq u e n t . By means of pulsed-echo techniques, the ultrasound beam emitted from the transducer was reflected perpendicular to the catheter's longitudinal axis by a distal mirror set at a 45 degree angle (Fig. 1). The signal returning from the tissue was reflected back to the transducer and transmitted by a flexible torque shaft to an 8 bit analog to digital (A/D) converter. The A / D converter transformed the radial ultrasound data into rectangular format for computer processing and viewing on a 640 x 480 pixel grey scale video monitor. Circumferential rotation of the torque shaft inside the catheter produced cross-sectional images of the vessel wall. Positional orientation of the catheter was defined by two potentiometers attached to a torque knob at the proximal end of the torque shaft. Heparinizcd saline (5000 units per liter of normal saline) was infused through the catheter lumen to provide lubrication for the torque shaft, purge air from the system, and prevent blood from entering the catheter lumen where it could impair the movement of the torque shaft. Images were displayed on the monitor and recorded in digital format on floppy disks for subsequent analysis. Animal in vivo studies. Catheter-based intravascular ultrasonography was used to study 23 normal arteries (10 femoral, 13 carotid) in 11 sheep (58 + 4 kg). The animals were cared for in accordance with the "Principles of Laboratory Animal
Care" and the "Guide for the Care and Use of Laboratory Animals" (NIH Publication 80-23). After administration of intramuscular thiopental sodium, each animal was intubated and anesthetized with halothane. Each artery was exposed surgically, and the ultrasound catheter was inserted through a 10F introducer sheath. Manual rotation of the catheter produced images at various sites with fluoroscopic localization of the catheter tip. In 14 arteries (11 carotid, 3 femoral) arteriograms were obtained before ultrasound imaging. Radiopaque markers were used to indicate the specific imaging location for the comparison of arteriographic and ultrasound images from the same arterial segment. The catheter was then fluoroscopically guided to the marked arterial segment for ultrasound imaging. A postprocedure arteriogram was obtained in each case. Intraoperative studies. Ten patients (6 men, 4 women) were studied in the operating room during vascular surgical procedures. Eleven procedures were performed in the 10 patients including aortobifemoral bypass (2), iliac artery angioplasty (1), superficial femoral artery laser-assisted angioplasty (1), Cimino shtmt revision (2), brachial artery-axillary vein shunt construction (4), and excision of a brachial artery pseudoaneurysm (1). In each case appropriate vascular control was obtained, and the ultrasound catheter was introduced through an arteriotomy necessary for the vascular procedure being performed. Vessel loops were used to maintain hemostasis while the ultrasound catheter was in the vessel. During angio-
276
Journal of VASCULAR SURGERY
Neville et al.
Fig. 2. An in vivo ultrasound image of a sheep carotid artery generated by the flexible catheter. The cross-sectional image demonstrates three concentric layers around an echo-free lumen; intima (I), media (M), and adventitia (A).
Fig. 3. This intraoperative image of a radial artery obtained during construction of a Cimino shunt delineates the echolucent lumen, intima (I), media (M), and adventitia (A).
plasty the catheter was inserted through the standard guide sheath. Manual rotation of the torque shaft in the catheter produced intraoperative ultrasound images. Images were obtained from the radial artery, brachial artery, axillary artery, iliac artery, superficial femoral artery, axillary vein, antecubital veins, polytetrafluoroethylene (PTFE) graft, and Dacron graft material. Lo-
calization of the catheter tip was possible by direct palpation in most cases because of the surgical exposure necessary for the operative procedure. When palpation was not possible, the length of catheter inserted into the vessel was measured to localize the tip. In the laser-assisted angioplasty, an intraoperative radiograph was obtained to delineate catheter position. In five procedures preoperative arteriography or intraoperative completion arteriography was performed. Quantitative measurements. Lumen diameter and cross-sectional area were measured for those ul;.. trasound images with corresponding arteriogramsl The ultrasound measurements were performed by use of computer video planimetry. Arteriographic lumen diameter of each corresponding arterial segment was measured with standard caliper techniques based on comparison with a radiopaque object of known size in the field. Arteriographic cross-sectional area was then calculated from the measured diameter (area = *rr2). Arteriographic and ultrasound measurements of corresponding arterial segments were compared by means of Student's t test analysis of paired data. Analysis of variance (ANOVA) was used to test for significant differences among sample means and variances. A p value less than 0.05 was considered a statistically significant difference. Correlations of paired data were determined for two variables by use of Pearson's correlation coefficient. The regression lines were compared to the line of identity (slope =.
Volume 13 Number 2 February 1991
Intravascular ultrasonography 277
Fig. 4. Intraoperative images of an axillary vein (A) and PTFE hemoaccess graft (B) both depict luminal configuration and the lumen-intima/graft interface.
Fig. 5. Intraoperative images of a superficial femoral artery, (A) and iliac arteU (B) demonstrate the luminal irregularity of circumference (arrows) consistent with these stenotic arteries. Intraluminal echo-reflections indicative of calcium are noted in the superficial femoral artery. 1, y intercept = 0) for each correlation to test for the level of significance (p < 0.05). RESULTS Animal in vivo images. Using the flexible ultrasound catheter, 206 images were generated from 10 normal femoral arteries (85 images) and 13 normal carotid arteries (121 images). These in vivo images demonstrated three distinct concentric layers around a central echo-free lumen; a highly reflective inner layer, an echo-free middle layer, and an echodense outer layer. Corresponding to the intima, media, and adventitia of the vessel wall, these layers clearly delineated the acoustic transitions at the lumcn-intima, intima-media, and media-adventitia interfaces (Fig. 2). Intraoperative images. One hundred sixty images were obtained in the operating room from arterial segments (n = 97), venous segments (n = 18), graft material (n = 34), and anastomotic sites (n = 11). Images were generated and recorded as
two-dimensional, 360 degree cross-sectional representations of the arterial wall, venous wall, or graft. One hundred percent of the images demonstrated the vascular luminal configuration by clearly defining the lumen-intima interface. Sixty-seven percent (65 / 97) of arterial images allowed delineation of the arterial wall architecture as echo-bright intima, echolucent media, and echo-dense adventitia (Fig. 3). The remaining arterial images did not provide resolution capable of differentiating media from adventitia but did demonstrate luminal configuration. The venous and PTFE images also depicted the lumcn-intima interface (Fig. 4). The intraoperative study provided an analysis of images obtained in a variety of clinical settings. Arterial stenoses examined in diseased superficial femoral and external iliac arteries displayed a reduction in lumen area and an alteration of the circular crosssectional appearance of the normal arterial image (Fig. 5). Intimal hyperplasia was noted in radial artery segments at the anastomotic sites ofhemodialysis
Journal of VASCULAR
278 Neville et al.
SURGERY
Fig. 6. A hyperreflective lumen-intima interface consistent with a hyperplastic intimal response is noted in the venous portion of the anastomotic site of a failed Cimino shunt.
Fig. 7. Intraoperative image of superficial femoral artery after laser-assisted balloon angioplasty. A central lumen (12) the size of the laser probe (2.5 mrn) is surrounded by a thickened layer of echo-reflective material before the arterial media is seen (M).
12"-~
-
o ANIMAL STUDY
m
E lo lad t~-,I
8
r = o.91 ---
/ o ~ n o /
p
