Mesenteric flow velocity variations function of angle of insonation

Transcription

1 Mesenteric flow velocity variations function of angle of insonation as a Robert J. Rizzo, MD, ~ GaB Sandager, RN, RVT, 2 Patricia Astleford, BSN, RVT, a Kathleen Payne, RN, RVT, BS, 3 Linda Peterson-Kennedy, RN, RVT, 3 William R. Flinn, MD, 1 and James S.T. Yao, MD, PhD, 1 Chicago, Ill. This study was designed to quantitate variations in duplex ultrasound arterial flow velocities (cm/sec) in the common carotid artery and the superior mesenteric artery that were produced by changes in the angle of pulsed Doppler insonation. Duplex scanning was used to measure peak systolic flow velocity and mean velocity at angles from 30 degrees to 80 degrees; individual measurements were made at 10-degree increments in both the common carotid artery and the superior mesenteric artery in normal subjects. Peak systolic velocity in the common carotid artery varied from 86 cm/sec at 30 degrees to 168 m/see at 80 degrees. Over the same transducer angle variation mean velocity ranged from 28 to 53 cm/sec. Similar changes in the superior mesenteric artery flow velocities were observed by varying the angle of insonation, where peak systolic velocity varied from 108 cm/sec (30 degrees) to 280 cm/sec (80 degrees), and mean velocity ranged from 29 cm/sec (30 degrees) to 71 cm/sec (80 degrees). Measurements taken from 70 to 80 degrees produced the most dramatic deviation from those taken at 60 degrees. In the common carotid artery the 70- and 80-degree angles produced 14% and 59% increases, respectively, in peak systolic velocity and 16% and 63% increases, respectively, in mean velocity. In the superior mesenteric artery 70-degree and 80-degree angles produced 16% and 120% increases, respectively, in peak systolic velocity and 17% and 111% increases, respectively, in mean velocity. At 80 degrees the percent increases in measured flow velocities for the superior mesenteric artery were significantly greater than those for the common carotid artery. The noninvasive investigation of normal mesenteric vessels may be dependent on accurate absolute flow velocity values, especially if these values are used to calculate volume flow. This study demonstrated that significant variations in resting flow velocities may be introduced by variations in the angle of pulsed Doppler insonation, particularly at angles above 70 degrees. Strict control of this variable will be essential in future studies, mainly those addressing normal visceral flow variations in response to physiologic stimuli. (J VAsc SVRG 1990;11: ) Duplex ultrasound scanning has been used routinely for the diagnosis of extracranial carotid artery occlusive disease. More recently it has been recognized that this technique could be used for the assessment of peripheral and visceral vessels including the celiac artery and superior mesenteric artery (SMA). Duplex scanning, the combination of realtime B-mode ultrasound imaging and pulsed Doppler spectral analysis, allows measurement of blood From the Department of Surgery, Division of Vascular Surgery, Northwestern University Medical School, ~ the Vascular Laboratory, Columbus Hospital, 2 and the Blood Flow Laboratory, Northwestern Memorial Hospital? Presented at the Thirteenth Annual Meeting of the Midwestern Vascular Surgical Society, Chicago, IlL, Sept , Reprint requests: William R. Flirm, MD, Associate Professor of Surgery, Division of Vascular Surgery, 251 East Chicago Ave., Suite 628, Chicago, IL /6/ flow velocities in the vessels examined, and the measurement of splanchnic blood flow has been a topic of keen interest to physicians in a variety of specialties including gastroenterology, gastrointestinal surgery, and vascular surgery. Cases of mesenteric arterial occlusive disease are rare, and thus investigative evaluation of the technique of mesenteric duplex scanning has largely involved measurement of arterial flow velocities in norreal volunteers and visceral flow responses to a variety of stimuli. These studies have confirmed the utility and applicability of the technique of mesenteric duplex scanning, and normal physiologic responses have been well characterized. Nevertheless, although there have been strict flow velocity criteria for the assessment of carotid artery disease, current studies have reported an unusually wide range of normal mesenteric flow velocities. Fasting S/vIA peak systolic flow velocities (PSV) have ranged from 103 to 196

2 Volume 11 Number 5 May 1990 esenteric flow velocity variations 689 cm/sec l~s (Table I). Reports of normal fasting SMA velocities have varied up to 90% even from the same institution. 3's It is clear that hemodynamically significant mesenteric arterial occlusive disease would be expected to produce flow velocity changes well in excess of this range and thus be detectable. Nevertheless, these discrepancies are disturbing especially if measured flow velocities are used to calculate volume flow as has been done in some studies. 6,7 It is recognized that blood flow measurement by Doppler frequency spectral analysis is affected by the angle between the ultrasound beam and the axis of flow in the vessel examined. The present study was designed to assess the magnitude of change in the measured arterial flow velocities in the SMA and common carotid artery (CCA) in normal subjects by varying the angle of Doppler examination. The CCA was chosen as a control in this study because of its predictable linear anatomy, which made precise measurement of the angle of insonation much more certain. It was also felt that comparative measurements in two separate arterial systems would help clarify whether observations were a function solely of the variations in examination technique. MATERIAL AND METHODS Twenty healthy volunteers free of symptoms and with no evidence of atherosclerotic occlusive disease underwent CCA duplex scanning, and 10 of these same subjects had serial SMA scans. All duplex scans in this study were conducted with commercially available duplex units (Advanced Technology Laboratories, Bothel, Wash.). All carotid scans were performed with the patient at rest in the supine position with a 7.5 MHz mechanical sector scanner for B-mode imaging and a 5 MHz pulsed Doppler with a 1.5 mm midstream sample volume for flow velocity measurements. After longitudinal imaging of the CCA approximately 4 cm below the carotid bifurcation, the sample volume flow cursor was adjusted parallel to the long axis of the vessel. The M-line cursor (Fig. 1) was then used to sequentially vary the angle of Doppler insonation. Individual PSV measurements were obtained at each 10-degree interval from 30 degrees to 80 degrees from the longitudinal vessel axis. The mean a~erial flow velocity (MV) at each interval was subsequently calculated by integration of the velocity waveform recorded. At least two velocity waveforms were obtained at each angle of insonation on every CCA examined, which resulted in a total of 40 complete sets of varying angle measurements. Results are expressed at each interval as mean _+ SD for each velocity parameter measured. Table I. Previously reported peak systolic velocities (PSV) for the superior mesenteric artery Author PSV (cm/sec) jager et al Flinn et al. s Lilly et al Nichols et al _ 68.8 Moneta et al _+ 3.9 Superior mesenteric artery scanning was performed with the patient in the supine position at rest after an overnight fast. Sagittal imaging of the abdominal aorta was performed by use of a 3 MHz sector scanner to identify the ventral origin of the SMA from the aorta (Fig. 2). The sample volume flow cursor was similarly placed parallel to the long axis of the SMA. An attempt was made to place the sample volume several centimeters distally on the vessel where its course was more linear. Pulsed Doppler flow velocity measurements were made by means of a 3 MHz probe, insonation angle was varied at the same 10-degree increments as for CCA sampling, and PSV was measured twice at each angle, and MV was calculated by integration of each velocity waveform. Twenty sets of varying angle measurements from the SMA were available for analysis. Since in the past, performance of Doppler examinations at 60 degrees has been recommended, the percentage changes in PSV and MV for each individual angle increment were compared to velocities obtained at 60 degrees in both the CCA and SMA. An analysis of variance that controlled for variation within subjects and between subjects was used for statistical analysis of these percentage differences. RESULTS The resting PSV and MV obtained from the SMA and CCA at angles varying from 30 degrees to 80 degrees are presented in Table ii. In the SMA, PSV ranged from 108 _ 30 cm/sec at 30 degrees to 280 _ 76 cm/sec at 80 degrees (159% increase), and MV varied from 29 _+ 10 cm/sec at 30 degrees to cm/sec at 80 degrees (145% increase). Across the same 30-degree to 80-degree angle variation in the CCA, PSV ranged from 86 _+ 12 cm/sec to cm/sec (95% in-i crease), and MV ranged from cm/sec to 53 _+ 12 cm/sec (89% increase). For both the SMA) and CCA, the greatest variation in PSV and MV occurred between 70 degrees and 80 degrees. Since 60 degrees is commonly accepted as the optimum angle for Doppler insonation, the PSV and

3 690 Rizzo et al Journa~ of VASCULAR SURGERY Fig. 1. B-mode image of a duplex scan of the common carotid artery illustrates the M-line of the Doppler ultrasound beam, the sample volume (SV) (arrow) with the flow cursor parallel to the blood vessel walls, and the 60-degree angle ofinsonation between the Doppler ultrasound beam and flow cursor. MV measurements obtained at other angles were compared to those obtained at 60 degrees in each respective vessel. In the SMA the percentage change for PSV ranged from - 15% at 30 degrees to 120% at 80 degrees (Fig. 3), and the percentage change for MV in similar fashion ranged from -23% at 30 degrees to 111% at 80 degrees. The SMA velocity changes at 80 degrees for both PSV and MV reached statistical significance atp < In the CCA, the percentage change for PSV ranged from -20% at 30 degrees to 56% at 80 degrees and for MV ranged from -15% at 30 degrees to 61% at 80 degrees. The CCA velocity changes reached statistical significance at 70 degrees (p < 0.01) and 80 degrees (p < ). The percentage changes in SMA velocities relative to the 60-degre e optimum angle tended to be greater than those for the CCA at all angles. However, only at 80 degrees were the angle-dependent velocity changes for the SMA significantly different from those obtained for the CCA, and this was true for comparisons of both the PSV (p < ) and MV (p < ) for these vessels. DISCUSSION The Doppler ultrasonic flow detector has been an invaluable device for evaluating patients with vascular diseases and has been the very foundation of noninvasive vascular testing. The Doppler effect is basically the observation that the frequency of a transmitted (sound) wave will be shifted after reflection off a moving target (in this case blood cells) in proportion to the speed (velocity) of the target. Most simply stated, with the target moving directly toward an ultrasound probe at a (real) velocity Vk, the measured Doppler frequency shift (fd) would be: fd = 2V~/X (1) The factor of 2 arises from the sound waves trip to and from the target, and k is the wavelength of the transmitted ultrasound. Further calculations are simplified by expressing this in terms of the known transmission frequency F of the ultrasound probe used, since X = c/f, where c is the speed of sound in blood and thus: fd = 2vR F/c (2) It is important to remember that the Doppler frequency shift is the basic information extracted from any Doppler examination, and increasingly sophisticated technology has simply developed more useful ways of processing this basic information. Additionally, real examinations are not conducted with blood flowing directly at the probe, but usually transcutaneously with the probe at an angle, 0, to the axis of flow in the vessel examined. The measured Doppler frequency shift then reflects only the effective velocity

4 Volume 11 Number 5 May 1990 Mesenteric flow velocity variations 691 Fig. 2. B-mode image of a duplex scan of the SMA (arrow) illustrates placement of the sample volume within center stream of the SMA at an insonadon angle of 60 degrees. Table II. The effect of angle of insonation on PSV and MV in the SMA and CCA SMA Degrees of angle PSV A/IV PSV CCA A//V rn +_ SD m + SD m ± SD m ± SD _ ± ± _ ±27 30 ~8 96_ ± ± _ ± ± ± All velocity measurements are expressed as cm/sec. vector Ve perceived at that angle, but the true velocity VR can be expressed by the known trigonometric relationship of these two vectors: Vo = Vr. cos 13 (3) Thus for normal conditions of noninvasive Doppler measurement, equation 2 should be expressed: fd = 2Vv. cos 0 F/c (4) It is clear from this expression that the measured Doppler frequency shift is proportional to the velocity of blood flow and directly related to the angle of Doppler insonation. Arterial blood flow is of course complex, and instantaneous velocities vary in different areas of the arterial lumen and at different times in the cardiac cycle, and thus frequency analysis must be dynamic. The microprocessors on most modern duplex scanners use digital techniques based on fast Fourier transform for instantaneous output of a frequency spectral waveform descriptive of Doppler frequency shifts from the sample volume throughout the cardiac cycle. Most duplex units also have the microprocessor capability to solve equation 4 to output blood flow velocity: V~ = fd c/2f cos t3 (5) It is now evident that blood flow velocities calculated from the measured Doppler frequency shift are proportional to the inverse of the cos 0. The accuracy of these calculated velocities is dependent on an accurate knowledge of the angle of Doppler insonation, especially as 0 approaches 90 degrees where cos e approaches zero. This relationship alone may largely explain the findings in this study, but there may also be aspects of the conduct of deep Doppler visceral examinations that account for these and pre-

5 692 Rizzo et al.,~ournai of VASCULAR SURGERY viously reported discrepancies in measured mesenteric flow velocities. One of the most serious limitations of early transcutaneous Doppler arterial examinations was the fact that the angle ofinsonation could only be determined indirectly. One could assume that the artery ran parallel to the skin surface and control the probe: skin angle. Another technique involved manipulation of the probe angle until the flow signal disappeared (presumed to be 90 degrees, cos 0 = 0, fd = 0) and rotate the probe a fixed angle from this point. 8 None of these indirect solutions were suitable for accurate quantitative flow measurements, even in easily accessible vessels. The great advantage of the duplex scanner is the real-time image of the vessel under examination. This allows at least a two-dimensional localization of the Doppler sample volume and identification of the angle of insonation. Much attention has been focused on the "angle correction" features of duplex ultrasound equipment, but it should be remembered that this is not an automated feature but must be actively used by the technologist performing the examination. The sample volume flow cursor must be placed midstream in the vessel visualized by the B-mode image. The machine will then display the insonation angle of the M-line, and the velocity waveform displayed will be based on this angle. Experienced investigators have suggested that an examination angle of 60 degrees is ideal for signal processing, 9'1 but examination angles of 30 to 60 degrees are both realistic and accurate. However, Burns n has noted that at 70 degrees an uncertainty of the measured angle of even 5 degrees will cause a 25% error in the calculated velocities, In the present study when the angle of insonation exceeded 70 degrees, significant increases in the measured blood flow velocities were observed. Also in the present study the percentage increases in measured flow velocities in the SMA at 80 degrees were significantly greater than those observed in the CCA at the same angle. This difference in the measured flow velocities between the SMA and CCA may result from their different depths and thus difficulty in insonation and angle measurement. Any errors in angle measurement would result in larger changes in velocity calculations at the higher angles. Measurement of arterial flow velocities should probably only be performed at examination levels below 70 degrees with active angle control. It is interesting to note that some newer duplex equipment automatically stops signal processing if the examination angle exceeds 70 degrees. The clinical interest in the measurement of rues- enteric blood flow is self evident since syndromes of mesenteric ischemia continue to challenge physicians caring for patients with these disorders. Ever since the report ofdunphy 12 the association ofprodromal gastrointestinal symptoms (or "abdominal angina") with thrombosis of the mesenteric vessels has been well recognized. Nevertheless, these cases frequently go undiagnosed since definitive diagnosis of mesenteric arterial disease has required arteriography with its risks and discomforts, and this examination was relegated to the very end of the diagnostic spectrum in patients with symptoms. Patency or occlusion of individual vessels can reliably be established by arteriography, but it offers no objective physiologic assessment of the adequacy of splanchnic blood flow. Dye-dilution, or video-dilution techniques 1316 have been used for quantitative measurement of splanchnic blood flow but have not been widely accepted because of their invasive requirements and complexity. More recently it was recognized that duplex ultrasound scanning could be used for assessment of visceral vessels including the mesenteric arteries in patients suspected of having intestinal ischemia. 17 Initial studies of mesenteric scalming were designed to measure fasting and postprandial arterial flow velocities in the celiac artery (CA) and the SMA.1,2"s7 Additional studies have been performed to assess the mesenteric flow response to a variety of physiologic and pharmacologic stimuli, a variable meal composition, 4 and exercise.18 In an early study s we observed a fasting peak systolic SMA flow velocity of 103 _ cm/sec in healthy volunteers. In a later study by Lilly et al,3 from our institution we observed a fasting SMA velocity of _+ 19 cm/sec in healthy subjects before a test meal. There were no identifiable age or sex differences in the two groups to possibly explain these differences, and it is possible that inaccurate angle correction may have played some role. Nevertheless, similar observations have been noted in separate reports from other groups at one institution, 2,4 so other factors may play a role in these observations. These studies demonstrated significant, reproducible changes in SMA blood flow and have substantiated the usefitlness of the duplex teclmique for evaluation of the splanchnic circulation, but they did not specifically address the question of the absolute SMA blood flow velocities. Burns ~9 has noted that in areas of significant arterial curvature, the momentum of the fastest moving blood at center stream may cause it to accelerate to the outer edge of the curve thereby skewing the ve-

6 Volume 11 Number 5 May 1990 Mesenteric flow velodty variations % AV 100, " -40 3o i I t I t ANGLE OF INSONATION Fig. 3. Graph illustrating the effect of angle of insonation on the percentage change in PSV in the SMA relative to the PSV measured at 60 degrees. lodty profile. A potential area of such arterial curvature almost routinely exists in the first few centimeters of the SMA beyond its origin. As the vessel makes an almost 90 degree turn in this region it may not be possible to reliably place the sample volume flow cursor parallel to the vessel axis and accurately correct the angle. For velocity measurement in healthy subjects it would appear to be more reasonable to sample several centimeters distally where the vessel becomes more linear in its course and reproducible angle correction is possible for serial examinations (Fig. 2). It is well known that atherosclerotic occlusive disease of the visceral vessels occurs most frequently at their ostia, and diagnosis by duplex scan will be most accurate when the sample volume is near the lesion. As noted above, however, velocities produced by hemodynamically significant stenoses will be significantly higher than normal and easily distinguishable in these,cases. Arterial curvature may also have the effect of causing a double rotation of flow so that the velocity has a helical component by the end of the curve. Beach et al.9 observed that standard velocity criteria for detection of significant carotid artery stenosis were less reliable when examinations were conducted at various angles and angle-corrected velocities were calculated. These authors speculated that the helical flow patterns in normal arteries may make the standard assumptions of the Doppler equation unreliable in these cases. They recommended that a standard examination angle be used in all examinations and this angle remain consistent for repeated examinations of the same artery. It has always been an intriguing physiologic possibility, especially in the splanchnic circulation, that accurate blood flow velocity measurements would allow calculation of volumetric blood flow. This could be done (if the arterial diameter were known) by multiplying velocity by cross-sectional area. Duplex scanners first allowed noninvasive arterial imaging (where vessel diameter could be measured) and image-directed velocity measurements. Quamar et al. 7 first reported the use of this technique to noninvasively measure volumetric SMA blood flow in man. Their preliminary results showed a resting SMA flow of ml/min and appeared to be in agreement with more invasive dye-dilution techniques. The authors cautioned, however, about sources of error inherent in these calculations, which included the angle of insonation and the estimate of vessel diameter. In our previous study 3 measured SMA vessel diameter (0.6 to 0.7 cm) was almost identical to that reported by Quamar et al. Although it was not stated in their report, calculation from their data would suggest a measured average SMA flow velocity in the range of 30 cm/sec, similar to our observations in the range of acceptable examination angles. Their highest flow calculations (890 ml/min) would have resulted from an average velocity over 50 cm/sec, which may have reflected measurement at higher, less accurate angles as was noted in the current study. Quamar et al. noted that overestimarion of the velocity may be compensated by underestimation of vessel diameter characteristic of ultrasound measurement. It would not appear to be scientifically sound to arrive at accurate information through a process where separate sources of error fortuitously cancel one another. Reproducible noninvasive volume flow measurements in the splanchnic circulation wilt require a tighter control of these vari-

7 694 Rizzo et al. ~o~nai of VASCULAR SURGERY ables in future studies. This appears less critical to the detection of significant mesenteric arterial occlusive disease since velocity waveform parameters are more reliable indicators of occlusive disease than volume flow. 2 The use of the duplex ultrasound scanner has been a major breakthrough in the noninvasive evaluation of visceral blood flow in man since previously available diagnostic techniques were invasive or prohibitively complex. Previous studies have confirmed the usefulness of this technique for characterizing the normal physiologic responses of the mesenteric circulation. The present study suggests that critical attention must be paid to the conduct of serial examinations in this area if results are to be legitimately standardized. Further clinical experience will be necessary to determine whether the anatomic configuration of the SMA, or the flow patterns within it may adversely effect the accurate quantitation of blood flow in this vessel. REFERENCES 1. Jager K, Bollinger A, VaUi C, Ammarm R. Measurement of mesenteric blood flow by duplex scan. J VASC SURG 1986;3: Nicholls SC, Kohler TR, Martin RS, Strandness DE. Use of hemodynamic parameters in the diagnosis of mesenteric insuflidency. J VASC SURG 1986;3: Lilly MP, Harward TRS, Flinn WR, Blackburn DR, Astleford PM, Yao JST. Duplex ultrasound measurement of changes in mesenteric flow velocity with pharmacologic and physiologic alteration of intestinal blood flow in man. J VAsc SUttG 1989;9: Moneta GL, Taylor DC, Helton WS, et al. Duplex ultrasound measurement of postprandial intestinal blood flow: effect of meal composition. Gastroenterology 1988;95: Flinn WR, Sandager GP, Lilly MP, Yao JST, Bergan JJ. Duplex scan of mesenteric and celiac arteries. In: Bergan JJ, Yao JST, eds. Arterial surgery: new diagnostic and operative techniques. Orlando: Grime & Stratton, 1988:367-75, 6. Quamar MI, Read AE, Skidmore R, Evans JM, WiUiarnson RCN. Transcutaneous Doppler ultrasound measurement of coeliac axis blood flow in man. Br j Surg 1985;72:391,3. 7. Quamar MI, Read AE, Skidmore R, Evans JiM, Wells PNT. Transcutaneous Doppler ultrasound measurement of superior mesenteric artery blood flow in man. Gut 1986;27: Fish P, Wakers D. Beam/vessel angle problem in Doppler flow measurement. In: Taylor DEM, Whamond D, eds. Noninvasive clinical measurement. Turnbridge-Wells UK: Pitman Publishing Co, 1977: Beach KW, Lawrence R, PhilLips DJ, Primozich J, Strandness DE. The systolic velocity criterion for diagnosing significant internal carotid artery stenoses. J Vasc Tech 1989;13: Taylor DC, Strandness DE. Carotid artery duplex scanning. J Clin Ultrasound 1987;15:635-44, 11. Burns PN. Interpretation and analysis of Doppler signals: In: Taylor KJW, Bums PN, Wells PNT, eds. Clinical applications of Doppler ultrasound. New York: Raven Press, 1988: Dunphy JE. Abdominal pain of vascular origin. Am J Med Sci 1936;192: Buchardt-Hansen HJ, Engell HC, Ring-Larsen H, Ranek L. Splanclmic blood flow in patients with abdominal angina before and after arterial reconstruction. Ann Surg 1977;186: Norryd C, Dencker H, Lunderquist A, Olin T. Superior ruesenteric blood flow in man studied with a dye-dilution technique. Acta Chir Scand 1975;141: Norryd C, Dencker H, Lunderquist A, Olin T, Tylen U. Superior mesenteric blood flow during digestion in man. Acta Chir Scand 1975;141: Lanz B, Link D, Holcroft J, Foerster J. Video dilution technique: angiographic determination of splanchnic blood flow. In: Granger D, Bulldey G, eds. Measurement of blood flow, application to the splanchnic circulation. Baltimore: Williams & Wilkins, 1981: Jager KA, Former GS, Thiele BL, Strandness DE. Noninvasive diagnosis of intestinal angina. J Grin Ultrasound 1984;12: Quamar MI, Read AE. Effects of exercise on mesenteric blood flow in man. Gut 1987;28: Bums PN. Hemodynamics. In: Taylor KJW, Bums PN, Wells PNT, eds. Clinical applications of Doppler ultrasound. New York: Raven Press, 1988:46-75.