Noninvasive detection of lilac artery stcnosis in the presence of superficial £emoral artery obstruction Steven J. Burnham, MD, RVT, Paul Jaques, MD, and Cynthia B. Burnham, BSN, R2q, RVT, Chapel Hill, N.C. In patients with superficial femoral artery obstruction, iliac disease may be difficult to diagnose by commonly used noninvasive techniques. We studied common femoral artery acceleration time (onset of systole to peak systole), as measured from a Doppler spectral display and expressed in milliseconds. Previous work has suggested that an acceleration time of 144 msec or greater is abnormal and is associated with iliac stenosis (-> 75% diameter reduction) or occlusion. During a 2-year period, in 139 limbs with superficial femoral artery obstruction, acceleration times were measured immediately before angiography. The overall test accuracy was 94.2% (131/139). In the 112 sides with normal angiograms, the acceleration time correctly identified no disease in 109 patients (97.3% specificity), and in the 27 iliac stenoses or occlusions the test detected disease in 22 patients (81.5% sensitivity). This appears to be a good test that can be done with equipment usually available in most vascular laboratories. (J VASC SURG 1992;16:445-52.)
The history of noninvasive detection of lilac occlusive disease is one marked by l~mitations. Indirect measurements of blood pressure "with cuffs placed at the upper thigh have two potential sources of error. One related to the cuff-to-hmb size discrepancy causes a falsely elevated measurement (compared with the intraluminal pressure). The other is that arterial obstruction beneath the cuff (superficial femoral artery [SFA]) and iliac artery obstruction cause similar decrements in upper thigh blood pressure measurements, preventing the differentiation of SFA and iliac obstruction. Other means of assessing the iliac segment include hyperemia testing with papaverine injection, 1,2 direct interrogation with duplex scanning, 3,4 various forms ofradiologic imaging, and intraarterial pressure measurements, s These tests are either invasive or limited by presence of intestinal gas or body habitus. From the Departments of Surgery and Radiology~ School of Medicine, University of North Carolina, Chapel Hill, and The Peripheral Vascular Laboratory, University of North Carolina Hospitals, Chapel Hill. Presented at ~e Sixteenth Annual Meeting of the Southern Association for Vascular Surgery, St. Thomas, Virgin Islands, ][an. 22-25, 1992. Reprint requests: Steven J. Burnham, MD, Department of Surgery CB 7210, 210 Burnett-Womack Bldg., Chapel Hill, NC 27599-7210. 24/6/39927
Mathematic analyses of the common femoral artery (CFA) velocity signal have shown some degree o f success 6"9 but have not gained widespread clinical use because of their complexity. Subjective assessment of the CFA velocity waveform based on the presence or absence of reverse flow in early diastole is not reliable because either iliac or SFA obstruction may produce loss of flow reversal. 1° Previous work from our institution has shown that the Doppler spectral pattern from the CFA does have differences compared with normal limbs that may be helpful in discriminating between iliac and SFA obstructions.Ii,12 Comparison with single-plane lilac angiograms showed that most lilac stenoses greater than 75% were associated with systolic rise times equal to or greater than 144 msec. This study is an attempt to validate this measurement in patients with SFA obstruction by correlation with anglograms.
MATERIAL AND METHODS All patients from a roster of diagnostic anglograms done at the University of North Carolina Hospitals between January 1988 and October 1990 were considered for inclusion in the data base. Limbs were included when noninvasive studies were done within the 2 months preceding diagnostic angiography, provided that no form of intervention or change in clinical status occurred during the interim. 445
446
Journal of VASCULAR SURGERY
Burnham, Jaques, a n d B u r n h a m
OVER STENOSIS
•
.L
~
¢,.,
Fig. 1. CFA Doppler spectral tracing from patient with stenosis at site of Doppler sampiing. Table I. Relationship of acceleration time in iliac artery disease: SFA obstruction
Table III. Relationship of acceleration time in iliac artery disease: Both groups combined*
lliac artery angiogram
lliac artery angiogram
Acceleration time (msec)
Normal or mild
Severe or occluded
Short times Long times
109 3
5 22
114 25
112
27
139
Total
Sensitivity 81.5% (22/27) Specificity 97.3% (109/112) Positive predictive value 88.0% (22/25) Negative predictive value 95.6% (109/114)
Acceleration time (msec)
Normal or mild
Severe or occluded
Short times Long times
315 5
9 75
324 8O
320
84
404
Total
Sensitivity 89.3% (75/84) Specificity 98.4% (315/320) Positive predictive value 93.8% (75/80) Negative predictive value 97.2% (315/324) *Values are number of extremities or limbs.
Table II. Relationship of acceleration time in iliac artery disease: SFA not obstructed Iliac artery angiogram Acceleration time (msec)
Normal or mild
Severe or occluded
Short times Long times
206 2
4 53
210 55
208
57
265
Total
Sensitivity 93.0% (53/57) Specificity 99.0% (206/208) Positive predictive value 96.4% (53/55) Negative predictive value 98.1% (206/210)
Four hundred eighty-one sides had clear views of both the aorto/iliac segment and the SFA. In this study, sides were excluded if inflow to the groin was provided by a bypass graft (29 sides) or the inflow vessels had aneurysms (15 sides). When the CFA was
stenotic (Fig. 1) or occluded (33 sides), no acceleration time was measured. Of the remaining 404 sides, 265 sides had patent and 139 sides had obstructed SFAs. Registered nurses/registered vascular technologists with at least 5 years of experience acquired and interpreted the velocity data. A 5 MHz continuouswave Doppler (Medasonics, Inc., Mountain View, Calif.) was used to obtain representative spectral waveforms from the CFA. The transducer was placed over the palpated CFA at the inguinal crease and moved cephalad to the inguinal ligament to optimize and clarify the signal. When Doppler data were confusing or ambiguous, a pulsed Doppler in a duplex system (Acuson Computed Sonography, Acuson, Inc., Mountain View, Calif.) was used to identify the CFA through direct visualization. (This was necessary in five limbs or 3.6% of the study
Volume 16 Number 3 September 1992
Doppler spectralpatte~ of iliac stenoses 447
O~12L1
4000
~000
2000
'0r.l,'.
A
. 78 mse0iiI
5~00
| l l
i..~
| |
..i~ - .~ -.~
-• •
Zl
iF.-
|
.~4.
,~ ,
,
~-
~" ,
_ ~
~
i, :.
.
°
i~, ~
""
"":
'
"
. . . . . .
"'I
B Fig. 2. A, CFA Doppler spectral tracing from patient with both SFA obstruction and iliac stenosis. Acceleration time is prolonged to 168 msec, and the tracing is broad based. Upslope "leans" to right. Prolongation of upslope time seemed to be independent of disease in SFA. B, CFA Doppler spectral tracing from patient with normal iliac segment by angiography but obstruction of proximal SFA. Tracing shows brief upstroke with most of widening of base occurring after peak systole. In both Figs• 1 and 2 there is loss of reversed-flow component. group.) Electronic cursors were placed over the point of onset of systolic upslope and the point of maximum velocity. The resulting interval was identiffed as the acceleration time or systolic rise time (Fig. 2). Measurements with the continuous-wave Doppler were made in 11 msec increments and the pulsed Doppler in 16 msec intervals. Times measuring 144 msec or longer were considered abnormal.
For the purposes of this study, angiograms without detectable iliac lesions (normal) were grouped with those showing less than 75% (diameter reduction) stenosis (mild), whereas greater degrees of stenosis (severe) were grouped with occlusions. Multiple stenoses were labeled according to the most severe lesion, Sides with mild and severe stenosis were reviewed by at least two investigators who had
Journal of VASCULAR SURGERY
448 Burnham, Jaques, and Burnham Number of Limbs 25
:::::::::::::::::::::: 20
ii:?i!!iii
[] [] [] []
'
15
!
Normal Mild Severe Occluded
! 0
,
i
•
56
67
78
89
00111122133144156167178189200211222 Acceleration Times in Milliseconds
Fig. 3. Distribution of sides based on iliac status and acceleration time for sides with obstructed SFAs.
[
178 msec t
."
•
Fig. 4. Doppler spectral tracing of left CFA corresponding to anglogram in Fig. 5. no knowledge of the associated acceleration times. When there was disagreement, resolution was by a third opinion. One hundred thirty-nine sides showed obstruction of the SFA. The SFA was occluded at the origin in 87 sides (62.6%), at the adductor canal in 46 sides (33.1%), and showed stenosis (> 50%) in 6 sides (4.3%). All six of the stenotic SFAs had multiple lesions. R E S U L T S (Tables I to III) The aorta and ipsilateral iliac artery were normal in 98 sides (70.5%), had less than 75% stenosis in 14 sides (10.1%), had 75% or greater stenosis in 17 sides (12.2%), and were occluded in 10 sides (7.2%). Fig. 3 shows the distribution of iliac disease by acceleration time. The accuracy of this criterion is 94.2% (121/139). The sensitivity is 81.5% (22/27),
and the specificity is 97.3% (109/112). The predictive power of a positive test result is 88.0% (22/25), and the predictive value of a negative test result is 95.6% (109/114). The sides without SFA obstruction are similar in distribution of iliac lesions and acceleration times. O f the 265 sides that showed no obstruction of the SFA, the iliac artery was normal in 188 sides (71.0%), had less than 75% stenosis in 20 sides (7.5%), had 75% or greater stenosis in 37 sides (14.0%), and was occluded in 20 sides (7.5%).
DISCUSSION In the group of patients with SFA obstruction there were three sides with prolonged acceleration times that were judged by image to have mild iliac disease; two patients had iliac stenosis of 50%, and the third had 30% stenosis. However, two of these
Volume t6 Number 3 September 1992
three patients with false-positive results underwent iliac angioplasty because of significant pressure gradients across the lesions, with reduction in the acceleration times from 178 to 100 msec and 144 to 111 msec. This suggests that these were not falsepositive acceleration times but false-negative anglograms, a recognized limitation of single-plane angiography. In our clinical practice the preangiographic finding of an abnormal acceleration time modifies the performance of the arterial study. Single-plane pelvic arteriography is supplemented by oblique views or direct intraluminal arterial pressure measurements with or without vasodilator drugs to resolve the significance of mild/moderate arterial stenoses when CFA acceleration times are prolonged. Five sides had either 75% or 80% iliac stenosis but did not have a prolonged acceleration time. Four of these five false-negative results had short-segment stenoses in the proximal common lilac artery (Figs. 4 and 5). All five stenoses were associated with resting pressure gradients made worse by peripheral vasodilation and were corrected by balloon catheter angioplasty. Therefore a normal acceleration time should not be used to exclude the presence of a significant iliac lesion. Rather, an acceleration time less than 144 msec may indicate that any significant lilac lesion would likely be amenable to angioplasty. Review of the eight errors revealed only one tracing (Fig. 6) that would, in retrospect, have been measured differently. If the third waveform, rather than the second, had been chosen for analysis, the acceleration time would have been abnormal, and the lilac stenosis would have been detected. This study supports the work of Thiele et al.,13 who have noted that, depending on the degree of stenosis, velocity patterns may resume a normal contour within three vessel diameters downstream from the stenosis. Therefore the presence of net flow reversal in a biphasic or triphasic waveform does not rule out the presence of significant proximal disease. However, our study suggests that some', of these lesions may be detected by a prolonged systolic rise time (Fig. 7). In 1976 Nicolaides et al.10 recognized that loss of reverse flow in the CFA waveform could be the result of either isolated inflow or outflow disease. They noted that a femoral waveform characterized by a rapid ascent, an initial rapid descent, and a prolonged late descent was found in patients with SFA obstruction. This finding was supported in our study (Fig. 2). Furthermore, in our study these changes were present in patients with SFA obstruction, whereas
Doppler spectralpattern of iliac stenoses 449
Fig. 5. Iliac angiogram of patient with short-segment 75% stenosis of proximal left common lilac artery-. the prolongation of the ascent of the waveform was specLfic for iliac disease. The receiver-operator characteristic curve for different criteria for a positive test result is displayed in Fig. 8. Although the point representing 144 msec is favorable, the points representing shorter acceleration times are also favorable; tile proximity of these points probably relates to the small proportion of sides with intermediate grades of lilac stenosis. However, an acceleration time of 133 msec should alert the interpreter to scrutinize the waveform and the clinician to consider more than single-plane pelvic angiograms, Fig. 9 suggests that more than hak¢ of these sides will be associated with severe iliac disease, and at 144 msec nearly all are associated with severe disease.
In clinical practice there are a number of circum stances in which acceleration time is useful in patient management. The detection and treatment of lilac artery stenosis is particularly important when a single iliac artery is expected to provide adequate arterial
450
Journal of VASCULAR SURGERY
Burnham,Jaques, and Burnham 5lIl'~lO
100 msec
III CI I'_l
m'_
• I
'
I
:
~|~~"~"" 4 "
L ,
~" " "
:.
C¢.':.¢_:
v
. . .~.-,e¢
Ic
',,~
[J
'?
Fig. 6. Incorrect placement of second cursor incorrectly suggesting absence of iliac disease.
,144 msec t!
_''°" j ::--~
ti "-I
Fig. 7. Tracing shows net flow reversal in triphasic waveform with prolonged acceleration time. Iliac angiogram shows severe iliac stenosis.
inflow to both lower extremities (femoral crossover donor vessel). An abnormal acceleration time may be the critical indicator that the donor side inflow is jeopardized. When angioplasty o f an iliac stenosis is carried out proximal to a SFA obstruction, ankle blood pressure may not show prompt increase because of the distal disease; however, improvement in acceleration time is prompt and reassuring.
REFERENCES
1. van Asten WNIC, Beijneveld WJ, Pieters BR, van Lier HIJ, Wijn PFF, Skomicki SH. Assessment of aortoiliac obstructive
2.
3.
4.
5.
disease by Doppler spectrum analysis of blood flow velocities in the common femoral artery at rest and during reactive hyperemia. Surgery 1991;109:633-9. Flanigan DP, Williams LR, Schwartz •A, Schuler 5~, Gray B. Hemodynamic evaluation of the aortoiliac system based on pharmacologic vasodilatation. Surgery 1983;93:70%14. Langsfeld M, Nepute J, Hershey FB, et al. The use of deep duplex scanning to predict hemodynamically significant aortoiliac stenoses. • VAsc SURG 1988;7:395-9. Haenen JH, van Asten WN~C, Theloosen-Kerstens AWP, Skomicki SH. The value of duplex scanning in assessing aorto-iliac disease. J Vasc Technol 1991;15:231-4. Thiele BL, Bandyk DF, Zierler RE, Strandness DE Jr. A systematic approach to the assessment of aortoiliac disease. Arch Surg 1983;118:477-81.
Volume 16 Number 3 September 1992
Doppler spectrgl patten~ of lilac stenoses 451
100 111 - 133 msec
90 100 msec 80
144 mse-~c
70
°m
60
t."
cO 50 40
156 msec
30
.. 70
I
|
80
90
100
Specificity Fig. 8. Receiver-operator characteristic curve for different criteria for positive test result.
Per cent 100
80 60 40 20 o 56 67 78 89 1 0 0 1 1 1 1 2 2 1 3 3 1 4 4 1 5 6 1 6 7 1 7 8 1 8 9 2 0 0 2 1 1 2 2 2 Acceleration Times in milliseconds
Fig. 9. Proportion o f each acceleration time's values wkh significant lilac disease.
6. Clifford PC, Slddmore R, Bird D, Woodcock JP, Lusby RJ, Baird RN. Femoral artery doppler signal analysis in lower limb ischemia. J Cardiovasc Surg 1982;23:6974. 7. Johnston KW, Kassam M, Koers J, Cobbold RSC, MacHattie D. Comparative study of four methods for quantifying Doppler ttltrasound waveforms from the femoral artery. Ultrasound Med Biol 1984;10:1-12. 8. Archie IP lr, Feltman RW. Determination of the hemodynamic significance of iliac artery stenosis by noninvasive Doppler ultrasonography. Surgery 1982;91:41%24. 9. Harward TRS, Bernstein EF, Fronek A. The value of power frequency spectrum analysis in the identification of aortoiliac artery disease. J Vasc SURG 1987;5:803-13. 10. Nicolaides AN, Gordon Smith IC, Dayandas J, Eastcott HH. The value of Doppler blood velocity tracings in the detection
of aortoiliac disease in patients with intermittent claudication. Surgery 1976;80:774-8. 11. Kupper CA, Young L, Keagy BA, Burnham S]. Spectra1 analysis of the femoral artery for identification of iliac lesio~so Bruit 1984;8:157-63. 12. Burnham CK, Dewhirst N, Bumhan~ SJ. Doppler spectral waveforms for recording peripheral arterial signals: the preferred method. J Vase Technol 1989;8:69-73. 13. Thiele BL, Hutchinson KJ, Greene FM, Forster FK, Strandness DE lr. Pulsed Doppler waveform patterns produced by smooth stenosis in the dog thoracic aorta. In: Taylor DEM, Stevens AL, eds. Blood flow: theory and practice. Orlando: Academic Press, 19834:85-104. Submitted Jan. 29, t992; accepted )rune 2, 1992.
452
Burnham, Jaques, and Burnham
Journal of VASCULAR SURGERY
DISCUSSION
Dr. Robert W. Barnes (Little Roclq Ark.). I agree that the so-called acceleration rime between onset and peak systole of the femoral arterial Doppler frequency waveform is one of the most valid indicators of the status of the iliac artery, particularly in the presence of superficial femoral artery occlusion. One must understand that these measurements require a Doppler sound spectrum analyzer. They cannot be obtained from simple analog tracings on a strip chart recorder. Accurate measurements are readily available on most duplex scanners in current use. You have correlated the acceleration time with contrast arteriography. Unfortunately, arteriograms may belie the hemodynamic significance of an arterial stenosis, a fact that you invoke to explain your three false-positive Doppler study results, two of which were accompanied by so-called significant pressure gradients. Unforttmately, you do not define the threshold of significant resting or hyperemic pressure gradients. Furthermore, all five false-negative Doppler study results were accompanied by such significant pressure gradients and a greater than 75% stenosis on arteriography. I remain concerned about the following three points: First, it would be helpful if you could provide us with correlations between the Doppler studies and aortofemoral resting or hyperemic pressure gradients obtained at the time of arteriography for all patients because such physiologic data may be a better gold standard than arteriography for validating acceleration times. Second, have you considered evaluating acceleration times during some form of reactive hyperemia, which may lead to better correlation with pressure gradients ? Finally, will you reiterate the value of this test in routine clinical situations, particularly to convince those cynics who might point out that most of us obtain arteriograms and, when needed, aortofemoral pressure gradients for patients requiring interventions for peripheral arterial occlusive disease. Dr. Steven J. Burnham. Since beginning this analysis, we have measured some pressure gradients in isolated patients. I do not have those data with me. I think that this
test offers to our patients, before hospitalization, an opporumity to understand which segment is likely to show disease and also offers to our radiology colleagues an additional tool to help plan the angiography. I think that for those clinicians who do routine biplane angiography of the lilac vessels, this test has very little value. For those who do single-plane angiography and are sometimes undecided about the angiographic significance, I think that this piece of information may represent helpful additional informarion. Dr. Timothy R. S. H a r w a r d (Gainesville, Fla.). As you demonstrated in one of your slides, in the same patient the tracings varied with each cardiac cycle. If you picked the wrong tracing, you get the wrong diagnosis. Have you looked at the reproducibility of this particular test? I would agree with you, as I showed in my own work, the arteriogram is not the gold standard. I tended to think of the noninvasive test being positive and the arteriogram as being a false-posirive or false-negarive examination. Also, I agree that some form of stress testing or pressure measurement is nice to know when making the final interpretation. The last question I have pertains to a problem that I had when I looked at this subject: the effect of a deep femoral artery stenosis on the common femoral artery Doppler tracing. I had two patients with occlusion of the superficial femoral artery with a stenosis at the origin of the deep femoral artery that was significant. This hemodynamic disease led to a false-positive examination because it acted the same as a proximal upstream stenosis. Have yon seen this in your own work? Dr. Burnham. I do not have an answer about internal consistency with Doppler angle. The angiogram is not the gold standard if you use only a single plane. The deep femoral artery was not analyzed separately; however, I can tell you that no patients in this group had total occlusion of the deep femoral artery, and a small minority of parients had stenosis of the deep femoral artery.
The history of noninvasive detection of lilac occlusive disease is one marked by l~mitations. Indirect measurements of blood pressure "with cuffs placed at the upper thigh have two potential sources of error. One related to the cuff-to-hmb size discrepancy causes a falsely elevated measurement (compared with the intraluminal pressure). The other is that arterial obstruction beneath the cuff (superficial femoral artery [SFA]) and iliac artery obstruction cause similar decrements in upper thigh blood pressure measurements, preventing the differentiation of SFA and iliac obstruction. Other means of assessing the iliac segment include hyperemia testing with papaverine injection, 1,2 direct interrogation with duplex scanning, 3,4 various forms ofradiologic imaging, and intraarterial pressure measurements, s These tests are either invasive or limited by presence of intestinal gas or body habitus. From the Departments of Surgery and Radiology~ School of Medicine, University of North Carolina, Chapel Hill, and The Peripheral Vascular Laboratory, University of North Carolina Hospitals, Chapel Hill. Presented at ~e Sixteenth Annual Meeting of the Southern Association for Vascular Surgery, St. Thomas, Virgin Islands, ][an. 22-25, 1992. Reprint requests: Steven J. Burnham, MD, Department of Surgery CB 7210, 210 Burnett-Womack Bldg., Chapel Hill, NC 27599-7210. 24/6/39927
Mathematic analyses of the common femoral artery (CFA) velocity signal have shown some degree o f success 6"9 but have not gained widespread clinical use because of their complexity. Subjective assessment of the CFA velocity waveform based on the presence or absence of reverse flow in early diastole is not reliable because either iliac or SFA obstruction may produce loss of flow reversal. 1° Previous work from our institution has shown that the Doppler spectral pattern from the CFA does have differences compared with normal limbs that may be helpful in discriminating between iliac and SFA obstructions.Ii,12 Comparison with single-plane lilac angiograms showed that most lilac stenoses greater than 75% were associated with systolic rise times equal to or greater than 144 msec. This study is an attempt to validate this measurement in patients with SFA obstruction by correlation with anglograms.
MATERIAL AND METHODS All patients from a roster of diagnostic anglograms done at the University of North Carolina Hospitals between January 1988 and October 1990 were considered for inclusion in the data base. Limbs were included when noninvasive studies were done within the 2 months preceding diagnostic angiography, provided that no form of intervention or change in clinical status occurred during the interim. 445
446
Journal of VASCULAR SURGERY
Burnham, Jaques, a n d B u r n h a m
OVER STENOSIS
•
.L
~
¢,.,
Fig. 1. CFA Doppler spectral tracing from patient with stenosis at site of Doppler sampiing. Table I. Relationship of acceleration time in iliac artery disease: SFA obstruction
Table III. Relationship of acceleration time in iliac artery disease: Both groups combined*
lliac artery angiogram
lliac artery angiogram
Acceleration time (msec)
Normal or mild
Severe or occluded
Short times Long times
109 3
5 22
114 25
112
27
139
Total
Sensitivity 81.5% (22/27) Specificity 97.3% (109/112) Positive predictive value 88.0% (22/25) Negative predictive value 95.6% (109/114)
Acceleration time (msec)
Normal or mild
Severe or occluded
Short times Long times
315 5
9 75
324 8O
320
84
404
Total
Sensitivity 89.3% (75/84) Specificity 98.4% (315/320) Positive predictive value 93.8% (75/80) Negative predictive value 97.2% (315/324) *Values are number of extremities or limbs.
Table II. Relationship of acceleration time in iliac artery disease: SFA not obstructed Iliac artery angiogram Acceleration time (msec)
Normal or mild
Severe or occluded
Short times Long times
206 2
4 53
210 55
208
57
265
Total
Sensitivity 93.0% (53/57) Specificity 99.0% (206/208) Positive predictive value 96.4% (53/55) Negative predictive value 98.1% (206/210)
Four hundred eighty-one sides had clear views of both the aorto/iliac segment and the SFA. In this study, sides were excluded if inflow to the groin was provided by a bypass graft (29 sides) or the inflow vessels had aneurysms (15 sides). When the CFA was
stenotic (Fig. 1) or occluded (33 sides), no acceleration time was measured. Of the remaining 404 sides, 265 sides had patent and 139 sides had obstructed SFAs. Registered nurses/registered vascular technologists with at least 5 years of experience acquired and interpreted the velocity data. A 5 MHz continuouswave Doppler (Medasonics, Inc., Mountain View, Calif.) was used to obtain representative spectral waveforms from the CFA. The transducer was placed over the palpated CFA at the inguinal crease and moved cephalad to the inguinal ligament to optimize and clarify the signal. When Doppler data were confusing or ambiguous, a pulsed Doppler in a duplex system (Acuson Computed Sonography, Acuson, Inc., Mountain View, Calif.) was used to identify the CFA through direct visualization. (This was necessary in five limbs or 3.6% of the study
Volume 16 Number 3 September 1992
Doppler spectralpatte~ of iliac stenoses 447
O~12L1
4000
~000
2000
'0r.l,'.
A
. 78 mse0iiI
5~00
| l l
i..~
| |
..i~ - .~ -.~
-• •
Zl
iF.-
|
.~4.
,~ ,
,
~-
~" ,
_ ~
~
i, :.
.
°
i~, ~
""
"":
'
"
. . . . . .
"'I
B Fig. 2. A, CFA Doppler spectral tracing from patient with both SFA obstruction and iliac stenosis. Acceleration time is prolonged to 168 msec, and the tracing is broad based. Upslope "leans" to right. Prolongation of upslope time seemed to be independent of disease in SFA. B, CFA Doppler spectral tracing from patient with normal iliac segment by angiography but obstruction of proximal SFA. Tracing shows brief upstroke with most of widening of base occurring after peak systole. In both Figs• 1 and 2 there is loss of reversed-flow component. group.) Electronic cursors were placed over the point of onset of systolic upslope and the point of maximum velocity. The resulting interval was identiffed as the acceleration time or systolic rise time (Fig. 2). Measurements with the continuous-wave Doppler were made in 11 msec increments and the pulsed Doppler in 16 msec intervals. Times measuring 144 msec or longer were considered abnormal.
For the purposes of this study, angiograms without detectable iliac lesions (normal) were grouped with those showing less than 75% (diameter reduction) stenosis (mild), whereas greater degrees of stenosis (severe) were grouped with occlusions. Multiple stenoses were labeled according to the most severe lesion, Sides with mild and severe stenosis were reviewed by at least two investigators who had
Journal of VASCULAR SURGERY
448 Burnham, Jaques, and Burnham Number of Limbs 25
:::::::::::::::::::::: 20
ii:?i!!iii
[] [] [] []
'
15
!
Normal Mild Severe Occluded
! 0
,
i
•
56
67
78
89
00111122133144156167178189200211222 Acceleration Times in Milliseconds
Fig. 3. Distribution of sides based on iliac status and acceleration time for sides with obstructed SFAs.
[
178 msec t
."
•
Fig. 4. Doppler spectral tracing of left CFA corresponding to anglogram in Fig. 5. no knowledge of the associated acceleration times. When there was disagreement, resolution was by a third opinion. One hundred thirty-nine sides showed obstruction of the SFA. The SFA was occluded at the origin in 87 sides (62.6%), at the adductor canal in 46 sides (33.1%), and showed stenosis (> 50%) in 6 sides (4.3%). All six of the stenotic SFAs had multiple lesions. R E S U L T S (Tables I to III) The aorta and ipsilateral iliac artery were normal in 98 sides (70.5%), had less than 75% stenosis in 14 sides (10.1%), had 75% or greater stenosis in 17 sides (12.2%), and were occluded in 10 sides (7.2%). Fig. 3 shows the distribution of iliac disease by acceleration time. The accuracy of this criterion is 94.2% (121/139). The sensitivity is 81.5% (22/27),
and the specificity is 97.3% (109/112). The predictive power of a positive test result is 88.0% (22/25), and the predictive value of a negative test result is 95.6% (109/114). The sides without SFA obstruction are similar in distribution of iliac lesions and acceleration times. O f the 265 sides that showed no obstruction of the SFA, the iliac artery was normal in 188 sides (71.0%), had less than 75% stenosis in 20 sides (7.5%), had 75% or greater stenosis in 37 sides (14.0%), and was occluded in 20 sides (7.5%).
DISCUSSION In the group of patients with SFA obstruction there were three sides with prolonged acceleration times that were judged by image to have mild iliac disease; two patients had iliac stenosis of 50%, and the third had 30% stenosis. However, two of these
Volume t6 Number 3 September 1992
three patients with false-positive results underwent iliac angioplasty because of significant pressure gradients across the lesions, with reduction in the acceleration times from 178 to 100 msec and 144 to 111 msec. This suggests that these were not falsepositive acceleration times but false-negative anglograms, a recognized limitation of single-plane angiography. In our clinical practice the preangiographic finding of an abnormal acceleration time modifies the performance of the arterial study. Single-plane pelvic arteriography is supplemented by oblique views or direct intraluminal arterial pressure measurements with or without vasodilator drugs to resolve the significance of mild/moderate arterial stenoses when CFA acceleration times are prolonged. Five sides had either 75% or 80% iliac stenosis but did not have a prolonged acceleration time. Four of these five false-negative results had short-segment stenoses in the proximal common lilac artery (Figs. 4 and 5). All five stenoses were associated with resting pressure gradients made worse by peripheral vasodilation and were corrected by balloon catheter angioplasty. Therefore a normal acceleration time should not be used to exclude the presence of a significant iliac lesion. Rather, an acceleration time less than 144 msec may indicate that any significant lilac lesion would likely be amenable to angioplasty. Review of the eight errors revealed only one tracing (Fig. 6) that would, in retrospect, have been measured differently. If the third waveform, rather than the second, had been chosen for analysis, the acceleration time would have been abnormal, and the lilac stenosis would have been detected. This study supports the work of Thiele et al.,13 who have noted that, depending on the degree of stenosis, velocity patterns may resume a normal contour within three vessel diameters downstream from the stenosis. Therefore the presence of net flow reversal in a biphasic or triphasic waveform does not rule out the presence of significant proximal disease. However, our study suggests that some', of these lesions may be detected by a prolonged systolic rise time (Fig. 7). In 1976 Nicolaides et al.10 recognized that loss of reverse flow in the CFA waveform could be the result of either isolated inflow or outflow disease. They noted that a femoral waveform characterized by a rapid ascent, an initial rapid descent, and a prolonged late descent was found in patients with SFA obstruction. This finding was supported in our study (Fig. 2). Furthermore, in our study these changes were present in patients with SFA obstruction, whereas
Doppler spectralpattern of iliac stenoses 449
Fig. 5. Iliac angiogram of patient with short-segment 75% stenosis of proximal left common lilac artery-. the prolongation of the ascent of the waveform was specLfic for iliac disease. The receiver-operator characteristic curve for different criteria for a positive test result is displayed in Fig. 8. Although the point representing 144 msec is favorable, the points representing shorter acceleration times are also favorable; tile proximity of these points probably relates to the small proportion of sides with intermediate grades of lilac stenosis. However, an acceleration time of 133 msec should alert the interpreter to scrutinize the waveform and the clinician to consider more than single-plane pelvic angiograms, Fig. 9 suggests that more than hak¢ of these sides will be associated with severe iliac disease, and at 144 msec nearly all are associated with severe disease.
In clinical practice there are a number of circum stances in which acceleration time is useful in patient management. The detection and treatment of lilac artery stenosis is particularly important when a single iliac artery is expected to provide adequate arterial
450
Journal of VASCULAR SURGERY
Burnham,Jaques, and Burnham 5lIl'~lO
100 msec
III CI I'_l
m'_
• I
'
I
:
~|~~"~"" 4 "
L ,
~" " "
:.
C¢.':.¢_:
v
. . .~.-,e¢
Ic
',,~
[J
'?
Fig. 6. Incorrect placement of second cursor incorrectly suggesting absence of iliac disease.
,144 msec t!
_''°" j ::--~
ti "-I
Fig. 7. Tracing shows net flow reversal in triphasic waveform with prolonged acceleration time. Iliac angiogram shows severe iliac stenosis.
inflow to both lower extremities (femoral crossover donor vessel). An abnormal acceleration time may be the critical indicator that the donor side inflow is jeopardized. When angioplasty o f an iliac stenosis is carried out proximal to a SFA obstruction, ankle blood pressure may not show prompt increase because of the distal disease; however, improvement in acceleration time is prompt and reassuring.
REFERENCES
1. van Asten WNIC, Beijneveld WJ, Pieters BR, van Lier HIJ, Wijn PFF, Skomicki SH. Assessment of aortoiliac obstructive
2.
3.
4.
5.
disease by Doppler spectrum analysis of blood flow velocities in the common femoral artery at rest and during reactive hyperemia. Surgery 1991;109:633-9. Flanigan DP, Williams LR, Schwartz •A, Schuler 5~, Gray B. Hemodynamic evaluation of the aortoiliac system based on pharmacologic vasodilatation. Surgery 1983;93:70%14. Langsfeld M, Nepute J, Hershey FB, et al. The use of deep duplex scanning to predict hemodynamically significant aortoiliac stenoses. • VAsc SURG 1988;7:395-9. Haenen JH, van Asten WN~C, Theloosen-Kerstens AWP, Skomicki SH. The value of duplex scanning in assessing aorto-iliac disease. J Vasc Technol 1991;15:231-4. Thiele BL, Bandyk DF, Zierler RE, Strandness DE Jr. A systematic approach to the assessment of aortoiliac disease. Arch Surg 1983;118:477-81.
Volume 16 Number 3 September 1992
Doppler spectrgl patten~ of lilac stenoses 451
100 111 - 133 msec
90 100 msec 80
144 mse-~c
70
°m
60
t."
cO 50 40
156 msec
30
.. 70
I
|
80
90
100
Specificity Fig. 8. Receiver-operator characteristic curve for different criteria for positive test result.
Per cent 100
80 60 40 20 o 56 67 78 89 1 0 0 1 1 1 1 2 2 1 3 3 1 4 4 1 5 6 1 6 7 1 7 8 1 8 9 2 0 0 2 1 1 2 2 2 Acceleration Times in milliseconds
Fig. 9. Proportion o f each acceleration time's values wkh significant lilac disease.
6. Clifford PC, Slddmore R, Bird D, Woodcock JP, Lusby RJ, Baird RN. Femoral artery doppler signal analysis in lower limb ischemia. J Cardiovasc Surg 1982;23:6974. 7. Johnston KW, Kassam M, Koers J, Cobbold RSC, MacHattie D. Comparative study of four methods for quantifying Doppler ttltrasound waveforms from the femoral artery. Ultrasound Med Biol 1984;10:1-12. 8. Archie IP lr, Feltman RW. Determination of the hemodynamic significance of iliac artery stenosis by noninvasive Doppler ultrasonography. Surgery 1982;91:41%24. 9. Harward TRS, Bernstein EF, Fronek A. The value of power frequency spectrum analysis in the identification of aortoiliac artery disease. J Vasc SURG 1987;5:803-13. 10. Nicolaides AN, Gordon Smith IC, Dayandas J, Eastcott HH. The value of Doppler blood velocity tracings in the detection
of aortoiliac disease in patients with intermittent claudication. Surgery 1976;80:774-8. 11. Kupper CA, Young L, Keagy BA, Burnham S]. Spectra1 analysis of the femoral artery for identification of iliac lesio~so Bruit 1984;8:157-63. 12. Burnham CK, Dewhirst N, Bumhan~ SJ. Doppler spectral waveforms for recording peripheral arterial signals: the preferred method. J Vase Technol 1989;8:69-73. 13. Thiele BL, Hutchinson KJ, Greene FM, Forster FK, Strandness DE lr. Pulsed Doppler waveform patterns produced by smooth stenosis in the dog thoracic aorta. In: Taylor DEM, Stevens AL, eds. Blood flow: theory and practice. Orlando: Academic Press, 19834:85-104. Submitted Jan. 29, t992; accepted )rune 2, 1992.
452
Burnham, Jaques, and Burnham
Journal of VASCULAR SURGERY
DISCUSSION
Dr. Robert W. Barnes (Little Roclq Ark.). I agree that the so-called acceleration rime between onset and peak systole of the femoral arterial Doppler frequency waveform is one of the most valid indicators of the status of the iliac artery, particularly in the presence of superficial femoral artery occlusion. One must understand that these measurements require a Doppler sound spectrum analyzer. They cannot be obtained from simple analog tracings on a strip chart recorder. Accurate measurements are readily available on most duplex scanners in current use. You have correlated the acceleration time with contrast arteriography. Unfortunately, arteriograms may belie the hemodynamic significance of an arterial stenosis, a fact that you invoke to explain your three false-positive Doppler study results, two of which were accompanied by so-called significant pressure gradients. Unforttmately, you do not define the threshold of significant resting or hyperemic pressure gradients. Furthermore, all five false-negative Doppler study results were accompanied by such significant pressure gradients and a greater than 75% stenosis on arteriography. I remain concerned about the following three points: First, it would be helpful if you could provide us with correlations between the Doppler studies and aortofemoral resting or hyperemic pressure gradients obtained at the time of arteriography for all patients because such physiologic data may be a better gold standard than arteriography for validating acceleration times. Second, have you considered evaluating acceleration times during some form of reactive hyperemia, which may lead to better correlation with pressure gradients ? Finally, will you reiterate the value of this test in routine clinical situations, particularly to convince those cynics who might point out that most of us obtain arteriograms and, when needed, aortofemoral pressure gradients for patients requiring interventions for peripheral arterial occlusive disease. Dr. Steven J. Burnham. Since beginning this analysis, we have measured some pressure gradients in isolated patients. I do not have those data with me. I think that this
test offers to our patients, before hospitalization, an opporumity to understand which segment is likely to show disease and also offers to our radiology colleagues an additional tool to help plan the angiography. I think that for those clinicians who do routine biplane angiography of the lilac vessels, this test has very little value. For those who do single-plane angiography and are sometimes undecided about the angiographic significance, I think that this piece of information may represent helpful additional informarion. Dr. Timothy R. S. H a r w a r d (Gainesville, Fla.). As you demonstrated in one of your slides, in the same patient the tracings varied with each cardiac cycle. If you picked the wrong tracing, you get the wrong diagnosis. Have you looked at the reproducibility of this particular test? I would agree with you, as I showed in my own work, the arteriogram is not the gold standard. I tended to think of the noninvasive test being positive and the arteriogram as being a false-posirive or false-negarive examination. Also, I agree that some form of stress testing or pressure measurement is nice to know when making the final interpretation. The last question I have pertains to a problem that I had when I looked at this subject: the effect of a deep femoral artery stenosis on the common femoral artery Doppler tracing. I had two patients with occlusion of the superficial femoral artery with a stenosis at the origin of the deep femoral artery that was significant. This hemodynamic disease led to a false-positive examination because it acted the same as a proximal upstream stenosis. Have yon seen this in your own work? Dr. Burnham. I do not have an answer about internal consistency with Doppler angle. The angiogram is not the gold standard if you use only a single plane. The deep femoral artery was not analyzed separately; however, I can tell you that no patients in this group had total occlusion of the deep femoral artery, and a small minority of parients had stenosis of the deep femoral artery.
