AJR 2000; 175:513-516
© American Roentgen Ray Society
Distal Arterial Flow in Patients Undergoing Upper Extremity Dialysis Shunting
A Prospective Study Using Doppler Sonography
M. Goldfeld1,
B. Koifman2,
N. Loberant1,
I. Krowll3 and
M. Haj4
1
Department of Radiology, Western Galilee Hospital, Meona St., Nahariya 22100,
Israel.
2
Department of Internal Medicine "B," Western Galilee Hospital,
Nahariya, Israel.
3
Department of Quality Assurance, Western Galilee Hospital, Nahariya,
Israel.
4
Department of General Surgery, Western Galilee Hospital, Nahariya,
Israel.
Received May 4, 1999;
accepted after revision January 4, 2000.
Address correspondence to M. Goldfeld.
Abstract
OBJECTIVE. The objective of this study was to document changes in
the distal circulation after creation of a proximal upper extremity dialysis
shunt and to correlate these findings with the patient's clinical
condition.
SUBJECTS AND METHODS. We prospectively examined 18 patients
scheduled for upper extremity shunt creation. We used color and spectral
Doppler sonography to examine flow in the radial and ulnar arteries, noting
flow direction and peak systolic velocity. After the shunt procedure, we
repeated the measurements and correlated them statistically with hand
symptomatology.
RESULTS. Six (33%) of 18 patients were symptomatic. The mean peak
systolic velocities in the radial and ulnar arteries were 52 and 61 cm/sec,
respectively, before surgery, and decreased to 12 cm/sec after surgery in the
radial artery and 44 cm/sec in the ulnar artery. The mean percentage of
decrease in peak systolic velocity was 77% in the radial artery and 28% in the
ulnar artery. Eight patients showed reversed flow. No statistical correlation
was found between change in peak systolic velocity values before and after
surgery and the presence of hand symptoms. Similarly, no correlation was found
between flow reversal and symptoms. The most consistent factor associated with
symptoms was diabetes; all symptomatic patients were diabetic, but only 54% of
the diabetic patients were symptomatic.
CONCLUSION. The difference in the peak systolic velocities in the
radial and ulnar arteries after shunt construction does not correlate with
symptoms. The hand can tolerate a significant decrease in the peak systolic
velocity and even flow reversal without symptomatology.
Introduction
Hand pain and sensorimotor disability after the creation of a vascular
access for hemodialysis may be caused by vascular insufficiency; venous
hypertension; diabetic, uremic, or ischemic neuropathy; carpal tunnel
syndrome; secondary hyperthyroidism; and embolic complications
[1,
2]. The differential diagnosis
may be challenging. Doppler sonography is a noninvasive method for hemodynamic
studies. We were interested in determining whether sonography could permit a
confident diagnosis in these patients.
Subjects and Methods
Eighteen patients who were scheduled for upper extremity hemodialysis shunt
surgery underwent baseline real-time gray-scale Doppler, color Doppler, and
spectral Doppler sonography 1 month before surgery. All patients were
reexamined 3-4 weeks after surgery. Half the patients were examined once again
6 months after the surgery, and the results of the two examinations were
averaged.
Examinations were performed using a 7-MHz linear transducer (128XP; Acuson,
Mountain View, CA) and a 5-MHz linear transducer (VST Master Series;
Diasonics, Santa Clara, CA). Examinations were performed with the patient
sitting and the hands on a pillow in supination. In addition, blood pressure
measurements were recorded before each sonographic examination.
We examined 10 men and eight women who were 24-69 years old. The mean and
median ages were both 58 years.
In addition to chronic renal failure, the patients also suffered from
hypertension (n = 16), diabetes mellitus (n = 11),
polycystic kidney (n = 1), and nephrolithiasis (n = 1). All
hypertensive patients were treated with 
- or ß-blockers.
Only patients with proximal arteriovenous fistulas were included in the
study. Sixteen patients underwent shunt placement in the left arm; two
patients underwent placement of a right-sided shunt. The types of shunts
placed are listed in Table
1.
Duplex Doppler studies were performed in the longitudinal orientation with
an insonation angle always smaller than 60°. The angle was kept relatively
constant for each vessel throughout the study. From each tracing, the peak
systolic velocity, Doppler waveform, and direction of the flow were recorded.
The radial and ulnar arteries of both hands were studied. The contralateral
arm served as control, and the peak systolic velocity obtained in the arm with
the shunt was normalized according to the values obtained in the control
side.
Statistical analysis included bivariate correlation (Kendall's tau). Values
of p equal to or less than 0.05 were considered statistically
significant.
Results
Six (33%) of 18 patients had hand symptoms: pain, coldness, numbness, color
change, or hand weakness (Table
2). All complaints started within 6 weeks of surgery, continued
during the 6 months of follow-up, and were treated conservatively. One
patient, who had presented with weakness of the hand, complained of severe
hand pain immediately after surgery; because of this, his shunt was closed
surgically.
The mean peak systolic velocities in the radial and ulnar arteries were 52
cm/sec and 61 cm/sec, respectively, before the shunt construction. After
surgery, the mean velocity decreased to 12 cm/sec in the radial artery and 44
cm/sec in the ulnar artery (normalized values) (Figs.
1 and
2). The mean percentage of
decrease was 77% in the radial artery and 28% in the ulnar artery
(Fig. 3).
View larger version (23K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —Graph shows peak systolic velocity in radial artery for each patient
before (dotted line) and after (dashed line) shunt
construction. Note that most values decrease after surgery. Negative values
indicate reversed flow.
|
|
View larger version (28K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2. —Graph shows peak systolic velocity in ulnar artery for each patient
before (dotted line) and after (dashed line) shunt
construction. Note that most values decrease after surgery. Negative values
indicate reversed flow.
|
|
View larger version (26K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3. —Graph shows percentage of change that occurs in peak systolic
velocity in ulnar (dotted line) and radial (dashed line)
arteries after shunt construction. Note that most values are less than zero,
indicating a decrease in velocity.
|
|
The greater decrease in the peak systolic velocity in the radial artery is
influenced by the higher prevalence of reversed flow in this artery. If we
exclude the cases of reversed flow, the mean velocities in the radial and
ulnar arteries after shunt construction were 47 cm/sec and 43 cm/sec,
respectively (normalized values). Two symptomatic patients had a peak systolic
velocity of less than 22 cm/sec in both the radial and ulnar arteries; in one
of these patients ulnar artery flow was 7 cm/sec and radial artery flow was
-23 cm/sec (reversed flow), and this patient required surgical closure of the
shunt because of hand weakness.
We did not find a statistical correlation between either the absolute peak
systolic velocity after the surgery or the magnitude of the decrease in peak
systolic velocity after surgery and the occurrence of hand symptoms. We also
calculated the numeric average of the peak velocity of the radial and ulnar
arteries in each examination and found no correlation between this average and
the development of hand symptoms.
In eight patients (44%) we observed reversed flow, in six during the entire
cardiac phase and in two only during diastole (Figs.
4A,4B,5,6).
In six patients the reversed flow was seen through the radial artery and in
two through the ulnar artery. Only one symptomatic patient had reversed flow.
This patient had hand weakness, and his shunt was surgically closed. The
occurrence of reversed flow in our study did not correlate statistically with
peripheral hand symptoms. We observed changes in the spectral Doppler
waveform, including increased length of systolic phase and continuous high
diastolic flow in nine patients (eight with reversed flow as already noted).
No correlation was seen between changes in the waveform of the spectral
Doppler signal and hand symptomatology.
View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —Development of reversed flow in 48-year-old asymptomatic man with
diabetes mellitus and hypertension who was taking -blockers. Doppler
sonogram of left radial artery obtained before shunt creation shows flow in
distal direction.
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B. —Development of reversed flow in 48-year-old asymptomatic man with
diabetes mellitus and hypertension who was taking -blockers. Doppler
sonogram obtained after shunt creation shows reversed flow through radial
artery. Note that waveform is above baseline because it was inverted
technically. Note also high systolic flow and continuous diastolic flow.
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5. —Doppler sonogram of left radial artery obtained after shunt creation
in 57-year-old asymptomatic woman with hypertension who was undergoing
-blocker treatment shows antegrade flow during systole and reversed
flow during diastole.
|
|
View larger version (97K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6. —Color Doppler sonogram obtained after shunt creation in 55-year-old
asymptomatic man with hypertension who was undergoing -blocker
treatment shows reversed flow on both color and spectral Doppler images.
|
|
All six patients with hand symptoms were diabetic, but just 54% of the
diabetic patients were symptomatic. Of the eight patients with reversed flow,
three were diabetic.
Blood pressure measurements did not change significantly after shunt
placement. The average change was 2.5 mm Hg systolic (median, 2.2 ± 5.0
mm Hg) and 0 mm Hg diastolic (median, -1.3 ± 5.5 mm Hg).
Discussion
Pain and sensorimotor complaints in the extremities are not uncommon after
dialysis shunt construction, and may be due to arterial insufficiency,
neuropathy (ischemic, diabetic, or uremic), carpal tunnel syndrome, secondary
hyperparathyroidism, or embolism
[1,
2]. We wanted to determine
whether the vascular complications might be predicted using Doppler sonography
in the pre- and postoperative periods.
Construction of hemodialysis fistulas diminishes the peripheral blood flow.
This iatrogenic decrease in the peripheral blood supply may be superimposed on
preexisting vascular disease, which is common in diabetic and hypertensive
patients and which may impair tissue perfusion. Haimov et al.
[3] and Zerbino et al.
[4] reported 1.6% and 2.5%
incidence, respectively, of symptomatic vascular insufficiency after shunt
creation.
The occurrence of reversed flow is a well-known phenomenon. Duncan et al.
[5] found 88% of reversed flow
in their study on patients with radiocephalic fistulas and concluded that this
phenomenon is not important as a cause of vascular insufficiency. In a series
of 14 patients, Valji et al.
[1] found six cases of reversed
flow. They found that digital ischemia was caused by underlying obstructive
arterial disease alone or in combination with a steal phenomenon.
In our study, the timing of the appearance of symptoms indicates that
vascular disturbance is the most probable cause of hand symptoms among the
possible diseases that occur in these patients. However, neither baseline peak
systolic velocity nor postoperative reduction in peak systolic velocity could
be statistically correlated with symptoms, and thus were not of predictive
value. This indicates the great tolerance of the hand for changes in
perfusion.
Reversed flow was a frequent phenomenon (44%) and was not statistically
correlated with peripheral symptoms.
All symptomatic patients were diabetic. No statistically significant
correlation exists between diabetes and hand symptoms; however, the
conditional probability of a diabetic patient to suffer from short-term hand
complications of a hemodialysis shunt is much greater than for a nondiabetic
patient. This finding is in accordance with the results published by Schanzer
et al. [6]. This greater risk
may be explained by preexisting peripheral vascular disease or by superimposed
diabetic neuropathy.
Our study included only two patients with a postoperative peak systolic
velocity of less than 22 cm/sec in both the radial and ulnar arteries, and
both patients were symptomatic. One of them also exhibited hand weakness and
reversed flow and required surgical closure of his shunt. We suggest that 22
cm/sec might be a limit below which ischemic symptoms are likely to occur. A
study of a larger population would be necessary to test this hypothesis.
References
-
Valji K, Hye RJ, Roberts AC, Oglevie SB, Ziegler T, Bookstein JJ.
Hand ischemia in patients with hemodialysis access grafts: angiographic
diagnosis and treatment. Radiology
1995;196:697
-701[Abstract/Free Full Text]
-
Lacombe M. Digital arterial embolism after thrombosis of
arteriovenous fistula [in French]. Ann Cardiol Angeiol
(Paris) 1993;42:199
-202[Medline]
-
Haimov M, Baez A, Neff M, Slifkin R. Complications of arteriovenous
fistulas for hemodialysis. Arch Surg
1975;110:708
-712[Abstract/Free Full Text]
-
Zerbino VR, Tice DA, Katz LA, Nidus BD. A 6-year clinical
experience with arteriovenous fistula and by-passes for hemodialysis.
Surgery
1974;76:1018
-1023[Medline]
-
Duncan H, Ferguson L, Faris I. Incidence of the radial steal
syndrome in patients with Brescia fistula for hemodialysis: its clinical
significance. J Vasc Surg
1986;10:144
-147
-
Schanzer H, Skladany M, Haimov M. Treatment of angioaccess induced
ischemia by revascularization. J Vasc Surg
1992;16:864
-866
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
