| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Racine Radiologist Group, 3803 Spring St., Rm. 208, Racine, WI 53405.
2
Department of Radiology, MC 2026, The University of Chicago Hospitals, 5841 S.
Maryland, Chicago, IL 60637.
Received April 6, 2000;
accepted after revision December 7, 2000.
Address correspondence to B. Funaki.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Between May 1993 and May 1997, 19 metallic stents were placed in 11 patients (seven women, four men; age range, 41-83 years) to treat elastic intragraft stenoses or graft dissections. Before stent placement, all patients had experienced multiple episodes of graft thrombosis, had very limited vascular access for hemodialysis, and were considered poor surgical candidates.
RESULTS. The technical success rate was 100%, and there were no procedural complications. Using life-table analysis, we found primary patency to be 36% at 6 months after stent placement, 12% at 12 months, and 12% at 18 months. Secondary patency was 91% at 6 months after stent placement, 71% at 12 months, and 47% at 18 months. The mean and median patencies per intervention were 4.2 and 3.6 months, respectively. Mean and median secondary graft patencies were both 14 months (range, 3 days-32 months). Puncture through the stents occurred during dialysis, causing stent distortion and fracture. Eight stents had a linear fracture suggesting compression contributed to the stent distortion. No clinically evident complications related to stent placement occurred.
CONCLUSION. Metallic stent deployment can salvage access in synthetic arteriovenous grafts by alleviating intragraft stenoses. Patency of intragraft stents is similar to venous stents used to treat other hemodialysis-related stenoses; however, fracture of Wall-stents occurs with prolonged graft use, especially in areas of needle punctures.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Nineteen Wallstents were depolyed in either the arterial limb (n = 6) or venous limb (n = 13) of the graft to maintain flow. Three patients had Wallstents placed in both the arterial and venous limbs of the graft. In five patients, two stents were required to completely cover the stenotic portion of graft and were considered as a single stent in patency calculations. The initial eight stents were placed in five patients with grafts unsuitable for surgical revision and who, in the opinion of the referring surgeons, had no remaining viable vascular access sites. In the subsequent six patients, surgical options were limited, and stents were placed after discussion with the surgical service. In the first six patients, a total of nine 10 x 42 mm Wallstents were used because smaller diameter stents were not available at that time. When smaller diameter Wallstents became available, 8 x 20 or 8 x 40 mm diameter Wallstents were deployed, depending on the length of the stenosis. Stents were chosen to closely approximate the length of the lesion. We preferred to place shorter stents to allow graft puncture around the stent and limit eventual stent distortion. Longer stents would limit the length of graft available for puncture. Puncture through the stent was generally avoided during dialysis because the segment was not as pulsatile as the remainder of the graft. However, punctures did occur. We did not mark the skin with tattoos [4]. In many patients, such marking would have significantly limited the length of graft available for vascular access.
Procedure
The skin was prepared in standard fashion with povidone-iodine scrub
(Clinidine; Climpad, Rocky Hill, CT). Radiologists were fully masked and
gowned for the entire procedure and observed standard scrub protocol.
Suspected graft infection (i.e., warmth, erythema, and tenderness of graft
site) or systemic infection was considered an absolute contraindication to
stent placement; although in this series, no patients had suspected graft or
systemic infection. Prophylactic use of 1 gm of IV ceftizoxime sodium
(Cefizox; Fujisawa, Deerfield, IL) was used in the most recent four patients
at the request of the clinical service.
Balloon angioplasty of intragraft stenoses was performed using a 7 x 40 mm high-pressure balloon catheter (Blue Max; Boston Scientific) inflated to a maximum of 25 atm (2.5 x 105 Pa) for 10-20 sec. If the waist of the stenosis did not resolve with a highpressure angioplasty balloon catheter, stenting was not performed. After angioplasty, digital subtraction angiography was performed to assess the results of dilatation. An unequivocal elastic stenosis was defined as luminal narrowing exceeding 50% after dilatation and was identified in four patients. In the seven patients with equivocal findings of stenosis, pressure measurements were taken across the stenosis, and a focal gradient of 10 mm Hg or greater was considered significant. A graft dissection was diagnosed when the stenosis had an associated linear flap revealed on angiography that appeared similar to an intimal dissection in an artery.
When a stenosis was identified, a Wallstent was deployed across the stenosis and dilated using a 7 x 40 mm balloon catheter as previously described. After stent placement, digital subtraction angiography was performed. Technical success was defined as complete resolution of the stenosis or a residual pressure gradient of less than 5 mm Hg. This outcome was achieved in all patients. In the patients with initially equivocal findings, angiograms obtained showed the stenoses to be completely resolved after stent placement. However, pressure measurements were repeated confirming the angiographic results in these grafts because a pressure catheter was already available. Pressure catheters were not routinely used in patients with obvious stenoses because of cost considerations.
The stenotic area within the graft had been incidentally punctured for the thrombolysis procedure in two patients. These patients returned within several days of the initial procedure, and a puncture away from the stenotic area was performed to reevaluate the intragraft lesion and to assess the need for placement of a stent. Stent restenosis was treated with repeated angioplasty. Stents were not deployed within preexisting Wallstents. Additional thrombolysis was not performed if the graft rethrombosed within 30 days of initial thrombolysis and intrafraft stent placement.
Follow-Up
Data on all patients were collected by the surgical and nephrology services
at the workup for the initial procedure and when patients returned for
venographic follow-up because of recurrent graft dysfunction. Venograms were
routinely scheduled when graft dysfunction occurred. At the time of data
analysis, eight patients were alive. The three remaining patients had died,
but their grafts had been functioning at the time of their deaths, which were
caused by factors unrelated to the grafts: motor vehicle collision,
tuberculous meningitis, and malignancy. Kaplan-Meier life-table analysis was
performed. Primary stent patency was defined as the period from stent
deployment until any intervention—either angioplasty or stent within the
graft or the venous outflow—or graft thrombolysis was performed.
Secondary patency was defined as the period from stent deployment until the
graft site was no longer used, regardless of the number of interventions.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
Distortion of stents occurred frequently. If the graft functioned well, the site was not abandoned because of stent distortion alone. However, if thrombosis occurred more than once during any 30-day period, and no alternative explanation (such as hypotension) could explain the recurrent thrombosis, the graft was abandoned. Generally, at 3 months, little stent distortion was evident (Fig. 3). However, progressive distortion of the stents did occur (Figs. 4 and 5). After 1 year, stent distortion was very severe in three patients (Fig. 6) and less severe in four patients (Fig. 7). Only two patients showed minimal or no distortion. Stents located at the puncture site had distortion attributable to multiple needle punctures, whereas stents farther from the puncture sites had distortion most likely caused by compression fatigue and stent fracture. Eight stents had a linear fracture suggesting compression contributed to the stent distortion rather than puncture. Two patients had separation of the intragraft matrix from the synthetic graft material akin to intimal dissection in arteries that could not be resolved with simple angioplasty and required intragraft stent placement. A review of available chest radiographs in five patients, including the three patients with the most severe stent deformities, showed no evidence of metallic fragment embolization. No clinically evident complications related to stent placement occurred.
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During the study period, we performed approximately 600 declotting procedures, and intragraft stenoses occurred in approximately 7% of the procedures. These lesions usually responded well to simple angioplasty. During our study, we placed 19 stents in 11 patients. Thus, stents were required in less than 2% of thrombolysis procedures. In patients that have no surgical options, placement of a stent within the graft will maintain graft patency.
The Dialysis Quality Outcomes Initiative Vascular Access Guidelines of the National Kidney Foundation [3] recommend treatment of a stenosis of 50% or greater. In our small series, we treated seven patients that had stenoses of less than 50% that were visible on one view of an angiogram because these lesions appeared to cause early rethrombosis; no other identifiable cause was found. We routinely measured gradients across the graft and venous outflow in any patient with early rethrombosis and equivocal stenoses. In the patients in this series, two equivocal lesions had gradients of greater than 30 mm Hg. Because many stenoses are asymmetric, a mild stenosis in one plane may still be important hemodynamically. Although pressure gradients occur normally from the arterial to the venous side of the graft [5], a focal gradient or change in the waveform typically indicates a stenosis. [6].
Cost is a major consideration in modern medicine. In patients who are surgical candidates for revision, the cost and durability of thrombolysis and stent deployment must be compared with surgical thrombectomy and revision because both alternatives restore graft function. One approach would compare the cost of surgical thrombectomy and revision to radiologic percutaneous thrombolysis; the estimates of the surgical costs range from approximately equal to or $6000 more than the radiologic procedure, depending on the reference [7, 8].
However, assuming equal costs, the practical situation should be considered. In many hospitals, percutaneous treatment of thrombosed hemodialysis grafts is the first-line treatment. If during thrombolysis, a flow-limiting intragraft stenosis is identified and is not adequately treated, the graft will rethrombose. In this situation, the decision to place an intragraft stent must be compared with the total cost of surgical thrombectomy, graft revision, and temporary catheter. Therefore, the practical cost of not placing an intragraft stent placement can be high. The hospital cost of a metallic stent is $1000-$1200, nearly identical to the Medicare reimbursement for the stent. Our local Medicare reimbursement for the procedural and interpretation components of vascular stent placement procedure is between $490 (nonparticipating fee schedule) and $515 (participating fee schedule).
Despite the added cost of performing a surgical thrombectomy and revision procedure after a radiologic declotting procedure, a surgical revision could still be preferable to intragraft stenting if the patency achieved by the surgical procedure would be much greater than the patency of the intragraft stent placed via radiologic procedure. The graft segment that is surgically revised does not deteriorate as rapidly as a stented portion of graft, and one would expect an overall longer patency. However, this benefit may not always be obtained.
Overall graft life is influenced by other factors and events, such as the occlusion of the outflow vein or venous anastomosis and subsequent loss of vascular access before the deterioration of the surgically replaced segment. Thus, mean patency after surgical graft revision is approximately 140 days or 4.7 months [9], which is similar to our mean patency between interventions of 4.2 months after intragraft stent placement. Furthermore, intragraft stenting was completed without the increased rate of infection and venous thrombosis inherent to central venous catheters [3, 10], which may be required before and sometimes after surgical graft revision. Overall, the use of intragraft Wallstents increased graft life by a mean of 1.2 years in grafts that already had a mean age of 3.3 years at the time of stent placement (Fig. 2).
Multiple prior graft revisions is a contraindication to subsequent graft revision. All grafts have a limited life expectancy. Patients do run out of access sites and must begin catheter hemodialysis with its greatly increased risk of lifethreatening infection—approximately 10% of patients with bacteremia die [11]. Percutaneous catheters are a very poor way to receive hemodialysis [3], and intragraft stent placement helps delay this inevitability.
Previous animal studies have shown no short-term complications or serious distortion in stents subjected to repeated puncture [12], although a recent case report did suggest stent deformity caused by needle puncture [4]. In our series, short-term distortion was rare; long-term distortion was more common. Stent distortion was rarely seen before 3 months but was definitely present in grafts functioning 1 year after stent placement. Even with the most severely distorted stents, no embolization of metallic fragments was observed. Although no fragment embolization was identified, we cannot assume that such embolization does not or could not occur. More important, no clinically important episodes of hemoptysis, chest pain, recurrent infections, or other signs of sequela of metallic fragment embolization occurred.
Stents placed in areas of needle puncture (Fig. 6) showed more distortion than stents placed in the periphery of the graft in areas away from the sites of frequent puncture (Fig. 5). Factors other than direct puncture, such as stent compression, appear to affect stent durability. We have observed lateral longitudinally oriented fractures in Wallstents that suggest compression of the stent during palpation actually caused the fractures rather than repeated needle punctures (Fig. 7). Stent distortion and fracture probably limited long-term patency in some patients. Puncture of the stent should be avoided if possible. We anticipate that stents made of nitinol alloy will be more puncture-and compression-resistant. Marking the graft to avoid palpation as well as needle puncture of the area is recommended.
Extensive restenosis did not occur in or near the stent. Significant intimal hyperplasia was not observed for intragraft Wallstents, unlike Wallstents deployed to treat venous stenoses [13]. Intuitively, one would not expect intimal hyperplasia to occur because there is no true intimal layer in grafts. In addition, the adherent matrix in a polytetrafluoroethylene graft is less metabolically active than the intima at the venous anastomosis [14] and thus less predisposed to react to the metallic stent to form a stenosis. In a surgical series consisting of 52 excised polytetrafluoroethylene grafts, a nonviable matrix was often identified in the grafts [15]. Thus, restenosis attributable to intimal hyperplasia was not a limiting factor. In our series, the stented portion of the graft usually had recurrent stenosis caused by Wallstent fragmentation and collapse, not by intimal hyperplasia.
In addition to elastic stenoses, we observed that the intragraft matrix separated from the graft itself in a fashion similar to an intimal dissection causing an obstructive flap [15]. Angioplasty may "tack" this stenosis down, but occasionally intragraft stents were required to fully reduce this narrowing and any associated gradient. We have found that metallic stents can be safely deployed within synthetic dialysis grafts with a stenosis unresponsive to angioplasty, but the Wallstent is likely not the ideal stent. A stent made of nitinol may be a better alternative. A stent made of nitinol may be more resistant to distortion due to direct puncture. Stents with wider interstices will also decrease the chance that a needle puncture will contact the stent and distort it.
We initially used long Wallstents because these were the only available stents. We now use a stent (Symphony; Boston Scientific) with wide interstices [16] and nitinol metal construction. A stent made of a continuous piece of alloy with wide interstices is even more desirable (the Symphony stent has solder points that may weaken the stent). Unfortunately, there is no stent available that meets all these criteria.
In conclusion, stent deployment will alleviate intragraft stenosis in patients with synthetic arteriovenous hemodialysis grafts who are poor candidates for surgery, and thus the stent can prolong graft life. Stents should be used judiciously because distortion and fracture occur with prolonged graft use, limiting secondary patency in some patients. For patients who are good candidates for surgery, the treatment of choice is likely to be bypass of the graft segment. In patients who are poor candidates for surgery, short-segment intragraft stent deployment is an acceptable, perhaps even preferable, technique to salvage graft function.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
=
primary patency. 
= secondary patency.
