GUIDELINES [ back to index ]
16. Management of graft thrombosis
*1 Postpone dialysis whenever possible
• As in patients with native A/V fistulas, postponement of dialysis may be possible in patients who are not expected to develop pulmonary oedema. Increase of body weight since the last dialysis session will indicate the amount of excess fluid retained. Furthermore, the patients must not show symptoms of uraemia, and the potassium level must not exceed 5.6 mmol/l.
*2 Treat graft thrombosis before next dialysis session
• Graft thrombosis should be treated without unnecessary delay and within 48 hours whenever possible. Early declotting allows for immediate use of the access without the need for central venous access. There is always a compact "arterial plug" present. Mature thrombi older than five days often are fixed to the vessel wall beyond the venous anastomosis, making surgical extraction more difficult. This is less of a problem for the interventional radiologist.
• Central venous catheters should be avoided whenever possible.
*3 Surgical thrombectomy
• Surgical thrombectomy is performed with a embolectomy catheter. A longitudinal incision at the venous anastomosis can be performed with the option of patch plasty. Alternatively an oblique incision of the graft within a reasonable distance of the venous anastomosis can be performed when intraoperative balloon dilatation of the suspected stenosis or surgical graft extension is planned. On-table completion angiography of the arterial and venous limbs of the graft as well as the central venous outflow is mandatory to exclude persistent thrombi and define the cause of thrombosis if it is not venous outlet stenosis. Identification and simultaneous correction of the underlying stenosis are integral parts of any surgical or interventional declotting procedure 1 2 (see Algorithm “Management of Graft Stenosis” *3, *4, *5).
*4 Interventional thrombolysis
• PTFE graft thrombosis is commonly treated using a range of percutaneous techniques including combinations of thromboaspiration, use of thrombolytic agents such as tissue plasminogen activator (tPA), mechanical thrombectomy and mechanical thrombectomy devices. In one study, success rate was 73 %, with primary patency rates of only 32 % and 26 % at one and three months, respectively 3.
• Comparing different mechanical devices for percutaneous thrombolysis, Smits et al. concluded, that “the treatment of the underlying stenoses was the only predictive value for graft patency” 4, which holds truth, as long as the thrombus is actually removed, although not true for all the techniques used.
• Surgery or interventional radiology? Each centre should choose the technique according to their expertise. Independent of the applied technique it is important to perform:
a) Thrombolysis or thrombectomy rapidly (within 48 hrs) to avoid the need for a temporary catheter.
b) Thrombolysis or thrombectomy as an outpatient procedure to decrease costs, whenever possible.
c) Post-procedural angiography to detect and correct inflow, intra-access or outflow venous stenosis 5. The latter is present in 85 % of thrombosed grafts.
d) Post-procedural documentation of residual stenosis and access blood flow.
• Quality indicators of successful thrombolysis / thrombectomy are 6.
a) Immediate patency (access usable for next haemodialysis treatment) in 85 % of the cases.
b) Unassisted patency at three months of at least 40 %.
• Dougherty et al. randomly assigned 80 patients to either surgical thrombectomy with patch plasty or interposition of a graft or to thrombolytic therapy with percutaneous transluminal angioplasty if indicated. No difference in outcome was seen, but thrombolytic therapy with PTA was more expensive than the surgical intervention 1. This study, however, has been criticised for its design (unblinded), for lacking sufficient information regarding the endovascular techniques used, and for incomplete cost analysis 7 .
• Marston et al. compared surgical (n = 56) versus endovascular (n = 59) management of thrombosed grafts. While immediate success rates were high in both groups (83 % and 72 %, respectively), patency rates were significantly higher with surgical intervention than with endovascular intervention (36 % versus 11 % after 6 months) 2. I t was pointed out, however, that the 31 % initial failure rate of the percutaneous approach was especially high in this study when compared to the close to 100 % success rates of many other series. This was possibly due to the lack of preoperative imaging before access creation, explaining the high rate of long outflow stenoses unmasked after declotting 8.
• In contrast, Turmel-Rodrigues et al. found higher patency rates after radiological intervention, with a 6 month primary patency rate of 32 % in thrombosed grafts 9.
• While the DOQI guidelines determined a threshold of 40 % for 1-year primary patency rate after surgical treatment of thrombosed grafts 6, none of the three most recent series reached that goal since rates ranged from 23 to 26 % 1 2 10.
• Bakran and McWilliams in an invited commentary on the radiology versus surgery debate pointed out the dearth of good quality of randomised controlled trials dealing with this subject. In some series both surgery and angioplasty had poor outcome 2 11, whilst others had better outcome for angioplasty 12, and yet others had good results from surgery 13. Their conclusion was that individual centres should audit their own results and choose the modality that produces the best results for that centre. Inevitably, in some centres this will mean angioplasty but in others surgery 14.
• In their recent meta analysis of randomised, controlled trials, Green et al. compared the results of surgical thrombectomy, mechanical thrombectomy and pharamacomechanical thrombolysis for thrombosed dialysis grafts 15. They found a clear superiority of surgery over endovascular procedures in terms of technical success and patency rates. No differences in complication rates between the groups were demonstrated.
*5 Assess for other causes of thrombosis
• The vast majority of access thromboses are due to graft stenosis (see Algorithm “Management of Graft Stenosis”). Other causes such as post-dialysis hypotension, excessive dehydration, hypercoagulability, trauma, or prolonged compression of puncture site are associated risk factors or triggers that often reveal an underlying stenosis.
• Infection can also cause or be associated with access thrombosis.
*6 Recurring graft thrombosis
• Recurring graft thrombosis can be repeatedly treated by interventional radiology. Mansilla found similar re-occlusion rates after the first, second and third radiological treatment 16. In cases of frequent or early re-stenosis of the venous graft anastomosis, stent implantation can be considered. Stents do not preclude re-stenosis, but slight over-dilating the stent (not more than 1 or 2 mm, see Appendix 4.1.1.2 PTA) may postpone clinical appearance of re-stenosis 17.
• Surgery (graft extension) should be considered when stenosis- or thrombosis-free intervals gradually become shorter, or when the stenotic segment gradually becomes longer.
• While old, clotted grafts are a potential source of silent infection 16, they should be removed only in cases of otherwise unexplained septicaemia or obvious sepsis.