Summary of main results
The available randomised trials contain different inclusion criteria, different outcome measures and relatively small numbers, which makes meaningful comparison very difficult.
Only one study looked at the use of intravenous rt-PA (Berridge 1991). This treatment was statistically significantly poorer in terms of 30 day and three month asymptomatic limb salvage rates and carried a significantly higher risk of haemorrhagic complications than intra-arterial rt-PA or intra-arterial streptokinase. In the same study, intra-arterial rt-PA offered significantly greater initially successful lysis (P < 0.04) and a significantly lower amputation rate than intra-arterial streptokinase. At 30 days, intra-arterial rt-PA gave a 20% increase in limb salvage over streptokinase, although this was not statistically significant.
Three studies compared intra-arterial urokinase to intra-arterial rt-PA. The study of Meyerovitz 1990 reported that intra-arterial rt-PA tended to cause more rapid initial lysis but clinical outcomes of limb loss, death and amputation were similar in both groups. Schweizer 1996 also demonstrated a significant reduction in treatment duration with intra-arterial rt-PA. Schweizer 1996 found that the initial recanalization rates were greater with intra-arterial rt-PA (P < 0.05) and this difference was sustained at six months. The study by Mahler 2001 was complicated by the use of two different techniques, end hole catheter versus microporous balloon infusion, in addition to urokinase and rt-PA. In the group treated by end hole catheter infusion the authors quote that rt-PA appeared to work more rapidly than urokinase, however it is difficult to extract the data from the paper that support this statement. The rates of limb loss or partial amputation were not significantly different between the groups.
A wide range of clinical presentations is seen within the trials reviewed, extending from patients with intermittent claudication to critical limb ischaemia; for example the study of Schweizer 1996 included a significant proportion of patients with claudication. Thrombolysis carries with it a small risk of bleeding complications including cerebral haemorrhage. Currently, in the United Kingdom, the risk to benefit ratio of thrombolysis means that fibrinolytic agents are not frequently employed for patients with claudication. It is questionable how representative the results, for example of Schweizer 1996, are for a population of patients with critical limb ischaemia.
The incidence of haemorrhagic complications with rt-PA varies significantly between the studies and this almost certainly reflects the very different fibrinolytic regimes employed. Berridge 1991 had the most conservative rt-PA regime (intra-arterial rt-PA 0.5 mg/hr) and found a lower haemorrhagic complication rate with intra-arterial rt-PA than with either intra-arterial streptokinase or intravenous rt-PA. Notably there was no significant difference in the duration of therapy between rt-PA or streptokinase. Meyerovitz 1990 used a 10 mg intrathrombic bolus followed by an infusion of 5 mg/hr. Major and minor haemorrhagic complications were more prevalent in the rt-PA group, however the difference was not statistically significant. Initial lysis was faster with rt-PA but this was not sustained at 24 hours. Schweizer 1996 used an intrathrombic 5 mg bolus with 5 mg/hr infusion and found no major haemorrhagic complications, however 15% of rt-PA patients had large groin haematomas compared with 8% in the urokinase group. Notably, this study incorporated a post-thrombolysis anticoagulation regime with heparin (20,000 IU/24 hrs for five days).
In terms of more novel fibrinolytic agents, limited information from phase II and phase III trials are available. Only one study described a comparison between intra-arterial urokinase and intra-arterial pro-urokinase (Ouriel 1999). Pro-urokinase has improved fibrin specificity and may therefore reduce the risk of bleeding complications related to thrombolysis. This trial was a phase II study designed to determine the optimum dose regimen for intra-arterial pro-urokinase and contained several different dose regimens for pro-urokinase versus a standard dose regimen for urokinase. A dose response curve for pro-urokinase was defined but there was no statistically significant difference between the highest dose of intra-arterial prourokinase and intra-arterial urokinase in terms of vessel patency, amputation rate, deaths or complication rate.
Results from the phase II and phase III trials with alfimperase have now been published and include safety and efficacy data for alfimperase and surgery free survival to 30 days against placebo (Han 2010). There was no significant difference on 30 day freedom from open vascular surgery between intrathrombus placebo and intrathrombus alfimperase. These trials have not been included in this review as they do not compare thrombolytic agents as described in the selection of types of studies. However, It should be noted that a significant improvement was demonstrated in the placebo group, possibly due to a mechanical effect of thrombolysis, and future studies may now need to consider a placebo control arm in any study design.
A number of randomised trials have compared the use of adjunctive agents such as glycoprotein IIb/IIIa receptor inhibitors with existing thrombolytic agents. Duda (Duda 2001) compared urokinase with and without abciximab and in a later trial Tepe (Tepe 2006) compared urokinase with abciximab against rt-PA with abxicimab. These trials have not been considered to represent a comparison of thrombolytic agents and are considered within a separate review examining infusion techniques (Kessel 2004).
Overall completeness and applicability of evidence
Overall, the evidence from these studies is very limited. The analysis suggests that intra-arterial rt-PA may be more effective than intra-arterial streptokinase or intravenous rt-PA but this is based on small numbers of studies. There is no evidence that intra-arterial rt-PA is more effective than intra-arterial urokinase in patients with critical limb ischaemia. Initial lysis may be more rapid with rt-PA depending on the regime. The incidence of haemorrhagic complications is not statistically significantly greater with intra-arterial rt-PA than with other equipotent regimes. The available evidence is insufficient to show any benefit in reduction of bleeding complications from the use of pro-urokinase.