Cardiovasc Intervent Radiol (1997) 20:369–376

CardioVascular and Interventional Radiology q Springer-Verlag New York Inc. 1997

Laboratory Investigation Local Intraarterial Thrombolysis: In Vitro Comparison of Various Infusion Catheters Jens J. Froelich,1 Martin Hoppe,1 Christina Freymann,1 Thomas Thiel,1 Hans-Joachim Wagner,1 Klemens H. Barth,2 Klaus J. Klose1 1

Medizinisches Zentrum fu¨r Radiologie, Abteilung fu¨r Strahlendiagnostik, Klinikum der Philipps-Universita¨t, D-35043 Marburg, Germany Division of Vascular and Interventional Radiology, Georgetown University Hospital, 3800 Reservoir Road, N.W., Washington, DC 20007, USA

2

Abstract Purpose: Catheters are compared in vitro to evaluate the efficacy of thrombolysis during urokinase infusion within the thrombus. Methods: Six catheters were introduced individually into human thrombus within a stenotic flow model. Urokinase was infused continuously into the thrombus. To quantify the efficacy of thrombolysis, pressure gradients were recorded proximal and distal to the thrombus and during the course of infusion. Uniformity of lysis was assessed radiographically. Results: The fastest and most homogeneous thrombolysis was achieved with the EDM and the straight-flush catheter, shown by decreasing transthrombotic pressure gradients. All other catheter designs showed less homogeneous and delayed thrombolysis (p ° 0.001, Friedmann-Test, Schaich-Hamerle). There was no significant difference in the efficacy of thrombus dissolution between the EDM and the straight-flush catheter (Wilcoxon, matched pairs, p ú 0.7). Conclusion: The EDM catheter and the straight flush catheter achieved the most homogeneous and fastest thrombolysis, apparently due to the best urokinase distribution within the thrombus. Key words: Catheters and catherization—Thrombolysis—Urokinase—Flow model—Infusion technique— Thrombus

Direct infusion of fibrinolytic agents into an intravascular thrombus results in higher success rates and less complications than systemic administration of thromCorrespondence to: J.J. Froelich, M.D.

bolytic drugs [1, 2]. Persistent problems of local intraarterial thrombolysis include significant failure rates, prolonged infusion times, hemorrhagic complications, early rethrombosis, and distal embolism [3–6]. Many efforts have been made to facilitate and improve the process of local intraarterial thrombolysis; among them are infusion catheter designs with different shapes and arrangements of side holes, with or without end-hole occlusion, and with single or multiple lumens. Forced intrathrombus injections, like pulsed-spray pharmacomechanical thrombolysis [7, 8], resulted in accelerated lysis in patients, confirming experimental results [9]. Even though various catheter designs are in widespread clinical use, little is known about differences in their effectiveness. We developed an in vitro flow model containing human thrombus within a stenotic graft segment to realistically compare different catheters for speed and homogeneity of thrombolysis during continuous infusion of urokinase. Materials and Methods Simulated Graft Thrombosis/Flow Model A human thrombus was generated within a knitted Dacron velour vascular prosthesis (DVP) with a length of 135 mm and a diameter of 10 mm. The DVP with an integrated stenosis was consequently introduced into a flow model with limited collateral flow around the thrombosed graft. Proximal and distal sideports within the flow models were used to measure pressure gradients across the occluded graft during thrombolysis. The model was filled with isotonic saline solution containing 100 IU heparin/L before being positioned on a fluoroscopic table to monitor progress of thrombolysis. A more detailed description of our model has previously been published [9].

Catheters and Infusion Regimen Catheters Six different catheters were compared; their descriptions, characteristics, and features are summarized in Table 1. All catheterization

4k10 0017 Mp 369 Wednesday Aug 06 04:11 PM SV-CVIR (v. 20, #4) 0017 (2367)

370

J.J. Froelich et al.: Catheter Comparison in Thrombolysis

Table 1. Description and features of the evaluated catheters Catheter (manufacturer)

Caliber and length

No. of sideholes

5 Fr/90 cm

Infusion (Cook, Bloomington, IN, USA)

3 Fr/100 cm

0

—

N

Jet-lysis (Angiomed, Karlsruhe, Germany)

5 Fr/90 cm

40

10 cm

Y

5 Fr/65 cm

20

10 cm

Y

18

9 cm

Y

4

9 cm

Y

Katzen wire (Boston Scientific Inc. Watertown, MA, USA) EDMa (Mallinckrodt, Hennef/Sieg, Germany) a

0.035 in./145 cm 4.7 Fr/90 cm

4 cm

End-occluded

Straight-flush High-flow (Angiomed, Karlsruhe, Germany)

Mewissen (Boston Scientific Inc., Watertown, MA, USA)

6

Infusion length

N

Multilumen catheter with four infusion channels, each communicating with a single sidehole

was performed through the model’s proximal sideport in an antegrade flow direction. The catheters were reproducibly centered within the thrombus with the aid of a thin caliber centering wire which was removed after catheter placement. The tips of multilevel infusion catheters were placed 10 mm proximal to the stenotic graft segment within the thrombus. The straight-flush and the infusion catheter were placed with their tips in the midportion of the thrombus. No catheter was advanced or repositioned during thrombolysis.

Infusion Regimen Thrombolysis was exclusively performed by continuous infusion of urokinase. Urokinase 50,000 IU in 4.5 ml H2O was mixed with 5000 IU heparin dissolved in 0.5 ml H2O and infused over 5 min, followed by 50,000 IU urokinase in 19.5 ml H2O mixed with 5000 IU heparin dissolved in 0.5 ml H2O for 275 min. Each experiment was repeated three times.

Measurements and Evaluation The progress of thrombolysis, including distribution of infusate, was monitored fluoroscopically and documented on radiographs every 15 min after contrast test injection through the sideport of the introducer sheath. System pressures were evaluated proximal and distal to the thrombus/stenosis (pre- and postthrombotic). The preset system pressures, as well as the flow rates within the model, changed during thrombolysis due to increasing antegrade flow through the partially lysed thrombotic occlusion. The mean pressure gradients and standard deviations for the three experiments were calculated (Figs. 1, 2). Further statistical evaluation of mean transthrombotic pressure gradients as an expression of increasing thrombolysis was performed using the Friedman test (multiple comparisons; Table 2) and the Schaich-Hamerle comparison for the different catheters.

Results The preset system pressures, as well as the flow rates within the model, changed due to increasing antegrade flow during thrombolysis. Preocclusion pressure decreased from a maximum of 58 mmHg to 35 mmHg and postocclusion pressure increased from 12 mmHg to a maximum of 25 mmHg as an effect of lysis. Flow rates increased from 500 ml/hr up to 1000 ml/hr at the end of thrombolysis. Due to the integrated stenosis, a pressure gradient of 10 mmHg (35 mmHg proximally

to 25 mmHg distally) remained across the stenosis when no thrombus was inserted into the flow model (baseline pressure gradient). None of the catheters achieved complete lysis of the thrombus, leaving residual amounts of thrombus within the DVP. Continuous infusion of urokinase resulted in inhomogeneous dissolution of the thrombus with only slow progress of lysis. The fastest thrombus dissolution and fastest thrombolysis was achieved with the EDM and with the straight-flush catheter (Friedman test, p ° 0.001; Figs. 1A, F, 2, 3A, F, Table 2). There was no significant difference with respect to thrombolysis between those two catheters (Wilcoxon test, matched pairs, p ú 0.7). Variations among the three experiments with each catheter were high for most catheters, expressed by high standard deviations (Fig. 2). However, the observed differences between catheters were significant or even highly significant with respect to the mean transthrombotic pressure gradients using the Friedmann test (multiple comparisons) and the Schaich-Hamerle comparison (p ° 0.001) (Figs. 1, 2, Table 2). Highly significant differences (p ° 0.001) were found between the following catheters (the catheter achieving faster thrombus dissolution is listed first): Straight-flush vs endhole catheter, straight-flush vs jet-lysis catheter, straight-flush vs Mewissen catheter, and straight-flush vs the Katzen wire. Highly significant differences between transthrombotic pressure gradients were also found between endhole and jet-lysis catheters, endhole and EDM catheters, Mewissen catheter and Katzen-wire, Mewissen and EDM catheters, jet-lysis catheter and Katzen wire, and jet-lysis and EDM catheters. Still significant (p ° 0.05) were differences between the straight-flush and jet-lysis catheters, whereas no significant difference was found (p ¢ 0.05) between the straight-flush and EDM catheters, endhole and Mewissen catheters, endhole catheter and Katzen wire, and jet-lysis and Mewissen catheters (Table 2). The largest amount of residual thrombotic material within the DVP was left behind by the Katzen wire and the infusion catheter (Fig. 1B, E; Fig. 2; Fig. 3B, E;

4k10 0017 Mp 370 Wednesday Aug 06 04:11 PM SV-CVIR (v. 20, #4) 0017 (2367)

J.J. Froelich et al.: Catheter Comparison in Thrombolysis

371

Fig. 1. Development of pre- and postthrombotic pressures during thrombolysis over time. A Straight-flush catheter. B Infusion catheter. C Jetlysis catheter. D Endoccluded Mewissen catheter. E Katzen infusion wire. F EDM catheter.

4k10 0017 Mp 371 Wednesday Aug 06 04:11 PM SV-CVIR (v. 20, #4) 0017 (2367)

372

J.J. Froelich et al.: Catheter Comparison in Thrombolysis

Fig. 1. Continued.

4k10 0017 Mp 372 Wednesday Aug 06 04:11 PM SV-CVIR (v. 20, #4) 0017 (2367)

J.J. Froelich et al.: Catheter Comparison in Thrombolysis

373

Fig. 2. Statistic evaluation of transthrombotic pressure gradients (D pre-/postthrombotic pressure) with means and standard deviations for the six different catheters.

Table 2. Statistic evaluation of transstenotic pressure gradients: Comparison of different catheters for significance (Friedman test) Straight-flush Straight-flush Endhole Jet-lysis Mewissen Katzen wire EDM-Pro

** * ** ** #

Endhole

Jet-lysis

Mewissen

Katzen wire

EDM-Pro

**

* **

** # #

** # ** **

# ** ** ** **

** # # **

# ** **

** **

**

Significance levels: * p ° 0.05; ** p ° 0.001; # p ¢ 0.05.

Table 2). As stated before, there is a highly significant difference between the EDM and the straight-flush catheter, characterized by the fastest and most complete thrombolysis. The mean number of distal emboli collected and the standard deviations for the different catheters are demonstrated in Figure 4. No statistical difference was found among the various catheters.

Discussion Local infusion of thrombolytic agents into an intravascular thrombus has resulted in higher success rates and fewer complications compared with their systemic administration [1, 2, 10]. Originally described by Dotter et al. in 1974 [11], a straight endhole catheter was introduced into the proximal portion of a thrombus and a fibrinolytic agent was infused continuously. As thrombolysis progressed, the tip of the infusion catheter was advanced stepwise to maintain delivery of the thrombolytic agent directly into the thrombus [1]. Many efforts were made to reduce catheter manipulations and to accelerate thrombolysis, resulting in a variety of infusion catheter designs with different calibers, single or multiple lumens, and arrangement of

sideholes (with or without endhole occlusion). Another innovation was the development of so-called pulsespray pharmacomechanical thrombolysis, suggested by Bookstein et al. [7, 8]. This technique injects a fibrinolytic agent by high-pressure jets to effectively induce clot disruption and to accelerate thrombolysis by enlarging the interface between thrombus and the fibrinolytic agent [7, 9]. Even though various catheter designs and infusion protocols are used in clinical practice, little has been published about their performance under controlled comparison. Our model was designed to compare six different catheters for continuous infusion of urokinase. In vivo conditions such as pressure gradients and collateral flow were simulated in the model to achieve close clinical correlation. We were able to demonstrate distinct differences in thrombolysis efficiency and homogeneity of clot dissolution among the six catheters, with results favoring the EDM and straight-flush catheter. All other catheters showed significantly less favorable results. Multilevel infusion catheters were originally designed to reduce catheter manipulations during local thrombolysis [12, 13]. However, under clinical conditions, intravascular thrombus does not consist of ho-

4k10 0017 Mp 373 Wednesday Aug 06 04:11 PM SV-CVIR (v. 20, #4) 0017 (2367)

374

J.J. Froelich et al.: Catheter Comparison in Thrombolysis

Fig. 3. A Radiograph of the thrombus (a) 15 min and (b) 280 min after continuous infusion of urokinase over the straight-flush catheter. Note the lytic area at the catheter tip after 15 min (a) and the distinct progress of thrombolysis after 280 min (b). B Radiograph of the thrombus (a) 15 min and (b) 280 min after continuous infusion of urokinase over the infusion catheter. A helical lytic tract has formed within the thrombus after 15 min (a) slightly enlarging after 280 min (b). C Radiograph of the thrombus (a) 15 min and (b) 280 min after continuous infusion of urokinase over the jet-lysis catheter. Almost no lytic activity is present after 15 min (a) but thrombolysis is apparent around the catheter shaft after 280 min (b). D Radiograph of the thrombus (a) 15 min and (b) 280 min after continuous infusion of urokinase over the end-occluded Mewissen catheter. A small lytic tract has formed around the catheter shaft after 15 min (a) slightly enlarging after 280 min (b). E Radiograph of the thrombus (a) 15 min and (b) 280 min after continuous infusion of urokinase over the Katzen wire. A lytic area has formed at the catheter tip after 15 min (a) without significant progress after 280 min (b). F Radiograph of the thrombus (a) 15 min and (b) 280 min after continuous infusion of urokinase over the EDM catheter. Note the lytic area around the catheter shaft after 15 min (a) and the distinct progress of thrombolysis after 280 min (b).

mogeneous material. With continuous infusion of a thrombolytic agent into such inhomogeneous material, its distribution will likely be inhomogeneous as the infusate will follow the path of least resistance. Pharmacomechanical thrombolysis [7, 8] offers a very elegant solution to this problem by overcoming the differences in thrombus resistance through forceful injection of short pulses of highly concentrated urokinase directly into the thrombus, thereby increasing the contact area between clot and lytic agent [9]. However, as demonstrated in our study, homogeneous thrombus dissolution may also be achieved with continuous throm-

bolysis when using certain catheter designs. The simply constructed straight-flush catheter (one through lumen with six large sideholes at the distal catheter end) and the relatively complex EDM catheter design (multilumen catheter with a sideport perfusing four independent infusion channels, each communicating with a single sidehole) seem to equalize sidehole resistance differences [13], providing more homogeneous distribution of the infused lytic agent within the thrombus. As far as generation of distal emboli is concerned, our study did not show statistically significant differences among the six catheters. This indicates that

4k10 0017 Mp 374 Wednesday Aug 06 04:11 PM SV-CVIR (v. 20, #4) 0017 (2367)

J.J. Froelich et al.: Catheter Comparison in Thrombolysis

375

Fig. 3. Continued.

use of efficient thrombolysis catheters is not associated with an increased risk of peripheral embolization. We conclude that continuous intraarterial infusion of lytic agents should be performed with either the EDM catheter or the straight-flush catheter in order to achieve the most homogenous and fastest thrombolysis, since all other catheter designs show significantly less favorable results. Because no significant difference was found between the EDM and the straight-flush catheter, cost factors could be decisive in the choice between the two catheters.

References 1. Matsumoto AH, Barth KH, Teitelbaum GP, Kremers PW (1990) Self contained, coaxial delivery system for intraarterial infusion of thrombolytic agents. Radiology 174:1056 – 1057 2. Katzen BT, van Breda A (1981) Low-dose streptokinase in the treatment of arterial occlusions. AJR 136:1171–1178 3. Mori KW, Bookstein JJ, Heeney DJ (1983) Selective streptokinase infusion: Clinical and laboratory correlates. Radiology 148:677–682 4. McNamara TO, Fischer JR (1985) Thrombolysis of peripheral arterial and graft occlusions: Improved results using high-dose urokinase. AJR 144:769–775

4k10 0017 Mp 375 Wednesday Aug 06 04:11 PM SV-CVIR (v. 20, #4) 0017 (2367)

376

J.J. Froelich et al.: Catheter Comparison in Thrombolysis

Fig. 4. Mean number of emboli and standard deviations for the six different infusion catheters. No statistical difference was found.

5. Hess H, Ingrisch H, Mietaschk A, Rath H (1982) Local low-dose thrombolytic therapy of peripheral arterial occlusions. N Engl J Med 307:1627–1630 6. van Breda A, Katzen BT, Scales FE (1987) Streptokinase vs urokinase in local thrombolytic therapy. Radiology 165:109– 111 7. Bookstein JJ, Fellmeth B, Roberts A, Valji K, Davis G, Machado T (1989) Pulsed-spray pharmacomechanical thrombolysis: Preliminary clinical results. AJR 152:1097–1100 8. Bookstein JJ, Valji K (1992) Pulse-spray pharmacomechanical thrombolysis. Cardiovasc Intervent Radiol 15:228–233 9. Froelich JJ, Freymann C, Hoppe M, Thiel T, Wagner HJ, Barth KH, Klose KJ (1996) Local intraarterial thrombolysis: In vitro

10. 11. 12. 13.

comparison between automatic and manual pulse-spray infusion. Cardiovasc Intervent Radiol 19:423–427 Gardiner GA Jr, Koltun W, Kandarpa K (1986) Thrombolysis of occluded femoropopliteal grafts. AJR 147:621 – 626 Dotter CT, Rosch J, Seaman AJ (1974) Selective clot lysis with low-dose streptokinase. Radiology 111:31–35 Hicks XE, Picus D, Darcy MD, Kleinhoffer MA (1991) Multilevel infusion catheters for use with thrombolytic agents. J Vasc Interv Radiol 2:73–75 Kaufman SL, Martin LG, Gilarsky BP, Finnegan MF (1991) Urokinase thrombolysis using a multiple side hole multilumen infusion catheter. Cardiovasc Intervent Radiol 14:334–337

4k10 0017 Mp 376 Wednesday Aug 06 04:11 PM SV-CVIR (v. 20, #4) 0017 (2367)