- Email: [email protected]
Reduced thrombus formation in vivo after administration of pentoxifylline (TrentalR)
THROMBOSIS RESEARCH 56; 359-368, 1989 0049-3848/89 $3.00 + .OO Printed in the USA. Copyright (c) 1989 Pergamon Press plc. All rights reserved.
REDUCED THROMBUS FORMATION IN VIVO AFTERR ADMINISTRATION OF PENTOXIFYLLINE (TRENTAL )
M. Michal*, N. Giessinger* and R. SchrGer+ * BATTELLE RESEARCH CENTRES Centre for Toxicology and Biosciences CH-1227 Carouge-Geneva, Switzerland ' HOECHST AG WERK ALBERT, Clinical Research D-6200 Wiesbaden 12, West-Germany (Received 8.8.1988; accepted in revised form 8.8.1989 by Editor M.B. Donati)
ABSTRACT
The hamster cheek pouch model of experimental thrombosis in which thrombi are induced in the microvasculature by iontophoretical administration of adenosine diphosphate (ADP) was uked to test the antithrombotic potential of pentoxifylline (Trental ). Single intraperitoneal injections of 5, 10 and 20 mg/kg pentoxifylline reduced thrombus formation by 20 to 50 % from 30 to 105 min following drug administration. The effect of a single application of 10 mg/kg was exceeded significantly (p < 0.05) by the higher rate of inhibition after repeated injections of the same dose given three times daily. This suggests a residual antithrombotic effect from the preceeding administrations.
INTRODUCTION Acute ischemic events such as myocardial infarction, transient ischemic attacks and strokes are caused by thrombus formation in arteries and/or ensueing thromboembolic complications (1). The identified events leading to thrombus formation are alterations in flow conditions (2), alterations of the vessel wall which interfere with its natural antithrombotic properties (3, 4) and activation of the haemostatic system, especially platelet activation (5). The rationale that a drug modifying platelet function might be effective in preventing clinical manifestations of thrombosis has prompted considerable research activity in recent years to find suitable antiplatelet drugs which potentially prevent thrombus formation in vivo. Reliable and valid in vivo models are therefore needed for identifying and evaluating potential antithrombotic drugs.
Key words: thrombosis, pentoxifylline, antithrombotic effect 359
PENTOXIFYLLINE AND THROMBOSIS
360
Vol. 56, No. 3
The haclscercneek pouch technique (6) offers a reliable and quantitative ti m r,lodel for testing potential antithrombotic drugs. Under in vivo haemodynamic conditions, this model closely simulates spontaneous thrombus formati0.lin the presence of all formed elements of flowing blood and the vessel wdll. Experimental thrombus formation is initially triggered by the iontopnore-cicapplication of ADP onto a microvessel. ADP diffused into the microvessel causes platelet adhesion and aggregation which in turn probably causes the release of intraplatelet material leading to the formation of a throclbus.kith this techniaue it is possible to mimic the pertinent physioevent - the formation of ah intravascular platelet plug or "white logica body" ( 7) under physiological haemodynamic conditions. Pentox fylline a drug well established for the treatment of peripheral vascular disease has been shown besides its beneficial effects on disturbed bloau f'luw (8) to reverse platelet hyperreactivity and hypercoagulability iil pa-ten.cs(9, 10). Furthermore, pentoxifylline has been reported to inhibit in vivo experimental thrombosis in stenosed femoral arteries of rab11) and laser-induced thrombus formation in vivo in mesenteric artebies ries in rats (12, 13). Therefore, it was of particular interest_to investi._ . gaze the antithrombotic effects of pentoxifylline on thrombus formation u vita using the hamster cheek pouch model.
(’
MATERIALS AND METHODS ?reoaration of the hamster cheek pouch The microcirculation of the hamster cheek pouch was studied by the method of Duling, Berne and Born (14). Male Syrian-golden hamsters, weighing 95-110 g, were anesthetized with pentobarbitone (60 mg/kg body weight) injacted intraperitoneally. Supplementary doses of anaesthetic were given when needed. The cheek pouch was everted, spread out and fixed to a special perspex stage equipped with fixing barbs. The top layer and fine connective tissue wsre carefully removed under a stereomicroscope, leaving the bottom layer of the pouch suitable for microscopic observation using a fixed stage microscooe (Leitz Laborlux) transilluminated with liaht from below. The mounted 'cheek was rinsed continuously at a fixed rate w th a physiological buffer solution at 37°C. Such preparations can be mainta ned in good conditio,l, as assessed by blood flow and reactivity of blood vessels and by abseuce of edema, for approximately 2 hours. inauction of thrombus formation Platelet thrornbiwere induced in venules (30-40 urnin diameter) by the iontophoretic application of ADP to the blood vessel wall (6). Glass electrodes (Clark Electromedical Instruments) were drawn out to form micropipettes (tip diameter 2-3 urn)with a pipette puller (Industrial Science AssDciates Inc.). A micropipette prefilled with ADP (10 mM in distilled water) was manipulated using a Leitz micromanipulator so that the tip was in contact with the venule. The ADP solution in the micropipette was connected to an electrical circuit and the reference electrode completing the circuit was immersed in the superfusing physiological buffer solution. Adenosine aipnosphate was applied to the vessel wall by passing a current between micropipette and reference electrode. From a battery power supply (36 V) a negative potential controlled by a modified digital voltmeter was app'lied to the micropipette; the resultant current of 300 nAMP was
Vol. 56, No. 3
PENTOXIFYLLINE AND THROMBOSIS
361
-14 calculated to eject ADP in the order of 2x10 mol/sec. Typically, after approximately 5 set a thrombus began to form on the inner side of the vessel wall, near the pipette tip. Once the current was switched off after 60 set, the thrombus embolized within approximately 10 sec. The same vessel can be used several times and thus can provide its own control. Rate of thrombus formation was based on size of thrombus formed at 30, 45 and 60 set from beginning of the application of ADP. Quantification of thrombus formation The formation of the experimental thrombus following the application of ADP to a venule was recorded using a Sony video camera (DXC 1800P) and a Sony video recorder (VO-5850P). The video tape was played back and the image frozen at various intervals during the progressive growth of the thrombus. The image was traced on the screen and analyzed using an image analyzer (MOP-AMOS, Carl Zeiss). The visualized surface area of the thrombus increased linearly with time. Therefore a plot of surface area against time provided the rate of growth of the thrombus. Administration of drug Pentoxifylling (10 mg/ml) stock solution was prepared in NaCl 0.9 % and stored at -18 C. Samples of this solution were thawed when required. Hamsters weighing 100 g were injected i.p. in volumes not exceeding 200 ul of pentoxifylline stock solution corresponding to the doses of 5, 10 and 20 mg/kg, respectively. For repeated administrations, a stock solution of pentoxifylline (5 mg/ml) was used. Three groups of eight hamsters each (weight 100 g) were injected 1, 4 and 3 times, respectively, with 200 ul/kg of pentoxifylline stock solution (5 mg/ml) which corresponds to 10 mg/kg. Repeated injections were made daily at 8h30, 12h30 and 16h30. Induction of thrombus formation was started 5 min after injection of the last dose in each case, i.e. in the group receiving 4 administrations, after the morning dose of day two and, in the group receiving 13 administrations, after the morning dose of day four. Injections of pentoxifylline were replaced by saline in the control group of hamsters. Statistical analysis of experimental data Data are expressed as mean + standard error of n observations. Experimental data of the pilot investigation were analyzed descriptively comparing values obtained 5-105 min following the injection of each dose of drug and those obtained in animals injected with saline. Statistical evaluation of experimental data of the second series (4 groups of 8 animals) was carried out on comparison of values obtained 5-90 min following the last injection of pentoxifylline in each of the 3 different drug regimens (1, 4 and 13 administrations) and control animals injected with saline. Statistical evaluation was performed using Student's t-test by means of an appropriate computer program on a Hewlett-Packard lOOOF. Materials Physiological buffer solution: composition (mM) Nahl 131,KCl 4.0, CaC12 3.0, MgSO 1.0 and Tris 5.0, pH 7.4, maintained at 37 C. Anaesthetfc: sodium pentobarbitone (Nembutal, Abbott). Thrombus inducer: adenosine diphosphate sodium salt (Sigma Chemicals). Testdrug: pentoxifylline (HOECHST AG Werk Albert, Wiesbaden).
362
PENTOXIFYLLINE AN0 THROMBOSIS
Vol. 56, No. 3
RESULTS Reoroducibilitv of thrombus formation The rate of thrombus formation was first established in the absence of pentoxifylline based on the visualized size of thrombus attained 15, 30, 45 and 60 set following the application of ADP. The data showed the rate of thrombus formation in control animals to be reproducible from one induction of thrombus to another at 15 min intervals over at least 90 min. Single administration of oentoxifvlline
In a pilot investigation with 3-4 experiments per group, three doses of pentoxifylline (5, 10 and 20 mg/kg) were tested after intraperitoneal injection of the drug. From 5-90 min after application, thrombus size was evaluated at 15, 30, 45 and 60 set of iontophoretic ADP application. As an example Figure 1 shows thrombus formation at different times after pentoxifylline 5 mg/kg i.p.
m Control 0 = 5min A = I5min ?? = 30min A = 45 min 0 = 90
15
30
45
x
??
0
??
min
75 min 60 min
60 set
TIME FROM START OF ADP IONTOPHORESIS FIG. 1
Thrombus formation in vivo at different times after intraperitoneal injection of 5 mg/kg pen-line. Visualized thrombus size was evaluated at various times of ADP iontophoresis before (control .) and 5 (o), 15 (A), 30 (a), 45 (A), 69 to), 75 (x) and 90 (0) min after injection of pentoxifylline. Values represent means of 3-4 experiments.
Vol. 56, No. 3
PENTOXIFYLLINE AND THROMBOSIS
363
From the data of visualized thrombus size, percent inhibition was calculated by comparing thrombus size before (control) and after drug application. In all cases the inhibitory effects were similar at 30, 45 and 60 set of ADP iontophoresis. Therefore, the experimental data obtained at 45 set were taken as representative for comparing the inhibitory effects obtained at different times and after the different doses tested. A summary graph showing the overall inhibitory action of the different doses of pentoxifylline on thrombus formation is given in Figure 2.
0
I
I
I
I
I
I
15
30
45
60
75
90
105
TIME AFTER ADMINISTRATION OF PENTOXIFYLLINE ( MlN 1
FIG. 2 Inhibition of thrombus formation after single intraperitoneal administration of 5 (o), 10 (0) and 20 (A) mg/kg pentoxifylline. Results are expressed as percent inhibition of thrombus formation at 45 set of ADP iontophoresis obtained by comparing values before (control) and after administration of pentoxifylline. Values represent mean + standard error of 3-4 experiments. Pentoxifylline in all three doses inhibited thrombus formation in vivo showing a similar time dependence, but only a weak dose dependence. While the inhibition after the dose of 5 mg/kg increased from 5 % at 15 min to 40 % maximal inhibition at 60-75 min after drug administration, 20 mg/kg resulted in only 10 % higher inhibition over the entire test period with a maximal inhibition of 50 % at the same time as the lower dose, In contrast, the results after 10 mg/kg differ quantitatively from this frame due to unknown reasons. - In all cases, the inhibitory effect on thrombus formation produced by pentoxifylline started to decrease after 75 min following drug administration but lasted for the full observation period.
PENTOXIFYLLINE AND THROMBOSIS
364
Repeated
administration
Vol. 56, No. 3
of oentoxifvlline
In the pharmacological dose range of 5-20 mg/kg, there was obviously no strong dose dependence thus suggesting no stronger inhibition with higher doses. Therefore a second series of experiments was undertaken to investigate the effect of repeated injections of pentoxifylline. Three regimens of 1, 4, or 13 administrations of 10 mg/kg pentoxifylline i.p. (see under "Administration") were investigated for their effect on thrombus formation jr_~ m. Percent inhibition was obtained by comparing visualized thrombus size at 45 set of ADP iontophoresis in control animals injected with saline and those treated with 10 mg/kg pentoxifylline and in case of the latter two regimens pretreated with the same dose three times daily for 1 or 3 days. Figure 3 is a summary graph showing the overall inhibitory effects of the three different regimens.
5 60 G r a 50 c '3 g 40 -
x
& F b
30 -
z 20 I= 5
-
F z t- loC!! zi no
I 0
15
I
I
1
30 45 60 75 TIME AFTER LAST ADMINISTRATION OF PENTOXIFYLLINE (MIN)
I 90
105
FIG. 3 Inhibition of thrombus formation after 1 (o), 4 (0) and 13 (A) intraperitoneal administrations of 10 mg/kg pentoxifylline. Percent inhibition was calculated by comparing size of thrombus attained 45 set following iontophoretic application of ADP in the absence of drug (control) and after 1, 4 and 13 administrations of pentoxifylline. Values represent mean + standard error. * p < 0.05 vs. 1 administration
t p < 0.05 vs. 4 administrations
Thrombus inhibition observed after 1, 4 or 13 administrations progressed with time reaching a maximum at 60-75 min following last administration of the drug. Inhibition obtained after a single application of 10 mg/kg was in agreement with the results after single application of 5 and 20 mg/kg, re-
Vol. 56, No. 3
PENTOXIFYLLINE AND THROMBOSIS
365
spectively. The maximal degree of inhibition (45.0 %) after a single injection of pentoxifylline (10 mg/kg) occurred at the same time (75 min) but was exceeded by that following 4 or 13 administrations producing 52.7 % and 52.5 % inhibition, respectively. The differences between single and repeated administrations were statistically significant (p < 0.05) with two exceptions (see Figure 3) whereas 4 and 13 administrations resulted in a similar inhibitory effect on thrombus formation. With all three regimens a decrease of the inhibitory effect was observed 90 min following the final administration of pentoxifylline. The major difference between single and repeated administrations of pentoxifylline was the significantly higher degree of inhibition observed in the early phases (up to 60 min) following either 4 or 13 administrations. After 13 administrations, thrombus formation was inhibited by approximately 33 % as early as 5 min after administration whereas a similar degree of inhibition was observed only 45 min after a single injection of pentoxifylline (Figure 2).
DISCUSSION The action of pentoxifylline on thrombus formation in vivo was tested quantitatively in the hamster cheek pouch. The results of a pilot series show that a single intraperitoneal injection of 5, 10 or 20 mg/kg pentoxifyline decreases thrombus formation in a time-dependent and slightly dose-dependent manner with a maximal inhibition of 40 % (5 mg/kg) to 50 % (10 and 20 mg/kg) at 60 min after drug administration. It was obvious, however, that significantly higher inhibition than with the low dose of 5 mg/kg was not to be expected in the pharmacological dose range. Therefore, single administration of 10 mg/kg pentoxifyline was compared to repeated administrations of the same dose as a model of chronic application. In this series inhibition of thrombus formation was more pronounced after repeated administration. This applies to the maximal degree of inhibition at 75 min after administration as well as to the slighty decreased inhibitory effect at 90 min after the last injection. The differences were most pronounced during the first 45 min following the last drug administration (see Figure 3) suggesting to reflect some longer lasting antithrombotic effect of the preceeding administrations being present even 16 hours later. This assumption has to be tested in additional investigations by measurements of thrombus formation before the last drug application, since thrombus formation under our experimental conditions cannot be controlled beyond a test period of approximately 2 hours. The findings of our study may be understood as a supplement to earlier investigations with other in vivo models of thrombosis. Pentoxifylline in doses of 10 to 30 mg/kg i.v. or p.o. caused significant inhibition of stenosis-induced thrombosis of rabbit femoral arteries (11) and of laser-induced thrombus formation in rat mesenteric arterioles (12, 13). Moreover, platelet aggregation in vivo tested in monkeys was also inhibited by pentoxifylline in a dose- and time-dependent manner (15). The antithrombotic activity of pentoxifyline as evidenced again in this study might be related in part to a stimulated prostacyclin release as demonstrated after pentoxifylline administration both in vitro (16) and ex vi~0 (17, 18). PGI itself was also found to be active in the hamster cheek pouch model of exp&imental thrombosis (19, 20). Interestingly, special in-
366
PENTOXIFYLLINE AND THROMBOSIS
Vol. 56, No. 3
vestigations on the time dependence of the pentoxifylline-stimulated PGI ’ 8 release in animals (18) and in human volunteers (21) demonstrated a tim course similar to that of the antithrombotic effect observed in this study. The increased inhibition of thrombus formation observed after repeated applications, however, is unlikely to be related to residual PGI or pentoxifylline on the basis of pharmacokinetic considerations. It might more likely be explained by a pharmacodynamic effect rendering the vascular wall more thromboresistant as far as thrombotic susceptibility or antithrombotic reactivity are concerned. In the hamster cheek pouch model, thrombi are induced under physiological haemodynamic conditions in the presence of the vascular wall. In this respect it is most suitable for testing antithrombotic effects of drugs acting on platelets as well as on the vessel wall. This in vivo model of experimental thrombosis has been used for the evaluation of a variety of agents affecting thrombus formation such as prostacyclin (19, 20), acetylsalicylic acid (22), calcium dobesilate (23), etofylline clofibrate (24) and an antimetastatic agent (25). In the present study a similar inhibition of thrombus formation was observed for pentoxifylline which besides its hemorheological action in improving blood flow properties and enhancing disturbed blood flow has also been reported to reverse hypercoagulability and hyperaggregability in patients suffering from vascular disease (for review see ref. 26). This suggests that alleviation of blood flow disturbances by pentoxifylline is accompanied by additional pharmacological effects preventing thromboembolic episodes in various prethrombotic states. First clinical studies with pentoxifylline have demonstrated a reduced incidence of shunt thrombosis in hemodialysis patients (27) as well as a distinct benefit for patients at risk of postoperative thrombosis (28), recurrent transient ischemic attacks (29, 30), and vascular graft occlusion (31, 32). Further investigations are needed to confirm these results and to define the limitations of the antithrombotic potential for preventing thromboembolic events.
REFERENCES 1. THOMPSON, A.R. and HARKER, A.L. Clinical approach to thrombosis and thromboembolism. In: Manual of Hemostasis and thrombosis, Edition 3. ;3!.I;iompson, L.A. Harmdsmhiladelphia, FA Davis Co., 1983, pp. . 2. MDLLER-MOHNSSEN, H., KRATZER, M. and BALDAUF, W. Microthrombus formation in models of coronary arteries caused by stagnation point flow arising at the predilection sites of atherosclerosis and thrombosis. In: The role of fluid mechanics in atheroaenesis. Proc. from a specialists meeting, Ohio, Aug. 24-25, 1978, pp. 12.1-12.8. 3. STEMERMAN, M.B. Relationship of thrombosis to atherosclerosis. In: &mostasis_and thrombosis, basic principles and clinical practa, R.W. Colman, il: Hirsh V.J. Marder, E.W. Salzman (Eds-.)Philadelphia, Toronto, JB Lippincott Co., 1982, pp. 781-788. 4. SINZINGER, H., FEIGL, W. and SILBERBAUER, K. Prostacyclin generation in atherosclerotic arteries. Lancet 2, 469, 1979.
Vol. 56, No. 3
PENTOXIFYLLINE ANDTHROMBOSIS
367
in a region of disturbed flow. Trans. Am. Sot. Artif. Intern. Oraans, 2_3,632-638, 1977.
5. GOLDSMITH, H.L. and KARINOT. Platelets 6. BEGENT, N.
and BORN, G.V.R. Growth rate in vivo of platelet thrombi, produced by iontophoresis of ADP, as a function of mean blood flow velocity. Nature (Lond.) 227, 926-930, 1970.
7. HONOUR, A.J.
and ROSS-RUSSEL, R.W. Brit. J. Exp. Path. 43, 350-362, 1962.
Experimental platelet embolism.
8. MULLER, R. Haemorheology and peripheral vascular diseases: A new therapeutic approach. J,~Med. 12, 209-235, 1981. 9. ANGELKORT,
B. and KIESEWETTER, H. Influence of risk factors and coagulation phenomena on the fluidity of blood in chronic arterial occlusive disease. Stand. ___-_I__ J. Clin. Lab. Invest. 41 (suppl. 156), 185-187, 1981.
10. STRANO, A.,
DAVI, G, NOVO, S., AVELLONE, G. and PINTO A. Clinical evaluation of the effects of pentoxifylline in patients with chronic peripheral circulatory disorders. Pharmatherapeutica -3 (suppl. l), 117-125, 1983.
11. SAKURAI, M., KOMINE I. and GOTO, M. Protective effect of pentoxifylline on experimental thrombosis induced by stenosis of femoral artery in rabbits. Yakuri to Chiryo (Jap. Pharmacol. Therap.) l3, 135-138, 1985.
12. SEIFFGE, D.
The antithrombotic effect of pentoxifylline in the laserinduced thrombus model in rat mesenteric arterioles. Proc. 3. Dtsch.-Jap. Kongr. Angiol., Heidelberg, H. Maurer (Ed.) Grlfelfing, Demeter-Verlag, 1985, pp. 86-90.
13. WEICHERT, W. and BREDDIN, H.K. Antithrombotic effects of acetylsalicylic acid (ASA) and pentoxifylline in laser-induced thrombosis in rat mesenteric vessels. V_ASA14, 280-284, 1985.
14. DULING,
B.R., BERNE, R.M. and BORN, G.V.R. Microiontophoretic application of vasoactive agents to the microcirculation of the hamster cheek pouch. Microvascular Res. 1, 158-175, 1968.
15. GASTPAR, H. Study of platelet aggregation in vivo III. oxifylline. _J%.. 8, 191-197, 1977
Effect of pent-
16. MATZKY, R., DARIUS H. and SCHROR, K. The release of prostacyclin (PGI ) by &.
pentoxifylline from human vascular tissue. 32, 1315-1318, 1982.
Arzneim.-Forsch./& 2
17. WEITHMANN, K.U.
Reduced platelet aggregation by pentoxifylline-stimulated prostacyclin release. VASA !O, 249-252, 1981. Pentoxifylline enhances formation of prostacyclin from rat vascular and renal tissue. Prostagl, Leukot~.___Med_.1_?, 217-228, 1983.
18. SINZINGER, H.
E.A., HIGGS, G.A., MONCADA, S. and VANE, J.R. Prostacyclin inhibits the formation of platelet thrombi in arterioles and venu!es of the hamster cheek pouch. &it. J. Pharmacol. 6_3,535-539, 1978.
19. HIGGS, (PGI )
368
PENTOXIFYLLINE AND THROMBOSIS
Vol. 56, No. 3
20. SHISHIDO, M., HIROSE, R., TANAKA, K. and KATORI, M. Manipulation of thrombus formation in the hamster pouch with drugs that interact with PG12 in vitro. Prostaqlandins 22, 907-917, 1982. 21. POHANKA, E. and SINZINGER H. Effect of a single pentoxifylline administration on platelet sensitivity, plasma factor activity, plasma 6-0x0PGFI and thromboxane B2 in healthy volunteers. Psstagl. Leukotr. Med. 22, 191-200, 1986. 22. MICHAL, M. and GIESSINGER, N. Standardization of animal models of thrombosis. Induction of thrombi by the iontophoretic application of ADP to venules of the hamster cheek pouch. Proc. XVII. Angiological Symposium, Kitzbuhel, K. Breddin, R. Zimmermann (Eds.), Stuttgart, FK Schattauer, 1983, pp.133-145. 23. MICHAL, M. and GIESSINGER, N. Effect of calcium dobesilate and its inRes. 4Q, teraction with aspirin on thrombus formation in vivo. nromb. 215-226, 1985. 24. SIM, A.K., DAVIES, M.E., McCRAW, A.P. and METZ, G. Effect of etofylline clofibrate on experimental thrombosis and platelet function. Arzneim.Forsch./Drug Res. 3, 2042-2045, 1980. 25. ATHERTON, A., BUSFIELD, D. and HELLMAN, K. The effects of an antimetastatic agent, 1,2-bis (3,5-dioxopiperazin-1-yl)propane on platelet behaviour. Cancer Res. 35, 953-957, 1975. 26. SCHRbER, R.H. Antithrombotic potential of pentoxifylline - a hemorheologically active drug. Angiology 36, 387-398, 1985. 27. RADMILOVIC, A., BORIC, Z., NAUMOVIC, T., STAMENKOVIC, M. and MUSIKIC, Shunt thrombosis prevention in hemodialysis patients - a doublebPiind randomized study: pentoxifylline vs. placebo. Anqiology 33, 499-566, 1987. 28. KOPPENHAGEN, K. Thromboembolie-Prophylaxe mit Trental nach chirurgischen Eangriffen im Bauchraum. In: Das antithrp_?o_ttschePotential von Trental . H. Hess, A. Bollinger (Eds.) Frankfurt, PMI, 1988, pp. 49-56. 29. HERSKOVITS, E., FAMULARI, A., TAMAROFF, L., GONZALES, A.M., VASQUEZ, A SMUD, R., FRAIMAN, H., VILA, J. and MATERA, V. Randomized trial of p&oxifylline versus acetylsalicylic acid plus dipyridamole in preventing transient ischemic attacks. Lancet I, 966-968, (1981). -_-. 30. HERSKOVITS, E., FAMULARI, A., TAMAROFF, L., GONZALES, A.M., VAZQUES, A DOMINGUEZ, R., FRAIMAN, H., and VILA, J. Preventive treatment of ceiebral transient ischemia: Comparative randomized trial of pentoxifylline vs. conventional antiaggregants. Eur. -__-_Neurol. 24, 73-81, 1985. 31. LUCAS, M.A. Prevention of post-operative thrombosis in peripheral arteriopathies. Pentoxifylline vs. conventional antiaggregants: A six-month randomized follow-up study. Angiology 35, 443-449, 1984. 32. RAITHEL, D. Prevention of reocclusion after prosthetic bypass operations in the femoro-popliteal region: A comparative study of pentoxifylline versus acetylsalicylic acid. Vast --'- Surg. 2_1,208-214, 1987.
REDUCED THROMBUS FORMATION IN VIVO AFTERR ADMINISTRATION OF PENTOXIFYLLINE (TRENTAL )
M. Michal*, N. Giessinger* and R. SchrGer+ * BATTELLE RESEARCH CENTRES Centre for Toxicology and Biosciences CH-1227 Carouge-Geneva, Switzerland ' HOECHST AG WERK ALBERT, Clinical Research D-6200 Wiesbaden 12, West-Germany (Received 8.8.1988; accepted in revised form 8.8.1989 by Editor M.B. Donati)
ABSTRACT
The hamster cheek pouch model of experimental thrombosis in which thrombi are induced in the microvasculature by iontophoretical administration of adenosine diphosphate (ADP) was uked to test the antithrombotic potential of pentoxifylline (Trental ). Single intraperitoneal injections of 5, 10 and 20 mg/kg pentoxifylline reduced thrombus formation by 20 to 50 % from 30 to 105 min following drug administration. The effect of a single application of 10 mg/kg was exceeded significantly (p < 0.05) by the higher rate of inhibition after repeated injections of the same dose given three times daily. This suggests a residual antithrombotic effect from the preceeding administrations.
INTRODUCTION Acute ischemic events such as myocardial infarction, transient ischemic attacks and strokes are caused by thrombus formation in arteries and/or ensueing thromboembolic complications (1). The identified events leading to thrombus formation are alterations in flow conditions (2), alterations of the vessel wall which interfere with its natural antithrombotic properties (3, 4) and activation of the haemostatic system, especially platelet activation (5). The rationale that a drug modifying platelet function might be effective in preventing clinical manifestations of thrombosis has prompted considerable research activity in recent years to find suitable antiplatelet drugs which potentially prevent thrombus formation in vivo. Reliable and valid in vivo models are therefore needed for identifying and evaluating potential antithrombotic drugs.
Key words: thrombosis, pentoxifylline, antithrombotic effect 359
PENTOXIFYLLINE AND THROMBOSIS
360
Vol. 56, No. 3
The haclscercneek pouch technique (6) offers a reliable and quantitative ti m r,lodel for testing potential antithrombotic drugs. Under in vivo haemodynamic conditions, this model closely simulates spontaneous thrombus formati0.lin the presence of all formed elements of flowing blood and the vessel wdll. Experimental thrombus formation is initially triggered by the iontopnore-cicapplication of ADP onto a microvessel. ADP diffused into the microvessel causes platelet adhesion and aggregation which in turn probably causes the release of intraplatelet material leading to the formation of a throclbus.kith this techniaue it is possible to mimic the pertinent physioevent - the formation of ah intravascular platelet plug or "white logica body" ( 7) under physiological haemodynamic conditions. Pentox fylline a drug well established for the treatment of peripheral vascular disease has been shown besides its beneficial effects on disturbed bloau f'luw (8) to reverse platelet hyperreactivity and hypercoagulability iil pa-ten.cs(9, 10). Furthermore, pentoxifylline has been reported to inhibit in vivo experimental thrombosis in stenosed femoral arteries of rab11) and laser-induced thrombus formation in vivo in mesenteric artebies ries in rats (12, 13). Therefore, it was of particular interest_to investi._ . gaze the antithrombotic effects of pentoxifylline on thrombus formation u vita using the hamster cheek pouch model.
(’
MATERIALS AND METHODS ?reoaration of the hamster cheek pouch The microcirculation of the hamster cheek pouch was studied by the method of Duling, Berne and Born (14). Male Syrian-golden hamsters, weighing 95-110 g, were anesthetized with pentobarbitone (60 mg/kg body weight) injacted intraperitoneally. Supplementary doses of anaesthetic were given when needed. The cheek pouch was everted, spread out and fixed to a special perspex stage equipped with fixing barbs. The top layer and fine connective tissue wsre carefully removed under a stereomicroscope, leaving the bottom layer of the pouch suitable for microscopic observation using a fixed stage microscooe (Leitz Laborlux) transilluminated with liaht from below. The mounted 'cheek was rinsed continuously at a fixed rate w th a physiological buffer solution at 37°C. Such preparations can be mainta ned in good conditio,l, as assessed by blood flow and reactivity of blood vessels and by abseuce of edema, for approximately 2 hours. inauction of thrombus formation Platelet thrornbiwere induced in venules (30-40 urnin diameter) by the iontophoretic application of ADP to the blood vessel wall (6). Glass electrodes (Clark Electromedical Instruments) were drawn out to form micropipettes (tip diameter 2-3 urn)with a pipette puller (Industrial Science AssDciates Inc.). A micropipette prefilled with ADP (10 mM in distilled water) was manipulated using a Leitz micromanipulator so that the tip was in contact with the venule. The ADP solution in the micropipette was connected to an electrical circuit and the reference electrode completing the circuit was immersed in the superfusing physiological buffer solution. Adenosine aipnosphate was applied to the vessel wall by passing a current between micropipette and reference electrode. From a battery power supply (36 V) a negative potential controlled by a modified digital voltmeter was app'lied to the micropipette; the resultant current of 300 nAMP was
Vol. 56, No. 3
PENTOXIFYLLINE AND THROMBOSIS
361
-14 calculated to eject ADP in the order of 2x10 mol/sec. Typically, after approximately 5 set a thrombus began to form on the inner side of the vessel wall, near the pipette tip. Once the current was switched off after 60 set, the thrombus embolized within approximately 10 sec. The same vessel can be used several times and thus can provide its own control. Rate of thrombus formation was based on size of thrombus formed at 30, 45 and 60 set from beginning of the application of ADP. Quantification of thrombus formation The formation of the experimental thrombus following the application of ADP to a venule was recorded using a Sony video camera (DXC 1800P) and a Sony video recorder (VO-5850P). The video tape was played back and the image frozen at various intervals during the progressive growth of the thrombus. The image was traced on the screen and analyzed using an image analyzer (MOP-AMOS, Carl Zeiss). The visualized surface area of the thrombus increased linearly with time. Therefore a plot of surface area against time provided the rate of growth of the thrombus. Administration of drug Pentoxifylling (10 mg/ml) stock solution was prepared in NaCl 0.9 % and stored at -18 C. Samples of this solution were thawed when required. Hamsters weighing 100 g were injected i.p. in volumes not exceeding 200 ul of pentoxifylline stock solution corresponding to the doses of 5, 10 and 20 mg/kg, respectively. For repeated administrations, a stock solution of pentoxifylline (5 mg/ml) was used. Three groups of eight hamsters each (weight 100 g) were injected 1, 4 and 3 times, respectively, with 200 ul/kg of pentoxifylline stock solution (5 mg/ml) which corresponds to 10 mg/kg. Repeated injections were made daily at 8h30, 12h30 and 16h30. Induction of thrombus formation was started 5 min after injection of the last dose in each case, i.e. in the group receiving 4 administrations, after the morning dose of day two and, in the group receiving 13 administrations, after the morning dose of day four. Injections of pentoxifylline were replaced by saline in the control group of hamsters. Statistical analysis of experimental data Data are expressed as mean + standard error of n observations. Experimental data of the pilot investigation were analyzed descriptively comparing values obtained 5-105 min following the injection of each dose of drug and those obtained in animals injected with saline. Statistical evaluation of experimental data of the second series (4 groups of 8 animals) was carried out on comparison of values obtained 5-90 min following the last injection of pentoxifylline in each of the 3 different drug regimens (1, 4 and 13 administrations) and control animals injected with saline. Statistical evaluation was performed using Student's t-test by means of an appropriate computer program on a Hewlett-Packard lOOOF. Materials Physiological buffer solution: composition (mM) Nahl 131,KCl 4.0, CaC12 3.0, MgSO 1.0 and Tris 5.0, pH 7.4, maintained at 37 C. Anaesthetfc: sodium pentobarbitone (Nembutal, Abbott). Thrombus inducer: adenosine diphosphate sodium salt (Sigma Chemicals). Testdrug: pentoxifylline (HOECHST AG Werk Albert, Wiesbaden).
362
PENTOXIFYLLINE AN0 THROMBOSIS
Vol. 56, No. 3
RESULTS Reoroducibilitv of thrombus formation The rate of thrombus formation was first established in the absence of pentoxifylline based on the visualized size of thrombus attained 15, 30, 45 and 60 set following the application of ADP. The data showed the rate of thrombus formation in control animals to be reproducible from one induction of thrombus to another at 15 min intervals over at least 90 min. Single administration of oentoxifvlline
In a pilot investigation with 3-4 experiments per group, three doses of pentoxifylline (5, 10 and 20 mg/kg) were tested after intraperitoneal injection of the drug. From 5-90 min after application, thrombus size was evaluated at 15, 30, 45 and 60 set of iontophoretic ADP application. As an example Figure 1 shows thrombus formation at different times after pentoxifylline 5 mg/kg i.p.
m Control 0 = 5min A = I5min ?? = 30min A = 45 min 0 = 90
15
30
45
x
??
0
??
min
75 min 60 min
60 set
TIME FROM START OF ADP IONTOPHORESIS FIG. 1
Thrombus formation in vivo at different times after intraperitoneal injection of 5 mg/kg pen-line. Visualized thrombus size was evaluated at various times of ADP iontophoresis before (control .) and 5 (o), 15 (A), 30 (a), 45 (A), 69 to), 75 (x) and 90 (0) min after injection of pentoxifylline. Values represent means of 3-4 experiments.
Vol. 56, No. 3
PENTOXIFYLLINE AND THROMBOSIS
363
From the data of visualized thrombus size, percent inhibition was calculated by comparing thrombus size before (control) and after drug application. In all cases the inhibitory effects were similar at 30, 45 and 60 set of ADP iontophoresis. Therefore, the experimental data obtained at 45 set were taken as representative for comparing the inhibitory effects obtained at different times and after the different doses tested. A summary graph showing the overall inhibitory action of the different doses of pentoxifylline on thrombus formation is given in Figure 2.
0
I
I
I
I
I
I
15
30
45
60
75
90
105
TIME AFTER ADMINISTRATION OF PENTOXIFYLLINE ( MlN 1
FIG. 2 Inhibition of thrombus formation after single intraperitoneal administration of 5 (o), 10 (0) and 20 (A) mg/kg pentoxifylline. Results are expressed as percent inhibition of thrombus formation at 45 set of ADP iontophoresis obtained by comparing values before (control) and after administration of pentoxifylline. Values represent mean + standard error of 3-4 experiments. Pentoxifylline in all three doses inhibited thrombus formation in vivo showing a similar time dependence, but only a weak dose dependence. While the inhibition after the dose of 5 mg/kg increased from 5 % at 15 min to 40 % maximal inhibition at 60-75 min after drug administration, 20 mg/kg resulted in only 10 % higher inhibition over the entire test period with a maximal inhibition of 50 % at the same time as the lower dose, In contrast, the results after 10 mg/kg differ quantitatively from this frame due to unknown reasons. - In all cases, the inhibitory effect on thrombus formation produced by pentoxifylline started to decrease after 75 min following drug administration but lasted for the full observation period.
PENTOXIFYLLINE AND THROMBOSIS
364
Repeated
administration
Vol. 56, No. 3
of oentoxifvlline
In the pharmacological dose range of 5-20 mg/kg, there was obviously no strong dose dependence thus suggesting no stronger inhibition with higher doses. Therefore a second series of experiments was undertaken to investigate the effect of repeated injections of pentoxifylline. Three regimens of 1, 4, or 13 administrations of 10 mg/kg pentoxifylline i.p. (see under "Administration") were investigated for their effect on thrombus formation jr_~ m. Percent inhibition was obtained by comparing visualized thrombus size at 45 set of ADP iontophoresis in control animals injected with saline and those treated with 10 mg/kg pentoxifylline and in case of the latter two regimens pretreated with the same dose three times daily for 1 or 3 days. Figure 3 is a summary graph showing the overall inhibitory effects of the three different regimens.
5 60 G r a 50 c '3 g 40 -
x
& F b
30 -
z 20 I= 5
-
F z t- loC!! zi no
I 0
15
I
I
1
30 45 60 75 TIME AFTER LAST ADMINISTRATION OF PENTOXIFYLLINE (MIN)
I 90
105
FIG. 3 Inhibition of thrombus formation after 1 (o), 4 (0) and 13 (A) intraperitoneal administrations of 10 mg/kg pentoxifylline. Percent inhibition was calculated by comparing size of thrombus attained 45 set following iontophoretic application of ADP in the absence of drug (control) and after 1, 4 and 13 administrations of pentoxifylline. Values represent mean + standard error. * p < 0.05 vs. 1 administration
t p < 0.05 vs. 4 administrations
Thrombus inhibition observed after 1, 4 or 13 administrations progressed with time reaching a maximum at 60-75 min following last administration of the drug. Inhibition obtained after a single application of 10 mg/kg was in agreement with the results after single application of 5 and 20 mg/kg, re-
Vol. 56, No. 3
PENTOXIFYLLINE AND THROMBOSIS
365
spectively. The maximal degree of inhibition (45.0 %) after a single injection of pentoxifylline (10 mg/kg) occurred at the same time (75 min) but was exceeded by that following 4 or 13 administrations producing 52.7 % and 52.5 % inhibition, respectively. The differences between single and repeated administrations were statistically significant (p < 0.05) with two exceptions (see Figure 3) whereas 4 and 13 administrations resulted in a similar inhibitory effect on thrombus formation. With all three regimens a decrease of the inhibitory effect was observed 90 min following the final administration of pentoxifylline. The major difference between single and repeated administrations of pentoxifylline was the significantly higher degree of inhibition observed in the early phases (up to 60 min) following either 4 or 13 administrations. After 13 administrations, thrombus formation was inhibited by approximately 33 % as early as 5 min after administration whereas a similar degree of inhibition was observed only 45 min after a single injection of pentoxifylline (Figure 2).
DISCUSSION The action of pentoxifylline on thrombus formation in vivo was tested quantitatively in the hamster cheek pouch. The results of a pilot series show that a single intraperitoneal injection of 5, 10 or 20 mg/kg pentoxifyline decreases thrombus formation in a time-dependent and slightly dose-dependent manner with a maximal inhibition of 40 % (5 mg/kg) to 50 % (10 and 20 mg/kg) at 60 min after drug administration. It was obvious, however, that significantly higher inhibition than with the low dose of 5 mg/kg was not to be expected in the pharmacological dose range. Therefore, single administration of 10 mg/kg pentoxifyline was compared to repeated administrations of the same dose as a model of chronic application. In this series inhibition of thrombus formation was more pronounced after repeated administration. This applies to the maximal degree of inhibition at 75 min after administration as well as to the slighty decreased inhibitory effect at 90 min after the last injection. The differences were most pronounced during the first 45 min following the last drug administration (see Figure 3) suggesting to reflect some longer lasting antithrombotic effect of the preceeding administrations being present even 16 hours later. This assumption has to be tested in additional investigations by measurements of thrombus formation before the last drug application, since thrombus formation under our experimental conditions cannot be controlled beyond a test period of approximately 2 hours. The findings of our study may be understood as a supplement to earlier investigations with other in vivo models of thrombosis. Pentoxifylline in doses of 10 to 30 mg/kg i.v. or p.o. caused significant inhibition of stenosis-induced thrombosis of rabbit femoral arteries (11) and of laser-induced thrombus formation in rat mesenteric arterioles (12, 13). Moreover, platelet aggregation in vivo tested in monkeys was also inhibited by pentoxifylline in a dose- and time-dependent manner (15). The antithrombotic activity of pentoxifyline as evidenced again in this study might be related in part to a stimulated prostacyclin release as demonstrated after pentoxifylline administration both in vitro (16) and ex vi~0 (17, 18). PGI itself was also found to be active in the hamster cheek pouch model of exp&imental thrombosis (19, 20). Interestingly, special in-
366
PENTOXIFYLLINE AND THROMBOSIS
Vol. 56, No. 3
vestigations on the time dependence of the pentoxifylline-stimulated PGI ’ 8 release in animals (18) and in human volunteers (21) demonstrated a tim course similar to that of the antithrombotic effect observed in this study. The increased inhibition of thrombus formation observed after repeated applications, however, is unlikely to be related to residual PGI or pentoxifylline on the basis of pharmacokinetic considerations. It might more likely be explained by a pharmacodynamic effect rendering the vascular wall more thromboresistant as far as thrombotic susceptibility or antithrombotic reactivity are concerned. In the hamster cheek pouch model, thrombi are induced under physiological haemodynamic conditions in the presence of the vascular wall. In this respect it is most suitable for testing antithrombotic effects of drugs acting on platelets as well as on the vessel wall. This in vivo model of experimental thrombosis has been used for the evaluation of a variety of agents affecting thrombus formation such as prostacyclin (19, 20), acetylsalicylic acid (22), calcium dobesilate (23), etofylline clofibrate (24) and an antimetastatic agent (25). In the present study a similar inhibition of thrombus formation was observed for pentoxifylline which besides its hemorheological action in improving blood flow properties and enhancing disturbed blood flow has also been reported to reverse hypercoagulability and hyperaggregability in patients suffering from vascular disease (for review see ref. 26). This suggests that alleviation of blood flow disturbances by pentoxifylline is accompanied by additional pharmacological effects preventing thromboembolic episodes in various prethrombotic states. First clinical studies with pentoxifylline have demonstrated a reduced incidence of shunt thrombosis in hemodialysis patients (27) as well as a distinct benefit for patients at risk of postoperative thrombosis (28), recurrent transient ischemic attacks (29, 30), and vascular graft occlusion (31, 32). Further investigations are needed to confirm these results and to define the limitations of the antithrombotic potential for preventing thromboembolic events.
REFERENCES 1. THOMPSON, A.R. and HARKER, A.L. Clinical approach to thrombosis and thromboembolism. In: Manual of Hemostasis and thrombosis, Edition 3. ;3!.I;iompson, L.A. Harmdsmhiladelphia, FA Davis Co., 1983, pp. . 2. MDLLER-MOHNSSEN, H., KRATZER, M. and BALDAUF, W. Microthrombus formation in models of coronary arteries caused by stagnation point flow arising at the predilection sites of atherosclerosis and thrombosis. In: The role of fluid mechanics in atheroaenesis. Proc. from a specialists meeting, Ohio, Aug. 24-25, 1978, pp. 12.1-12.8. 3. STEMERMAN, M.B. Relationship of thrombosis to atherosclerosis. In: &mostasis_and thrombosis, basic principles and clinical practa, R.W. Colman, il: Hirsh V.J. Marder, E.W. Salzman (Eds-.)Philadelphia, Toronto, JB Lippincott Co., 1982, pp. 781-788. 4. SINZINGER, H., FEIGL, W. and SILBERBAUER, K. Prostacyclin generation in atherosclerotic arteries. Lancet 2, 469, 1979.
Vol. 56, No. 3
PENTOXIFYLLINE ANDTHROMBOSIS
367
in a region of disturbed flow. Trans. Am. Sot. Artif. Intern. Oraans, 2_3,632-638, 1977.
5. GOLDSMITH, H.L. and KARINOT. Platelets 6. BEGENT, N.
and BORN, G.V.R. Growth rate in vivo of platelet thrombi, produced by iontophoresis of ADP, as a function of mean blood flow velocity. Nature (Lond.) 227, 926-930, 1970.
7. HONOUR, A.J.
and ROSS-RUSSEL, R.W. Brit. J. Exp. Path. 43, 350-362, 1962.
Experimental platelet embolism.
8. MULLER, R. Haemorheology and peripheral vascular diseases: A new therapeutic approach. J,~Med. 12, 209-235, 1981. 9. ANGELKORT,
B. and KIESEWETTER, H. Influence of risk factors and coagulation phenomena on the fluidity of blood in chronic arterial occlusive disease. Stand. ___-_I__ J. Clin. Lab. Invest. 41 (suppl. 156), 185-187, 1981.
10. STRANO, A.,
DAVI, G, NOVO, S., AVELLONE, G. and PINTO A. Clinical evaluation of the effects of pentoxifylline in patients with chronic peripheral circulatory disorders. Pharmatherapeutica -3 (suppl. l), 117-125, 1983.
11. SAKURAI, M., KOMINE I. and GOTO, M. Protective effect of pentoxifylline on experimental thrombosis induced by stenosis of femoral artery in rabbits. Yakuri to Chiryo (Jap. Pharmacol. Therap.) l3, 135-138, 1985.
12. SEIFFGE, D.
The antithrombotic effect of pentoxifylline in the laserinduced thrombus model in rat mesenteric arterioles. Proc. 3. Dtsch.-Jap. Kongr. Angiol., Heidelberg, H. Maurer (Ed.) Grlfelfing, Demeter-Verlag, 1985, pp. 86-90.
13. WEICHERT, W. and BREDDIN, H.K. Antithrombotic effects of acetylsalicylic acid (ASA) and pentoxifylline in laser-induced thrombosis in rat mesenteric vessels. V_ASA14, 280-284, 1985.
14. DULING,
B.R., BERNE, R.M. and BORN, G.V.R. Microiontophoretic application of vasoactive agents to the microcirculation of the hamster cheek pouch. Microvascular Res. 1, 158-175, 1968.
15. GASTPAR, H. Study of platelet aggregation in vivo III. oxifylline. _J%.. 8, 191-197, 1977
Effect of pent-
16. MATZKY, R., DARIUS H. and SCHROR, K. The release of prostacyclin (PGI ) by &.
pentoxifylline from human vascular tissue. 32, 1315-1318, 1982.
Arzneim.-Forsch./& 2
17. WEITHMANN, K.U.
Reduced platelet aggregation by pentoxifylline-stimulated prostacyclin release. VASA !O, 249-252, 1981. Pentoxifylline enhances formation of prostacyclin from rat vascular and renal tissue. Prostagl, Leukot~.___Med_.1_?, 217-228, 1983.
18. SINZINGER, H.
E.A., HIGGS, G.A., MONCADA, S. and VANE, J.R. Prostacyclin inhibits the formation of platelet thrombi in arterioles and venu!es of the hamster cheek pouch. &it. J. Pharmacol. 6_3,535-539, 1978.
19. HIGGS, (PGI )
368
PENTOXIFYLLINE AND THROMBOSIS
Vol. 56, No. 3
20. SHISHIDO, M., HIROSE, R., TANAKA, K. and KATORI, M. Manipulation of thrombus formation in the hamster pouch with drugs that interact with PG12 in vitro. Prostaqlandins 22, 907-917, 1982. 21. POHANKA, E. and SINZINGER H. Effect of a single pentoxifylline administration on platelet sensitivity, plasma factor activity, plasma 6-0x0PGFI and thromboxane B2 in healthy volunteers. Psstagl. Leukotr. Med. 22, 191-200, 1986. 22. MICHAL, M. and GIESSINGER, N. Standardization of animal models of thrombosis. Induction of thrombi by the iontophoretic application of ADP to venules of the hamster cheek pouch. Proc. XVII. Angiological Symposium, Kitzbuhel, K. Breddin, R. Zimmermann (Eds.), Stuttgart, FK Schattauer, 1983, pp.133-145. 23. MICHAL, M. and GIESSINGER, N. Effect of calcium dobesilate and its inRes. 4Q, teraction with aspirin on thrombus formation in vivo. nromb. 215-226, 1985. 24. SIM, A.K., DAVIES, M.E., McCRAW, A.P. and METZ, G. Effect of etofylline clofibrate on experimental thrombosis and platelet function. Arzneim.Forsch./Drug Res. 3, 2042-2045, 1980. 25. ATHERTON, A., BUSFIELD, D. and HELLMAN, K. The effects of an antimetastatic agent, 1,2-bis (3,5-dioxopiperazin-1-yl)propane on platelet behaviour. Cancer Res. 35, 953-957, 1975. 26. SCHRbER, R.H. Antithrombotic potential of pentoxifylline - a hemorheologically active drug. Angiology 36, 387-398, 1985. 27. RADMILOVIC, A., BORIC, Z., NAUMOVIC, T., STAMENKOVIC, M. and MUSIKIC, Shunt thrombosis prevention in hemodialysis patients - a doublebPiind randomized study: pentoxifylline vs. placebo. Anqiology 33, 499-566, 1987. 28. KOPPENHAGEN, K. Thromboembolie-Prophylaxe mit Trental nach chirurgischen Eangriffen im Bauchraum. In: Das antithrp_?o_ttschePotential von Trental . H. Hess, A. Bollinger (Eds.) Frankfurt, PMI, 1988, pp. 49-56. 29. HERSKOVITS, E., FAMULARI, A., TAMAROFF, L., GONZALES, A.M., VASQUEZ, A SMUD, R., FRAIMAN, H., VILA, J. and MATERA, V. Randomized trial of p&oxifylline versus acetylsalicylic acid plus dipyridamole in preventing transient ischemic attacks. Lancet I, 966-968, (1981). -_-. 30. HERSKOVITS, E., FAMULARI, A., TAMAROFF, L., GONZALES, A.M., VAZQUES, A DOMINGUEZ, R., FRAIMAN, H., and VILA, J. Preventive treatment of ceiebral transient ischemia: Comparative randomized trial of pentoxifylline vs. conventional antiaggregants. Eur. -__-_Neurol. 24, 73-81, 1985. 31. LUCAS, M.A. Prevention of post-operative thrombosis in peripheral arteriopathies. Pentoxifylline vs. conventional antiaggregants: A six-month randomized follow-up study. Angiology 35, 443-449, 1984. 32. RAITHEL, D. Prevention of reocclusion after prosthetic bypass operations in the femoro-popliteal region: A comparative study of pentoxifylline versus acetylsalicylic acid. Vast --'- Surg. 2_1,208-214, 1987.