555
335
MEDICAL SCIENCE
Trial of iloprost versus aspirin treatment for critical limb ischaemia of thromboangiitis obliterans
FOR THE TAO
152
patients with thromboangiitis obliterans (Buerger’s disease) and pain from critical leg ischaemia were randomly allocated to receive iloprost, a chemically stable prostacyclin analogue, or low-dose aspirin, for 28 days in a double-blind trial. On review, 19 patients did not fulfil the stringent entry criteria. Of the other 133 patients, 98 also had leg ulcers. After 21-28 days, 58 (85%) of 68 iloprost-treated patients showed ulcer healing or relief of ischaemic pain, compared with 11 (17%) of 65 in the aspirin-treated group. 43 (63%) on iloprost treatment had complete relief of pain, compared with 18 (28%) on aspirin. Ulcers healed completely in 18 of 52 (35%) who received iloprost compared with 6 of 46 (13%) who received aspirin. 6 months after the start of treatment, the response rate was 45 of 51 (88%) patients treated with iloprost compared with 12 of 44 (21%) patients treated with aspirin.
STUDY*
but is
prevalent in Israel and parts of Asia. Thromboangiitis obliterans is typically observed in young men who smoke cigarettes heavily and, other than advice to stop smoking, its treatment remains controversial. Although rarely fatal, the incidence of major amputation is much higher than for atherosclerotic ischaemia-amputation rates of up to 70% have been reportedl-s-and because the disease mainly affects distal arteries, reconstructive surgery is rarely possible or successful. A beneficial effect of prostanoid treatment has been reported in some small studies.6-8 We have compared the effect of treatment by iloprost, a prostacyclin analogue, to that of low-dose aspirin in a randomised, controlled, double-blind, multicentre trial. more
Patients and methods Patients had
to
fulfil criteria for both critical limb ischaemia and
thromboangiitis obliterans. Critical ischaemia was defmed as continuous ischaemic rest pain in a limb for which continuous
Lancet 1990; 335: 555-57.
Introduction
Thromboangiitis rare cause *
obliterans (TAO; Buerger’s disease) is a of critical limb ischaemia in Europe and the USA,
Full list of participants and
centres at
end of article.
ADDRESSES: Service de Pathologie Vasculaire, et Centre Claude Bernard de Recherches sur les Maladies Vasculaires Périphériques, Hôpital Broussais, 75674 Paris, France (Prof J. N. Fiessinger, MD), and Clinical Research Department of Internal Medicine 1, Schering AG, Berlin, West Germany (Dr M. Schäfer, MD). Correspondence to Prof J. N. Fiessinger, Service de Pathologie Vasculaire, Médecine IV, Hôpital Broussais, 75674 Paris Cedex 14, France.
556
analgesia in hospital was required for at least 7 days, with or without tissue necrosis. For a diagnosis of thromboangiitis obliterans the following criteria had to be met: the patient had to be under 50 years of age, a smoker, with angiographic evidence of non-arteriosclerotic arterial occlusion below the knee, and at least 2 of the following 3 features GM: superficial thrombophlebitis; ischaemia of the arm; and/or typical arteriographic findings of skip lesions, direct collaterals (Martorell’s sign), and corkscrew collaterals. Patients with diabetes mellitus, other inflammatory vascular diseases, and amputation or lumbar sympathectomy in the preceding 3 months excluded. Patients were randomised within each centre to receive either oral aspirin (100 mg(day) and a 6-h daily placebo infusion, or a placebo tablet identical to aspirin and a 6-h infusion of iloprost. For the first 3 days, infusion was started at a rate equivalent to 0-5 ng/kg per min iloprost and increased every 30 min by a further 0-5ng/kg per min to a maximum tolerated dose or 2-0 ng/kg per min, which was maintained for the rest of the 6 h. The maximum dose used on day 3 was repeated on days 4-28 or until the patient’s symptoms and signs had completely resolved. Patients were followed for a further 5 months, when treatment was not controlled. Rest pain and analgesic consumption were recorded daily. In patients with trophic lesions the course of healing was assessed by surface area estimations. Distal peripheral pressures were measured weekly. An overall analysis, which included assessment of claudication, thrombophlebitis, and vasospastic symptoms, was made by the physician in charge of the patient. To investigate whether the appearance of symptoms caused by a prostacyclin analogue might influence the physician’s assessment, ischaemic lesions before and after treatment were also assessed by a physician not involved in the patient’s management and who was unaware of intervening
TABLE I-BASELINE CHARACTERISTICS OF STUDY GROUPS I
were
events.
At the end of treatment, patients were classified as responders or non-responders. Trophic lesions, rest pain, analgesic consumption, and overall improvement were rated by a 4-scale classification. Patients who entered the study with rest pain but without trophic lesions were considered to have responded if pain had been eliminated or strikingly reduced, with no or much reduced need for analgesics, and a striking overall improvement. Patients with
trophic lesions were considered to have responded if over 50% of the affected surface area had healed, with a striking overall improvement. Responder rates were compared by use of the X2 test. It was calculated that, to detect a 30% difference in response rate, a sample size of 53 patients would be needed if the type I error was (X = 0-05 and type II error was p = 0-1. Between January, 1986, and February, 1988, 152 patients were recruited at 26 centres in 9 countries. 19 patients (8 treated with iloprost, 11treated with aspirin) had to be excluded on review because they did not fulfil the stringent criteria for diagnosis of thromboangiitis obliterans. The baseline characteristics of the other 133 patients (68 treated with iloprost, 65 treated with aspirin) are shown in table i.
Results The two treatment groups had similar age distribution, incidence of ulceration or gangrene, duration of symptoms, previous therapy, concomitant diseases, other medications, and angiographic appearances (table i), and a similar distribution of symptoms of thrombophlebitis, claudication, and vasospasm. Smoking habits were also similar in both treatment groups: all patients were smokers, most of whom said that they had smoked 21-40 cigarettes per day for 11-20 years. Most gave up cigarettes at the start of treatment but, by day 28, 16 patients treated with iloprost and 13 treated with aspirin were still smoking. Headache, flushing, nausea, and abdominal cramps were more common in patients treated with iloprost but no patient in either group had to be withdrawn because of side-effects. 4 patients treated with iloprost and 13 treated
Values shown
with
as
numbers
aspirin
had
(%)
to
be withdrawn due
to
a
clinical
deterioration; of these, 3 treated with iloprost and 8 treated with
aspirin required above or below knee amputation. 64 (94%) of the iloprost group and 52 (80%) of the aspirin group completed the study; 8 patients treated with iloprost, but none from the aspirin group, fmished the trial between 21 and 28 days because of complete resolution of symptoms and signs. The results on an intention-to-treat basis are summarised in table 11. 58 patients in the iloprost group (85%) and 11 patients in the aspirin group (17%) were considered to be responders by an independent observer (p < 005). The difference in response was similar in patients with and without trophic changes. Complete relief of pain was achieved in 63% of patients treated with iloprost compared with 28% on aspirin; of patients with trophic changes, 35 % on iloprost treatment achieved complete healing compared with 13% on aspirin. Analgesic consumption was reduced in line with rest pain: 78% of the iloprost group and 25% of the aspirin group did not require any analgesia by the end of treatment. Mean ulcer area decreased from 3-0 cm to 0 cm at day 28 in iloprost-treated patients but did not change with aspirin therapy (30 cm before and after treatment). During the study, 1 patient on iloprost and 5 on aspirin had acute thrombophlebitis. Improvement of symptoms and signs of vasospasm was slightly greater in patients who received iloprost, but the differences were not significant; for example, incidence of colour change was reduced by 43% in the iloprost group and 40% in the aspirin group. There were no significant changes in distal doppler pressure in either treatment group.
treated with iloprost and 44 who received followed up 6 months after the start of treatment. As before, 88% in the iloprost group reported improvement, compared with 21 % in the aspirin group (increased from 17% at 28 days). 14 of the 18 patients who 51
patients aspirin were
TABLE II-RESULTS ON INTENTION-TO-TREAT ANALYSIS
557
achieved complete healing of trophic changes
on
iloprost
observed at 6 months, 13 of whom were still completely healed. 3 of the 5 patients who achieved complete healing of trophic changes on aspirin treatment were observed at 6 months, and were still completely healed. During treatment 2 patients from both groups required amputation; during the next 5 months, 1 patient who received iloprost, and 6 who were treated with aspirin,
treatment
were
required amputation. Independent assessment showed good agreement with that of the physician in charge of the care: for example, in assessment of ulcer healing and necrosis there was disagreement in only 2 patients treated with iloprost and 4 patients who received aspirin. Discussion Intravenous iloprost was significantly more effective than low-dose oral aspirin for relief of rest pain and healing of trophic lesions in patients with thromboangiitis obliterans. These results are more encouraging than those reported for prostanoids in arteriosclerosis obliterans 8--13 where placebo response rates can be as high as 60% .12,13 Ofthe few reports of prostanoids in the treatment of thromboangiitis obliterans, Olsson observed a woman with ischaemic ulcers which healed after treatment with prostaglandin Ep7 and Nyzankowski and co-workers,8 in a randomised study in 30 patients-15 of whom had Buerger’s disease--found a beneficial effect after intra-arterial prostacyclin infusion. Several observations should be made about the study design. Thromboangiitis obliterans is unusual in Europe, and many centres were required to obtain enough patients; although 1 centre included 26 patients, only 2 others had more than 10. To reduce any potential bias, patients were centrally randomised within each centre. Although the trial was designed to be double-blind, and parallel treatments were given, the side-effects of a prostanoid infusion may be obvious to both patient and physician, and the definition of response was partly subjective. To reduce the risk of physician bias, a second evaluation was made by an independent medical observer who only saw the patient before and after the course of treatment: there was good agreement between these medical opinions. The rate of response to aspirin was unexpectly low (16-9% at 28 days), the reasons for which are unclear. It is possible that aspirin might enhance the inflammatory process of thromboangiitis obliterans by an effect on leukotriene production, but as aspirin and other antiplatelet drugs are commonly prescribed in patients with arteriopathy, one would suspect such an effect to have already been observed. Moreover, Bollinger et ap4 have described a beneficial effect of high-dose aspirin in thromboangiitis obliterans. Another explanation could be the stringent classification that we used: to be considered a responder, a patient had to have reduced pain and analgesic requirement, an overall clinical improvement, and healing of trophic lesions, when present. In both groups, however, the amputation rate is lower than in most published series 5 Only 95 patients were available for follow-up at 6 months, so 38 (29%) patients were lost to long-term review. If it is assumed that all 17 missing patients who received iloprost died, whereas all 21 who received aspirin achieved full response, the final response rates at 6 months would become
iloprost versus 51 % for aspirin. This eventuality unlikely, but should be considered. In thromboangiitis obliterans, requirement for major amputation is the nearest to an entirely objective guide to therapeutic efficacy. In this study, the need for major amputation was low compared to most other reports and was not statistically different between the treatment groups. However, in the 2 patients on iloprost and 2 on aspirin who required amputation during treatment (days 1-28), it is possible that amputation was inevitable because of the initial severity of ischaemia. During the next 5 months, only 1 patient treated with iloprost required amputation, compared with 6 who received aspirin-a tendency that may become statistically significant with longer follow-up or more patients. Our findings indicate that intravenous iloprost for 21-28 days is well tolerated and has a greater effect than low-dose aspirin on symptoms and signs in patients with thromboangiitis obliterans. These encouraging results require further investigation, perhaps with amputation rate as a primary endpoint.
66% for seems
very
The principal investigators and participating centres in the TAO study were: Prof N. S. Angelides (Nicosia), Prof A. Vardar (Istambul), Prof A. Szczeklik (Krakow), Prof J. P. Barroy (Brussels), Dr D. Duprez (Gent), Prof H. Denck (Wien), Prof H. Ehringer (Wien), Prof G. Flora (Innsbruck), Prof H. Boccalon (Toulouse), Prof J. N. Fiessinger (Paris), Prof J. L. Guilmot (Tours), Dr D. Asteriadis (Kavala), Prof P. Balas (Athens), Prof K. Papadimitriou (Thessaloniki), Dr V. Colanceskj (Skopje), Dr L. Pirnat (Sarajevo), Dr P. Poredos (Ljubljana), ProfB. Radevic (Tuzla), Prof A. Sepic (Rijeka), Dr G. Baitsch (Bad Sackingen), Dr C. Diehm (Heidelberg), Dr B. Hagen (Berlin), ProfD. Loose (Hamburg), ProfD. Raithel (Nurmberg), Prof F. Spengel (Miinchen), Dr M. Vogelpohl (Ulm). Study coordinator: Dr M. D. Schafer; statistician: Ms U. Muller, monitoring: Ms K. Matz (all
Schering AG, Berlin). We thank Mr J.
Dormandy
for continued interest and review of the
manuscript. REFERENCES
1. Spittell JA. Thromboangiitis obliterans: an autoimmune disorder? N Engl J Med 1983; 308: 1157-58. 2. McPherson JR, Juergens JL, Gifford RW. Thromboangiitis obliterans and arteriosclerosis obliterans. Clinical and prognostic differences. Ann Intern Med 1963; 59: 288-96. 3. Hill GL. A rational basis for management of patients with the Buerger syndrome. Br J Surg 1974; 61: 476-81. 4. Nielubowicz J, Rosnowski A, Pruszynski B, et al. Natural history of Buerger’s disease. J Cardiovasc Surg 1980; 21: 529-40. 5. Mills JL, Taylor LM, Porter JM. Beurger’s disease in the modem era. Am J Surg 1987; 154: 123-29. 6. Sakaguchi S, Kusaba A, Mishima Y, et al. A multi-clinical double-blind study with PGE1 (cyclodextrin clathrate) in patients with ischemic ulcer of the extremities. Vasa 1978; 7: 263-66. 7. Olsson AG, Thyresson N. Healing of ischaemic ulcers
by intravenous prostaglandin E1 in a woman with thromboangiitis obliterans. Acta Derm Venereol (Stockh) 1978; 58: 467-72. 8. Nizankowski R, Krolikowski W, Bielatowicz J, et al. Prostacyclin for ischemic ulcers in peripheral arterial disease. A random assignment, placebo controlled study. Thromb Res 1985; 37: 21-28. 9. Carlson LA, Olsson AG. Intravenous prostaglandin E1 in severe peripheral vascular disease. Lancet 1976; ii: 810. 10. Szczeklik A, Nizankowski R, Skawinski S, Szczeklik J, Gluszko P, Gryglewski RJ. Successful therapy of advanced arteriosclerosis obliterans with prostacyclin. Lancet 1979; i: 1111-14. 11. Belch JJF, McKay A, McArdle B, et al. Epoprostenol (prostacyclin) and severe arterial disease. A double-blind trial. Lancet 1983; i: 315-17. 12. Cronenwett JL. The use of prostaglandins PGE1 and PGI2 in peripheral arterial ischemia. J Vasc Surg 1986; 2: 370-74. 13. Cronenwett JL, Zelenock GB, Whitehouse WM, et al. Prostacyclin treatment of ischemic ulcers and rest pain in unreconstructable peripheral arterial occlusive disease. Surgery 1986; 100: 369-75. 14. Bollinger A, Hollmann B, Schneider E, et al. Thromboangiitis obliterans: Diagnose und Therapie im licht neuer immunologischer Befunde. Schweiz Med Woschenschr 1979; 109: 537-43.
335
MEDICAL SCIENCE
Trial of iloprost versus aspirin treatment for critical limb ischaemia of thromboangiitis obliterans
FOR THE TAO
152
patients with thromboangiitis obliterans (Buerger’s disease) and pain from critical leg ischaemia were randomly allocated to receive iloprost, a chemically stable prostacyclin analogue, or low-dose aspirin, for 28 days in a double-blind trial. On review, 19 patients did not fulfil the stringent entry criteria. Of the other 133 patients, 98 also had leg ulcers. After 21-28 days, 58 (85%) of 68 iloprost-treated patients showed ulcer healing or relief of ischaemic pain, compared with 11 (17%) of 65 in the aspirin-treated group. 43 (63%) on iloprost treatment had complete relief of pain, compared with 18 (28%) on aspirin. Ulcers healed completely in 18 of 52 (35%) who received iloprost compared with 6 of 46 (13%) who received aspirin. 6 months after the start of treatment, the response rate was 45 of 51 (88%) patients treated with iloprost compared with 12 of 44 (21%) patients treated with aspirin.
STUDY*
but is
prevalent in Israel and parts of Asia. Thromboangiitis obliterans is typically observed in young men who smoke cigarettes heavily and, other than advice to stop smoking, its treatment remains controversial. Although rarely fatal, the incidence of major amputation is much higher than for atherosclerotic ischaemia-amputation rates of up to 70% have been reportedl-s-and because the disease mainly affects distal arteries, reconstructive surgery is rarely possible or successful. A beneficial effect of prostanoid treatment has been reported in some small studies.6-8 We have compared the effect of treatment by iloprost, a prostacyclin analogue, to that of low-dose aspirin in a randomised, controlled, double-blind, multicentre trial. more
Patients and methods Patients had
to
fulfil criteria for both critical limb ischaemia and
thromboangiitis obliterans. Critical ischaemia was defmed as continuous ischaemic rest pain in a limb for which continuous
Lancet 1990; 335: 555-57.
Introduction
Thromboangiitis rare cause *
obliterans (TAO; Buerger’s disease) is a of critical limb ischaemia in Europe and the USA,
Full list of participants and
centres at
end of article.
ADDRESSES: Service de Pathologie Vasculaire, et Centre Claude Bernard de Recherches sur les Maladies Vasculaires Périphériques, Hôpital Broussais, 75674 Paris, France (Prof J. N. Fiessinger, MD), and Clinical Research Department of Internal Medicine 1, Schering AG, Berlin, West Germany (Dr M. Schäfer, MD). Correspondence to Prof J. N. Fiessinger, Service de Pathologie Vasculaire, Médecine IV, Hôpital Broussais, 75674 Paris Cedex 14, France.
556
analgesia in hospital was required for at least 7 days, with or without tissue necrosis. For a diagnosis of thromboangiitis obliterans the following criteria had to be met: the patient had to be under 50 years of age, a smoker, with angiographic evidence of non-arteriosclerotic arterial occlusion below the knee, and at least 2 of the following 3 features GM: superficial thrombophlebitis; ischaemia of the arm; and/or typical arteriographic findings of skip lesions, direct collaterals (Martorell’s sign), and corkscrew collaterals. Patients with diabetes mellitus, other inflammatory vascular diseases, and amputation or lumbar sympathectomy in the preceding 3 months excluded. Patients were randomised within each centre to receive either oral aspirin (100 mg(day) and a 6-h daily placebo infusion, or a placebo tablet identical to aspirin and a 6-h infusion of iloprost. For the first 3 days, infusion was started at a rate equivalent to 0-5 ng/kg per min iloprost and increased every 30 min by a further 0-5ng/kg per min to a maximum tolerated dose or 2-0 ng/kg per min, which was maintained for the rest of the 6 h. The maximum dose used on day 3 was repeated on days 4-28 or until the patient’s symptoms and signs had completely resolved. Patients were followed for a further 5 months, when treatment was not controlled. Rest pain and analgesic consumption were recorded daily. In patients with trophic lesions the course of healing was assessed by surface area estimations. Distal peripheral pressures were measured weekly. An overall analysis, which included assessment of claudication, thrombophlebitis, and vasospastic symptoms, was made by the physician in charge of the patient. To investigate whether the appearance of symptoms caused by a prostacyclin analogue might influence the physician’s assessment, ischaemic lesions before and after treatment were also assessed by a physician not involved in the patient’s management and who was unaware of intervening
TABLE I-BASELINE CHARACTERISTICS OF STUDY GROUPS I
were
events.
At the end of treatment, patients were classified as responders or non-responders. Trophic lesions, rest pain, analgesic consumption, and overall improvement were rated by a 4-scale classification. Patients who entered the study with rest pain but without trophic lesions were considered to have responded if pain had been eliminated or strikingly reduced, with no or much reduced need for analgesics, and a striking overall improvement. Patients with
trophic lesions were considered to have responded if over 50% of the affected surface area had healed, with a striking overall improvement. Responder rates were compared by use of the X2 test. It was calculated that, to detect a 30% difference in response rate, a sample size of 53 patients would be needed if the type I error was (X = 0-05 and type II error was p = 0-1. Between January, 1986, and February, 1988, 152 patients were recruited at 26 centres in 9 countries. 19 patients (8 treated with iloprost, 11treated with aspirin) had to be excluded on review because they did not fulfil the stringent criteria for diagnosis of thromboangiitis obliterans. The baseline characteristics of the other 133 patients (68 treated with iloprost, 65 treated with aspirin) are shown in table i.
Results The two treatment groups had similar age distribution, incidence of ulceration or gangrene, duration of symptoms, previous therapy, concomitant diseases, other medications, and angiographic appearances (table i), and a similar distribution of symptoms of thrombophlebitis, claudication, and vasospasm. Smoking habits were also similar in both treatment groups: all patients were smokers, most of whom said that they had smoked 21-40 cigarettes per day for 11-20 years. Most gave up cigarettes at the start of treatment but, by day 28, 16 patients treated with iloprost and 13 treated with aspirin were still smoking. Headache, flushing, nausea, and abdominal cramps were more common in patients treated with iloprost but no patient in either group had to be withdrawn because of side-effects. 4 patients treated with iloprost and 13 treated
Values shown
with
as
numbers
aspirin
had
(%)
to
be withdrawn due
to
a
clinical
deterioration; of these, 3 treated with iloprost and 8 treated with
aspirin required above or below knee amputation. 64 (94%) of the iloprost group and 52 (80%) of the aspirin group completed the study; 8 patients treated with iloprost, but none from the aspirin group, fmished the trial between 21 and 28 days because of complete resolution of symptoms and signs. The results on an intention-to-treat basis are summarised in table 11. 58 patients in the iloprost group (85%) and 11 patients in the aspirin group (17%) were considered to be responders by an independent observer (p < 005). The difference in response was similar in patients with and without trophic changes. Complete relief of pain was achieved in 63% of patients treated with iloprost compared with 28% on aspirin; of patients with trophic changes, 35 % on iloprost treatment achieved complete healing compared with 13% on aspirin. Analgesic consumption was reduced in line with rest pain: 78% of the iloprost group and 25% of the aspirin group did not require any analgesia by the end of treatment. Mean ulcer area decreased from 3-0 cm to 0 cm at day 28 in iloprost-treated patients but did not change with aspirin therapy (30 cm before and after treatment). During the study, 1 patient on iloprost and 5 on aspirin had acute thrombophlebitis. Improvement of symptoms and signs of vasospasm was slightly greater in patients who received iloprost, but the differences were not significant; for example, incidence of colour change was reduced by 43% in the iloprost group and 40% in the aspirin group. There were no significant changes in distal doppler pressure in either treatment group.
treated with iloprost and 44 who received followed up 6 months after the start of treatment. As before, 88% in the iloprost group reported improvement, compared with 21 % in the aspirin group (increased from 17% at 28 days). 14 of the 18 patients who 51
patients aspirin were
TABLE II-RESULTS ON INTENTION-TO-TREAT ANALYSIS
557
achieved complete healing of trophic changes
on
iloprost
observed at 6 months, 13 of whom were still completely healed. 3 of the 5 patients who achieved complete healing of trophic changes on aspirin treatment were observed at 6 months, and were still completely healed. During treatment 2 patients from both groups required amputation; during the next 5 months, 1 patient who received iloprost, and 6 who were treated with aspirin,
treatment
were
required amputation. Independent assessment showed good agreement with that of the physician in charge of the care: for example, in assessment of ulcer healing and necrosis there was disagreement in only 2 patients treated with iloprost and 4 patients who received aspirin. Discussion Intravenous iloprost was significantly more effective than low-dose oral aspirin for relief of rest pain and healing of trophic lesions in patients with thromboangiitis obliterans. These results are more encouraging than those reported for prostanoids in arteriosclerosis obliterans 8--13 where placebo response rates can be as high as 60% .12,13 Ofthe few reports of prostanoids in the treatment of thromboangiitis obliterans, Olsson observed a woman with ischaemic ulcers which healed after treatment with prostaglandin Ep7 and Nyzankowski and co-workers,8 in a randomised study in 30 patients-15 of whom had Buerger’s disease--found a beneficial effect after intra-arterial prostacyclin infusion. Several observations should be made about the study design. Thromboangiitis obliterans is unusual in Europe, and many centres were required to obtain enough patients; although 1 centre included 26 patients, only 2 others had more than 10. To reduce any potential bias, patients were centrally randomised within each centre. Although the trial was designed to be double-blind, and parallel treatments were given, the side-effects of a prostanoid infusion may be obvious to both patient and physician, and the definition of response was partly subjective. To reduce the risk of physician bias, a second evaluation was made by an independent medical observer who only saw the patient before and after the course of treatment: there was good agreement between these medical opinions. The rate of response to aspirin was unexpectly low (16-9% at 28 days), the reasons for which are unclear. It is possible that aspirin might enhance the inflammatory process of thromboangiitis obliterans by an effect on leukotriene production, but as aspirin and other antiplatelet drugs are commonly prescribed in patients with arteriopathy, one would suspect such an effect to have already been observed. Moreover, Bollinger et ap4 have described a beneficial effect of high-dose aspirin in thromboangiitis obliterans. Another explanation could be the stringent classification that we used: to be considered a responder, a patient had to have reduced pain and analgesic requirement, an overall clinical improvement, and healing of trophic lesions, when present. In both groups, however, the amputation rate is lower than in most published series 5 Only 95 patients were available for follow-up at 6 months, so 38 (29%) patients were lost to long-term review. If it is assumed that all 17 missing patients who received iloprost died, whereas all 21 who received aspirin achieved full response, the final response rates at 6 months would become
iloprost versus 51 % for aspirin. This eventuality unlikely, but should be considered. In thromboangiitis obliterans, requirement for major amputation is the nearest to an entirely objective guide to therapeutic efficacy. In this study, the need for major amputation was low compared to most other reports and was not statistically different between the treatment groups. However, in the 2 patients on iloprost and 2 on aspirin who required amputation during treatment (days 1-28), it is possible that amputation was inevitable because of the initial severity of ischaemia. During the next 5 months, only 1 patient treated with iloprost required amputation, compared with 6 who received aspirin-a tendency that may become statistically significant with longer follow-up or more patients. Our findings indicate that intravenous iloprost for 21-28 days is well tolerated and has a greater effect than low-dose aspirin on symptoms and signs in patients with thromboangiitis obliterans. These encouraging results require further investigation, perhaps with amputation rate as a primary endpoint.
66% for seems
very
The principal investigators and participating centres in the TAO study were: Prof N. S. Angelides (Nicosia), Prof A. Vardar (Istambul), Prof A. Szczeklik (Krakow), Prof J. P. Barroy (Brussels), Dr D. Duprez (Gent), Prof H. Denck (Wien), Prof H. Ehringer (Wien), Prof G. Flora (Innsbruck), Prof H. Boccalon (Toulouse), Prof J. N. Fiessinger (Paris), Prof J. L. Guilmot (Tours), Dr D. Asteriadis (Kavala), Prof P. Balas (Athens), Prof K. Papadimitriou (Thessaloniki), Dr V. Colanceskj (Skopje), Dr L. Pirnat (Sarajevo), Dr P. Poredos (Ljubljana), ProfB. Radevic (Tuzla), Prof A. Sepic (Rijeka), Dr G. Baitsch (Bad Sackingen), Dr C. Diehm (Heidelberg), Dr B. Hagen (Berlin), ProfD. Loose (Hamburg), ProfD. Raithel (Nurmberg), Prof F. Spengel (Miinchen), Dr M. Vogelpohl (Ulm). Study coordinator: Dr M. D. Schafer; statistician: Ms U. Muller, monitoring: Ms K. Matz (all
Schering AG, Berlin). We thank Mr J.
Dormandy
for continued interest and review of the
manuscript. REFERENCES
1. Spittell JA. Thromboangiitis obliterans: an autoimmune disorder? N Engl J Med 1983; 308: 1157-58. 2. McPherson JR, Juergens JL, Gifford RW. Thromboangiitis obliterans and arteriosclerosis obliterans. Clinical and prognostic differences. Ann Intern Med 1963; 59: 288-96. 3. Hill GL. A rational basis for management of patients with the Buerger syndrome. Br J Surg 1974; 61: 476-81. 4. Nielubowicz J, Rosnowski A, Pruszynski B, et al. Natural history of Buerger’s disease. J Cardiovasc Surg 1980; 21: 529-40. 5. Mills JL, Taylor LM, Porter JM. Beurger’s disease in the modem era. Am J Surg 1987; 154: 123-29. 6. Sakaguchi S, Kusaba A, Mishima Y, et al. A multi-clinical double-blind study with PGE1 (cyclodextrin clathrate) in patients with ischemic ulcer of the extremities. Vasa 1978; 7: 263-66. 7. Olsson AG, Thyresson N. Healing of ischaemic ulcers
by intravenous prostaglandin E1 in a woman with thromboangiitis obliterans. Acta Derm Venereol (Stockh) 1978; 58: 467-72. 8. Nizankowski R, Krolikowski W, Bielatowicz J, et al. Prostacyclin for ischemic ulcers in peripheral arterial disease. A random assignment, placebo controlled study. Thromb Res 1985; 37: 21-28. 9. Carlson LA, Olsson AG. Intravenous prostaglandin E1 in severe peripheral vascular disease. Lancet 1976; ii: 810. 10. Szczeklik A, Nizankowski R, Skawinski S, Szczeklik J, Gluszko P, Gryglewski RJ. Successful therapy of advanced arteriosclerosis obliterans with prostacyclin. Lancet 1979; i: 1111-14. 11. Belch JJF, McKay A, McArdle B, et al. Epoprostenol (prostacyclin) and severe arterial disease. A double-blind trial. Lancet 1983; i: 315-17. 12. Cronenwett JL. The use of prostaglandins PGE1 and PGI2 in peripheral arterial ischemia. J Vasc Surg 1986; 2: 370-74. 13. Cronenwett JL, Zelenock GB, Whitehouse WM, et al. Prostacyclin treatment of ischemic ulcers and rest pain in unreconstructable peripheral arterial occlusive disease. Surgery 1986; 100: 369-75. 14. Bollinger A, Hollmann B, Schneider E, et al. Thromboangiitis obliterans: Diagnose und Therapie im licht neuer immunologischer Befunde. Schweiz Med Woschenschr 1979; 109: 537-43.
