Effect of Amezinium Metilsulfate on the Finger Skin Vasoconstrictor Response to Cold Stimulation and Venoconstrictor Response to Noradrenaline

Jpn Circ J 1998; 62: 824–828

Kazuhiro Harada, MD; Masami Ohmori, MD; Akio Fujimura, MD; Kyoichi Ohashi, MD*

Orthostatic hypotension can be caused by an inadequate vasoconstrictor response. The effects of amezinium on vasoconstrictor response to sympathetic stimulation and to exogenous noradrenaline were investigated and compared with those of midodrine. In 8 healthy men, the following experiments were performed after a single oral dose of 10 mg of amezinium, 2 mg of midodrine or a placebo. First, finger-tip blood flow (FTBF) was recorded using a laser Doppler flowmeter before and during the contralateral hand cooling and a reduction ratio of FTBF was calculated as an index of the vasoconstrictor response. Second, dose-response curves to increasing doses (1–512ng/min) of noradrenaline infused locally to the dorsal hand vein were determined using a linear variable differential transformer. The reduction ratio of FTBF was significantly increased (p<0.05) by amezimium [placebo, 75.9±9.8(mean±SD)%; amezinium, 85.1±7.9%; midodrine, 78.1±9.3%]. The infusion rate of noradrenaline producing a half-maximum venoconstriction was significantly decreased (p<0.05) by amezinium (placebo, 40.6±33.9ng/min; amezinium, 21.0±21.3ng/min; midodrine, 33.2±31.5ng/min). These findings indicate that amezinium increases the vasoconstrictor response to sympathetic stimulation and to noradrenaline in normal subjects, and this mechanism might contribute to the improvement by amezinium of the symptoms of orthostatic hypotension. (Jpn Circ J 1998; 62: 824–828) Key Words: Amezinium; Dorsal hand vein method; Laser Doppler, Midodrine; Vasoconstrictor response

hen a person assumes an upright posture, an methods were thought to be suitable and were used in the increase of sympathetic outflow and an inhibi- present study, because they are sensitive and noninvasive W tion of parasympathetic activity occur re- and are used for the evaluation of autonomic disturbances flexly,1,2 which increases the heart rate and vascular resis- or orthostatic hypotension. First, the finger-tip vasocon- tance in order to maintain systemic blood pressure. strictor response to cold stimulation on the contralateral Orthostatic hypotension may be caused by failure of these hand was examined using a laser Doppler flowmeter, which responses,3,4 and for the treatment it is important to select a is an appropriate means for the study of the vasomotor drug considering its pharmacological properties and the response mediated byα-adrenoceptor stimulation.7,8 Second, pathophysiology of the disease. At present, vasoconsticting the venoconstrictor response to locally infused noradrena- agents are often used for the treatment of orthostatic line was examined using a linear variable differential trans- hypotension. To understand the mechanism of these drugs, former (LVDT), which provides a sensitive and easy repro- their effects not only on systemic blood pressure but also ducible means of assessing the effect of a drug on periph- on the vasoconstrictor response to sympathetic stimulation eral veins.9 As the decrease in venous blood return to the should be examined. However, to our knowledge, such heart is involved in the pathogenesis of orthostatic information is limited. hypotension,3 it is interesting to also examine the effect of Amezinium metilsulfate (amezinium) is one of the most these drugs on veins. popular drugs in Japan for the treatment of orthostatic In the present study, we examined, using these methods, hypotension. Amezinium has a unique mechanism of vaso- the effects of amezinium on the vasoconstrictor responses, constriction; that is, it inhibits monoamine oxidase (MAO) and compared them in healthy subjects with those of activity and uptake of noradrenaline,5,6 which prompted us midodrine hydrochloride (midodrine), anα1-adrenoceptor to examine whether it influences the vasoconstrictor agonist,10 which is also a popular drug for orthostatic hypo- response to sympathetic stimulation and toα-adrenoceptor tention. stimulation by noradrenaline in humans. To examine the effects of the drug, the following 2 Methods (Received April 10, 1998; revised manuscript received July 9, 1998; Subjects accepted July 14, 1998) Eight healthy Japanese males (21–26 years, 64–78 kg) Department of Clinical Pharmacology, Jichi Medical School, Tochigi were selected. They were normotensive and nonsmokers, and *Department of Clinical Pharmacology, Hamamatsu University without concomitant medication for at least 2 weeks before School of Medicine, Shizuoka, Japan Mailing address: Kazuhiro Harada, MD, Department of Clinical the study. The study was approved by the Ethical Pharmacology, Jichi Medical School, Minamikawachi-machi, Committee of Jichi Medical School and all subjects gave Kawachi-gun, Tochigi 329-0498, Japan their informed consent.

Japanese Circulation Journal Vol.62, November 1998 Effect of Amezinium on Vasoconstrictor Response 825

Table 1 Blood Pressure and Pulse Rate in the Supine Position Table 2 Emax and ED50 in Response to Noradrenaline Infusion and 1min After Standing Up After the Oral Intake of a Placebo, After the Oral Dosing With Placebo, Amezinium or Midodrine Amezinium and Midodrine (Mean±SD, n=8) Placebo Amezinium Midodrine Placebo Amezinium Midodrine Emax (%) 100.8±9.9 99.2±11.4 102.6±12.0 SBP (mmHg) ED50 (ng/min) 40.6±33.9 21.0±21.3* 33.2±31.5 Supine 113±10 117±11 115±11 Standing 122±13 128±15 127±15 Mean±SD, n=8. Emax, maximum venoconstrictory effect; ED50, infusion rate producing a half-maximum response. *p<0.05 vs placebo. DBP (mmHg) Supine 67±6 69±9 68±8 Standing 71±8 74±12 72±10 PR (beats/min) which are the recommended starting doses in Japan for 11,12 Supine 62±8 61±7 62±7 these drugs, or a placebo. Two and half hours after Standing 79±16 74±12 77±13 amezinium, around the tmax of the drug,13 1.5h after midodrine, around the tmax of its active metabolite, dimethoxyphenyl- SBP, systolic blood pressure; DBP, diastolic blood pressure; PR, pulse rate. aminoethanol,14 or 2 h after placebo, the finger-tip skin vasoconstrictor response to cold stimulation was assessed, followed by assessment of the venoconstrictor response to Finger-Tip Blood Flow (FTBF) Measurement noradrenaline infusion. The measurements started around The subjects were comfortably seated with their right 10.30h. Between the preceding assessments (ie around 2h arm resting on a table at the heart level. Cutaneous FTBF and 50min after amezinium, 1hour and 50min after mido- was recorded from the pad of the right midfinger by a laser drine, and 2h and 20min after the placebo) blood pressure Doppler flowmeter (Peri Flux PF3, Perimed, Sweden) with and pulse rate were measured both in the supine position an integrating probe with 7 efferent fibers. Laser Doppler and 1min after standing up, using a semiautomated sphyg- flow was expressed in arbitrary units. The recording of momanometer (BP-103iII, Nihon Colin, Komaki, Japan). FTBF was started after the subjects rested for more than 10 The experiments were performed in a room with the min, and when the value of FTBF stabilized, cold stimula- ambient temperature controlled at 24–26°C. tion was performed by immersing the left hand in ice water (1–3°C) for 15sec. FTBF rapidly decreased with a latent Data Evaluation time of a few seconds and gradually returned to the level FTBF: Data were expressed in arbitrary units. Effects before the stimulation. of contralateral hand cooling on FTBF were quantitated using the following equation: Dorsal Hand Vein Method The LVDT used (Schaevits type 025MHR, Pennsauken, Reduction Ratio (RR) NJ, USA) has a freely movable core that rests over the =[(FTBFbef –FTBFmin)/FTBFbef]×100% center of the vein to be studied. There is a linear relation- ship between the vertical movement of the core and the where FTBFbef is the blood flow recorded just before the change in voltage output. During the study, the subject was cold stimulation and FTBFmin is the minimum blood flow in the supine position with an arm placed on a support recorded during the cold stimulation. sloping upward at an angle of about 30 degrees from the Dorsal Hand Vein Constriction: The individual dose- horizontal. In this position, emptying of the superficial response curve was analyzed by the following equation: hand vein occurred. A 25-gauge needle was inserted into a dorsal hand vein and a continuous infusion of physiological Effect=[Emax ×ex/(ED50 +ex)]×100% saline was started. The core was placed over the center of the vein about 10mm downstream from the tip of needle. where x is the infusion rate of noradrenaline, Emax is the The voltage output was recorded on a strip chart recorder maximum venoconstrictor effect, and ED50 is the infusion (Linearcorder Mark VII; Graphtec, Yokohama, Japan). Ten rate producing a half- maximum response. The data were minutes after the insertion of the needle, increasing doses fitted to this equation by the least-squares method, provid- (from 1 to 512ng/min) of noradrenaline (Nor-Adrenalin®, ing estimates of ED50 and Emax. Sankyo, Tokyo) were infused into the dorsal hand vein and Data are expressed as mean±SD. Statistical analysis for recordings of the position of the core were made before and the differences between the drugs and placebo were per- after inflation of a sphygmomanometer cuff on the same formed by analysis of variance (ANOVA) followed by arm to 45mmHg. Drug infusion lasted for 5min and the cuff Scheffe’s multi-range test. A probability value of p<0.05 was inflated for the last 1min during each infusion period. was considered as significant.

Study Protocol Subjects were instructed to refrain from caffeine- Results containing and alcoholic beverages from 12h before the Blood Pressure and Pulse Rate start of the study. No breakfast was taken on the study Blood pressure and pulse rate in the supine position and days. The study was carried out as a single, blind, placebo- 1min after standing up are shown in Table1. No signiﬁcant controlled, 3 period cross-over study with an interval of differences were observed between placebo and drug treat- 1–2 weeks. The subjects received a single oral dose of 10 ments in any of these parameters. mg of amezinium metilsulfate (amezinium) (Rismic®, Dainippon, Osaka, Japan), 2 mg of midodrine hydrochlo- FTBF Measurements ride (midodrine) (Metligine®, Taisyo, Tokyo, Japan), There were no significiant differences in the FTBF

Japanese Circulation Journal Vol.62, November 1998 826 HARADA K et al.

Fig1. Reduction ratio of finger tip blood flow after oral dosing with Fig2. Dose-response curves to noradrenaline after oral dosing with placebo, amezinium and midodrine (n=8). Values are expressed as placebo (○), amezinium (●) and midodrine (■) (n=8). Values are mean with standard deviation. *p<0.05 vs placebo. expressed as mean. before the cold stimulation between placebo and drug grating probe employing a bundle of 7 efferent fibers, treatments (198±17 units for placebo, 201±20 units for which increases the total measuring volume in order to amezinium, and 191±18 units for midodrine). The minimum allow a more precise detection of the blood flow.24 FTBF during the cold stimulation after amezinium, but not Although measurements obtained with a laser Doppler midodrine, treatment was significantly less (p<0.05) than flowmeter are of a relative nature, the method is particu- that after placebo (48±12 units for placebo, 31±13 units for larly appropriate for the study of local skin vasomotor amezinium, and 42±11 units for midodrine). The RR of responses in human subjects.9,25 Cold stimulation rapidly FTBF after the cold stimulation for amezinium, but not for increases an endogenous noradrenaline concentration, midodrine, was significantly (p<0.05) increased (75.9±9.8% which, in turn, causes α-adrenoceptor-mediated vasocon- for placebo, 85.1±7.9% for amezinium and 78.1±9.3% for striction and a reduction in blood flow.26 This method is midodrine) (Fig1). used for the evaluation of autonomic failure,25,27,28 including Shy-Drager syndrome, Parkinson’s disease, amyloido- Venoconstrictor Response to Noradrenaline sis and diabetic neuropathy. We demonstrated that α1- Dose-response curves for the effect of noradrenaline adrenoceptor antagonists, which could induce orthostatic infusion on dorsal hand vein constriction after the 3 treat- hypotension, suppress the finger skin vasoconstrictor ments are shown in Fig2. The values of the Emax for the 3 response and the degree of the suppression is correlated treatments were comparable. However, after amezinium but with the drug plasma concentration.8 not after midodrine treatment, ED50 was significantly (p< Second, the venoconstrictor response to noradrenaline 0.05) decreased compared with that of placebo (Table2). was examined using LVDT. The dorsal hand vein technique provides a sensitive, easy and reproducible means of assessing the effect of drugs on peripheral veins. The Discussion effects of oral drugs on venoconstrictor response caused by Drugs used for the treatment of orthostatic hypotension the infusion of a vasoconstricting agent can be detected include fludrocortisone acetate, a synthetic mineralcorti- without affecting systemic hemodynamics,9 because the coid,15 inhibitors of prostaglandin synthesis, such as measurement is made within the vicinity of the infusion indomethacin,16 phenylpropanolamine, a sympathomimetic site, allowing the amount of the agent to be extremely agent,17 and caffeine.18β-Blockers19 and disopyramide20 are small. The venoconstrictor response of dorsal hand veins to effective for vasovagal syncope, a common cause of faint- phenylephrine or noradrenaline is reported to reflect the ness. In Japan, amezinium11,21 and midodrine,22,23 both of responsiveness of the foot vein29 and the overall vascular which are vasoconstricting agents, are widely used for system.30 Several investigators have applied this method to various type of orthostatic hypotension because of their the evaluation of the pathophysiological conditions of excellent safety and good efficacy. patients with orthostatic hypotension31 or diabetic auto- Orthostatic hypotension is induced by a variety of disor- nomic neuropathy.32 ders, including inadequate vasoconstrictor response to Reproducibility of the finger-tip vasoconstrictor response sympathetic stimulation on standing.3,4 Therefore, it is to cold stimulation33 and the venoconstrictor response to interesting to investigate the effects of the drugs on vaso- noradrenaline34 were reported to be excellent. In the present constrictor reponse to sympathetic or adrenoceptor stimula- study, we examined the vascular responses under standard- tions. In the present study, we examined whether the drugs ized conditions including room temperature, time of the have such effects or not in normal subjects using 2 nonin- measurements and effect of food, to ensure the reliability of vasive methods. This is a new approach to evaluating the the experiments. drugs used for orthostatic hypotension. The 10mg of amezinium and 2mg of midodrine used in First, the finger-tip vasoconstrictor response to cold this study were the recommended starting oral doses in stimulation was examined using a laser Doppler flowmeter. Japan for the treatment of orthostatic hypotension. The Laser Doppler analysis allows real-time and noninvasive difference in pharmacokinetic properties of the drugs measurement of skin blood circulation.7 We used an inte- prompted us to select different time points for this single

Japanese Circulation Journal Vol.62, November 1998 Effect of Amezinium on Vasoconstrictor Response 827 dose comparative study. As the tmax of amezinium13 and nously infused noradrenaline in healthy subjects. Further that of dimethoxyphenyl-aminoethanol (DMAE),14 an active studies are needed to evaluate this drug in patients with metabolite of midodrine, are around 2.5h and 1.5h, respec- orthostatic hypotension. tively, the measurements were scheduled at 2.5h, 1.5h and 2 h after amezinium, midodrine, and placebo administra- References tion, respectively. In the present study, the increase in blood pressure by 1. Burke D, Sundlö G, Wallin BG: Postura effects on muscle nerve sympathetic activity in man. J Physiol (Lond) 1977; 272: 399–414 amezinium or midodrine was not statistically significant. 2. Eckberg DL: Parasympathetic cardiovascular control in human Other studies in healthy volunteers13,14 also showed that disease: a critical review of methods and results. Am J Physiol 1980; neither amezinium nor midodrine at the doses used in this 239: H581–H593 study caused significant change in blood pressure or pulse 3. Streeten DHP, Anderson GH Jr, Richardson R, Thomas FD: Abnormal orthostatic changes in blood pressure and heart rate in rate. However, pretreatment with amezinium at the dose subjects with intact sympathetic nervous function: evidence of exces- used in this study increased the finger-tip vasoconstrictor sive venous pooling. J Lab Clin Med 1988; 111: 326–335 response to cold stimulation and venoconstrictor response 4. Schapel GJ, Betts WH: The effect of a single oral dose of prazosin on to noradrenaline. These findings indicate that amezinium venous reflex response, blood pressure and pulse rate in normal enhances vasoconstriction induced by sympathetic stimula- volunteers. Br J Clin Pharmacol 1981; 12: 873–881 5. Steppler A, Pfandler R, Hedler L, Starke K: An analysis of the tion and venoconstriction induced by noradrenaline, and effects of amezimium on postganglionic sympathetic neurones. this mechanism by amezinium might contribute to the Naunyn-Schmiedeberg’s Arch Pharmacol 1980; 314: 1–11 improvement of the symptoms of orthostatic hypotension. 6. Lenke D, Gries J, Kretzschman R: Pharmacology of amezinium, a On the other hand, midodrine did not exert such an effect. novel antihypotensive drug/III. Studies on the mechanism of action. Arzneimittel forschung/Drug Res 1981; 31: 1558–1565 The mechanism of these findings might be explained by the 7. Holloway GA Jr, Watkins DW: Laser doppler measurement of cuta- pharmacological properties obtained in in-vitro studies. neous blood flow. J Invest Dermatol 1977; 69: 306–309 Amezinium is shown to inhibit MAO activity inside nora- 8. Harada K, Ohashi K, Fujimura A, Kumagai Y, Ebihara A: Effects of drenergic neurons and the uptake of noradrenaline in the α1-adrenoceptor antagonists, prazosin and urapidil, on a finger skin 5,6 vasoconstrictor response to cold stimulation. Eur J Clin Pharmacol peripheral monoamine nervous system. Therefore, it can 1996; 49: 371–376 be speculated that the activity of noradrenaline released by 9. Aellig WH: Clinical pharmacology, physiology and pathophysiology cold stimulation and exogenously infused noradrenaline, of superficial veins: 1. Br J Clin Pharmacol 1994; 38: 181–196 which is inactivated by MAO, is enhanced by amezinium 10. Pittner H, Stormann H, Enzenhofer R: Pharmacodynamic action of Midodrine, a new α-adrenergic stimulating agent, and its main pretreatment. On the other hand, this study showed that metabolite, ST1059. Arzneimittelforschung/Drug Res 1976; 26: midodrine did not affect the responsiveness of vasocon- 2145–2154 striction to the stimulation, and this might be because it 11. Nagasawa K, Ohta M, Kawaguchi N: Clinical evaluation of directly stimulatesα1-adrenoceptors on blood vessels.10 It amezinium metilsulfate (LU-1631) on orthostatic dysregulation would enhance vascular tone irrespective of the stimula- syndrome and hypotension: Optimal Dosage. J Clin Therap Med 1988; 4: 1103–1122 (in Japanese) tion, although this phenomenon is hard to detect by the 12. Nagata K, Kure S, Komiyama H: Influence on hemodynamics and methods used in the present study. According to studies pharmacokinetics of healthy volunteers in multiple oral administra- using the doral hand vein method,31,32 some patients with tion of TS-701 (midodrine hydrochloride). Jpn Pharmacol Ther orthostatic hypotension show increased venoconstriction to 1987; 15: 1225–1238 (in Japanse) 13. Nakashima M, Kanamaru M: Phase I study of amezinium metilsul- exogenous noradrenaline due to denervation. Midodrine fate. Antihypotensive Drug (No. 1): Single dose study in healthy might be preferable in such patients, and this assumption is volunteers. J Clin Therap Med 1988; 4: 495–507 (in Japanese) supported by a study in patients with diabetic orthostatic 14. Tsutsui S: Safety and pharmacokinetics study of TS-701 in healthy hypotension.35 voluntees. Clin Rep 1984; 21: 694–712 (in Japanese) 15. Campbell, IW, Ewing, DJ, Clarke, BF: 9-alpha-fluorohydrocortisone in the treatment of postural hypotension in diabetic autonomic Study Limitations neuropathy. Diabetes 1975; 24: 381 This study was performed in 8 healthy young subjects, 16. Kochar MS, Itskovitz HD: Treatment of idiopathic orthostatic using a single oral dose to characterize the properties of the hypotension (Shy-Drager syndrome) with indomethacin. Lancet 1978; 1: 1011–1014 drugs. Further studies are needed to elucidate whether the 17. Biaggioni I, Onrot J, Steward CK, Robertson D: The potent pressor findings are similar after long treatment with the drugs and effect of phenylpropanolamine in patients with autonomic impair- to evaluate the effects of the drugs on the vasoconstrictor ment. JAMA 1987; 258: 236–239 responses as well as on blood pressure or symptoms due to 18. Onrot J, Goldberg MR, Biaggioni I, Hollister AS, Kincaid D, hypotension in older subjects or in patients with various Robertson D: Hemodynamic and humoral effects of caffeine in autonomic failure: therapeutic implications for postprandial hypotension. types of orthostatic hypotension. N Engl J Med 1985; 313: 549–555 From the findings observed in this study, amezinium 19. Sra JS, Murthy VS, Jazayeri MR, Shen YH, Troup PJ, Avitall B, et might be effective for the treatment of patients with an al: Use of intravenous esmolol to predict efficacy of oral beta-adren- impaired vasoconstrictor response. Applying the methods ergic blocker therapy in patients with neurocardiogenic syncope. J Am Coll Cardiol 1992; 19: 402–408 used in the present study in further investigations of the 20. Milstein S, Buetikofer J, Dunnigan A, Benditt DG, Gornick C, Reyes pathophysiology of orthostatic hypotension, in order to WJ: Usefulness of disopyramide for prevention of upright tilt- characterize the effects of the drugs in such patients, might induced hypotension-bradycardia. Am J Cardiol 1990; 65: 1339 – give suggestions for the selection of the most appropriate 1344 21. Oloff SH: A synopsis of the therapeutic trials of Amezinium. drug in clinical practice. Arzneimittel forschung/Drug Res 1981; 31: 1657–1671 22. Kaufmann H, Brannan T, Kradoff L, Yahr MD, Mandeli J: Treatment of orthostatic hypotension due to autonomic failure with a Conclusion peripheral alpha-adrenergic agonist (midodrine). Neurology 1988; 38: 951–956 In the present study, we showed, using a new approach, 23. Koyama S, Matsubara T, Yoshifusa A, Ohshima M, Yamaguchi T, that a single oral dose of amezinium increases the vasocon- Yamazaki Y, et al: A case of vasovagal syncope with convulsions: strictor response to sympathetic stimulation and exoge- The effects of midodrine hydrochloride. Jpn Circ J 1992; 56: 950–

Japanese Circulation Journal Vol.62, November 1998 828 HARADA K et al.

954 systemic vascular responses. Clin Pharmacol Ther 1992; 51: 68–75 24. Salerud EG, Nilsson GE: Integrating probe for tissue laser doppler 31. Miller JW, Streeten DHP: Vascular responsiveness to norepinephrine flowmeters. Med Biol Eng Comput 1986; 24: 415–419 in sympathicotonic orthostatic intolerance. J Lab Clin Pharmacol 25. Low PA, Neumann C, Dyck PJ, Fealey RD: Evaluation of skin vaso- 1990; 5: 549–558 motor reflexes by using laser doppler velocimetry. Mayo Clin Proc 32. Eichler HG, Blaschke TF, Kraemer FB, Ford GA, Blochl-Daum B, 1983; 58: 583–592 Hoffman BB: Responsiveness of superficial hand veins toα-adreno- 26. Ekenvall L, Lindblad LE, Norbeck O, Etzell BM:α-Adrenoceptors ceptor agonists in insulin-dependent diabetic patients. Clin Sci 1992; and cold-induced vasoconstriction in human finger skin. Am J 82: 163–168 Physiol 1988; 255: H1000–H1003 33. Müller PH, Keller R, Imhof P: Laser doppler flowmetry, a reliable 27. Baron R, Feldmann R, Lindner V: Small fibre function in primary technique for measuring pharmacologically induced changes in cuta- autonomic failure. J Neurol 1993; 241: 87–91 neous blood flow? Methods Find Exp Clin Pharmacol 1987; 9: 409– 28. Hatanaka H, Matsumoto S, Ishikawa K, Kawasaki T, Kunota S, 420 Takagi K: Fundamental studies on the measurement of skin blood 34. Alradi AO, Carruthers SG: Evaluation and application of the linear flow by periflux laser doppler flowmeter and its clinical application. variable differential transformer technique for the assessment of Rinshyo-Byori 1984; 9: 1025–1028 (in Japanese) human dorsal hand vein alpha-receptor activity. Clin Pharmacol 29. Aellig WH: Superficial hand and foot veins show no difference in Ther 1985; 38: 495–502 sensitivity to constrictor agents. Clin Pharmacol Ther 1990; 48: 96– 35. Takahashi Y, Oowada K, Kawagoe C, Inoue Y, Hirata Y: 101 Classification and treatment of orthostatic hypotension in diabetes. 30. Vincent J, Blaschke TF, Hoffman BB: Vascular reactivity to Ann Jpn Soc Res Diabet Complicat 1990; 3: 99–104 (in Japanese) phenylephrine and angiotensin II: comparison of direct venous and

Japanese Circulation Journal Vol.62, November 1998