Effects of SR 49059, a New Orally Active and Specific Vasopressin V1 Receptor Antagonist, on Vasopressin-Induced Vasoconstriction in Humans
Abstract We have evaluated the efficacy of SR 49059, a new orally active and specific vasopressin V1 receptor antagonist (arginine-vasopressin [AVP]), in the blockade of the vascular effects of exogenous AVP in healthy subjects. In preliminary experiments, two procedures to measure the V1 vascular effects of AVP were assessed. First, the AVP-induced changes in skin blood flow were investigated by the injection of increasing doses of AVP intradermally, with or without a previous local vasodilation with calcitonin gene–related peptide (CGRP). In a second protocol, AVP was infused intra-arterially, and the changes in radial artery diameter and blood flow were measured. The intradermal injection of AVP caused significant decreases in skin blood flow, and the use of CGRP increased the sensitivity of the method by a factor of 102 to 103. AVP infused intra-arterially caused dose-dependent decreases in the radial artery diameter and blood flow. In the main study, the potency and efficacy of SR 49059 to block the AVP-induced changes in skin blood flow were assessed in 12 healthy men with a double-blind, triple crossover study design. The subjects were randomized to receive a placebo orally and 30 mg and 300 mg of the antagonist at a 1-week interval. The subjects were then further randomized to evaluate the efficacy of the same doses of the antagonist to block the vasoconstriction of the radial artery induced by an intra-arterial infusion of AVP. SR 49059 inhibits, dose-dependently and significantly, the AVP-induced changes in skin blood flow, with a peak effect occurring between 2 and 6 hours after injection. In addition, the 300-mg dose of SR 49059 completely blocked the vasoconstriction of the radial artery induced by AVP. In conclusion, skin blood-flow measurement, after intradermal injection of AVP on a skin area vasodilated with CGRP, is an effective method to investigate the V1 vascular effect of AVP in humans. SR 49059 is a potent and specific antagonist of V1 receptors, which blocks the AVP-induced vasoconstriction.
A recent trend in cardiovascular drug research is the development of new compounds that act on very specific targets, such as enzymes or well-defined cell surface receptors. Several receptor antagonists have become available that block various receptors involved in the regulation of blood pressure or flow. These receptor antagonists include angiotensin AT1 and AT2,1 endothelin,2 bradykinin,3 and vasopressin V1 and V2 receptors.4 5 6
Several specific nonpeptide antagonists of the V1 vasopressin receptor have been characterized and shown in experimental animals to block the increase in blood pressure induced by exogenous AVP.4 5 7 8 The initial goal of investigations of such new receptor antagonists in humans is to demonstrate their efficacy and specificity and to establish a dose-response curve of the inhibition of the AVP-induced effects. In many cases, the intravenous injection of the agonist peptide is feasible with little, if any, risk, even if the peptide has vasoconstrictor properties. In the case of AVP, the peptide has potent vasoconstrictor properties, which may cause transient myocardial ischemia and transmural infarction with dysrhythmia when infused intravenously.9 10 11 This complication is explained by the high responsiveness of the coronary artery tree to the vascular effects of vasopressin.9 10 11 Because of their potential role in the pathophysiology of hypertension and congestive heart failure, several V1 receptor antagonists have been evaluated.4 5 6 7 8 So far, only one of the nonpeptide, orally active V1 antagonists has been evaluated in humans with the administration of exogenous AVP.12 In that study, AVP was infused intra-arterially to achieve lower concentrations of the peptide in the general circulation, and the changes in forearm blood flow were measured by strain-gauge plethysmography with a venous occlusion technique.12
We have evaluated, with a double-blind, placebo-controlled study design, the potency and efficacy of SR 49059, a new orally active V1 receptor antagonist, in normotensive subjects. SR 49059 shows a particularly high affinity for the V1a vasopressin receptor (Ki in the range of 1.1 to 6.3 nmol in human tissues) and no in vitro effect on the V2 receptor.5 In this study, two technical approaches were used. With the first, increasing doses of AVP were administered intradermally, and the changes in skin capillary blood flow were measured by laser Doppler flowmetry. With the second approach, AVP was infused into the brachial artery, and the AVP-induced changes in the radial artery diameter and blood flow were measured directly with a high-resolution echotracking device combined with a continuous-wave Doppler.
The goal of the preliminary study was to establish methods and was conducted in 10 healthy male volunteers aged 21 to 34 years. The protocols were approved previously by the Hospital Ethics committee. The outline of the protocols and the potential risks were explained, and a signed informed consent was obtained from each participant. Both protocols followed procedures that were in accord with our institutional guidelines. The volunteers were asked to abstain from alcohol, caffeine, and cigarette smoking for a minimum of 24 hours before the study day. All experiments were performed with subjects in the supine position.
In this first protocol, the AVP-induced changes in capillary skin blood flow were investigated by the ID injection of increasing doses of AVP on the internal face of the forearm. Before injection of the peptide, four to five injection sites were determined on the forearm, and a baseline measurement of skin blood flow was determined on each site by laser Doppler flowmetry (Periflux PF 4D, Perimed KB). AVP (Pitressin, Parke-Davis Inc, 1 mL=20 IU) was then injected strictly intradermally with a 0.3-mm-diameter Microlance needle (Becton-Dickinson) to produce a well-defined papula, without spreading outside the site. The AVP injections were always initiated on the most proximal sites. Each site was injected with 10 μL of a solution that contained a 2.5-, 12.5-, or 25-ng dose of AVP. In control sites, 10 μL of 0.9% NaCl was injected. Skin blood flow was measured with the application of the laser Doppler on the papula. Measurements were repeated up to 180 minutes after the AVP injection. The results were expressed in arbitrary units, and the percentage of changes from baseline were considered.
In the second part of the protocol, a pronounced local skin vasodilation was first induced on each injection site by the ID injection of 10 μL of a 10−11 mol/L solution of CGRP (Clinalfa AG) before the injection of AVP. Under these conditions, considerably smaller doses of AVP were injected to produce a dose-response curve (ie, 0.01, 0.05, 0.1, 0.25, and 2.5 ng per site).
For the IA protocol, the subjects were placed on a bed in the morning. Under local anesthesia with 2% xylocaine, the left brachial artery was cannulated with a 20-gauge polyethylene catheter (Leader cath 115.09) for drug infusion by use of a modified Seldinger method. The arterial line was kept open by the infusion of heparinized 0.9% saline at 3 mL/h. Thereafter, the volunteers remained supine for the entire morning in a room with a constant temperature of 22°C.
When the arterial diameter was measured, the left arm was placed in a splint to prevent involuntary movements. A description of the measuring device (NIUS 02), which is an upgraded version of the apparatus that used hardware tracking,13 14 and reproducibility data have been reported previously.15 Forearm blood flow velocity was measured by a continuous-wave Doppler with an 8-MHz transducer angle (Doptek 2003) at a 45° angle distal to the 10-MHz probe used for diameter measurements, with both beams focused on the same spot of the radial artery. Resting blood flow (mL/min) is the product of time-averaged mean flow velocity and arterial cross-sectional area obtained continuously and simultaneously from the arterial diameter.
After a stable hemodynamic baseline was established, vehicle (NaCl 0.9%) or AVP was infused at 0.04, 0.08, 0.2, and 0.8 ng · kg−1 · min−1 for 2.5 minutes. Each dose was separated from the previous one by 30 minutes. Infusion rate was maintained constant (0.2 mL/min) with a high-precision pump (Harvard Apparatus Co, Inc). Radial artery diameter and blood flow were recorded continuously for 10 minutes. Determination of the radial artery blood flow with the echotracking continuous-wave Doppler coupling was performed with free-flowing blood into the hand or with a wrist occlusion. Heart rate and blood pressure were monitored at the right middle finger by a photoplethysmographic instrument (Finapres). Skin capillary blood flow was measured by laser Doppler flowmetry.
Preparation of Drugs
AVP and CGRP were diluted in physiological isotonic saline and prepared in sterile conditions by the hospital pharmacist on the day of the experiment immediately before use.
The effects of SR 49059 were evaluated in 12 healthy volunteers aged 18 to 35 years. The protocol of the study was accepted by the local ethics committee. The study was conducted according to a double-blind, triple crossover design, in which each subject was randomized to receive, at 1-week intervals, a single oral dose of a placebo, 30 mg of SR 49059, and 300 mg of SR 49059. On these three occasions, 2.5 ng AVP (Pitressin, Parke Davis) was injected intradermally into a 2-cm2 area of skin prevasodilated with 10 μL of a 10−11 mol/L CGRP solution. The decrease in skin blood flow induced by AVP was measured, as described above, twice before the administration of the antagonist (at times −1 hour and 0) and 1, 2, 4, 6, and 8 hours after drug intake. At each time, a new site was chosen on the forearm. The seven sites were distributed on both forearms; the early sites were always the most proximal. The AVP-induced vasoconstriction was measured continuously for 15 minutes after the injection of AVP.
One to 2 weeks after the last ID evaluation, the same subjects were readmitted into the Division and randomized to receive in double-blind fashion one dose of a placebo (n=4) or 30 mg (n=4) or 300 mg (n=4) of SR 49059. Because the response to AVP varies among individuals, the dose of AVP was chosen to obtain at least an 8% to 10% decrease in radial artery diameter, which was measured as described above (eg, 0.8 to 1.2 ng · kg−1 · min−1). AVP was infused into the humeral artery before the oral administration of the drug and again at 1, 2, 3, and 4 hours after dosing. The AVP-induced changes in radial artery diameter and blood flow and the changes in skin blood flow were measured simultaneously and continuously for 3 minutes before and for 20 minutes after the IA injection of AVP. From preliminary results obtained in humans, single oral doses of 25 and 300 mg SR 49059 were sufficient to inhibit 50% and 90%, respectively, of the AVP-induced in vitro platelet aggregation, which is mediated by the V1a receptor.16 These doses were chosen to assess the dose-response effect of SR 49059.
All results are presented as mean±SEM. The statistical comparisons of the doses were performed with a one- or two-way ANOVA and followed by a least-significant difference test or Student’s t test, when appropriate, with a value of P<.05 as the minimum level of significance. The continuous variations noted in radial artery diameter and blood flow during the IA protocol were analyzed at each time point by the calculation of a 3-minute area under the curve before the injection of AVP and at 5, 10, and 15 minutes after the injection of AVP.
The top panel of Fig 1⇓ shows the percent changes in skin blood flow obtained with the ID administration of AVP in the absence of CGRP. Without the preinjection of CGRP, only the ID injection of the highest dose of AVP caused significant decreases in skin blood flow (up to −30% from baseline at maximum), which was expressed clinically by a marked blanching of the skin. However, the dose-response curve of the AVP-induced change in skin blood flow was somewhat flat with doses of AVP between 2.5 and 25 ng.
When CGRP is injected intradermally, basal skin blood flow increases at least 7-fold, and the vasodilation is sustained for at least 2 hours. ID injection of sodium chloride also increases skin blood flow 4-fold, but the increase is only transient. The response to AVP on sites vasodilated previously with CGRP is shown in the middle panel of Fig 1⇑. Under these experimental conditions, the responses to ID AVP are clearly dose-dependent and last for at least 1 hour with the 3 highest doses of 0.1, 0.25, and 2.5 ng. The use of CGRP shifts the dose-response curve considerably to the left, which makes possible the use of much lower concentrations of AVP to demonstrate a significant reduction in skin blood flow (Fig 1⇑, bottom panel).
The effects of the IA infusion of AVP on radial artery blood flow and diameter and on forearm skin blood flow are shown in Fig 2⇓. In these experiments, blood flow to the hand is left intact. AVP caused slight decreases in radial artery blood flow at doses ranging between 0.04 and 0.8 ng · kg−1 · min−1. Significant reductions in radial artery diameter were observed with the infusion of 0.2 and 0.8 ng · kg−1 · min−1 AVP. The decrease in radial artery diameter occurred during the infusion of AVP and lasted up to 45 minutes in some subjects. As shown in the bottom panel of Fig 2⇓, decreases in forearm skin blood flow were observed also with the IA administration of AVP. These changes appeared almost immediately after the AVP infusion was started but were of short duration, as the skin blood flow was back to baseline within 3 to 5 minutes after the end of the AVP infusion.
Fig 3⇓ shows the changes in radial artery blood flow and diameter and forearm skin blood flow induced by the IA infusion of AVP when the perfusion of the hand was interrupted by a wrist occlusion. The wrist occlusion considerably reduced the variability of the measurements of the radial artery and skin blood flow under control conditions (saline infusion). With the hand excluded, no significant change in radial artery blood flow was observed, although it fell slightly with the highest dose. Significant decreases in radial artery diameter were seen again with the two highest doses of AVP, but the reduction occurred later, 1 minute after the end of the infusion. With the occlusion of the wrist, the changes in skin blood flow induced by the IA administration of AVP were more marked and sustained.
AVP injected intradermally was well tolerated and caused no significant local or general discomfort. With the IA infusion, one subject experienced a marked and sustained vasoconstriction of the radial artery (from 2.45 to 2.05 mm), which was associated with a feeling of oppression in the chest. Both the constriction of the radial artery and the thoracic oppression were relieved with the administration of sublingual nitroglycerin. Additional clinical investigations of this subject failed to reveal any cardiac abnormality.
Effect of SR 49059: A Vasopressin V1 Receptor Antagonist
As expected from the preliminary study, the ID injection of 2.5 ng AVP after dilatation by CGRP decreased skin blood flow by 80% at peak (ie, 5 minutes after the injection). This effect was reproducible during the 8 hours of the study in the placebo group (Fig 4⇓, top panel). Fifteen minutes after the AVP injection, the vasoconstrictive effect was already slightly attenuated (Fig 4⇓, bottom panel). Oral administration of the V1 receptor antagonist SR 49059 at time 0 induced a significant, dose-dependent inhibition of the AVP-induced vasoconstriction. The inhibition was significant for each level of dose versus placebo. The peak effect was observed between 2 and 6 hours, depending on the subjects.
The effects of the antagonist on the AVP-induced changes in radial artery diameter are shown in Fig 5⇓. The baseline mean diameter was comparable in the three groups (ie, 2876±89 μmol/L in the placebo group and 2744±201 and 2799±175 μmol/L in the 30- and 300-mg SR 49059 groups, respectively). In the placebo group, the mean diameter at peak effect of AVP did not vary significantly during the 4-hour observation period. In contrast, a significant inhibition of the AVP-induced vasoconstriction was observed with 300 mg of the antagonist (P<.02), whereas 30 mg induced only a partial inhibition.
The IA infusion of AVP reduced the radial artery flow by ≈40% in all three groups. The response to IA AVP was reproducible during the 4 hours of investigation in the placebo group. Surprisingly, neither the AVP-induced decrease in radial artery flow nor the reduction in skin blood flow was affected significantly by the antagonist, although in some subjects the fall in radial artery flow was blunted considerably. Overall, SR 49059 was well tolerated and no side effect was reported by any of the subjects.
The results of the present study demonstrate that the vascular effects of a specific AVP V1 receptor antagonist can be evaluated in humans with two technical approaches that involve either the ID injection of AVP, after local vasodilation with CGRP, and the measurement of skin blood flow or a short-lasting IA infusion of AVP combined with a direct, noninvasive measurement of radial artery diameter and blood flow. With these techniques, we demonstrated that SR 49059 is an orally active, effective, and well-tolerated V1 antagonist that inhibits dose-dependently the vascular response to exogenous AVP in normotensive subjects.
AVP is a potent vasoconstrictor peptide that can cause serious side effects such as myocardial ischemia and arrhythmias when administered intravenously.10 11 To investigate safely the new AVP receptor antagonists, which become available for clinical research, it is necessary to develop alternative techniques that make the injection of exogenous AVP safe and sufficiently sensitive to discriminate a drug-induced effect. In addition to the coronary and the gastrointestinal vasculatures, which exhibit a high responsiveness to AVP, the skin microvasculature is another AVP-sensitive vascular bed.17 Applied intradermally, AVP causes a local vasoconstriction that leads to a blanching area; the size of the area is dose-proportionate. In our experimental setting, ID AVP decreased skin blood flow, and the vasoconstriction lasted up to 2 hours, depending on the concentration of the peptide. However, no clear-cut dose-response curve could be obtained with AVP alone. This was probably caused by the difficulty of measuring a reduction from an already low basal flow and by the oscillatory pattern of basal cutaneous circulation.18 To improve the technique, CGRP was injected intradermally to markedly enhance skin blood flow locally before the injection of AVP. CGRP is a 37–amino acid peptide known to be localized in sensory neurons in many organs, including the skin.19 20 CGRP is one of the most potent vasodilators.21 In humans, an ID injection of CGRP causes a prolonged increase in local blood flow that can last for several hours.21 CGRP has been used recently to investigate the microvascular effect of endothelin, another potent vasoconstrictor hormone.22 With CGRP, we have obtained a clear dose-response curve to exogenous AVP with a range of changes in skin blood flow broad enough to study subsequently a drug-induced effect. CGRP increases the sensitivity of the method by a factor of 102 to 103.
The technique described above was used to investigate the efficacy of SR 49059, which is a new, potent, and selective orally active nonpeptide AVP antagonist. With human platelets in vitro, SR 49059 has been shown to induce a dose-related inhibition of AVP-induced human platelet aggregation.5 In the present study, SR 49059 inhibited dose-dependently the AVP-induced decrease in skin blood flow. The maximal effect of the drug was observed 2 hours after drug intake, and the inhibition persisted up to 6 hours after drug intake. With the higher dose of 300 mg, the AVP-induced vasoconstriction of the skin was still not abolished completely. The lack of complete blockade of the AVP-induced vasoconstriction could be a matter of submaximal dosing. However, considering the individual responses, a complete inhibition of the vasoconstrictor effect of exogenous AVP was observed in 9 of the 12 subjects, in whom the maximal inhibitory effect occurred 2 hours after drug intake in 5 volunteers and 4 to 6 hours after drug administration in 3 other subjects. Thus, the individual variation in time to maximal effect may be linked to differences in gastrointestinal absorption, which may explain the apparent incomplete blockade of the effect of exogenous AVP observed with the 300-mg dose. In support of this hypothesis, earlier observations have shown that 200 mg of SR 49059 completely blocks the AVP-induced aggregation of platelets.16
Previous investigations of the vascular effects of AVP and a V1 antagonist in humans have been conducted with 2- to 5-minute IA infusions of AVP and forearm blood flow and vascular resistance determined with venous occlusion plethysmography.23 24 25 The direct infusion of AVP into the brachial artery is usually chosen to enhance the safety of the methodology, because low doses can be administered locally without increasing plasma AVP systemically. A similar method of administering AVP has been used in our second protocol. Even though small doses of AVP were infused, one of our subjects experienced an episode of chest pain with a sensation of hotness and breathlessness, which suggests angina pectoris. The subject had no cardiac abnormality and was perfectly healthy on careful reexamination. This episode demonstrates that the risk of coronary vasoconstriction persists, even with the administration of low IA doses of AVP, and that careful cardiac monitoring is mandatory when this potent vasoconstrictor hormone is infused.
In contrast to previous studies,23 24 25 the changes in radial artery diameter and blood flow were measured directly with a noninvasive methodology. AVP-induced dose-dependent decreases in radial artery diameter and blood flow occurred at doses between 0.04 and 1.2 ng · kg−1 · min−1. Even when injected into the brachial artery, AVP causes a dose-dependent reduction of skin blood flow measured simultaneously. These results are expected from the known vasoconstrictor effects of AVP and tend to confirm previous observations which suggest that the skin microvasculature is particularly sensitive to AVP.17 26 However, our findings contrast with earlier reports that describe a decrease in forearm blood flow with low doses of AVP and an apparent increase in flow with higher doses of AVP.23 24 25 Although comparable doses of AVP have been administered, we have never observed an increase in flow, either in the radial artery or in the skin microvasculature. Our results, therefore, suggest that the apparent increase in forearm blood flow obtained with high doses of AVP with venous occlusion plethysmography is neither caused by a vasodilation of the radial artery nor an increase in skin perfusion but more likely by a selective increase in skeletal muscle blood flow. Our observation is, however, in agreement with the finding of Hirsch et al,23 who found that AVP causes a decrease in blood flow only at the digital level, probably because digital arteries have a larger population of V1 receptors.
With the 300-mg dose of the antagonist, the AVP-induced vasoconstriction of the radial artery is blocked completely, therefore confirming that the effect of AVP on the radial artery is mediated entirely by the V1 receptor. SR 49059 had no influence on the AVP-induced changes in radial artery and skin blood flow. The persistent decrease of radial blood flow, when the arterial diameter is normalized, is likely caused by the persistence of some peripheral vasoconstriction of resistance vessels; these resistance vessels are important determinants of radial artery flow. The lack of effect of the antagonist on skin blood flow, when AVP is given intra-arterially, is more surprising. The reason that V1 receptors are blocked completely in the radial artery, but not in the skin, could be caused by differences in receptor density or affinity at the two sites. On nonvasodilated skin, however, the determination of skin blood flow is less sensitive to demonstrate small changes.
The present results demonstrate that the measurement of skin blood flow after ID injection of exogenous AVP on a skin area previously vasodilated with CGRP is an effective method to investigate the vascular V1 effect of specific AVP receptor antagonists. This approach is not only easy but also free of any risk. In our opinion, the use of small IA doses of AVP does not appear to be entirely safe. We also demonstrated that SR 49059 is a potent and specific antagonist of V1 receptors that blocks the AVP-induced vasoconstriction of skin microvasculature and the radial artery.
Selected Abbreviations and Acronyms
|CGRP||=||calcitonin gene–related peptide|
Reprint requests to M. Burnier, MD, Division of Hypertension, CHUV, 1011 Lausanne, Switzerland.
- Received February 20, 1997.
- Revision received March 17, 1997.
- Accepted April 18, 1997.
Clozel M, Breu V, Gray GA, Kalina B, Löffler BM, Burri K, Cassal JM, Hirth G, Müller M, Neidhart W, Ramuz H. Pharmacological characterization of bosentan, a new potent orally active nonpeptide endothelin receptor antagonist. J Pharmacol Exp Ther. 1994;270:228-235.
Serradeil-Le Gal C, Wagnon J, Garcia C, Lacour C, Guiraudou P, Christophe B, Villanova G, Nisato D, Maffrand JP, Le Fur G, Guillon G, Cantau B, Barberis C, Trueba M, Ala Y, Jard S. Biochemical and pharmacological properties of SR 49059, a new potent, nonpeptide antagonist of rat and human vasopressin V1a receptors. J Clin Invest. 1993;92:224-231.
Yamamura Y, Ogawa H, Yamashita H, Chihara T, Miyamoto H, Nakamura S, Onogawa T, Yamashita T, Hosokawa T, Mori T, Tominaga M, Yabuuchi Y. Characterisation of a novel aquaretic agent, OPC-31260, as an orally effective, nonpeptide vasopressin V2 receptor antagonist. Br J Pharmacol. 1992;105:787-791.
Burrell LM, Phillips PA, Stephenson JM, Risvanis J, Rolls KA, Johnston CI. Blood pressure-lowering effect of an orally active vasopressin V1 receptor antagonist in mineralocorticoid hypertension in the rat. Hypertension. 1994;23(part 1):737-743.
Yamada Y, Yamamura Y, Chihara T, Onogawa T, Nakamura S, Yamashita T, Mori T, Tominaga M, Yabuuchi Y. OPC-21268, a vasopressin V1 antagonist, produces hypotension in spontaneously hypertensive rats. Hypertension. 1994;23:200-204.
Kelly KJ, Stang JM, Mekhjian HS. Vasopressin provocation of ventricular dysrythmia. Ann Intern Med. 1980;92:205-206.
Imaizumi T, Harada S, Hirooka Y, Masaki H, Momohara M, Takeshita A. Effects of OPC-21268, an orally effective vasopressin V1 receptor antagonist in humans. Hypertension. 1992;20:54-58.
Hayoz D, Tardy Y, Rutschmann B, Mignot JP, Achakri H, Feihl F, Meister JJ, Waeber B, Brunner HR. Spontaneous diameter oscillations of the radial artery in humans. Am J Physiol. 1993;264:H2080-H2084.
Hayoz D, Rutschmann B, Perret F, Niederberger M, Tardy Y, Mooser V, Nussberger J, Waeber B, Brunner HR. Conduit artery compliance and distensibility are not necessary reduced in hypertension. Hypertension. 1992;20:1-6.
Brouard R, Chassard D, Hediard N, Pignol R, Leenhardt AF, Serradeil-Le Gal C, Thebault J, Kusmierek J. The advantage of surrogate markers during phase 1 for new exploratory compounds: application for SR 49059, an orally active V1a vasopressin receptor antagonist. Thérapie. 1995;50:S34. Abstract.
Waeber B, Schaller MD, Nussberger J, Bussien JP, Hofbauer KG, Brunner HR. Skin blood flow reduction induced by cigarette smoking: role of vasopressin. Am J Physiol. 1984;247:H895-H901.
Brain SD, Crossman DC, Buckley TL, Williams TJ. Endothelin-1: demonstration of potent effects on the microcirculation of humans and other species. J Cardiovasc Pharmacol. 1989;13(suppl 5):S147-S149.
Hirsch AT, Dzau VJ, Majzoub JA, Crager MA. Vasopressin-mediated forearm vasodilation in normal humans: evidence for a vascular V2 receptor. J Clin Invest. 1989;84:418-426.
Suzuki S, Takeshita A, Imaizumi T, Hirooka Y, Yoshida M, Ando S, Nakamura M. Biphasic forearm vascular responses to intraarterial arginine vasopressin. J Clin Invest. 1989;84:427-434.
Tagawa T, Imaizumi T, Endo T, Shiramoto M, Hirooka Y, Ando S, Tekeshita A. Vasodilatory effect of arginine-vasopressin is mediated by nitric oxide in human forearm vessels. J Clin Invest. 1993;92:1483-1490.
Montani JP, Liard JF, Schoun J, Mohring J. Hemodynamic effects of exogenous and endogenous vasopressin at low plasma concentrations in conscious dogs. Circ Res. 1980;47:346-355.