View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector 920 JACC Vol. 30, No. 4 October 1997:920–6

Nitric Oxide-Mediated Flow-Dependent Dilation Is Impaired in Coronary Arteries in Patients With Coronary Spastic

KIYOTAKA KUGIYAMA, MD, MASAMICHI OHGUSHI, MD, TAKESHI MOTOYAMA, MD, SEIGO SUGIYAMA, MD, HISAO OGAWA, MD, MICHIHIRO YOSHIMURA, MD, YOSHITO INOBE, MD, OSAMU HIRASHIMA, MD, HIROAKI KAWANO, MD, HIROFUMI SOEJIMA, MD, HIROFUMI YASUE, MD Kumamoto City, Japan

Objectives. This study sought to examine whether flow- Results. Flow-dependent dilation of the proximal LAD was dependent dilation is impaired at the site of coronary artery found to be less in spasm arteries than in control arteries. G spasm in patients with coronary spastic angina. Infusion of N -monomethyl-L-arginine (L-NMMA) in the proxi- Background. Physiologic stimuli such as exercise and exposure mal LAD suppressed flow-dependent dilation in control arteries to cold have been shown to cause an increase in coronary blood but had no significant effect on spasm arteries. The dilator flow, leading to flow-dependent dilation of coronary arteries in response to nitroglycerin was not impaired in spasm coronary normal subjects, but cause coronary constriction in patients with arteries. coronary spastic angina. Conclusions. Our results indicate that flow-dependent coronary Methods. A maximal increase in blood flow was induced dilation is impaired in spasm arteries, partly due to a deficiency in selectively in the left anterior descending coronary artery (LAD) endothelial nitric oxide bioactivity, which in turn may contribute by infusion of adenosine through a Doppler flow catheter tip in the to the increase in coronary tone during physiologic stimuli in midportion of the LAD in 10 patients with coronary spastic patients with coronary spastic angina. angina, all with angiographically demonstrated spasm of the LAD, (J Am Coll Cardiol 1997;30:920–6) and in 11 control patients. Coronary artery diameter at the ©1997 by the American College of Cardiology proximal site of the LAD (exposed to increased flow but not to adenosine) was measured by quantitative angiography.

Coronary artery spasm has been shown (1,2) to play an continuously generates NO, which has been shown (8–14) to important role not only in the pathogenesis of variant angina, maintain basal vascular tone in animals and humans. We but also of ischemic disease in general, including other recently showed (14) that there is a deficiency in endothelial forms of angina pectoris, acute and NO bioactivity at the sites of coronary artery spasm and that sudden death. However, the precise mechanism by which this deficiency plays an important role in the pathogenesis of coronary spasm occurs remains unknown. Nitric oxide (NO) is coronary spasm. Flow-dependent is well known to an important vasodilator and is produced from L-arginine by have an important role in the regulation of arterial tone way of the enzyme NO synthase (NOS) (3–7). Analogues of (15–19). Physiological stimuli, such as exercise and the cold G L-arginine, such as N -monomethyl-L-arginine (L-NMMA), pressor test, which induce an increase in coronary blood flow, competitively inhibit NO production and have been widely have been shown (20–23) to dilate coronary arteries in normal used to examine the NO pathway in animals and humans subjects. However, we and others have shown (1,2,24,25) that (6–13). The constitutive NOS in the arterial endothelium spastic coronary arteries have increased basal tone and do not dilate but constrict during exercise and exposure to cold. An increase in blood flow or shear stress has been shown (3– From the Division of Cardiology, Kumamoto University School of Medicine, Kumamoto City, Japan. This study was supported in part by Grant-in-Aid for 13,26,27) to produce endothelial release of NO, which causes Scientific Research C05670622 from the Ministry of Education, Science and flow-dependent vasodilation. Thus, the present study aimed to Culture and a Smoking Research Foundation Grant for Biochemical Research, examine whether flow-dependent dilation is impaired at the Tokyo, Japan. Manuscript received February 12, 1997; revised manuscript received May 27, sites of coronary artery spasm in patients with coronary spastic 1997, accepted June 20, 1997. angina and to determine the possible relevance of deficient NO Address for correspondence: Dr. Kiyotaka Kugiyama, Division of Cardiol- ogy, Kumamoto University School of Medicine, 1-1-1 Honjo, Kumamoto City, bioactivity to the abnormal response of coronary tone to blood 860 Japan. E-mail: [email protected]. flow in spasm coronary arteries.

©1997 by the American College of Cardiology 0735-1097/97/$17.00 Published by Elsevier Science Inc. PII S0735-1097(97)00236-2 JACC Vol. 30, No. 4 KUGIYAMA ET AL. 921 October 1997:920–6 FLOW-INDUCED DILATION IN SPASM CORONARY ARTERY

Table 1. Clinical Characteristics of Study Patients* Abbreviations and Acronyms Patients Control ACh ϭ acetylcholine With CSA Patients ECG ϭ electrocardiogram (n ϭ 10) (n ϭ 11) LAD ϭ left anterior descending coronary artery Age (yr) 59 11 57 10 LCx ϭ left circumflex coronary artery Ϯ Ϯ G Women/men 4/6 3/8 L-NMMA ϭ N -monomethyl-L-arginine NO ϭ nitric oxide Total serum cholesterol (mg/dl) 181 Ϯ 36 191 Ϯ 33 NOS ϭ nitric oxide synthase HDL cholesterol (mg/dl) 39 Ϯ 846Ϯ14 Serum triglyceride (mg/dl) 127 Ϯ 69 140 Ϯ 41 Hypercholesterolemia† 0 1 Hypertension‡ 3 2 Diabetes mellitus§ 0 2 Methods Current smokerʈ 24 *p ϭ NS for all comparisons. †Total serum cholesterol Ն240 mg/dl. ‡Blood Study patients. The study included 10 patients with coro- pressure Ն140/90 mm Hg or taking antihypertensive medication. §Fasting blood nary spastic angina (mean age 59 years, range 43 to 71; six men, glucose Ն140 mg/dl or taking antidiabetic medication. ʈՆ10 cigarettes/day for four women) who fulfilled the following inclusion criteria: 1) Ն10 years. Data presented are mean value Ϯ SD or number of patients. CSA ϭ spontaneous chest pain in association with ST segment eleva- coronary spastic angina. tion or depression on the 12-lead electrocardiogram (ECG) or the ambulatory ECG at rest, usually in the middle of the night or early morning and more than two times a day during the spasm (Ͻ50% decrease in coronary diameter from baseline) study; 2) coronary spasm (total or subtotal occlusion) of the after intracoronary injection of ACh. Neither nitrates nor left anterior descending coronary artery (LAD) demonstrated calcium channel blockers were effective in relieving chest pain angiographically during chest pain, with ST segment changes in any control patient. after intracoronary injection of acetylcholine (ACh), as previ- The clinical characteristics of the study patients are shown ously reported (28,29); 3) no organic coronary artery stenosis in Table 1. All medications except sublingual nitroglycerin on angiography. Five patients had transient ST segment ele- were withdrawn at least 3 days before the study, and no patient vation and five ST segment depression during the spontaneous had taken nitroglycerin within6hofthestudy. No patient had attacks. All 10 patients with coronary spastic angina performed a previous myocardial infarction, congestive or treadmill exercise and hyperventilation tests. Anginal attacks other serious diseases. Written informed consent was obtained of chest pain with ST segment changes occurred in two patients from all patients before the study. The study was approved by during the treadmill test (ST segment elevation in one, ST the ethics committee of our institution. segment depression in the other) and in four during the Study protocol. Coronary angiography was performed with hyperventilation test (ST segment elevation in three, ST seg- the Judkins technique using contrast material (Ioxaglate, ment depression in one). The spontaneous attacks were Guerbet S.A.) in the morning when the patients were fasting. promptly resolved by the administration of sublingual nitro- After and measurements, control glycerin and were almost completely prevented by calcium angiography of the left main and right coronary arteries was channel blocking agents in all 10 patients with coronary spastic performed. Thereafter, incremental doses of ACh were in- angina. In patients with coronary spastic angina, the proximal jected into the left main coronary artery (20, 50 and 100 ␮g) LAD diameter at baseline angiography was Ͼ1.5 mm, so the and subsequently into the right coronary artery (20 and 50 ␮g). Doppler catheter was allowed to avoid catheter-induced coro- The method of injecting ACh for provocation of coronary nary spasm in these patients. All study patients had ambulatory spasm has been detailed elsewhere (28,29). Coronary spasm ECG monitoring for at least 48 h. ST segment changes on resolved spontaneously without the use of nitroglycerin in all 12-lead and ambulatory were considered patients with coronary spastic angina in this study. significant if transient ST segment elevation (Ͼ0.2 mV) or Flow-dependent dilation in human coronary arteries in vivo depression (Ͼ0.1 mV) versus baseline levels occurred at 60 to was assessed similar to previously reported methods (18,19). In 80 ms after the J point. brief, 15 min after completion of the intracoronary injection of This study also included 11 control patients who underwent ACh, when systemic hemodynamic variables and angiographic diagnostic cardiac catheterization for evaluation of atypical coronary artery diameters had returned to baseline levels, an chest pain (mean age 57 years, range 39 to 70; eight men, three additional 5,000 U of heparin was given intravenously, and an women) and were matched for risk 8F guiding catheter was positioned in the ostium of the left factors to patients with coronary spastic angina. The control main coronary artery. A 2.5F Doppler flow velocity catheter patients fulfilled the following inclusion criteria: 1) no signifi- (20-MHz pulsed Doppler catheter, Miller Instruments) was cant ST segment changes during chest pain on 12-leads and positioned in the midportion of the LAD through the guiding ambulatory electrocardiography; 2) no chest pain or ST seg- catheter over a guide wire. Flow velocity recordings were ment changes during treadmill and hyperventilation testing; 3) referenced to zero and calibrated before introduction of the angiographically normal coronary arteries and no coronary Doppler catheter into the guiding catheter. The Doppler flow 922 KUGIYAMA ET AL. JACC Vol. 30, No. 4 FLOW-INDUCED DILATION IN SPASM CORONARY ARTERY October 1997:920–6 velocity catheter was connected to a velocimeter (model MDV-20, Miller Instruments) and then to a multichannel oscillographic recorder (Nihon Kohden, Japan) and was care- fully positioned to obtain a stable flow velocity signal. There- after, the position and range gate control were not changed during the study. Throughout the study, phasic and mean coronary blood flow velocities, heart rates, (through the guiding catheter) and 12-lead ECGs were contin- uously monitored and were recorded just before each coronary angiogram. Baseline angiography of the left coronary artery was per- formed after injection of 8 ml of contrast medium through the guiding catheter. Five minutes later, adenosine (40 ␮g/kg body Figure 1. Schematic representation of assessment for flow-dependent dilation in the LAD. weight per min at a rate of 2 ml/min) was infused into the mid-portion of the LAD through the Doppler catheter tip. This dose was chosen on the basis of a previous report (30) showing that infusion of adenosine into the left main coronary and stored in a cardiac image analysis system (Cardio 500, artery at a concentration Ͼ40 ␮g/kg per min elicited a maximal Kontron Instruments). Automated contour detection was per- increase in coronary blood flow with no change in systemic formed by the geometric edge differentiation technique de- hemodynamic variables. Angiography was again performed scribed in previous validation studies (14,29,33). The size of 60 s after the maximal increase in coronary blood flow (90 to the Judkins or guiding catheter was used to calibrate arterial 120 s after the start of the infusion), according to previous diameters (in mm); coronary segments analyzed were ϳ10 mm reports (31,32) showing the time course of the maximal long. increase in coronary blood flow and maximal flow-dependent The sites at which the coronary artery diameter was mea- coronary dilation. Special care was taken to avoid uninten- sured were chosen according to previous reports (18,19) and tional injection of adenosine into the proximal portion of the are depicted in the diagram of the LAD shown in Figure 1. In LAD. The adenosine infusion was stopped just after angiog- brief, a segment of the LAD that was positioned between the raphy, and a 10-min rest period was allowed for return of orifices of the first and second septal branches and was at least coronary artery diameter and coronary blood flow to basal 10 mm proximal to the Doppler catheter tip was analyzed for levels, as confirmed by angiography and measurement of blood flow-mediated effects (proximal LAD). This segment was ex- flow. Thereafter, the guiding catheter was filled with L-NMMA posed to L-NMMA and changes in blood flow but was not solution, and L-NMMA was infused into the ostium of the left exposed to adenosine. A segment of the LAD 20 mm distal to main coronary artery through the guiding catheter (50 ␮mol/ the Doppler catheter tip (distal LAD) was analyzed for the min for 4 min at a rate of 2 ml/min). Measurement of coronary effects of adenosine and L-NMMA. A midsegment of the left and systemic hemodynamic variables and coronary angiogra- circumflex coronary artery (mid-LCx) was exposed to L- phy were performed during the last 30 s of the infusion. NMMA but not to changes in blood flow or adenosine and Subsequently, adenosine infusion was repeated, and coronary served as the control vessel. A segment of the LAD at the angiography was performed in the same manner as during the Doppler catheter tip was analyzed to allow estimation of first adenosine infusion, before L-NMMA. After an additional volume flow from flow velocity measurements. All segments 10 min, coronary angiography was performed before and 2 min were analyzed at each angiogram, and the responses of after an intravenous bolus injection of nitroglycerin (250 ␮g) in the coronary diameters to the first adenosine infusion, the multiple projections in all study patients. All drug solutions L-NMMA infusion and the nitroglycerin injection were ex- were kept at 37°C. pressed as the percent change in coronary diameter from that Quantitative coronary angiography. Quantitative coronary at angiography performed just before each infusion. The angiography was performed according to a previously validated response to the second adenosine infusion was expressed as technique (14,29,33). Coronary angiograms were recorded in the percent change in diameter from that after the L-NMMA the right and left anterior oblique positions with adequate infusion. Analysis of intraobserver and interobserver variability angulation to allow clear visualization of the left main coronary for measurement of coronary artery diameter showed high artery during the study period. The relations among focal spot, reproducibility (r ϭ 0.99, SEE ϭ 0.05 mm, p Ͻ 0.001 and r ϭ patient and height of the image tube were kept constant during 0.99, SEE ϭ 0.04 mm, p Ͻ 0.001, respectively). the angiographic study. Measurement of the lumen diameter Coronary blood flow measurements. Estimates of blood of the coronary artery was performed quantitatively with the flow were calculated by multiplying the mean Doppler-derived use of a computer-assisted coronary angiographic analysis blood flow velocity by the cross-sectional area of the arterial system by two observers (M.O., S.S.) who had no knowledge of segment at the Doppler catheter tip. Cross-sectional area was the study protocol. The end-diastolic cine films that most calculated from measurements of the coronary diameter, as- clearly visualized the coronary segments were videodigitized suming a circular shape. The response of the coronary blood JACC Vol. 30, No. 4 KUGIYAMA ET AL. 923 October 1997:920–6 FLOW-INDUCED DILATION IN SPASM CORONARY ARTERY

Table 2. Systemic Hemodynamic Variables

L-NMMA ϩ Baseline Adenosine L-NMMA Adenosine NTG HR (beats/min) Pts with CSA 74 Ϯ 10 72 Ϯ 10 69 Ϯ 967Ϯ872Ϯ5 Control pts 74 Ϯ 14 76 Ϯ 13 72 Ϯ 11 69 Ϯ 11 81 Ϯ 17 MAP (mm Hg) Pts with CSA 104 Ϯ 15 108 Ϯ 23 111 Ϯ 17 119 Ϯ 26 83 Ϯ 25* Control pts 93 Ϯ 11 95 Ϯ 11 96 Ϯ 10 104 Ϯ 11 81 Ϯ 15* *p Ͻ 0.01 versus baseline. Data presented are mean value Ϯ SD. CSA ϭ coronary spastic angina; HR ϭ heart rate; L-NMMA ϭ NG-monomethyl-L- arginine; NTG ϭ nitroglycerin; Pts ϭ patients; MAP ϭ .

Figure 2. Flow-dependent dilation of the proximal LAD (mean Ϯ SD flow to the intracoronary infusion of various drugs was ex- percent increase in arterial diameter from control values) before and pressed as the percent change in blood flow from that just after L-NMMA infusion in spasm (solid circles) and control coronary arteries (open circles). before each infusion. Drugs. L-NMMA and adenosine were obtained from Wako Chemicals (Osaka, Japan) and were dissolved in physiologic saline and sterilely filtered before use. flow in both spasm and control arteries, but there was no Statistical analysis. Results are expressed as mean value Ϯ significant difference in percent increase in flow from baseline SD, unless otherwise indicated. Differences between two mean values between spasm and control coronary arteries (percent values were compared by a paired or unpaired Student t test. increase in flow: 540 Ϯ 350 vs. 520 Ϯ 260, respectively). Changes in systemic hemodynamic variables, proximal and L-NMMA infusion had no effect on the increase in flow in distal LAD diameters and coronary blood flow in response to response to the second adenosine infusion in both spasm and the various agents were statistically analyzed by repeated coronary arteries, and there was no significant difference in measures analysis of variance, followed by the Bonferroni percent increase in flow from values measured just before the comparison test. Correlation between the increase in blood second adenosine infusion, after L-NMMA, between spasm flow and the dilator response of the proximal LAD during the and control coronary arteries (percent increase in flow: 550 Ϯ adenosine infusion was examined by linear regression analysis. 410 vs. 520 Ϯ 370, respectively). A value of p Ͻ 0.05 was considered statistically significant. Flow-dependent dilation of the proximal LAD. The base- line diameter of the proximal LAD in spasm arteries was significantly smaller than that in control arteries (2.0 Ϯ 0.2 vs. Results 2.5 Ϯ 0.5 mm, respectively, p Ͻ 0.02). The proximal segment of Provocation of coronary spasm. Coronary spasm was de- the LAD dilated in response to the increase in flow in both fined as total or subtotal occlusion of the epicardial coronary spasm and control coronary arteries, but the dilator response arteries associated with signs of myocardial ischemia such as was significantly less in spasm than in control arteries (Fig. 2). chest pain and ischemic ST segment changes. Spasm was There was a significantly positive correlation between the angiographically demonstrated by intracoronary injection of increase in coronary blood flow and the dilator response of the ACh in the LAD of all patients with coronary spastic angina. proximal LAD in control but not in spasm coronary arteries Coronary spasm was also demonstrated in four LCx and seven (Fig. 3). Infusion of L-NMMA alone decreased the basal right coronary arteries in patients with coronary spastic angina. diameter in control coronary arteries but had less effect in Coronary spasm in the LAD was documented at the proximal spasm than in control arteries (Fig. 4). L-NMMA inhibited the segment in three patients with coronary spastic angina, and dilator response of the proximal LAD to the increase in flow subtotal occlusion occurred diffusely from the proximal to the during the second adenosine infusion in control coronary distal segments in the remaining seven. In contrast, intracoro- arteries but had minimal effect in spasm coronary arteries (Fig. nary injection of ACh did not induce coronary spasm in any 2). The mid-LCx did not change significantly in response to control patient. adenosine infusion in both patients and control subjects (2.0 Ϯ Systemic hemodynamic variables and coronary blood flow. 0.6 and 2.2 Ϯ 0.4 mm before adenosine vs. 2.1 Ϯ 0.7 and 2.1 Ϯ As shown in Table 2, infusion of adenosine and L-NMMA did 0.4 mm after adenosine, respectively), excluding the potential not affect systemic hemodynamic variables either in patients effect of the injected contrast material itself and that of the with coronary spastic angina or in control patients. There was recirculating adenosine. no significant difference in basal coronary blood flow between Effect of adenosine on the distal LAD. The distal LAD spasm and control arteries (22 Ϯ 8 vs. 25 Ϯ 10 ml/min, dilated in response to infusion of adenosine in both spasm and respectively). Adenosine infusion increased coronary blood control arteries, but there was no significant difference in 924 KUGIYAMA ET AL. JACC Vol. 30, No. 4 FLOW-INDUCED DILATION IN SPASM CORONARY ARTERY October 1997:920–6

Figure 3. Correlation between percent in- crease from control values in coronary blood flow and the proximal LAD diameter of con- trol (left panel) and spasm (right panel) cor- onary arteries during adenosine infusion and before L-NMMA infusion.

percent increase in coronary diameters from control values crease in the flow response to adenosine was comparable before and after L-NMMA infusion in the spasm and control between two groups of patients, impairment of flow-dependent coronary arteries (before L-NMMA: 19.6 Ϯ 10.1 vs. 22.2 Ϯ dilation appears to be due to endothelial vasodilator dysfunc- 6.5 mm; after L-NMMA: 20.5 Ϯ 12.7 vs. 21.8 Ϯ 9.0, respectively tion. The present study also demonstrated that L-NMMA p ϭ NS). suppressed flow-dependent dilation in control coronary arter- Dilator response to nitroglycerin. Nitroglycerin dilated the ies but had minimal effect in spasm coronary arteries. It has proximal and distal segments of the LAD in both spasm and been shown (3–13,26,27) that an increase in blood flow or control arteries. Figure 4 shows that the dilator response to shear stress induces endothelial release of NO, which causes nitroglycerin was significantly greater in spasm arteries than in flow-dependent vasodilation. The present results therefore control arteries. After nitroglycerin administration, the spasm indicate that endothelial-derived NO bioactivity may contrib- artery diameters were not different from those of the control ute at least in part to flow-dependent dilation in human arteries (2.9 Ϯ 0.6 vs. 3.1 Ϯ 0.7 mm for the proximal LAD; coronary arteries and that the NO bioactivity in response to an 2.1 Ϯ 0.5 vs. 1.9 Ϯ 0.5 mm for the distal LAD, respectively, p ϭ increase in flow may be decreased in spasm coronary arteries. NS). Basal tone and NO activity. Moncada et al. (9) demon- strated that removal of basal NO-mediated vasodilation of the Discussion vascular system in an in vivo rat model leads to supersensitivity to an exogenous . The present study showed The present study shows that dilation of the epicardial that infusion of L-NMMA alone constricted control coronary coronary arteries in response to increased blood flow was arteries at baseline but had less effect on spasm coronary impaired in patients with coronary spastic angina compared arteries. Furthermore, the vasodilator response to nitroglyc- with that in control subjects. Because the response to nitro- erin was increased in the spasm arteries compared with control glycerin, a smooth muscle dilator, was not impaired in patients arteries. The present study also showed that the diameter of with coronary spastic angina, and the magnitude of the in- the spasm coronary arteries was smaller at baseline but was comparable to control arteries after nitroglycerin administra- Figure 4. Percent changes (mean Ϯ SD) from control values in tion, indicating that basal tone may be increased in spasm proximal LAD diameter in response to intracoronary infusion of coronary arteries, although the present study excluded patients L-NMMA alone (left panel) and nitroglycerin injection (right panel)in with coronary spastic angina who had highly increased basal spasm (solid circles) and control coronary arteries (open circles). tone of the LAD. These results are in agreement with those of previous reports by us (14) and Moncada et al. (9) and suggest that basal NO bioactivity may be decreased in spasm coronary arteries, leading to a supersensitive dilator response to exoge- nous and to increased basal coronary tone. Egashira et al. (34) recently reported that NO bioactivity was preserved in the spasm coronary arteries of eight patients with coronary spastic angina, which does not agree with our present and previous (14) results. As shown in our and other previous reports (1,35–37), the hyperreactivity of not only the spasm sites but also the entire coronary arteries to the constrictor effects of ACh as well as the dilator effects of nitrates is a JACC Vol. 30, No. 4 KUGIYAMA ET AL. 925 October 1997:920–6 FLOW-INDUCED DILATION IN SPASM CORONARY ARTERY characteristic feature of spasm coronary arteries without or- avoid this influence by matching the risk factors of patients ganic stenosis. Spasm frequently occurs in multiple coronary with coronary spastic angina with those of control subjects. It arteries in patients with coronary spastic angina (36). However, was recently reported (42) that flow-dependent coronary dila- in the report by Egashira et al. (34), the coronary artery tion in patients with normal coronary angiograms was not diameter response to L-NMMA at the spastic sites was com- significantly affected by infusion of L-NMMA, in contrast to the pared with that at adjacent sites in the same spasm artery in present results. The report (42) did not describe the coronary each patient. Furthermore, they did not compare the response risk factors of the study patients or specify the coronary artery to L-NMMA in each segment of spasm coronary arteries with segments analyzed for flow response, both of which impor- that in the corresponding segment of the control coronary tantly affect NO bioactivity in the coronary arteries (11,14), arteries, even though the coronary artery diameter response to and did not examine the coronary arteries studied for the L-NMMA has been reported (11,14) to vary within segments of presence of spasm. Because angiograms underestimate the the coronary artery. Thus, differences in the method of analysis extent of atherosclerosis (43), the pathologic presence of and the limited number of the study patients in the report by coronary atherosclerosis cannot be excluded as an explanation Egashira et al. (34) may explain in part the discrepancies for the failure to show the contribution of NO to flow- between our results and theirs. dependent dilation (42). Possible mechanisms. It has been shown (3–13,26,27) that Clinical implications. Physiologic stimuli, such as exercise several agonists, as well as shear stress or blood flow, induce and exposure to cold, which increase coronary blood flow, have production and release of NO in the normal vascular endothe- been shown to dilate coronary arteries in normal patients lium. We previously showed (14) that NO bioactivity is also (20–23) but cause coronary constriction in patients with coro- deficient during stimulation with ACh in the spasm coronary nary spastic angina (1,2,24,25). The present results indicate arteries of patients with coronary spastic angina. Therefore, that flow-dependent coronary dilation is impaired in spasm both our present and previous studies suggest that the defi- arteries at least in part because of a deficiency in endothelial ciency in NO bioactivity is unlikely to be dependent on the type NO bioactivity, which in turn may contribute in part to an of stimulation; rather, dysfunction of NOS itself or the intra- increase in coronary tone during physiologic stimuli in patients cellular regulatory pathway to endothelial NOS activation, with coronary spastic angina. mediated by different types of stimulation, may possibly con- Conclusions. The present study shows that endothelium- tribute to deficient NO bioactivity in the spasm coronary dependent flow-mediated dilation is impaired in spasm coro- artery. Increases in blood flow have been shown (38) to induce nary arteries compared with that in control coronary arteries endothelial release of vasodilators such as prostacyclin in and that L-NMMA inhibits flow-dependent dilation in control addition to NO. It is thus possible that a decrease in prosta- coronary arteries but has minimal effect in spasm coronary cyclin release may also be involved in the mechanism of the arteries. From these results, we conclude that endothelial impairment of flow-dependent dilation of the spasm coronary NO-mediated flow-dependent dilation is impaired in spasm arteries. Also, the possibility of an increased breakdown of NO coronary arteries and that this impairment may play an impor- by the increased presence of superoxide anions, as previously tant role in the pathophysiology of coronary artery spasm in reported (38), cannot be excluded. The present study shows patients with coronary spastic angina. that the dilator response of the distal LAD to adenosine infusion was comparable between the spasm and control coronary arteries, and L-NMMA infusion had no significant References effect on the increase in the flow response to adenosine in 1. Yasue H, Omote S, Takizawa A, Nagano M. Coronary arterial spasm in either the control or the spasm coronary arteries. Although it ischemic heart disease and its pathogenesis. Circ Res 1983;52 Suppl I:147– has been reported (40) that endothelial NO may partly con- 52. tribute to adenosine-induced vasodilation in vitro, smooth 2. Maseri A, Severi S, De Nes DM, et al. “Variant” angina: one aspect of a muscle dilation mediated by adenosine receptors present in continuous spectrum of vasospastic myocardial ischemia. Am J Cardiol 1978;42:1019–35. smooth muscle tissue may play a predominant role in 3. Moncada S, Palmer RMJ, Higgs EA. Nitric oxide: physiology, pathophysi- adenosine-induced coronary dilation (40,41). ology, and pharmacology. Pharmacol Rev 1991;43:109–42. NO deficiency does not seem to be caused by decreased 4. Nishida K, Harrison DG, Navas JP, et al. Molecular cloning and character- ization of the constitutive bovine aortic endothelial cell nitric oxide synthase. availability of L-arginine for NO synthesis. In our previous J Clin Invest 1992;90:2092–6. report (14), we showed that intracoronary infusion of an excess 5. Lamas S, Marsden PA, Li GK, Tempst P, Michel T. Endothelial nitric oxide of L-arginine had no effect on the diameter of spasm coronary synthase: molecular cloning and characterization of a distinct constitutive arteries. Flow-dependent dilation and NO bioactivity have enzyme isoform. Proc Natl Acad Sci USA 1992;89:6348–52. 6. Ignarro LJ. Biological action and properties of endothelium-derived nitric been reported (12,18,19) to be impaired in the presence of oxide formed and released from artery and vein. Circ Res 1989;65:1–21. coronary risk factors, even though apparent atherosclerotic 7. Vanhoutte, PM. The endothelium: modulator of vascular smooth muscle changes were angiographically absent. However, this possibil- tone. N Engl J Med 1988;319:512–3. ity is unlikely to explain the impairment of flow-dependent 8. Chu A, Chambers DE, Lin C, et al. Effects of inhibition of nitric oxide formation on basal vasomotion and endothelium-dependent responses of the dilation and the deficiency in NO bioactivity in patients with coronary arteries in awake dogs. J Clin Invest 1991;87:1964–8. coronary spastic angina. The present study was designed to 9. Moncada S, Rees DD, Schulz R, Palmer RMJ. Development and mechanism 926 KUGIYAMA ET AL. JACC Vol. 30, No. 4 FLOW-INDUCED DILATION IN SPASM CORONARY ARTERY October 1997:920–6

of a specific supersensitivity to nitrovasodilators after inhibition of vascular 27. Cooke JP, Rossitch E Jr, Andon NA, Loscalzo J, Dzau VJ. Flow activates an nitric oxide synthesis in vivo. Proc Natl Acad Sci USA 1991;88:2166–70. endothelial potassium channel to release an endogenous nitrovasodilator. 10. Vallance P, Collier J, Moncada S. Effects of endothelium-derived nitric oxide J Clin Invest 1991;88:1663–71. on peripheral arteriolar tone in man. Lancet 1989;2:997–1000. 28. Yasue H, Horio Y, Nakamura N, et al. Induction of coronary artery spasm 11. Lefroy DC, Crake T, Uren NG, Davies GJ, Maseri A. Effect of inhibition of by acetylcholine in patients with variant angina: possible role of the nitric oxide synthesis on epicardial coronary artery caliber and coronary parasympathetic nervous system in the pathogenesis of coronary artery blood flow in humans. Circulation 1993;88:43–54. spasm. Circulation 1986;74:955–63. 12. Quyyumi AA, Dakak N, Andrews NP, et al. Nitric oxide activity in the 29. Okumura K, Yasue H, Matsuyama K, et al. Effect of H1 receptor stimulation human coronary circulation: impact of risk factors for coronary atheroscle- on coronary artery diameter in patients with variant angina: comparison with rosis. J Clin Invest 1995;95:1747–55. effect of acetylcholine. J Am Coll Cardiol 1991;17:338–45. 13. Yamabe H, Okumura K, Ishizaka H, Tsuchiya T, Yasue H. Role of 30. Wilson RF, Wyche K, Christensen BV, Zimmer S, Laxson DD. Effects of endothelium-derived nitric oxide in myocardial reactive hyperemia. Am J adenosine on human coronary arterial circulation. Circulation 1990;82:1595– Physiol 1992;263:H8–14. 606. 14. Kugiyama K, Yasue H, Okumura K, et al. Nitric oxide activity is deficient in 31. Hintze TH, Vatner SF. Reactive dilation of large coronary arteries in spasm arteries of patients with coronary spastic angina. Circulation 1996;94: conscious dogs. Circ Res 1984;54:50–7. 266–272. 32. Canty JM Jr, Schwartz JS. Nitric oxide mediates flow-dependent epicardial 15. Pohl U, Holtz J, Busse R, Bassenge E. Crucial role of the endothelium in the coronary vasodilation to changes in frequency but not mean flow in vasodilator response to increase flow in vivo. Hypertension 1986;8:37–44. conscious dogs. Circulation 1994;89:375–84. 16. Rubanyi GM, Romero JC, Vanhoutte PM. Flow-induced release of 33. Nichols AB, Gabrieli CFO, Fenoglio JJ, Esser PD. Quantification of relative endothelium-derived relaxing factor. Am J Physiol 1986;250:H1145–9. coronary arterial stenosis by videodensitometric analysis of the coronary 17. Kuo L, Davis MJ, Chilian WM. Endothelium-dependent, flow-induced angiograms. Circulation 1984;69:512–22. dilation of isolated coronary arterioles. Am J Physiol 1990;259:H1063–70. 34. Egashira K, Katsuda Y, Mohri M, et al. Basal release of endothelium- 18. Cox DA, Vita JA, Treasure CB, et al. Atherosclerosis impairs flow-mediated derived nitric oxide at site of spasm in patients with variant angina. J Am Coll dilation of coronary arteries in humans. Circulation 1989;80:458–65. Cardiol 1996;27:1444–9. 19. Drexler H, Zeiher AM, Wollschlager H, Meinertz T, Just H, Bonzel T. 35. Okumura K, Yasue H, Matsuyama K, et al. Diffuse disorder of coronary Flow-dependent coronary artery dilation in humans. Circulation 1989;80: artery vasomotility in patients with coronary spastic angina: hyperreactivity 466–74. to the constrictor effects of acetylcholine and the dilator effects of nitroglyc- 20. Nabel EG, Ganz P, Gordon JB, Alexander RW, Selwyn AP. Dilation of erin. J Am Coll Cardiol 1996;27:45–52. normal and constriction of atherosclerotic coronary arteries caused by the 36. Okumura K, Yasue H, Horio Y, et al. Multivessel coronary spasm in patients cold pressor test. Circulation 1988;77:43–52. with variant angina: a study with intracoronary injection of acetylcholine. 21. Nabel EG, Selwyn AP, Ganz P. Paradoxical narrowing of atherosclerotic Circulation 1988;77:535–42. coronary arteries induced by increases in heart rate. Circulation 1990;81: 37. Hoshio A, Kotake H, Mashiba H. Significance of coronary artery tone in 850–9. patients with vasospastic angina. J Am Coll Cardiol 1989;14:604–9. 22. Zeiher AM, Drexler H, Wollschlager H, Just H. Modulation of coronary 38. Frangos JA, Eskin SG, McIntire LV, Ives CL. Flow effects on prostacyclin vasomoter tone in humans: progressive endothelial dysfunction with differ- production by cultured human endothelial cells. Science 1985;227:1477–9. ent early stages of coronary atherosclerosis. Circulation 1991;83:391–401. 39. Ohara Y, Peterson TE, Harrison DG. Hypercholesterolemia increases 23. Gordon JB, Ganz P, Nabel EG, et al. Atherosclerosis influences the endothelial superoxide anion production. J Clin Invest 1993;91:2546–51. vasomoter response of epicardial coronary arteries to exercise. J Clin Invest 40. Kuo L, Chancellor JD. Adenosine potentiates flow-induced dilation of

1989;83:1946–52. coronary arterioles by activating KATP channels in endothelium. Am J 24. Yasue H, Omote S, Takizawa A, Nagano M, Miwa K, Tanaka S. Circadian Physiol 1995;269:H541–9. variarion of exercise capacity in patients with Prinzmetal’s variant angina: 41. Headrick JP, Berne RM. Endothelium-dependent and -independent relax- role of exercise-induced coronary arterial spasm. Circulation 1979;59:938–48. ations to adenosine in guinea pig aorta. Am J Physiol 1990;259:H62–7. 25. Raizner AE, Chahine RA, Ishimori T, et al. Provocation of coronary artery 42. Shiode N, Morishima N, Nakayama K, Yamagata T, Matsuura H, Kajiyama spasm by the cold pressor test: hemodynamic, arteriographic and quantita- G. Flow-mediated vasodilation of human epicardial coronary arteries: effect tive angiographic observations. Circulation 1980;62:925–32. of inhibition of nitric oxide synthase. J Am Coll Cardiol 1996;27:304–10. 26. Joannides R, Haefeli WE, Linder L, et al. Nitric oxide is responsible for 43. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ. Com- flow-dependent dilation of human peripheral conduit arteries in vivo. pensatory enlargement of human atherosclerotic coronary arteries. N Engl Circulation 1995;91:1314–9. J Med 1987;316:1371–5.