Preservation of endothelium-dependent vasodilation in the spastic segment of the human epicardial coronary artery by substance P The objective of this study was to determine if endothelium-dependent vasodilation is preserved in the spastic segment of the epicardial coronary artery. Segmental responses of the coronary artery to substance P were examined by the use of a quantitative angiographic technique in 21 patients with variant angina. Coronary diameter at the basal state did not differ between the spastic and the nonspastic segments (2.3 +- 0.2 mm, 2.3 L 0.4 mm, p > 0.05). Changes in coronary diameter in response to substance P did not differ between segments with ergonovine-induced spasm and nonspastic segments. Maximal dilation averaged 27.1 + 9.5% in the spastic segment and 24.4 k 9.6% in the nonspastic segment (expressetl as a percent increase over the value before drug administration). It appears that both the potential of the endothelium to release endothellum-dependent relaxing factor and the dilating response of the smooth muscle to endothelium-dependent relaxing factor are preserved, even in the spastic segment. (AM HEART J 1992;123:298.)

Hideo Yamamoto, MD, Hitoshi Yoshimura, MD, Mitsuru Noma, MD, Hisashi Kai, MD, Satoshi Suzuki, MD, Tsukasa Tajimi, MD, Masayoshi Sugihara, MD, and Yutaka Kikuchi, MD. Fukuoka, Japan

Coronary vasospasm plays an important role in the pathogenesis of ischemic heart disease,lW3 but the related mechanisms remain controversial.4 Coronary tone is determined by the interplay of dilating and constricting forces. It is believed that localized disorders with constricting forces that exceed dilating forces are present in the spastic segment of coronary arteries in patients with vasospastic angina.5-7 The endothelium of the human epicardial coronary artery produces and releases endothelium-dependent relaxing factor (EDRF) in response to vasoactive substances such as acetylcholine, histamine, substance P, and bradykinins ’ and this EDRF relaxes the vascular smooth muscle. Thus a dysfunction of the endothelium may be present in the spastic segment, where the endothelium is not able to produce and release EDRF, and/or the smooth muscle in the spastic segment loses mechanisms of response related to

From the Section of Cardiology, Welfare Pension Hospital. Received Reprint Kyushu takyushu 411133957

298

for publication requests: Hideo Welfare Pension City, Fukuoka

March

Department 22, 1991;

of Medicine, accepted

Yamamoto, MD, Section Hospital, 2-1-I Kishinoura 806. Japan.

Aug.

The

Kyushu

2, 1991.

of Cardiology, Yahata Nishi-ku,

The Ki-

EDRF. Acetylcholine-induced vasospasm is believed to be related to a defect in endothelium-dependent vasodilation in the spastic segment. However, it should be considered that intrinsic alterations in the arterial smooth muscle itself might mask the normal functioning of the endothelium-dependent vasodilation. An endothelium-dependent vasodilator such as substance P, which has no direct effect on the smooth muscle,8-10 can thus be used to clarify the potential for the production of EDRF in the spastic segment of the epicardial coronary artery in conscious humans. Large atherosclerotic arteries from patients with coronary heart disease do not dilate normally in response to endothelium-dependent vasodilators.8, g, l1 In addition, ergonovine, which provokes coronary vasospasm, constricts the atherosclerotic artery to a significantly greater extent than does the nonatherosclerotic artery.12 To determine if a defect in the endothelium-dependent vasodilation exists in the spastic segment, patients with variant angina but with no apparent atherosclerotic lesions need to be studied. We examined endothelial function in the spastic segment after the intracoronary administration of substance P. The segment was angiographically normal after the intracoronary administration of isosorbide dinitrate (ISDN).

Volume

123

Number

2

Endothelium-dependent

vasodilation

in coronary spasm

299

I. Profiles of patients and responsesto intracoronary administration of ergonovine

Table

LCA

RCA ECG Patient NO.

Age (w)

1 2 3 4 5 6

59 64 46 54 56 54 55 69 48 53 53 52 46 43 46 63

7

8 9 10 11

12 13 14 15 16 17 18 19

20 21

69 55 58 58 49

Sex

M M M M M M M F F M F M F F M M M F M F M

The doses of intracoronary administration 16 to 21. RCA, Right coronary artery: WA, Left mittee Report.‘”

METHODS Definitions.

Chest pain

+ + + + + +

ST change

Lead

mV

III

+0.5 -0.2 +0.2 +0.2

aVf

III II NC II III NC NC

+ -

aVf

NC aVf

L

NC

NC III II

+ + + + +

aVf

of substance coronary

artery,

ECG Spastic segment

mV

Spastic segment

NC

-

Lead

Sl

-

NC

Sl

+ +

v3

+0.1

III

+0.2

Sll

NC

-

s3 s2

NC

-

+

V5

+0.2 +0.5

s2 s2

v5 v4

+0.‘2

-

+ +

NC -0.2 +0.02

+ +

II V4

Sl

+0.3 +0.2 NC -0.3 +0.3 -0.3 f0.2 NC +0.4 +0.2 -0.3 +0.3 +0.2 NC

NC NC

+

-

NC

NC

+

v4

+0.3 NC

s2

+ +

+0.5

NC

-

+0.2 +0.3 +0.8

s2 Sl Sl

-

NC v4 v6

+

NC +0.2

NC

NC

-

aVf

+0.3

s2

Sl

in patients

Sl to S15 refer

Coronary vasospasmwasdefined as a total or subtotal vasoconstrictionprovoked by the intracoronary administration of ergonovine in the angiographically normal segment.Evidence wasneeded of chest pain and/ or ischemicelectrocardiographic changes.An angiographically normal coronary artery was defined as a vessel having neither a discrete stenosis nor intimal irregularity. Patients. The protocol of this study wasapproved by the Kyushu Welfare PensionHospital Human Subject Review Committee, and informed consentwasobtained from each patient. Twenty-one Japanesepatients with variant angina who entered the Kyushu Welfare Pension Hospital were studied. Their age,sex, electrocardiographic changesduring attack, and results of coronary arteriography are listed in Table I. No patient had a history of myocardial infarction. All patients had experienced spontaneousattacks of angina in the month before the study. All drugs except nitroglycerin were discontinued for at least 48 hours before the study. Nitroglycerin was also discontinued at least 2 hours before the study. Protocol. Coronary arteriography wasperformed on the morning of the study by meansof the Judkins technique with a 5F Judkins catheter. Blood pressurewasmonitored through this catheter. A standard 12-lead electrocardiogram was recorded continuously with a six-channel recorder. Single-plane coronary cineangiogramswere obtained immediately before and after drug administration at

v4 v5

+ +

aVf

P were 20 pmol/min

V5

+

NC

NC, no change.

ST change

Chest pain

v4

+

V5

1 to 9,5 pmol/min to the segments

v4

+

NC in patients of the coronary

10 to 15, and arteries

S6

S7 Sll s12

s12 S7 Sll

S6 S7 Sll

S6 S6 s13 s13 s12 s13 -

1.5 pmol/min

as defined

in patients

by the AHA

Com-

30 frames/set by injection of 8 ml nonionic contrast material (Iopamidole 370, Schering AG Pharma, Japan). After the baselinecoronary angiogramwasobtained, physiologic saline solution (5 ml) and substanceP dissolved in 5 ml warmed physiologic saline solution were injected into the right and left coronary arteries every 2 minutes. One coronary arteriogram wastaken after eachinfusion. Each patient wasgiven only onedoseof substanceP (1.5pmol/min, n = 6; 5 pmol/min, n = 6; 20 pmol/min, n = 9) to avoid any influence on subsequent provocation of coronary vasospasmby ergonovine. Then, 40 pg ergonovine wasadministered into the right or left coronary artery to provoke coronary vasospasm. I4 A coronary arteriogram was taken immediately after anginaor ischemicelectrocardiographic changesappeared and/or 4 minutes after the administration of ergonovine. ISDN (1 mg) was then administered into the coronary artery and a coronary arteriogram was taken. The intracoronary infusion of ergonovineand ISDN wasdone in one coronary artery and repeated 30 minutes later in the other coronary artery. ST segmentelevation wasdefined assignificant whenthe ST segmentin the normally conducted beatswaselevated from the control level by more than 0.2 mV at 0.08seconds after the J point. ST segmentdepressionwasalsodefined as significant when it wasdepressedby more than 0.1 mV. Coronary vasospasmwas consideredto be induced by ergonovine if a severe transient coronary vasoconstriction wasdemonstratedangiographically in associationwith angina and/or ischemicST segmentdeviation.

February

300

Yamamoto

et al.

American

Heart

1992 Journal

Fig. 1. Photographs of four end-diastolic framesof the right coronary artery in left anterior oblique projection. A, Angiogram after physiologic salinesolution infusion. 6, Angiogram after 20 pmol/min substance P infusion. Significant epicardial coronary arterial dilation occurs.C, Angiogram after 40 pg ergonovine infusion. Coronary spasmwith complete occlusion of the right coronary artery occurs. D, Angiogram after ISDN administration. The coronary artery is angiographically normal.

Quantitative angiographic analysis. Coronary artery luminal diameterswere measuredwith the useof an automated edgecontour detection computer system (Densitometric analyzer CAD-98, ELK CINE-ANGIO SYSTEM, ELM0 Medical Products, Japan), from the end-diastolic film frames.The analyzed regionswerethe spasticsegment of the right and left coronary artery and (1) nonspastic segmentjust proximal to the spastic segmentor (2) nonspastic segment of another branch, the diameter after administration of ISDN being almost the sameas that of the spastic segment.The size of the Judkins catheter was used to calibrate the imagein millimeters, and correction was made for radiographic pincushion distortion. The arteries were analyzed by a technician who was unaware of the identity of the patient and of the working hypothesis and designof the study. Coronary arteries that were inadequately opacified and segmentsthat were overlapped by other structures were not analyzed. Statistical analysis. All values are given asmean +_SD. An unpaired t test wasusedfor intergroup comparisonsof drug-induced changesin relative coronary artery diameter and absolutecoronary artery diameter (expressedasa percent increaseover the value obtained before drug administration or a percent diameter of value after ISDN administration). A p value of lessthan 0.05wasconsideredto have statistical significance.

RESULTS

Intracoronary administrations of physiologic saline solution, substance P up to a maximal dose of 20 pmol/min, and ISDN had no apparent effect on heart rate and systemic arterial pressure. The mean heart rates and arterial pressures were 62 + 3 beats/min and 101 + 3 mm Hg during physiologic saline solution infusion, 62 s 2 beatsjmin and 102 -t 4 mm Hg with highest dose of substance P, and 64 -t 4 beats/ min and 100 +- 3 mm Hg after the administration of ISDN. Intracoronary administration of physiologic saline solution and substance P did not cause any electrocardiographic change or chest pain. In the 21 patients, diagnostic electrocardiographic changes occurred (ST elevation in 17, ST depression in 5), and these were accompanied by chest pain that occurred 1 to 4 minutes after the intracoronary administration of ergonovine. Right coronary arterial spasms occurred in 13 (62%) of 21 patients, left coronary arterial spasmsin 17 (81% ) of 21 patients, and spasms in both coronary arteries in 9 (43 % ) of 21 patients. Intracoronary administration of ISDN rapidly reversed the coronary vasospasm, electrocardiographic changes, and chest pain.

Volume Number

123 2

Endothelium-dependent

I

CONT

I

SALINE

vasodilation

I

I

I

1.5

5.0

20.0

SUBSTANCE-P

in coronary spasm

30 1

(pmWMln)

2. Dose-response relation of spasticsegment(closed circles) and nonspasticsegment(open circles) of epicardial coronary artery to intracoronary administration of substanceP. Changein coronary arterial diameter (percent increasein diameter) in this figure is expressedasa percent increaseover the value obtained before drug administration. Vertical bars in this and subsequentfigures indicate 1 SD. CONT, control; SALINE, physiologic saline solution. Fig.

Intracoronary administration of substance P dilated both spastic and nonspastic segments (Figs. 1 to 4). An example of the dilation of the epicardial coronary artery in response to substance P is shown in Fig. 1. Ergonovine provoked vasospasms in the right coronary artery but not in the proximal segment of the artery. Substance P dilated both the proximal, nonspastic segment and the distal, spastic segment (Fig. 1). The dose-response relationship is shown in Fig. 2. The dose-dependent rise was highly significant for both spastic and nonspastic segments, and there was no significant difference between them. Intracoronary administration of substance P at 20 pmol/min dilated both spastic and nonspastic segments to a degree comparable to that resulting from intracoronary administration of ISDN (Fig. 3). Responses of the coronary artery to substance P in spastic and nonspastic segments were virtually identical (96.3 f 11.5% vs 95.7 + 5.0%, p > 0.05, expressed as a percent diameter of the value seen after ISDN administration). When spastic and nonspastic segments were adjacent in the same epicardial coronary artery and the nonspastic segment was located just proximal to spastic segment, substance P dilated both segments to almost the same degree (Fig. 4). Those vasospastic coronary segments showed a significantly larger constriction in response to the intracoronary administration of ergonovine, compared with findings in the segments not vasospastic

(20.0 f 13.1% vs 80.1 +- lO.O%, p < 0.01, expressed as a percent diameter of the value seen after ISDN administration). The diameter after the intracoronary administration of ISDN did not differ significantly between the spastic and the nonspastic segments (2.3 _+0.2 mm vs 2.3 -+ 0.4 mm, p > 0.01). DISCUSSION

This study demonstrates that substance P can dilate both spastic and nonspastic segments of the human epicardial coronary artery. It is well known that substance P is comparable to nitrates in its ability to dilate epicardial coronary arteries in conscious humans, without peripheral effects.lO! i5 The degree of epicardial coronary arterial dilation by the intracoronary administration of substance P in this study was almost the same as that reported by Bossaller et al.1° and Crossman et al. i5 In vitro studies with isolated coronary arteries from patients who have received heart transplants showed that substance P is an endothelium-dependent vasodilating agent.% g In addition, substance P produces monotonic dose-response curves that are probably reflecting a single mechanism of action, that is, the release of EDRF from endothelial cells.8-10Because substance P does not appear to influence the tone of deendothelialized arteries, it doesnot seemto exert a direct effect on vascular smooth muscle. Hence, unlike acetylcholine, substance P can be used to evaluate only the endotheli-

February

302

Yamamoto

et al.

American

RIGHT

A

+ --o--

CORONARY

LEFT

Spastic segment Non-spastic segment

Heart

1992 Journal

CORONARY

6

T

80 -

60 -

P