Heart period variability in patients with variant angina

Heart

Period

Takeshi Tsuchiya,

Variability in Patients Variant Angina

MD, Ken Okumura,

MD, Hirofumi

and Hisao Ogawa, To clarify how cardiac autonomic control is affected in variant angina, we analyzed heart period variability in 35 patients with variant angina and in 19 control subjects. Patients with variant angina were divided into lvessel (group S, n = 17) and multivessel spasm groups (group M, n = 18) according to the site(s) of ST elevation on the electrocardiogram during attacks. The U-hour Holter electrocardiogram recorded 6 + 3 days after the treatment with calcium antagonist was analyzed to avoid the possible influence of spontaneous attacks. In 5 group M patients, the electrocardiogram recorded 1 month after the treatment was also analyzed. There was no difference in the number of spontaneous attacks between groups S and M. The standard deviation of all normal RR intervals (SDNN) and the percentage of differences between adjacent normal RR intervals >50 (pNN50) in variant angina were slightly but significantly

Yasue, MD, Kiyotaka

With Kugiyama,

MD,

MD

lower than those in controls. There were no differences in other indexes between variant angina and controls. When the data were analyzed separately in groups S and M, averaged RR intervals (MN), SDNN, pNN50, high-frequency power, and low-frequency power in group M were significantly lower than those in group S and controls, and the ratio of low- to high-frequency power in group M was significantly higher than that in group S and controls. There was no difference in any index between group S and controls. All abnormal indexes in group M recovered to levels similar to those in controls 1 month after the treatment. In conclusion, depressed cardiac vagal control and sympathetic-dominant sympathovagal interaction were present in patients with variant angina, especially in those with multivessel spasm. (Am J Cardiol 1996;77:932-936)

nalysis of heart period variability provides a useful, noninvasive, and quantitative evaluation A of the sympathovagal interaction modulating heart

angina had resting angina associated with ST elevation r2 mm on the electrocardiogram. The patients with variant angina were divided into 1-vesperiod in humans.lm3 Reduced time and frequency sel (group S, n = 17) and multivessel spasm domain measures of heart period variability assessed groups (group M, n = 18) according to the site(s) by 24-hour ambulatory electrocardiographic record- of ST elevation on the electrocardiogram during ings are associated with an increased mortality rate spontaneous and/or hyperventilation-induced anin patients with acute and old myocardial infarc- ginal attacks. Of the 17 group S patients, 8 had ST tion.3-6 Coronary artery disease due to organic fixed elevation noted in the anterior leads and 9 in the lesions is also associated with vagal-dominant im- inferior leads. In group M patients, ST elevation pairment in the cardiac autonomic function.‘.’ In pa- was in both the anterior and inferior leads, either tients with variant angina, transmural myocardial simultaneously or at different times. All controls ischemia due to coronary artery spasm occurs during exhibited atypical chest pain; angina was excluded angina1 attacks9,” and the extent of ischemia is par- by noninvasive examinations, including treadmill ticularly large in patients with multivessel coronary exercise testing and hyperventilation test and by spasm.‘l-13 Thus, cardiac autonomic activity may be the coronary angiographic study, including the affected in these patients, but this remains to be elu- provocation test for spasm with intracoronary acetylcholine injection. Table I shows the patient procidated. This study examines heart period variability in patients with variant angina and clarifies how the files, including age, sex, history of hypertension cardiac autonomic control is affected. (defined as systolic blood pressure >160 mm Hg and/or diastolic pressure >90 mm Hg), coronary METHODS arteriographic findings, left ventricular ejection Study patients: This study included 35 patients fraction, left ventricular end-diastolic pressure, with variant angina (3 1 men and 4 women; mean number of ischemic attacks detected in the 24-hour age 60 years, range 46 to 76) and 19 control sub- Holter recording, and calcium antagonists used for jects without any organic heart diseases (16 men the suppression of spasm. and 3 women; mean age 59 years, range 46 to 73) All subjects met the following criteria: (1) sinus who were admitted to our hospital for evaluation rhythm without atrioventricular conduction disturof their chest symptoms. Each patient with variant bance, (2) extrasystoles < 1,000 beats/day, and (3) normal electrolyte levels. The mean number of exFrom the Division of Cardiology, Kumamoto University School of Medtrasystoles in all of the present subjects was 322 -t icine, Kumamoto, Japan. Manuscript received October 3, 1995; re 282 beats/day (mean 5 SD). Patients with myocarvised manuscript received and accepted December 6, 1995. dial infarction, cardiomyopathy, valvular heart disAddress for reprints: Ken Okumura, MD, Division of Cardiology, ease, congestive heart failure, diabetes mellitus, and Kumamoto University School of Medicine, Honio l-1-1, Kumamoto, 860 Japan. renal failure were excluded from the study. Techni932

0 1996 by Excerpta Medica, All rights reserved.

Inc.

0002.9 149/96/$15.00 PII SOOO2.9149(96)00005-7

TABLE

I Patient

Characteristics Variant Control

Characteristics

(n =

Age (rd

58.6

Men/women

-t 6.6

16/3

S

(n = 17) 59.5

t 7.8

16/l

6 (32%)

Hypertension Calcium

Group 19)

3 (18%)

Angina Group

M

(n = 18) 60.2

+ 8.8

15/3

A (22%)

antagonists

A (24%)

Dihydropyridine Diltiazem

10 (59%)

3 (18%)

Combined

7 (39%) 10 (56%)

1 (6%)

administration to evaluate the extent of significant organic stenosis >50% of the luminal diameter. A left ventriculogram was also filmed in the 30” right anterior oblique projection, and the left ventricular ejection fraction was calculated using an arealength method. Holter recording system and analysis of heart period variability:

The Holter electrocardiogram was recorded using 2-channel 1 Vessel 0 4 (24%) 3 (17%) Del Mar Avionics recorders 22 Vessels 0 1 (6%) 1 (6%) (models 445B and 447 ) and anLeft ventricular ejection fraction (%) 75 k 8 73 + 8 74 i- 8 alyzed with the Del Mar AvionLeft ventricular enddiastolic 7.4 + 2.4 9.0 t 2.3 9.0 +- 3.7 ics analyzer (563 stratascan), pressure (mm Hg) Number of attacks (times/day) 11 2 11 11 +-9 Los Angeles, California. To prevent the processing of artifacts, Values ore expressed OSmean i I SD. we checked the quality of the sigCombined = dihydropyridine plus diltiozem; group M = patients with multivessel spasm; group S = patients with l-vessel spasm. nal by visual inspection of the computer display while sampling the RR intervals. Recordings in tally poor Holter electrocardiograms were excluded which artifacts were present in >5% of the recording from analysis. period were excluded from analysis. To avoid the Protocols: All medications, except for sublingual inclusion of artifacts and premature beats in the nitroglycerin, and smoking were stopped for ~3 count for heart period variability, we set the system days, after which the 24-hour Holter electrocardio- to remove all the cycle lengths > 120% or ~80% of gram was recorded. In all patients with variant an- the preceding RR interval. Time domain analysis and gina, at least 1 spontaneous attack was recorded on spectral analysis of heart period variability using a the 24-hour electrocardiogram. An ischemic attack fast-Fourier transform method based on a windowed due to coronary spasm was defined by ST segment periodogram technique were made using Del Mar elevation 22 mm from the baseline at 80 ms after Avionics software (system E5 ESHRV revision-A). From the entire Holter recording, we calculated the nadir of the S wave, lasting for > 1 minute. If 2 sequential episodes of ST segment elevation identi- averaged RR intervals (MN) and 2 conventional fied on the 24-hour Holter electrocardiogram were time domain indexes, including the global standard separated by 51 minute, these episodes were rec- deviation of all normal RR intervals over 24 hours ognized as 1 episode. From the 24-hour Holter elec- (SDNN) and the percentage of differences between trocardiogram, the number of ischemic attacks was adjacent normal RR intervals >50 ms computed during the entire 24-hour electrocardiogram (pNN50) , counted to elucidate the disease activity. The patients were then treated with diltiazem both of which are markers of cardiac vagal control. ( 120 to 360 mg; n = 20) or dihydropyridine com- Frequency domain analysis was performed on 5pound (nifedipine 40 to 80 mg or nisoldipine 10 to minute blocks of data, and the resulting 5-minute 20 mg; n = 11)) or a combination of diltiazem and power spectra were averaged over 24 hours. The dihydropyridine (n = 4). Some patients also re- areas under the low-frequency (0.04 to 0.15 Hz) and ceived long-acting nitrates. None of the patients were the high-frequency (0.15 to 0.4 Hz) regions of the treated with 0 blocker, digitalis, or angiotensin-con- curve were used for analysis.’ Low-frequency power vetting enzyme inhibitor. At 6 ? 3 days after the reflects cardiac vagal activity when assessedby the treatment with calcium antagonist, when angina1 at- 24-hour Holter electrocardiographic recording, I4and tack was completely suppressed, the 24-hour Holter high-frequency power is a pure cardiac vagal efferelectrocardiogram was recorded again in all patients ent signal.‘-” High- and low-frequency powers were with variant angina. To avoid the influence of spon- expressed in natural logarithmic units. The low- to taneous angina1 attack(s) on the cardiac autonomic high-frequency power ratio, a measure of sympacontrol, we analyzed heart period variability in the thovagal interaction, ‘J was calculated in absolute Holter electrocardiogram recorded after the suppres- values from every 5-minute block and averaged over sion of angina1 attacks. In controls, the 24-hour Hol- 24 hours. Statistical analysis: The data are shown as mean ?Z ter electrocardiogram was recorded without any medications. In 5 group M patients, Holter electro- 1 SD. The differences in the mean values between the cardiogram was performed again 1 month after the patient groups were assessedby analysis of variance treatment with calcium antagonist. followed by the Fisher’s protected least significant In all patients with variant angina and in controls, difference multiple comparisons test. The difference in the mean values within 1 group was assessedby 2coronary arteriography was done after nitroglycerin Organic

None

stenosis

(-50%)

19

CORONARY

12 (71%)

14 (78%)

ARTERY DISEASE/HEART PERIOD VARIABMTY IN VARIANT ANGINA

933

respectively, whereas those in the latter were 937.7 2 112.0 ms (p = NS), 149.2 ? 29.7 ms (p Control Group S Group M <0.05), and 9.4 t 4.1% (p (n = 19) (n = 17) Vorioble (n = 18) <0.05), respectively. There was 937.7 + 1 12.0 997.8 2 155.9 850.6 2 1 13.5* MN (ms) no difference in any of the freTime domoin quency domain variables be141.2 + 34.6 106.2 + 33.7’ 149.2 2 29.7 SDNN (ms) tween the patients with variant a.2 2 4.3 9.4 t 4.1 3.2 + 3.6’ pNN50 (%) angina and controls: high-freFrequency domain 5.3 + 0.7 5.1 2 0.6 4.2 + 0.9’ Ln (HF) (ms2) quency power, low-frequency 5.9 2 0.5 6.0 k 0.7 5.6 2 0.7’ Ln (LF) (m?) power, and the ratio of low- to 2.6 + 0.7 2.5 rf: 0.8 3.8 k 1.8’ LF/HF high-frequency power in the forValues ore expressed os meon + 1 SD. mer were 4.7 + 0.9 ms’, 5.6 + l p <0.05 vs control and p 50 milliseconds over 24 hours; SDNN = global standard deviation of RR intervals When the data of the patients over 24 hours. with variant angina were analyzed separately in groups S and M, all measures of MN, SDNN, tailed paired t test. The differences in clinical profiles pNN50, high-frequency power, and low-frequency and hemodynamic data among groups were analyzed power in group M were significantly different from by chi-square test with Yates’ correction. A p value those in controls and group S (Table II). There was <0.05 was considered significant. no difference in the variables between group S and controls. The distributions of MN and SDNN are shown in Figure 1, and those of high-frequency RESULTS power and the ratio of low- to high-frequency power There were no differences in clinical profiles, coronary angiographic findings, or hemodynamic in- are shown in Figure 2. There was no significant cordexes among controls, group S, and group M (Table relation between the number of ischemic attacks deI). There were also no differences in the kind of tected by the 24-hour Holter electrocardiogram and each of the values in group M. Among group S pacalcium antagonist used, the number of significantly tlents, there was no difference in the variables bestenotic coronary arteries, or the number of ischemic attacks detected before treatment between group S tween the patients with ST elevation in the anterior leads versus the inferior leads. and group M. The influence of the kind of calcium antagonist Comparison of heart period variability indexes: Time domain indexes in the patients with variant angina on the indexes of heart period variability was evalwere slightly but significantly lower than those in uated in the patients with variant angina. There were controls: MN, SDNN, and pNN50 in the former were no differences in any of the time and frequency do922.1 + 153.1 ms, 124.2 + 37.6 ms, and 5.6 2 4.6%, main indexes among the patients treated with dihyTABLE

Subjects

II

of Time and Frequency S and Group M Patients

Comparison

and Group

Domain

Variables

B

P-d.05 I

I

-n

NS

THE AMERICAN JOURNAL OF CARDIOLOGY@’

Control

p-co.01 I

III NS

pco.01

Conho1 G&p S Grdup M Variant angina

934

Among

300

1

I

pxo.01

OV Variant angina

VOL. 77

MAY 1, 1996

FIGURE 1. Distributions of MN (A) and SDNN (s) in control, group S, and group M obtained 6 z 3 days oE~“k treatment with calcium antagonist(s). MN and SDNN in roup M were significantly Power than those in controls and grou S. Group M = patients WIx muhivessel spasm; group S = patients with l-vessel spasm; MN = average RR interval over 24 hours; SDNN = global standard deviation over 24 hours.

r

A FIGURE 2. Distributions of the natural logarithms of high-frequency power (A) and low- to high+ uency power ratio (s) in contra7 , group S, and group M patients obtained 6 2 3 days after treatment with calcium antagonist(s). High-frequency power in group M was significantly lower than those in controls and group S. lowto high-frequency power ratio in roup M was significantly higher than those in controls and group S. HF = high-frequency power; LF/HF = lowto high-frequency power ratio.

I

p&o1

p
nn

A-

3.

NS

pd.01

l

Control Group S Group M Variant angina

Control Group S G&p M Variant anglna

tween group S and controls. In this study, heart period variability was assessedafter complete suppresChanges in heart period variability indexes in group sion of angina1 attacks with calcium antagonist(s) in M: All of the abnormal indexes documented in the all patients. Moreover, there was no difference in the Holter electrocardiogram recorded 6 + 3 days after kind of calcium antagonist used between group S and the treatment with calcium antagonist in group M group M. Thus, the influences of spontaneous angichanged significantly at 1 month after the treatment nal attacks and calcium antagonist on the present re(Figure 3) : MN, SDNN, pNN50, low-frequency sults are considered to be minimal. To the best of power, and high-frequency power increased, our knowledge, this is the first study to show abnorwhereas the ratio of low- to high-frequency power mal heart period variability indexes in patients with decreased. All of the indexes at 1 month after treat- variant angina. Thus, low magnitudes of SDNN, ment were similar to those observed in controls. pNN50, low-frequency power, and high-frequency power and a high magnitude of low- to high-frequency power ratio detected in the patients with mulDISCUSSION tivessel spasm strongly suggest that the cardiac vagal The present study showed that SDNN and pNN50 influence on the heart period is decreased and that in patients with variant angina were slightly but sig- cardiac sympathovagal interaction is predominantly nificantly lower than those in controls. When the data sympathetic in these patients. were analyzed with regard to the extent of ischemia, The present study could not clarify the mechaall of the evaluated heart period variability indexes nism for these abnormal cardiac autonomic controls in group M were significantly different from those in patients with variant angina. Many mechanisms in controls. Furthermore, all of them were also sig- are involved in the reduced cardiac vagal control.’ nificantly different from those in group S, whereas Decreased cardiac function has been reported to be no difference was found in any of the indexes be- associated with reduction of cardiac autonomic control, 3,15but all of the present patients showed normal ventricular TABLE III Comparison of Time and Frequency Domain Variables Among Patients function. None of the patients were Treated With Dihydropyridine (DHP) or Diltiazem (OIL] or Dihydropyridine Plus restricted during hospitalization, esDiltiazem (Combined) pecially after the treatment with calDHP DIL Combined cium antagonist. Medications with (n= 11) (n = 4) Variable (n = 20) cardiovascular actions are likely to 945.5 + 150.7 900.7 2 152.0 964.8 5 190.5 MN (ms) influence the autonomic nerve acTime domain tivity,” but all drugs except calcium 122.9 2 44.7 123.2 t 32.6 133.2 + 50.0 SDNN (ms) antagonist were withdrawn during 6.5 k 5.7 4.4 t- 3.2 9.0 t 6.3 pNN50 (Z) Frequency domain the study period. Reduced cardiac 4.7 t 1 .o 4.7 t 0.8 4.9 + 1.5 Ln (HF) [ms’) vagal function may be related to the 5.5 t 0.9 5.7 + 0.6 5.8 t 1.0 Ln (LF) (ms’) progression of coronary atheroscle2.9 k 1.5 3.3 -t 1.6 3.1 2 1.6 LF/HF rosis,’ but most of the present paData are shown (IS mecln t 1 SD. tients had normal or nearly normal Abbreviations OSin Table II. coronary arteries. Hypertension was dropiridine compound, those with diltiazem, and those with the combination (Table III).

CORONARY

ARTERY DISEASE/HEART PERIOD VARlAi3llliY IN VARIANT ANGINA

935

180-

B

A h N

Q

FIGURE 3. Changes in SDNN (A) and high-frequen power (6) after treatment WI9 calcium antagonist in the 5 group M patients. Abbreviations as in Figures 1 and 2.

6k3 days 1 month After treatment

6f3 days 1 month After treatment

reported to influence heart period variability, I6 but there was no difference in the incidence of hypertension among controls, group S, and group M. One mechanism that might explain the present results is damage of the intrinsic cardiac nerves and receptors caused by frequent transmural ischemia, which might affect the cardiac autonomic activity in patients with variant angina. Because the group S patients showed normal heart period variability indexes regardless of the site of ischemia, the extent of ischemia might be important in the abnormal cardiac autonomic activity. The present results further showed that the abnormal indexes detected in the group M patients all recovered to levels similar to those seen in controls after treatment with calcium antagonist for 1 month. This indicates that the decrease of cardiac vagal activity is reversible in patients with variant angina. It may be pointed out that in the present study, heart period variability was analyzed after the treatment with calcium antagonist, which may have affected the results obtained. However, there is no evidence in published reports that calcium antagonist alters heart period variability indexes as assessedby the 24-hour Holter electrocardiogram.‘4 We also analyzed the heart period variability indexes before treatment with calcium antagonist (data not shown) and found that all of the indexes in group M were abnormal and, moreover, were similar to those obtained after calcium antagonist. In conclusion, depressed cardiac vagal control and sympathetic-dominant sympathovagal interaction were present in patients with variant angina, especially in those with multivessel spasm. These abnormal autonomic activities were restored after treatment with calcium antagonist for 1 month. 1. Malliani A, Pagani M, Lombardi F, Cemtti S. Cardiovascular neural regulation explored in the frequency domain. Circulation 1991;84:482-492.

936

THE AMERICAN JOURNAL OF CARDIOLOGY@

VOL. 77

2. Pomeranz B, Macaulay R, Caudill M, Katz I, Adam D, Gordon D, Kilbom K, Barger A, Shannon D, Cohen R, Benson H. Assessment of autonomic function in humans by heart rate spectral analysis. Am J Physiol 1985;248( Heart Circ Physiol 17):H151-H153. 3. Bigger JT. Spectral analysis of R-R variability to evaluate autonomic pharmacology and to predict cardiovascular outcomes in humans. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology: From Cell’ to Bedside. Philadelphia: WB Saunders, 1994:1151- 1170. 4. Bigger JT, Fleiss JL, Steinman RC, Rolnitzky LM, Kleiger RE, Rottman JN. Frequency domain measures of heart period variability and mortality after myocardial infarction. @rulation 1992;85: 164-171. 5. Bigger JT, Fleiss JL, Rolnitzky LM, Steinman RC. Frequency domain measures of heart period variability to assess risk late after myocardial infarction. JAm Coil Cardiol 1993;21:729-736. 6. Kleiger RE, Miller JP, Bigger JT, Moss AJ, the Multicenter Post-Infarction Research Group. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 1987;59:256262. 7. Hayano J, Sakakibara Y, Yamada M, Ohte N, Fujinami T, Yokoyama K, Watanabe Y, Takata K. Decreased magnitude of heart rate spectral components in coronary artery disease: its relation to angiographic severity. Circulation 1990;81:1217-1224. 8. Bonaduce D, Petretta M, Piscione F, Indolfi C, Migaux ML, Bianchi V, Esposito N, Marciano F, Chiariello M. Influence of reversible segmental left ventricular dysfunction on heart period variability in patients with one-vessel coronary artery disease. JAm Coil Cardiol 1994;24:399-405. 9. Oliva RB, Potts DE, Pluss RG. Coronary arterial spasm in Prinzmetal angina. Documentation by coronary arteriography. N EngI J Med 1973;288:745-751. 10. Yasue H, Touyama M, Shimamoto M, Kato H, Tanaka S, Akiyama F. Role of autonomic nervous system in the pathogenesis of Ptintzmetal’s variant form of angina pectoris. Circukztion 1974;50:534-539. 11. Kugiyama K, Yasue H, Okumura K, Minoda K, Takaoka K, Matsuyama K, Kojima A, Koga Y, Takahashi M. Simultaneous multivessel coronary artery spasm demonstrated by quantitative analysis of thallium-201 single photon emission computed tomography. Am J Cardiol 1987;60:10091014. 12. Okumura K, Yasue H, Horio Y, Takaoka K, Matsuyama K, Kugiyama K, Fujii H, Morikami Y. Multivessel coronary spasm in patients with variant angina: a study with intracoronary injection of acetylcholine. Circulation 1988;77:535-542. 13. Dunn RF, Kelly DT, Sadick N, Uren R. Multivessel coronary artery spasm. Circulation 1979;60:45 l-455. 14. Cook JR, Bigger JT, Kleiger RE, Fleiss JL, Steinman RC, Rolnitzky LM. Effect of atenolol and diltiazem on heart period variability in normal persons. J Am Coil Cardiol 1991;17:480-484. 15. Saul JP, Arai Y, Berger RD, Lilly LS, Colucci WS, Cohen RJ. Assessment of autonomic regulation in chronic congestive heart failure by heart rate spectral analysis. Am J Cardiol 1988;61:1292- 1299. 16. Cbakko S, Mulingtapang RF, Huikuri HV, Kessler KM, Materson BJ, Myerburg RJ. Alterations in heart rate variability and its circadian rhythm in hypertensive patients with left ventricular hypertrophy free of coronary artery disease. Am Heart .I 1993;126:1364-1372.

MAY 1, 1996