Pindolol and propranolol in patients with angina pectoris and normal or near-normal ventricular function

Pindolol and Propranolol in Patients With Angina Pectoris And Normal or Near-Normal Ventricular Function Lack of Influence of Intrinsic Sympathomimetic Activity on Global and Segmental Left Ventricular Function Assessed by Radionuclide Ventriculography DANTE E. MANYARI, MD, WILLIAM J. KOSTUK, MD, S. GEORGE CARRUTHERS, MD, DONALD J. JOHNSTON, MD, and PAUL PURVES, RT To investigate the role of intrinsic sympathomimetic activity on left ventricular (LV) function during antianginal therapy with beta-adrenoreceptor antagonists, 23 patierds with chronic, exercise-induced angina pectoris and normal or near normal LV function underwent radionuclide ventriculography at rest and during exercise, during 3 randomly allocated periods: (1) treatment with oral propranolol, a drug without intrinsic sympathomimetic activity, 40 to 80 mg 4 times a day; (2) treatment with pindolol, a drug with marked intrinsic sympathomimetic activity, 5 to 10 mg 2 times a day; and (3) a control period. During the control period, the LV ejection fraction decreased from rest (58.9 -t- 8.2 % ) to exercise (54.3 -I- 10.7%), and the wall motion score decreased from 0.57 -I- 1.08 at rest to 2.39 -I- 2.10 during exercise, p <0.001. After propranolol, the ejection fraction did not change significantly at rest (57.2 -I- 8.1% ) but improved during exercise (56.8 ± 11.8%), compared with control values. After

pindolol, the ejection fraction did not change at rest (57.9 -4- 8.6 % ) but improved during exercise (56.9 -I- 8.1% ), compared with control values. Similarly, the wall motion score after administration of both agent s did not change significantly at rest, but improved during exercise (p <0.001). The number of anginal episodes, nitroglycerin tablets consumed, and magnitude of S-T segment depression decreased significantly with both pindolol and propranolol. With both drugs, a similar improvement in exercise tolerance and a similar decrease in exercise heart rate and blood pressure were obtained. it is concluded that pindolol and propranolol, betaadrenoreceptor antagonists with and without intrinsic sympathomimeUc activity, respectively, have similar effects on global and regional LV function in patients with angina pectoris, at doses producing equal suppression of exercise heart rate and similar antianginal effect.

Comprehensive, carefully designed studies have failed to show any significant antianginal differences between various beta-adrenoreceptor antagonists with different ancillary pharmacologic properties, when administered on a short- or long-term basis to patients with stable, exertional angina pectoris. 1,2 The influence of intrinsic sympathomimetic activity on global or regional LV

function in patients with angina is not known. This investigation compared the effects of pindolol and propranolol, drugs with and without intrinsic sympathomimetic activity, respectively, on LV performance at rest and during exercise in patients with exercise-induced angina. LV function was assessed using electrocardiogram-multigated blood pool cardiac scintigraphy, a noninvasive technique which has been shown to reliably assess global and segmental LV function. 3-5

From the Cardiac Investigation Unit, University Hospital, and the Department of Medicine, University of Western Ontario, London, Ontario, Canada. This study was supported by the Ontario Heart Foundation and Sandoz (Canada)Limited, Dorval, Quebec, Canada. Manuscript received April 12, 1982; revised manuscript received September 15, 1982, accepted September 24, 1982. Address for reprints: William J. Kostuk, MD, Cardiac investigation Unit, University Hospital, Box 5339, Terminal A, London, (J~btar~o,N6A 5A5 Canada.

Methods Patient group: Twenty-six men, aged 39 to 68 years (mean 52), with exercise-induced angina pectoris were studied. The duration of angina ranged from 3 months to 6 years (average 17 months). No patient had systemic hypertension, valvular

427

428

LEFT VENTRICULAR FUNCTION AND BETA-ADRENERGIC BLOCKING AGENTS

heart disease, intermittent claudication, or any condition that could limit exercise performance. None had diabetes mellitus, obstructive airway disease, or any history of previous adverse reaction to beta-blocking agents. The cardiothoracic ratio was normal in each. There was no clinical or roengtenographic evidence of heart failure. All had sinus rhythm. Ischemic electrocardiographic changes developed during treadmill exercise in all 26. Patients seen consecutively at the outpatient clinic, who met the foregoing criteria, were selected for this study solely on the basis of their willingness to participate. Coronary cineangiography performed within 1 year of study in 16 of the 26 patients showed evidence of 3-vessel disease in 10, 2-vesset disease in 4, and 1-vessel disease in 2. Two patients had a history of myocardial infarction with abnormal Q waves (>40 ms duration) on the electrocardiogram. Six patients were taking long-acting oral nitrates, which were kept constant during the study period. Study design: This investigation assessed objective measures of LV function; thus an open crossover design was used. Each patient underwent 3 consecutive study periods: (1) a control period in which no beta-adrenoreceptor antagonists were administered; (2) treatment with oral propranolol 4 times a day; and (3) treatment with oral pindolol 2 times a day. The order of the 3 treatment periods was allocated randomly for each patient. The duration of the control period varied between 1 and 4 weeks; the duration of treatment periods with propranolol and pindolol varied from 4 to 6 weeks. The dose of propranolol was 40 mg 4 times a day and the dose of pindolol was 5 mg 2 times a day, a 16:1 ratio shown to exert similar beta-adrenergic blockade. 6 These doses were increased to 80 mg 4 times a day and 10 mg 2 times a day, respectively, in patients in whom clinical improvement did not occur (compared with the control period) or whose exercise heart rate was >120 beats/min with the initial dose, indicating that betaadrenoreceptor blockade was inadequate. The initial doses were not increased in patients in whom clinical improvement was achieved or in whom the heart rate at peak exercise was lower than 120 beats/rain. Improvement was defined by a reduction in the number of anginal attacks of at least 40% compared with that in the control stage. Patients requiring dosage increases underwent more than 3 study periods, but only data during drug treatment with the larger doses are reported. Patients kept a daily record of the number of anginal episodes and the number of therapeutic glyceryl trinitrate tablets consumed. Change in therapy when switching from one beta-adrenoreceptor antagonist to another was made abruptly from 1 day to the next. When treatment with 1 of the 2 drugs was followed by the control period, the drug was discontinued gradually over 1 week, At the end of each of the 3 treatment periods, all patients underwent, successively, radionuclide ventriculography at rest and during bicycle exercise 120 to 150 minutes after the morning dose of either drug or at a similar hour of the day after the control period, a treadmill exercise test 45 minutes after termination of bicyeJe exercise, and clinical assessment. In the 6 patients taking long-acting nitrates, doses were kept constant throughout the entire investigation, but were withheld for 24 hours before exercise studies; the same dose was given again after clinical assessment. The clinical assessment was made by 2 participating cardiologists. The number of anginal episodes experienced, glyceryl trinitrate tablets consumed, and adverse effects were recorded. The number of remaining tablets were counted to assess compliance. New drugs, instructions, or both were given. The treadmill exercise test was performed using a modified Bruce protocol, with elevation of the speed and incline at 3-

minute intervals. 7 All patients were exercised to angina during the control stage, and to angina or fatigue after drug treatment. An ischemic electrocardiographic response was a fiat or downward sloping depression of the S-T segment of at least 0.1 mV (1 mm) in any lead, persisting for >_0.08 second and present in 3 consecutive beats with stable isoelectric baseline. Cuff blood pressure measured over the right brachial artery and a 12-lead electrocardiogram were obtained at rest, at each exercise stage, and immediately and at 2, 4, 6, and 8 minutes after exercise. The treadmill walking time was recorded to the nearest 5-second period. Exercise heart rate was measured continuously on a cardiotachometer and the maximal heart rate was measured from 10 consecutive beats on the electrocardiogram recorded during the last 15 seconds of peak exercise. S-T segment depression was measured as the linear depression of the J point from the isoelectric line joining 2 consecutive P-R segments. Measurements were made in the lead that showed maximal S-T segment depression and were averaged over 4 consecutive complexes in which the isoelectric line was stable. The symptoms limiting exercise tolerance were recorded. Assessment of LV function: Radionuclide ventriculography, to assess global and regional LV function, was performed using the electrocardiographic-multigated method described previously. 3-~ After in vivo labelling of the red blood cells with technetium-99m, blood pool cardiac scintigraphy was performed in the modified left anterior oblique projection that best separated both ventricles and atria, at rest and during graded supine bicycle ergometer exercise. Details of the exercise protocol and the instrumentation used in our laboratory have previously been described. 4 Analysis of the radionuclide studies was carried out blindly by an experienced observer. Global LV ejection fraction and semiquantitative measurements of segmental LV function (wall motion score) were obtained at rest and during peak exercise. In our laboratory, radionuclide LV ejection fraction and wall motion analysis correlated closely with contrast studies, r = 0.87.5 Inter- and intraobserver variability of ejection fraction measurements showed correlation coefficients of 0.97 and 0.92, respectively. 4,5 A radionuclide ventriculogram at rest was considered normal if ejection fraction was >55% and the wall motion score was equal to zero. Failure of ejection fraction to increase by at least 7% (absolute value), or any increase in wall motion score from resting values, or both was considered an abnormal LV function response to peak exercise. 4 Statistical methods: Analysis of variance was used to determine the significance of changes during treatment periods. If this analysis indicated a significant difference in the mean values for a variable, then Tukey's multiple comparison procedure wasperh)rmed. In addition, changes in exercise duration, electrocardiographic S-T segment depression, and wall motion score were also analyzed by a nonparametric method using Wilcoxon's rank sum test. Probability of _<5% was considered significant.

TABLE I

Order of the Therapy Periods in 23 Patients With Angina Pectoris in This Study

Patients (n)

First Therapy

Second Therapy

Third Therapy

6 4

C C

Pr Pi

Pi

3 2

Pr Pr

C Pi

Pr Pi C

2 6

Pi Pi

Pr C

C Pr

C = control; Pi = pindolol; Pr = propranolol.

February 1983 THE AMERICANJOURNAL OF CARDIOLOGY Volume 51

p(.J

H CONTROL o.----o PROPRANOLOL

E)

o rr tt

T

250

uJ u3 u') uJ

200 t L,)

TM

150 CO I uJ

I00

rr

N0[ W i

50 REST-I

EX-T

REST-2

EX-B

FIGURE 1. Comparison of the effects of propranolol and pindolol during sustained therapy in 23 patients with angina pectoris. Data represent mean 4- standard deviation. REST-1 = data standing, immediately before treadmill exercise. REST-2 = data supine, just before bicycle exercise. EX-T and EX-B = data at peak exercise on the treadmill and the bicycle, respectively. * p <0.05; t p <0.01; t p <0.001, versus control values.

Results Three of the initial 26 patients did not complete the entire investigation: i had hypothyroidism during the initial treatment period; 2 became asymptomatic while on treatment with the first beta-adrenoreceptor antagonist and insisted on receiving the same medication thereafter. Thus, these 3 patients were excluded, leaving a total of 23 patients who completed all phases of this study and form the basis of this report. Random allocation of the order of treatment periods resulted in fairly even distribution of patients among the 6 possible combinations (Table I). Compliance was good because <3% (average value) of the tablets were returned. No patient experienced adverse effects serious enough to warrant discontinuation of medications during the study period. Complaints such as fatigue,

TABLE II

Drug Control Propranolol Pindolol

429

drowsiness, vivid dreams, and cold extremities were reported with similar frequency during therapy with propranolol and pindolol. Clinical improvement of similar degree was observed during propranolol therapy in 17 patients taking 40 mg 4 times a day, in 5 patients taking 80 mg 4 times a day, and during pindolol therapy in 14 patients taking 5 mg twice each day, and in 7 patients taking 10 mg twice a day. In the remaining patient taking propranolol 80 mg 4 times a day and 2 patients taking pindolol 10 mg twice a day, the frequency of angina also was reduced but by <40% despite a significant reduction in the exercise heart rate (to <120 beats/min). Effect of t r e a t m e n t on the f r e q u e n c y of angina, number of glyceryl trinitrate tablets consumed, and exercise duration: During the control period, the mean number of anginal episodes reported was higher (p <0.001) than those during therapy with propranolol or pindolol (19.4 4- 10.4, 7.0 4- 5.7, and 8.6 4- 6.4 anginal episodes/month, respectively). The average number of glyceryl trinitrate tablets consumed was lower (p <0.001) during propranolol and pindolol therapy than during the control period (4.0 4. 3.0, 5.2 4. 5.4, and 11.0 4. 8.3 tablets of glyceryltrinitrate/month, respectively), in keeping With the number of anginal episodes reported. Exercise duration on the treadmill (Table II) was significantly higher during therapy with both drugs than during the control period. Patients exercised for shorter periods of time on the bicycle, but the mean exercise duration showed trends similar to those of exercise on the treadmill (duration of bicycle exercise: 5.61 4- 2.08 minutes during control, 6.52 4- 1.73 minutes after propranolol, and 6.00 4- 1.81 minutes after pindolol). The treadmill walking time, frequency of angina, and number of glyceryl trinitrate tablets consumed after propranolol were similar (p = 0.20) to those after pindolol administration. Circulatory changes: No significant differences were observed between supine and standing resting heart rates. The heart rate increased significantly (p <0.001) from resting values with treadmill and bicycle exercise during the 3 study periods. Resting and exercise heart rates (Table II) were lower during treatment with both beta-adrenoreceptor antagonists when compared

Effects of Propranolol and Pindolol on Exercise Tolerance, S-T Segment Depression, Heart Rate, Systolic Blood Pressure, and LV Function During Sustained Therapy in Patients With Angina Pectoris

TWT (rain)

S-T Segment Depression (mm)

6.74 4-2.30 7.94 4-2.70* 7.97 4-2.31"

2.4 4,0.9 1.8 4-0.8 t 1.5 4-0.8 t

Heart Rate (beats/min) Rest 78 4-12 65 4-11" 70 4-125

SBP (mm Hg)

EF ( % )

Exercise

Rest

Exercise

Rest

137 4-25 113 4-20* 110 4-17"

128 4-12 t22 4-10 125 -I-9

169 4-24 151 4-21" 148 4-19"

59

WMS

Exercise 54

4-8 57

4-11 57

4-8 58

4-12 57

4-9

4-8

Rest

Exercise

0.57 4-1.08 0.61 4-1.16 0.70 4-1.29

2.39 4-2.10 1.78 4-2.15 t 1.61 4-2.31 t

Data expressed as mean 4- standard deviation. EF = LV ejection fraction; SBP = systolic blood pressure; TWT = treadmill walking time; WMS = left ventricular wall motion score. Data for , " segment depression, exercise heart rate, and systolic blood pressure represent values during peak treadmill exercise. * p <0.001; t p <0.01; P <0 05 versus control values.

430

LEFT VENTRICULAR FUNCTION AND BETA-ADRENERGIC BLOCKING AGENTS

=CONTROL ~

-@

" z

3o

w

c~

20

t

= PROPRANOLOL ~ 1 ~ =PINDOLOL

, i

--tqI--NS'--~

li,

TREADMILL EX

- -

t---H---NS~

BICYCLE EX

FIGURE 2. Comparison of the effects of propranolol and pindolol during sustained therapy in 23 patients with angina pectoris. Data represent mean + standard deviation at peak exercise (EX) on the treadmill and the bicycle. No significant differen?e in S-T segment depression was observed between the 2 modes of exercise. 1-p <0.01; NS = p >0.05.

with heart rates during the control period. Both supine and standing heart rates at rest were somewhat lower during propranolol than during pindolol administration but the difference between means did not reach statistical significance. Heart rates during peak exercise were similar with the 2 drugs. Systolic blood pressure increased (p <0.001) from rest to exercise during control and during administration of propranolol and pindolol (Table II). At rest, supine and standing systolic blood pressure after both beta-adrenoreceptor antagonists were similar to those during the control period. After administration of both drugs, systolic blood pressures at peak exercise, on the treadmill and the bicycle, were significantly lower (p <0.001) than those during control. The differences in systolic blood pressure, at rest or during exercise, after administration of the 2 drugs were not significant. Compared with the control period, the product of heart rate and systolic blood pressure (Fig. 1), with the patients at rest in both supine and standing positions and during peak treadmill and bicycle exercise, were lower during treatment with both drugs. The products of heart rate and systolic blood pressure were similar after administration of both propranolol and pindolol. E l e c t r o c a r d i o g r a p h i c changes: An ischemic electrocardiographic response to exercise was present in all patients during the control stage, in 14 patients after propranolol, and in 10 patients after pindolol. During the control period, S-T segment depression at peak exercise was significantly greater than after propranolol or pindolol administration (Table II). The difference between mean S-T segment depression after both beta-adrenoreceptor antagonists was not statistically significant (Fig. 2). E f f e c t s of t r e a t m e n t on global LV function: During the control stage, the LV ejection fraction at rest was normal (>55%) in 14 patients and abnormal (<55%) in 9 patients, but <50% in only I patient. After administration of propranolol, resting ejection fraction was found within the normal range in 14 patients and <50%

in 3, whereas after administration of pindolol, the ejection fraction at rest was normal in 18 patients a~d <50% in the 3 (Fig. 3). There was no significant difference between the means of resting ejection fraction during the 3 study periods (Table II). An abnormal response of ejection fraction to peak exercise, as defined in the Methods section, was observed in 22 patients during the control period, in 18 patients after propranolol, and in 21 patients after pindolol. During the control period, the mean ejection fraction during peak exercise was significantly lower (p <0.05) than that at rest, whereas it was not significantly different .from the resting ejection fraction after administration of both beta-adrenoreceptor antagonists (Fig. 3). The average ejection fraction at peak exercise during therapy with propranolol and pindolol was somewhat greater than during control (Table II), but the differences did not reach statistical significance. No significant differences were observed in rest or exercise mean ejection fraction between patients taking 5 and l 0 mg of pindolol twice a day. E f f e c t s of t r e a t m e n t on regional LV function: During the control period, 20 patients showed a normal or near-normal segmental LV function (wall motion score values of 0 to 1), the remaining 3 patients had wall motion score values of 2 or higher; in only 1 patient was this higher than 3. After propranolol therapy, the resting wall motion score was normal or near normal in 20 patients, higher than 1 in 3 patients, and higher than 3 in I patient. After pindolol, the wall motion score at rest was normal or near-normal in 18 patients, higher than 1 in 5 patients, and higher than 3 in i patient (Fig. 4). No significant differences during any of the 3 study periods were found between the means of wall motion score at rest (Table II). An abnormal response of regional LV function to exercise, as defined previously, was observed in 18 patients during the control stage, in 12 patients after propranolol therapy, and in 10 patients after pindolol administration. The mean wall motion score during exercise was significantly higher (p <0.001) than the respective resting values during the control period, after propranolol, and after pindolol (Fig. 4). However, the mean exercise wall motion score during the control stage was significantly higher than that during therapy with both drugs (Table II). At peak exercise, the mean wall motion score after propranolol was similar (p = 0.30) to that after pindolol administration. No significant differences were observed in rest or exercise mean wall motion score among patients taking 5 and 10 mg of pindolol twice a day. E f f e c t s of t r e a t m e n t in relation to the basal status of L ¥ function: Resting LV function during the control period was considered normal (ejection fraction >55% and wall motion score, 0) in 13 patients and below in 10 patients. In the former group, the mean ejection fraction was 64 4- 6% and the mean wall motion score was zero; those in the latter group were 52 4- 5% and 1.3 4- 1.3, respectively. In the 2 subgroups, both propranolol and pindolol produced an improvement in global and regional LV function during exercise to an extent similar to that observed in the entire group.

February 1983 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 51

REST

8O

EX

70

REST

EX

REST

EX

:

"

V- 6 0

431

0

rr LL

5 I-L) UJ --3 UJ

FIGURE 3. Effects of propranolol and pindolol on global LV function. Data represent individual and group (mean :1: standard error of the mean) values, at rest and during peak exercise (EX). Exercise ejection fraction after both beta-adrenoreceptor antagonist drugs were similar. * p <0.05; NS = p >0.05 versus rest values.

> J

50

\.

40

30 25-

CONTROL

Discussion This study shows that pindolol and propranolol, adrenergic beta-receptor antagonist agents with and without intrin.sic sympathomimetic activity, respectively, have similar effects on global and regional LV function in patients with angina pectoris. At the doses used, both drugs exerted an equal suppression of the exercise heart rate and had significant and similar antianginal effects. Differences in the pharmacodynamic and pharmacokinetic properties s of the many beta-adrenoreceptor antagonists available have led investigators to postulate the theoretical advantages of some over others when treating patients with angina pectoris. 9 11 One such ancillary property is the intrinsic sympathomimetic activity or partial agonist activity. Drugs with this characteristic cause a measurable agonistic response when they interact with beta-adrenergic receptors in the absence of a primary agonist such as adrenaline or isoprenaline; this can be demonstrated in the adrenalectomized animal treated beforehand with reserpine. 9 Although drugs with and without partial agonist activity

oL~

PINDOLOL

have been available for patient treatment for several years, the clinical significance of this property remains uncertain. 1°,12 It has been suggested that drugs with partial agonist activity induce less subjective and objective respiratory impairment in patients with bronchospastic disease and lessen exercise cardiac output to a reduced degree in patients with hypertension 13,14 than drugs without partial agonist activity. Thadani et aP ,2,15 and others 16-1s found that betaadrenoreceptor antagonist drugs with and without partial agonist activity are equally effective in the treatment of angina pectoris. However, no information is available regarding the role of the partial agonist activity on global and regional LV function. We therefore elected first to compare pindolol and propranolol because pindolol is the drug with the most pronounced partial agonist activity. 19 Second, because myocardial oxygen supply and demand in patients with chronic, stable angina pectoris are at equilibrium at rest and myocardial ischemia develops only during stress, we studied the effects of both drugs on ventricular function at rest as well as during exercise. Third, we used the

REST

uJ

PROPRANOLOL

EX

REST

6

EX

REST

eo

EX

.--

U3 Z

FIGURE 4. Effects of propranolol and pindolol on regional LV function. Data represent individual and group (mean :l: standard error of the mean) values, at rest and during Peak exercise (EX). LV exercise wall motion score imProved significantly (p <0.01) after both beta-adrenoreeptor antagonists, There was no significant difference P ~0.05) in exercise LV wall motion score between the 2 drugs. $ p <0.001 versus rest values.

~

o Co

4

d >,

2

•

i.: '.7 t

.. CONTROL

~> iiiiI ~"

.. -4..

PROPRANOLOL

.; .iii!.I[":~!i: P INDOLOL

•

432

LEFT VENTRICULAR FUNCTION AND BETA-ADRENERGIC BLOCKING AGENTS

electrocardiographic-nmltigated blood pool cardiac scintigraphy, a reliable, noninvasive technique wellsuited for repeated measurements at rest and during exercise. 3-5 On the basis of previous experience with the drugs tested, 1°,16,2° we used 2 dose levels. With use of only 2 dose levels, it is possible that the optimal dosage may not have been reached in a few patients; however, a significant beta-adrenoreceptor blocking effect was obtained at the doses used. Our results are in agreement with more comprehensive dose-response studies in which pindolol, 10 to 20 mg/day, and propranolol, 160 to 320 mg/day, were found to be well within the therapeutic range. 15,2~,22 Several aspects of our study need further comment. Open trials do not constitute worthwhile evidence of treatment efficacy. However, our principal aim was to assess the effects of pindolol and propranolol on objective measurements of LV function. The antianginal efficacy of the drugs used was not in question and a double-blind study was not vital. All radionuclide studies were analyzed blindly at the end of the trial. Although we did not have formal run-in and trial-proper periods, and we were not concerned about possible carry-over effects, we believe that the random distribution of the order of treatment and control sequences (Table I) and the existence of 2 dose levels reconciled these deficiencies in design. In those subjects who received 320 mg/day of propranolol or 20 mg of pindolol/day, an actual run-in period was present. Moreover, rest-exercise studies were carried out after a minimum of 4 weeks of treatment and the carry-over effect by then should have been nonexistent. Lastly, patients were carefully selected and all had uncomplicated, stable angina pectoris and an ischemic electrocardiographic response to exercise. This may impose limitations in applying the results to a broader clinical population, but is necessary for meaningful comparison of the drug effects. At the doses used, pindolol and propranolol were efficacious antianginal agents. Both drugs reduced clinical and electrocardiographic signs of myocardial ischemia, and improved exercise tolerance to a similar degree. No further comments are needed regarding these effects because many previous well-designed studies 16,23have shown similar results. LV f u n c t i o n changes at rest: In patients with normal or near normal LV function, previous studies have shown that oral administration of propranolol did not significantly influence resting global or segmental systolic LV function as assessed by radionuclide methods. 23-26 In patients with moderate to marked LV dysfunction, 26 or when studies were performed in the upright position, 27 a decreased LV ejection fraction was observed after propranolol administration. The results of the present study, which largely involved patients with normal or near normal LV function studied in the supine position, are in keeping with these observations. LV ejection fraction and wall motion patterns at rest were also minimal and were not significantly affected by oral pindolol at clinically effective doses. Moreover,

with a few exceptions, individual responses to propranolol were closely correlated with those to pindolol. The comparative angiographic response to propranolol and pindolol, or between beta-adrenergic receptor antagonists with and without partial agonist activity, has not previously been reported. Frishman et aP ° found that the LV ejection fraction improved significantly after pindolol but was significantly reduced compared with control values after propranolol. The differences between the results in that study 1° and ours are unlikely to reflect differences in study design because the clinical, electrocardiographic, and circulatory changes were similar to those in the present study. Two factors may account for the divergent results: (1) The ejection fraction in Frishman's study was derived from the M-mode echocardiogram. This method has recognized limitations, especially in subjects with coronary heart disease in whom localized areas of LV dysfunction may go undetected. The left ventricle was assumed to have the shape of a cube. 1° With the radionuclide method, counts in the entire ventricle are included to calculate ejection fraction, and no assumptions are made to the LV shape. Several investigations using radionuclide methods 23-26 have shown that resting ejection fraction is not significantly altered by oral administration propranolol at the doses used in Frishman's study. (2) We administered pindolol at approximately one third of the dose given by Frishman et al. 1° Future studies should investigate whether larger doses of pindolol may increase resting ejection fraction. LV f u n c t i o n changes during exercise: The beneficial effects of oral propranolol on LV function during exercise in patients with stable angina pectoris has previously been documented. 24-26 In our study, both propranolol and pindolol prevented the ejection fraction t¥0m decreasing significantly with exercise, and both drugs improved exercise wall motion score to the same extent. The mechanisms whereby beta-adrenoreceptor antagonists improve global and segmental systolic function during exercise in patients with coronary artery disease are not known. Possible mechanisms include a decrease in myocardial oxygen requirements, an improved total myocardial blood flow, or a favorable redistribution of myocardial blood flow. We conclude that this mechanism is independent of the presence or absence of intrinsic sympathomimetic activity. A major shortcoming of the present study was failure to measure LV volumes. Possible differences in the effects of propranolol and pindolol on ventricular volumes were not assessed. 2s,29 E f f e c t s of t r e a t m e n t in relation to the basal status of LV function: If partial agonist activity has any beneficial effect on LV function, this effect would be more important in patients with impaired or borderline LV performance. 9,1°,12 The patient group in this investigation included, with 1 exception, subjects with normal or near normal LV function. Thus, we cannot make any valid conclusions regarding the comparative effects of propranolol and pindolol on LV function in patients with poor LV performance.

February 1983 THE AMERICAN JOURNAL OF CARDIOLOGY

Acknowledgment: We acknowledge the secretarial help of Mary Collier and the technical assistance of the electrocardiogram technicians of our institution.

References 1. Thadani U, Davidson C, Singleton W, Taylor S. Comparison of the tmmediate effects of five beta-adrenoreceptor-blocking drugs with different ancillary properties in angina pectoris. N Engl J Med 1979;300"750-755. 2. Thadani U, Davidson C, Singleton W, Taylor S. Comparison of five betaadrenoreceptor antagonists with different ancillary properties during sustained twice daily therapy Jn angina peotorJs. Am J Med 1980;68:243250 3. Borer JS, Bacharach SL, Green MV, Kent KM, Epstein SE, Johnston GS. Real-time radionuclide cineangiography Jn the noninvasive evaluation of global and regional left ventricular function at rest and during exercise in patients wtth coronary artery disease N Engl J Med 1977;296:839-844. 4. Manyari DE, Nolewajka A, Purves P, Donner A, Kostuk WJ. Comparative value of the cold-pressor test and supine bicycle exercise to detect subjects w=th coronary artery dJsease using radionuclide ventriculography. Ctrculation 1982;65:571-579 5. Kostuk WJ, Chamberlain M, Sang H. Radionucllde angiocardiography, a noninvas=ve method for evaluating left ventricular ejection fraction and reg=onal wall motion: comparison wtth contrast left ventricular angiography. Can Med Assoc J 1978;119:877-883. 6. Aellig WH. Beta-adrenoceptor blocking activity and duration of action of pmdolol and propranolol tn healthy volunteers. Br J Clin Pharmaoo11976; 3:251-257 7. Robertson D, Kostuk WJ, Ahuja S. The localization of coronary artery stenosts by 12 lead ECG response to graded exercise test: support for intercoronary steal. Am Heart J 1976;91:437-444 8. Gibson DG. Pharmacodynamio properttes of beta-adrenergic receptor blockmg drugs in man. Drugs 1974;7:8-38. 9. Barrett AM, Nunn B. Intrinsic sympathomimetic activity in relation to the precipitation of heart failure by beta adrenoceptive blockade Arch Int Pharmacodyn Ther 1971; 189:168-179. 10. Frishman W, Kostis J, Strom J, Hossler M, Elkayam U, Goldner S, Silverman R, Davis R, Weinstein J, Sonnenblick E. Clinical pharmacology of the new beta-adrenergic blocking drugs. Part 6. A comparison of pindolol and propranolol in treatment of patients with angina pectoris. The role of intrinsic sympathomimetJc activity. Am Heart J 1979;98:526-535. 11. Parratl JR, Marshall RJ. Beta adrenoreceptor blocking drugs in acute experimental myocardial ischemia (abstr). Presented to the Eighth World Congress of Cardiology, Tokyo, September 1978 12. Frishman W, Silverman R, Strom J, Elkayam U, Sonnenblick E. Clinical pharmacology of the new beta-adrenergic blocking drugs. Part 4. Adverse effects. Choosing a beta-adrenoreceptor blocker. Am Heart J 1979;98: 256-262.

Volume 51

433

13. Beumer HM, Hardunk HJ. Effects of beta-adrenergic blocking drugs on ventilatory function in asthmabcs. Eur J Clin Pharmacol 1972;5:77-85. 14. Franciosa JA, Johnson SM, Tobian LJ. Exercise performance in mildly hypertensive patients. Impairment by propranolol but not oxprenoloL Chest 1980;78:291-299. 15. Thadani U, Sharma B, Meeran MK, Majid PA, Whilaker W, Taylor SH. Comparfson of adrenergic beta-receptor antagonists in angina pectoris. Br Med J 1973;1:138-142. 18. Boakes AJ, Prichard BNC. The effect of AH5158, pindolol, propranolol, and D-propranolol on acute exercise tolerance m angina pectoris. Br J Pharmacol 1973;47:673P-674P 17. Sowlon E, DasGupta DS, Baker I. Comparative effects of beta adrenergic blockmg drugs. Thorax 1975;30:9-18. 18. Jackson G, Atkinson L, Cram S. Double-blind comparison of timolot, propranolol, practolol, and placebo Jn the treatment of angina pectoris. Br Med J 1975;1:708-712. 19. Barrett AM, Carter J. Comparative chronotropic actiwty of beta-adrenocaptive antagonists. Br J Pharmacol 1970;40"373-480 20. Nair DV. A double blind trial of visken (LB-46), in the treatment of angina pectoris. Indtan Heart J 1972;2(Suppl I):183-191. 21. Aravanis C, Michaelides G. Results of the use of a new beta-adrenergic blocker (LB-46) in angina pectoris, Acta Cardtol 1970;25:501-509. 22. Alderman EL, Davies RO, Crowley JJ, Lopez MG, Brooker J, Friedman J, Graham A, Matlof H, Harrison DC. Dose response effectiveness of propranolol for the treatment of angina pectoris. Ctrculation 1975;51: 964-975. 23. Marshall RC, Berger HJ, Reduto LA, Cohen LS, Gottschalk A, Zaret BL. Assessment of cardiac performance with quantitative radionuclide anglocardiography. Effects of oral propranolol on global and regional left ventricular function an coronary artery disease. Circulation 1978;58:808814. 24. Battler A, Ross J Jr, Slutsky R, Pfisterer M, Ashburn W, Froelicher V. Improvement of exerctse-mduced left ventricular dysfunction with oral propranolol in patients with coronary heart disease. Am J Cardiol 1979; 44:318-324 25. Marshall RC, Wisenberg G, Schelberi HR, Henze E. Effect of oral propranolol on rest, exercise and postexerclse left ventricular performance in normal subjects and patients with coronary artery d~sease. Circulation 1981;63"572-583. 26. Rainwater J, Steele P, Kirch D, LeFree M, Jensen D, Vogel R. Effect of propranolol on myocardial perfusion images and exercise ejection fraction in men with coronary artery disease. Circulation 1982;65:77-81, 27. Port S, Cobb F, Jones RJ. Effects of propranolol on left ventncular function in normal men. Circulation 1980;61:358-366. 28. Slutsky R, Karllner J, Ricci D, Kaiser R, Pfisterer M, Gordon D, Peterson K, Ashburn W. Left ventricular volume by gated equihbrtum radionuchde angiography: a new method. Circulation 1979;60:558-564. 29. Moniz de Bettencourt J, Comes-Pinto B. Les modifications de la dynamique cardiaque dans I'angine de poitnne et Faction du propranolol et du pindolol chez le sujet normal etches le malade angineux. Ann CardJol Angeiol 1975;24:453-460