Trends and determinants of calcium antagonist usage after acute myocardial infarction (the GISSI experience)

Trends and Determinants of Calcium Antagonist Usage After Acute Myocardial Infarction (The GISSI Experience)* Giulio Zuanetti, MD, Roberto Latini, MD, Faust0 Avanzini, MD, Maria Grazia Franzosi, PhD, Aldo P. Maggioni, MD, Fabio Colombo, PhilD, Enrico Nicolis, BS, and Francesco Mauri, MD, on behalf of the GISSI Investigators* In the last decade, several clinical trials in patients with, or recovering from, acute myocardial infarction (AMI) have evaluated the role of calcium antagonists in affecting patients’ prognosis. Results have been disparate, with evidence of possible harm, no effect, or some benefit, depending on the agent used. We evaluated how the evidence from these trials has influenced the pattern of prescription of calcium antagonists and assessed the important determinants of use of these agents in patients after AMI. We analyzed retrospectively the prescription of calcium antagonists at discharge in all patients recovering from AMI enrolled in 3 large randomized clinical trials (Gruppo Italian0 per lo Studio della Soprawivenza nell’lnfarto1 (GISSI- 1 I, GISSI-2, and GISSI-3) during the last 10 years. A progressive decrease in prescriptions for calcium antagonists was evident, from 47.2% in GISSI-1

to 35.1% in GISSI-2 to 19.0% in GISSI-3 (p
n the last decade, several clinical trials in patients with, or recovering from, acute myocardial infarction (AMI) have been performed to evaluate the role of calcium antagonists in affecting patients’ prognosis. Nifedipine, ‘-’ diltiazem, 1s and verapamil’,’ have been the 3 drugs used in the large multicenter trials, with mixed results. Nifedipine tended to increase mortality lm3; diltiazem decreased the incidence of reinfarction, but not mortality, in non-Q-wave infarction’ and had a dichotomous effect on mortality depending on the presence of congestive heart failure.’ Finally, a positive effect of verapamil in patients without heart failure and a neutral effect in patients with heart failure were observed.6.7 In the present study, we evaluated how the growing evidence from these clinical trials has influenced the pattern of prescription of calcium antagonists and assessed the important determinants of use of these agents in patients after AMI.

METHODS

From the Department of Cardlovoscular Research, lstituto di Ricerche Formocologiche Mario Negri, Milan, Ital.,). GISSI studies are endorsed by Associozione Nazionale Medlci Cardiologi Ospedalieri ond Mario Negri Institute. This study has been supported in port by Knoll Formaceutici Italy. G.Z. is a recipient of o Mourelio Caniato senior fellowship. Manuscript received September 20, 1995; revised manuscript received ond accepted January 2, 1996. Address for reprints: Giulio Zuonetti, MD, Department of Cardiovascular Research, lstituto di Ricerche Farmacologiche hlario Negri, via Eritl-eo, 62 201 57 Milan, Italy. *A list of all investigators involved in the GISSI studies con be found in the appendix of references 8, 9, and 10.

0 1996 by Excerpta Medico, All rights reserved.

Inc.

We retrospectively analyzed prescriptions for calcium antagonists at discharge in all patients recovering from AMI enrolled in 3 large randomized clinical trials (Gruppo Italian0 per lo Studio della Sopravvivenza nell’Infarto- 1 [ GISSI- 11, GISSI-2, and GISSI-3) during the last 10 years.Ym’OAbout 75% of all coronary care units in Italy were involved in the GISSI studies, the main characteristics of which are summarized in Table I. To be included in the present analysis, patients had to have survived the acute phase, with the diagnosis of AM1 confirmed by enzyme elevation or electrocardiographic changes indicative of myocardial necrosis. Thus, all patients with suspected, but not confirmed, AM1 were excluded from the analysis. We analyzed the overall prescription of calcium antagonists and of the most commonly prescribed agents-nifedipine, diltiazem, and verapamil. No data were available on dosage. Definition of variables included in the analysis: On the basis of pharmacodynamics of calcium antagonists, their clinical indications, and characteristics of patient populations enrolled in the clinical trials published in the literature, we tested the following variables for their association with usage of calcium antagonists: ( 1) patient age ( 5 vs > 70 years), (2) sex, (3 ) history of hypertension, (4) history of angina, (5) previous AMI, (6) signs of congestive heart failure at randomization (Killip class 2 2), (7) type of AM1 (Q-wave vs non-Q-wave AM1 and anterior AM1 vs other sites), (8 ) occurrence of angina 0002.9149/96/$15.00 PII 50002.9 149(96)00249-4

153

TABLE

I Main

Characteristics

of GISSI

Studies GISSI-2

GISSI-1 2/84-6/85

2/88-7/89

Study

Recruitment design

period

open

open

Study

treatment

Streptokinase

vs

open

criteria

(interval

from

symptoms

to

cl2

vs

Lisinopril

vs control

alteplase Heporin

Inclusion

6/91-7/93

Streptokinase

control

GISSI-3

56

h

vs control

Nitrates

h

524

vs control h

randomization) Exclusion

Contraindications

criteria

to

Contraindications thrombolysis

thrombolysis Randomized Patients

potients discharged

acute

alive

myocardial

GISSI = Gruppo

with

confirmed

to

Killip

class

pressure

11,806

12,490

19,394

9,452

10,407

16,958

IV; systolic < 100

blood

mm Hg

infarction lbliono

per lo Studio

della

Suprawivenza

nell’lnfarto.

This usage was higher in the GISSI-3 study, which was initiated after the introduction of several new agents, such as nisoldipine, nitrendipine, felodipine, amlodipine, and lacidipine. Role of concomitant diseases: Table II summarizes the prescription of calcium antagonists in the presence of preexistent or concomitant diseases. As expected, a history of hypertension or angina after AM1 was associated with a higher usage of calcium antagonists. This is indicated in the multivariate analysis performed for the GISSI-2 and GISSI-3 data base (Table III), in which the most powerful predictors of calcium antagonist usage for both studies were angina after AM1 and a history of hypertension. However. even in these patients, the trend toward decreasing usage is evident (p 90% (OR 0.075 and 0.09 1, re1. In all 3 studies, a varying percentage of patients spectively) . Thus, independent of other variables, was treated with other calcium antagonists (3.7% in the decision to use a p blocker was a major determinant of the choice of calcium antagonist. GISSI-1, 1.9% in GISSI-2, and 5.3% in GISSI-3).

after AMI, (9) reinfarction, ( 10) presence of clinical or instrumental left ventricular dysfunction at discharge, ( 11) presence of ventricular arrhythmias, and (12) use of p blockers, digitalis, and antiarrhythmic agents at discharge. This allowed us to investigate the trends in use of calcium antagonists in a selected population of patients. All the variables were available for the more recent GISSI-2 and GISSI-3 studies, whereas variables (3)) (4)) (5), (9)) ( lo), and ( 11) were not collected in the GISSI1 data base. To analyze the use of calcium antagonists for secondary prevention after AM1 (i.e., in the absence of specific indications for their use), patients given calcium antagonists were analyzed after excluding those with a history of hypertension or angina, those who had reinfarction, and those who developed angina or had a positive exercise stress test for ischemia after infarction. Statistical analysis: Differences between trials were assessed with chi-square analysis for trend. Within each GISSI study, the influence of the clinical variables (as defined earlier) on the prescription of calcium antagonists was assessedusing univariate analysis; then, all the variables were tested for their independent contribution to the use of overall and individual calcium antagonists by multivariate logistic analysis (SAS statistical package; SAS Institute, Cary, North Carolina).

154

THE

AMERICAN

JOURNAL

OF

CARDIOLOGY”

VOL.

78

JULY

15,

1996

TABLE

II

Percentage

of Patients

Receiving

Calcium

Antagonists

at Discharge

(n =

16,958)

18/22.9 17.6/23.1

34.9/35.9

AAh/

Yes/no

systemic

(n =

19.0%

34.4/38.1

47.1/47.7 47.2/47.2

c70/>70

Age (yrl Patients with

10,407) 35.1%

47.2%

Overall MIX&Omen

GISSI-3

GISSI-2

GISSI-1 (n = 9,452)

1.3

26.1/15.6

hypertension* Patients

with

history

Patients

with

previous

Patients

with

Killip

Site

of angina*

Yes/no

43.7/32.9

28.2/17.1

AMI*

Yes/no l/>l

42.9/33.8

26.3/l

47.6/46

Anterior/other

48.9/46.3

class

at entry

of AMI

Type

of AMI

*

Patients

with

angina

after

Patients

with

in-hospital

Yes/no

AMI*

15.7/20.3

32.7/36.3

34.4/38.7

47/53.7 67/44.4

Q/tlOn-Q

66.5/3

7.9

18.8/20.3.

35/35.2

17.7/23.9 1 .A

48/l

Yes/no

53.6/34.7

36.1/18.7

Yes/no

29.9/36.1

18.1/19.2

Yes/no

37.3/34

19.9/l

4.7

reinfarction* Left ventricular

dysfunction

at

discharge Ventricular

premature

(> 1 O/h)

during

fl blockers Digitalis

at discharge* at discharge*

Antiarrhythmics l

at discharge

Significant

AMI

complexes

(p ~0.01)

d‘ffI erences

myocordiol

infarction;

= acute

TABLE Ill

Determinants

Multivariate

between

patient

Yes/no

36.4/48.2

18.1/40.7

13.1/21.7

Yes/no

52.8/46.2

25.4/36.1

16.9/l

Yes/no

66.8/44.5

29.6/35.7

groups

other abbreviation

of Calcium

for each GISSI study for which

CIS in Table

Antagonist

Prescriptions

after

AMI

History

of hypertension

Reinfarction of angina AMI

Non-Q-AM1 Anterior AMI Age >70 y Left ventricular

dysfunction

OR

95%

A.73

4.06-5.51

Cl

1.94

1.75-2.15

1.66

1.19-2.32 1.22-l .52

1.87

1.36 1.34

1.18-1.53

1.45

OR 5.66 1.97 1 .A9

1.27

1 .12-l

1.06

0.96-l.

.45 17

1.24

.oo

0.88 1.09

0.89

0.79-l

0.75

0.65-0.87

0.74

0.70

0.60-0.82

0.72

0.52

0.43-0.62

0.66

0.23

0.20-0.26

at discharge Antiarrhythmics Digitalis fl blockers Treatment

at discharge

at discharge at discharge with

Cl = confidence

0.39

0.47

lisinopril interval;

OR = odds ratio;

other abbreviation

tls in Table

I.

In the GISSI-3 study, randomization to nitrates did not affect the use of calcium antagonists, whereas randomization to lisinopril significantly influenced it ( 14.7% vs 23.4%; p
antagonists

in secondary

prevention:

This subanalysis is of particular interest because all CORONARY

ARTERY

was available.

the major clinical trials have addressed the use of calcium antagonists in the secondary prevention of GISSI-3 morbidity and mortality after AMI. Reliable data could be obtained 95% Cl from the GISSI-2 and GISSI-3 da5.05-6.33 tabase because of the more detailed 1.79-2.16 clinical record form. Patients with1.39-2.5 1 1.31-1.69 out specific indications for calcium 1.30-l .62 antagonists represented 45.5% of 1.14-1.34 the total patients in GISSI-2 and 0.78-0.98 46.5% in GISSI-3 (p = NS). Data 0.98-l .22 0.64-0.86 are summarized in Table IV. Total prescriptions of calcium antagonists 0.59-0.88 decreased by approximately 60% 0.53-0.82 (from 26.1% to 10.3%), with simi0.35-0.44 lar reductions for all 3 main drugs. 0.43-0.52 - In this population, the increasing use of newer agents was particularly ev ident, so that they accounted for 30% of the total prescriptions in GISSI-3.

at Discharge:

GISSI-2

Angina

information

9.1

18/19.1

I.

Analysis

History Previous

8.8

Halter

DISEASE/CALCIUM

DISCUSSION The results of this study indicate that, during the last 10 years, the prescription of calcium antagonists in patients after AM1 has declined progressively, both overall and individually for all 3 major active drugs. Before discussing how the growing information available from clinical trials has affected prescription data, we will emphasize some limitations and advantages of the study. First, information on drug prescription at discharge was derived from the clinical record forms sent to the coordinating center from the individual cecters as part of the data collection of patients; analysis on this specific information was not predefined. Second, there are some differences in the study populations that may influANTAGONISTS

AFTER MYOCARDIAL

INFARCTION

155

trial (DAVIT) ,6 and DAVIT-II.’ Use of calcium antagonists in our population dropped from 26.1% in GISSI-3 GISSI-2 GISSI-2 to 10.3% of patients in (n = 4,739) (n = 7,887) GISSI-3. Thus, currently about 1 paOverall 26.1% 10.3% tient of 10 without any specific Men/women 9.9/l 2.4 26.2/25.4 indication for the use of calcium an<70/>70 9.7/l 2.6 26.1/25.8 Age (yrl tagonists is discharged with such Yes/no 14.3/l 0.0 29.0/25.7 Previous AMI drugs. This decreasing trend was as l/>l Killip class at entry 10.1/12.7 26.1/26.1 evident for verapamil as for nifediAnterior/other Site of AMI 24.8/26.9 8.2/l 1.3 Q/non-Q 9.4/l 4.3 Type of AMI 25.4/30.1 pine and diltiazem, despite the neuYes/no 9.6/l 0.5 Left ventricular dysfunction at 2 1.4/27.0 tral or even positive results obtained discharge with verapamil in trials of secondary Ventricular premature complexes Yes/no 27.0/24.0 1 1 S/9.2 prevention.7 This suggests an over(> 1 O/h) during Halter Yes/no 5.1 /12.5 fl blockers at discharge 9.8/3 1.4 whelming class effect on the attitude Yes/no 11.3/10.3 Digitalis at discharge 18.0/26.9 of clinicians toward this particular Yes/no 10.3/l 0.4 Antiarrhythmics at discharge 22.6/26.5 agent. Abbreviations CK in Tables I to III. In the GISSI-3 study, there was a more marked tendency toward using newer agents in secondary preence in part the results of the data analysis. For ex- vention after AMI, perhaps on the assumption based ample, patients with Killip class IV were included in on data from smaller studies that the newer agents the GISSI-2 study but excluded from the GISSI-3 do not adversely affect, or even improve, left venstudy (Table I), and randomization to lisinopril in tricular function.” These data are relevant in light of the GISSI-3 study affected the overall use of calcium the recent debate regarding the safety of calcium anantagonists. These differences are relatively small, tagonists, and, more specifically, of short-acting nihowever, and it seems improbable that they affected fedipine. in patients with ischemic heart disease.12,” It would have been misleading to correlate the use the data significantly because, for example, the decrease in calcium antagonist usage was evident and of nonrandomized treatments, such as calcium antagonists in the GISSI-3 study, to outcome, because significant in all groups. On the other hand, the representativeness of cen- the results of these post hoc analyses may suffer from ters involved in the three studies, the large number significant biases, and it was decided a priori not to of patients enrolled, and the number and quality of analyze outcome in patients according to nonranvariables recorded allowed a thorough interpretation domized treatment. However, the data on prescripof the data not possible in previous studies on the tions suggest that the newer agents are now preferred among dihydropyridines. same topic. Comparison with data from other countries: Data on Calcium antagonists in patients after AMI with hyperthe trends of usage of calcium antagonists in large tension, angina, and congestive heart failure: A history of hypertension and angina played an important role populations of patients after AMI are scanty. Rein the use of calcium antagonists, but the occurrence cently, Pashos et al I4 documented a decrease in the of angina after AM1 was the most influential factor usage of calcium antagonists in the U.S. (from 63% to 47%) concurrent with the increase in the usage of favoring the prescription of a calcium antagonist. Nifedipine and newer dihydropyridine agents 0 blockers (from 29% to 38%); data from GUSTOtended to be favored in the presence of hypertension, 1 investigators indicated current use of calcium anwhereas diltiazem was clearly the drug of choice in tagonists in 24% to 34% of patients at discharge.” the presence of ischemia after AMI. Signs of left In Spain, Agusti et alI6 reported data on 737 patients ventricular dysfunction decreased the use of calcium admitted to a single center; they demonstrated a deantagonists overall but did not play a major role in crease in calcium antagonists and an increase in the the choice among them, despite the growing evi- use of p blockers. The lack of data on the clinical dence of the detrimental effect of nifedipine and dil- characteristics of the patients in these studies does tiazem’.3’5 in patients with impaired left ventricular not allow a direct comparison with our study; however, these data suggest an overall trend of decreasfunction. ing use of calcium antagonists in western countries. Calcium antagonists in secondary prevention after Clinical implications: The overall analysis of the AMI: One of the most interesting results of this study was the analysis of prescriptions of calcium antag- data in our study indicates that the usage of calcium onists in the absence of an established indication for antagonists in patients after AMI has been strongly their use, such as hypertension or angina after AMI. affected by the results of other large multicenter triIndeed, this population is most similar to that en- als, mostly performed in the 1980s. In general, calrolled in clinical trials such as Trial of Early Nife- cium antagonists are now prescribed only if there is dipine Treatment, ’ Secondary Prevention Reinfarc- a specific indication for their use. Thus, the main tion Israeli Nifedipine trial, 3 Multicenter Diltiazem lesson learned from clinical trials has been transPost-Infarction trial,’ Danish Verapamil Infarction ferred rapidly into clinical practice. However, a sigTABLE IV Secondary

156

THE

Prescription

of

Prevention

After

AMERICAN

Calcium

Antagonists

at Discharge

for

AMI

JOURNAL

OF

CARDIOLOGY@

VOL.

78

JULY

15,

1996

suits of a double-blind, randomized, multicenter trial. N Engl J Med 1986;315:423-429. 5. The Multicenter Diltiazem Postinfarction Trial Research Group. The effect of diltiazem on mortality and reinfarction after myocardial infarction. N Engl J Med 1988;319:385-392. 6. The Danish Study Group on Verapamil in Myocardial Infarction. Effect of verapamil on mortality and major events after acute myocardial infarction (the Danish Verapamil Infarction Trial II-DAVIT II). A~I J Cwdiol 1990;66:779-

nificant fraction of patients with impaired left ventricular function still receive a calcium antagonist such as nifedipine and diltiazem, the use of which has been demonstrated to be detrimental.‘-3,5 On the other hand, the use of verapamil, the only calcium antagonist demonstrated to be neutral or beneficial in patients after AMI, has dropped to trivial figures. Finally, the trend toward use of newer dihydropyridines in patients after AMI, especially in those with signs of left ventricular dysfunction, needs to be reassessedfollowing the results of recent clinical trials in patients with heart failure. Acknowledgment: We thank Luigi Tavazzi, MD, for helpful criticism and Luisa Galbiati for secretarial assistance.

Lancet 1990;336:65-71. 10. Gruppo Italiano

1. Muller E. Morrison J, Stone PH, Rude RE, Rower B. Roberts R, Peale DL. Tori ZG, Schneider JF, Serfas DH. Tate C, Scheiner E. Sobel BE, Hennekens CH. Braunwald E. Nifedipine therapy for patients with threatened and acute myocardial infarction: a randomized, double-blind, placebo-controlled cornparison. Circulation 1984:69:740-747. 2. Wilcox RG. Hampton JR. Banks DC, B&head JS, Brooksby IAB. BurnsCox CJ. Hayes MJ, Joy MD. Malcolm AD. Mather HG, Rowley JM. Trial of early nifedipine in acute myocardial infarction: the Trent study. BbfJ 1986;293:1204-1208. 3. Goldbourt U, Behar S, Reichsr-Reiss H, Zion hl. Mandelzweig L, Kaplinsky E: for the SPRINT Study Group. Early administration of nifedipine in suspected acute myocardial infarction. The Secondary Prevention Reinfection Israel Nifedipine Trial 2 Study. Arch Inrem Med 1993;153:345-353. 4. Gibson RS. Bode” WE Theroux P, Strauss HD, Pratt CM, Gheorghiade M, Capone RJ, Crawford MH, Schlant RC, Kleiger RE, Young PM. Schechtman K, Perryman B, Roberts R, and the Diltiazem Reinfarction Study Group. Diltiazem and reinfarction in patients with non-Q-wave myocardial infarction. Re-

1994;46:95-99.

CORONARY

785. 7. Yusuf

S. Held P, Furberg C. Update of effects of calcium antagonists in myocardial infarction or angina in light of the Second Danish Verapamil Infarction Trial (DAVITII) and other recent studies. Am J Cnrdiol 1991;67:1295-1297. 8. Gruppo Italiano per lo Studio della Streptochinasi nell’lnfarto Miocardico (GISSI). Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. Lancer 1986:1:397-402. 9. Gruppo It&no per lo Studio della Sopravrivenza nell’Infarto Miocardico. GISSI-2: a factorial randomised trial of alteplase versus streptokinase and heparin versus no heparin among 12490 patients with acute myocardial infarction. per lo Studio della Sopravvivenza nell’Infarto Miocardico. GISSI-3: effects of lisinopril and uansdermal glycql trinitrate singly and together on 6.week mortality and ventricular function after acute myocardial infarction. Lnncet 199%333:1115- 1122. 11. Cohn JN. Vasodilators in heart failure. Conclusions from V-HeFf II and rationale for V-H&T III. Drugs 1994;47(Suppl4):47-58. 12. Furberg CD, Psaty BM, Meyer JV. Nifedipine. Dose-related increase in mortality in patients with coronary heart disease. Circulan’on 1995;92: 13261331. 13. Opie LH, Messerli FH. Nifedipine and mortality. Grave defects in the dossier. Circulation 1995;92:1068- 1073. 14. Pashos CL, Normand SLT, Garfinkle JB. Newhouse JP, Epstein AM, McNeil BJ. Trends in the use of drug therapies in patients with acute myocardial infarction: 1988 to 1992. JAm Co11 Cardiol 1994;23: 1023- 1030. 15. Pilote L, Califf RM. Sapp S, Miller DP, Mark DB. Weaver WD, Gore JM, Armstrong PW, Ohman EM, Top01 EJ, for the GUSTO-l Investigators. Regional variation across the United States in the management of acute myocardial infarction. N Engl J Med 1995:333:565-572. 16. Agusti A, Amau JM, Laporte J-R. Clinical trials versus clinical practice in the secondary prevention of myocardial infarction. Eur J Clin Phnmacol

ARTERY DISEASE,.‘CALCIUM ANTAGONISTS

AFTER MYOCARDIAL

INFARCTION

157