Nifedipine and Mortality

Abstract

HomeCirculationVol. 92, No. 5Nifedipine and Mortality Free AccessResearch ArticleDownload EPUBAboutView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticleDownload EPUBNifedipine and Mortality Grave Defects in the Dossier Lionel H. Opie and Franz H. Messerli Lionel H. OpieLionel H. Opie From the Medical Research Council, University of Cape Town, South Africa (L.H.O.), and the Ochsner Clinic, New Orleans, La (F.H.M.). Search for more papers by this author and Franz H. MesserliFranz H. Messerli From the Medical Research Council, University of Cape Town, South Africa (L.H.O.), and the Ochsner Clinic, New Orleans, La (F.H.M.). Search for more papers by this author Originally published1 Sep 1995https://doi.org/10.1161/01.CIR.92.5.1068Circulation. 1995;92:1068–1073When an approved treatment is considered for an unapproved indication, the physician must evaluate the safety of the medication, its value in related conditions, and the individual patient. What is asked is that he make a prudent decision based upon full knowledge of the available evidence.”1Furberg and his colleagues2 entitle their article “Nifedipine: Dose-Related Increase in Mortality in Patients With Coronary Heart Disease.” They provide a valuable service by making us think more carefully about the benefit/risk ratio of the calcium antagonists as a group. However, by touting calcium antagonists as being unsafe and possibly lethal, they also create a great deal of uncertainty and anxiety among physicians and patients. We feel that they overstate their case. The following is an attempt to show that (1) the allegations focus on short-acting nifedipine, so that even if correct, they cannot be applied to other calcium antagonists; (2) the accusation of a dose-related increase in mortality rests on a biased selection of data chosen for the meta-analysis; and (3) several of the mechanisms suggested for the proposed adverse effects are unlikely. Finally, we shall argue that pharmaceutical companies have been remiss in not obtaining outcome studies for short-acting nifedipine and for other calcium antagonists, especially in the case of ischemic heart disease. The Indicted Agent Is Short-Acting Nifedipine, and Evidence for Other Calcium Antagonists Is Scant Several aspects of the Furberg paper suggest that calcium antagonists in general are accused; for example, the statement that “the literature reviews strongly suggest that the problem may go beyond short-acting nifedipine. Troubling and major adverse cardiovascular events have been linked to other calcium antagonists, primarily dihydropyridines.” It should be clearly and unequivocally stated that Furberg et al2 provide no evidence whatsoever for any adverse effect on mortality or morbidity for any other agent, with one important exception. That exception is nimodipine, approved for use in subarachnoid hemorrhage. When nimodipine was given for potential cerebroprotective properties to patients undergoing major cardiovascular surgery, mortality increased.3 Hints that all calcium antagonists are toxic in some way need to be firmly set aside and the real evidence examined. The Meta-Analysis The major argument used by Furberg et al is that their meta-analysis shows a dose-related increase in mortality. In reality, their confidence limits for doses of 30 to 60 mg/d are not significant. For example, at 50 mg/d, the confidence limits are 0.44 to 2.37 and at 60 mg/d, 0.93 to 1.50. Thus, their case must rest on the data with a dose of ≥80 mg/d. For reasons that are not clear, Furberg et al took only short-term, 2-week mortality data for the two studies of Muller et al,45 with mortality rates of 7 of 93 for nifedipine versus 2 of 88 control subjects in the threatened infarct study4 and 4 of 68 versus 6 of 68 in the unstable angina study.5 Reanalysis using 6-month data allows the proposed statistical differences to disappear, with a value for the overall comparison of P=.13 (Table). In a recent report, Yusuf et al6 suggest that because of statistical multiplicity, the “results from overviews should only be accepted if the levels of statistical significance are extreme (eg, P<.001).” The present analysis by Furberg et al falls far short of their own ambitious goal by a factor of at least 10-fold; the best value that we could calculate from their listed data was P=.01. Regarding the study by Muller et al,4 the authors indicate a number of important reservations, including the unexpected absence of any mortality at all in the placebo patients in the first 2 weeks versus the expected range of 4% to 12%. Logically, an agent given in the early stages of acute myocardial infarction (AMI), an event found in 75% of the patients of Muller et al,4 especially during the developing phase, may influence the extent of infarct size, subsequent remodeling, left ventricular function, and long-term outlook. For example, in the ISIS-I study,7 even 1 week of early treatment with a β-blocker, atenolol, reduced mortality at 1 year. In the study by Muller et al,4 nifedipine altered neither infarct size nor the 6-month follow-up, observations that are internally consistent and logical. In contrast, the 2-week data are atypical, and this short-term follow-up should at the very least be considered together with the longer-term data (Table).The major case against nifedipine in the group given 80 mg/d hinges on the INTACT study, carried out by Lichtlen et al.8 It is, in fact, misleading to include INTACT in the same group as other studies that deal with either early-phase AMI or unstable angina. In the INTACT study, only one third of the patients had prior AMI, diagnosed by ECG abnormalities, and two thirds had mild effort angina. To be included, these patients had to have had coronary angiography that showed a mild degree of coronary stenosis, and only 12% of patients had total coronary occlusions. These patients did not, as a group, have myocardial infarction or unstable angina, as claimed in the legend to the Table in the article by Furberg et al.2 The ideal meta-analysis “should be restricted to trials done in populations with similar risks.”9Thus, the crucial data in the case that Furberg et al mount against short-acting nifedipine rest on (1) whether a long- or short-term follow-up is taken as appropriate for the patients studied by Muller et al4 ; (2) whether including the patients from the INTACT study in the same grouping as those with AMI or unstable angina is justified; and (3) whether the probability value in their analysis (P=.046 by our calculations; P=.01 from the listed data of Furberg et al2 ) can be matched with their own requirement in another publication for a value of P<.001. This inherent statistical contradiction attests to the shakiness of their meta-analysis. It is hardly reassuring that, on the basis of such a meta-analysis, Furberg et al call for regulatory authorities to consider whether moderate to high doses of nifedipine capsules should be excluded from approved labeling. This is especially so when we consider that the increase in mortality in the INTACT study was found when the drug was used for an experimental nonapproved indication, namely, retardation of the progress of coronary atherosclerosis. Mechanism of Alleged Adverse Effects of Short-Acting Nifedipine If short-acting nifedipine were to have any adverse effects (which is not yet established), how could they be manifested? Hypotensive Effect A fall in blood pressure is a major therapeutic aim in the treatment of hypertension. Nonetheless, the benefit can turn to harm in the presence of coronary artery disease. The danger of excess nifedipine-induced hypotension is that it increases myocardial ischemic damage and infarct size, as shown experimentally by Selwyn et al.10 Nifedipine, as given in several of the trials shown in the Table of the article by Furberg et al,2 appears to have been used without adequate cognizance of the fact that it is a powerful hypotensive agent. In the Israeli SPRINT Study11 and SPRINT-II,12 the effects of nifedipine on the blood pressure are not even reported, apart from hypotension as an exclusion criterion. SPRINT-II12 found that most of the adverse effects were in patients with initial blood pressures of ≤100 mm Hg and occurred within the first week. The adverse effects of excess early hypotension in early AMI are not limited to calcium antagonists, as can be seen from the CONSENSUS study,13 in which intravenous enalaprilat caused hypotension and cessation of the trial. Nonetheless, hypotension cannot explain the deaths in the nifedipine arm of the INTACT trial (P.R. Lichtlen, personal communication). Muller et al45 note the increased risk of cardiogenic shock, a likely side effect of excess hypotension. The dangers of acute hypotension seem more marked in the context of the potential hemodynamic instability of AMI, rather than in the presence of left ventricular hypertrophy caused by hypertension.14 Hypotension, a decreased coronary perfusion pressure, and reflex cardiac acceleration are logical explanations for the proischemic effects of calcium antagonists.15 These mechanisms may explain why the Muller et al4 data show that short-acting nifedipine may be associated with an excess early mortality in patients with AMI, which is clearly not an approved indication for nifedipine. Nonhypotensive Proischemic Effects That short-acting nifedipine can cause angina is well established; hence, careful clinicians have titrated the dose upward, as also advised in the package insert: “The dose needed to suppress angina and that can be tolerated by the patient must be established by titration. Excessive doses can result in hypotension.” In addition to the role of hypotension, two other factors need to be considered, namely, neurohumoral activation (see below) and coronary steal.16 The existence of these mechanisms means that even in patients in whom hypotension is avoided, proischemic effects could still occur, which reinforces the need for dose titration when short-acting nifedipine is introduced. Repetitive Neurohumoral Stimulation Acute vasodilation induced by short-acting nifedipine can elicit activation of both the autonomic nervous system and the renin-angiotensin cascade.17 As was reported for another short-acting dihydropyridine, felodipine, such reflex activation is thought to limit regression of left ventricular hypertrophy despite blood pressure control in hypertensive patients.18 The detrimental experiences with short-acting nifedipine used for patients with congestive heart failure19 could likewise be explained by adverse neurohumoral stimulation, a likely response to acute vasodilation. Baroreflex desensitization may occur when short-acting nifedipine tablets are given over the long term, with a decrease of the adrenergic-mediated reflexes.20 This hypothesis would explain why there may be a cluster of adverse effects of nifedipine within the early period of initiation of therapy, such as the proischemic effects and risk of increased danger in the very early postinfarct period.5Of interest are the recent preliminary data of Packer,21 not yet fully reported, in which another dihydropyridine calcium antagonist, amlodipine, was given to patients with cardiomyopathy already treated by diuretics, digoxin, and angiotensin-converting enzyme (ACE) inhibition. Amlodipine had neither beneficial nor harmful effects on mortality in patients with ischemic cardiomyopathy, yet it reduced mortality in those with dilated cardiomyopathy. Hypothetically, the use of the ACE inhibitor lessened neurohumoral activation. Also, the long half-life and constant blood levels of amlodipine might have contributed to the results. Postulated Proarrhythmic Effects The proarrhythmic mechanism proposed by Furberg et al2 seems extremely unlikely in view of the extensive experimental work showing that calcium antagonists have an antiarrhythmic effect on both ischemic and reperfusion arrhythmias.22 None of the studies quoted by Furberg et al relate to nifedipine, and none provide convincing evidence for a proarrhythmic effect. In fact, the Finnish study shows a clear antiarrhythmic effect, with ventricular fibrillation reduced to 1 of 75 in nimodipine-treated patients versus 12 of 80 in control subjects (P<.01).23 Furthermore, ventricular fibrillation recurred in 1 of the nimodipine and 12 of the placebo patients, a clear benefit for nimodipine (P=.006). The second trial concerns lidoflazine, a drug no longer used and not listed among calcium antagonists by the International Union of Pharmacologists.24Negative Inotropic Effect This property of itself is no great disadvantage, rather being part of the basis of the beneficial effect of β-blockers in ischemic heart disease. Furthermore, as suggested years ago by Katz,25 a negative inotropic effect may paradoxically explain the benefit of β-blockers when given long-term in congestive heart failure. Experimentally, the dihydropyridines have various effects on the inotropic state, with nifedipine and amlodipine having negative effects yet felodipine having a small stimulation.26 Limited evidence suggests that nondihydropyridines such as verapamil and diltiazem have a therapeutic range more closely resembling that of β-blockers and that their greater negative inotropic effect is part of the mechanism of their antianginal properties.27 Although nifedipine has a greater direct negative inotropic action than verapamil or diltiazem on the human ventricular myocardium,28 in clinical practice the reflex adrenergic activation induced by short-acting nifedipine often causes an increase in contractile activity.29Prohemorrhagic Effects Furberg et al correctly state that the calcium antagonists have various degrees of antiplatelet effects. So do many other agents, including aspirin and nitrates. The latter agents inhibit platelets and vasodilate, as do the calcium antagonists as a group. Logically, an antiplatelet effect of verapamil30 could be the basis of prevention of reinfarction in the DAVIT-II study.31 This antiplatelet effect is a potential double-edged sword and could explain the data of Becker et al32 showing that in certain circumstances diltiazem could promote bleeding (as could aspirin in other studies). This aspect of the article by Furberg et al is very valuable and alerts us to a possible new side effect and a new mode of action of the calcium antagonists. Nonetheless, it is difficult to be precise about prohemorrhagic effects in any of the clinical trials commented on by Furberg et al except those concerning nimodipine. The use of this drug in the perioperative period is now clearly contraindicated.3Two other studies require specific comment. The proposal that the calcium antagonist caused more cerebrovascular events than did an ACE inhibitor in the treatment of hypertension was based not even on an abstract but rather on a brief meeting report.33 In reality, as the full report on the GLANT study, now in press, shows, the incidences of cerebrovascular events in the calcium antagonist and ACE inhibitor groups (11 of 980 and 5 of 956, respectively) were not significantly different.34The observation that thrombolytic reperfusion was associated with increased hemorrhage in patients treated with calcium antagonists35 must be tempered by caution. One obvious explanation is that all of the eight patients involved had hypertension and therefore were candidates for cerebral bleeding. Also, the data given in that article do not exclude the possibility that the small number of patients given calcium antagonists who bled might all have been receiving the higher dose of recombinant tissue-type plasminogen activator that was associated with increased bleeding. Of the eight patients involved, only three were taking nifedipine. Therefore, the data remain soft and the interpretation reserved. More information on the comparative effects of various calcium antagonists on platelets and on the coagulation process is required to balance the potentially harmful prohemorrhagic effect versus the potentially beneficial antiplatelet effect. In the case of diltiazem, this question will be settled by a prospective study currently under way.36Does the Threat of Increased Mortality Apply to Calcium Antagonists Other Than Short-Acting Nifedipine? At present, there are no indications that these data, the proposed adverse effects of short-acting nifedipine on mortality, even if accepted, could be extrapolated to those calcium antagonists with a slow onset of action, such as verapamil, diltiazem, and amlodipine, or to any of the truly long-acting preparations of nifedipine and other dihydropyridines. An example of the importance of slow onset and long action is the case of nisoldipine. In the short-acting form, it was not antianginal,37 in contrast with positive results when the long-acting core-coat form was used.38 The most likely explanation is that the slow-release form induces less neurohumoral activation and hence fewer side effects.39Particular attention should be paid to the benefits of verapamil in postinfarct patients. Verapamil is an agent known not to cause neurohumoral activation but rather to decrease circulating norepinephrine30 and cardiac stores of norepinephrine.40 Regarding the postinfarct studies with verapamil, there have been two carefully designed series. In the first, DAVIT-I,41 verapamil was started soon after the onset of symptoms of AMI, within 48 hours. A late benefit in mortality was offset by earlier harm (reminiscent of the timing of the adverse effects of nifedipine in early AMI patients). In the second, DAVIT-II,42 therapy was initiated later and patients with overt congestive heart failure requiring >160 mg/d furosemide were excluded. There were several benefits in DAVIT-II, in particular, lessened reinfarction and angina and decreased development of heart failure. Thus, there is provisional evidence that the nondihydropyridine agent verapamil, which reduces circulating norepinephrine levels,30 has a beneficial effect on patients with prior myocardial infarction, although the data are still weak and require further corroboration,6 ideally by more prospective studies. Because DAVIT-II was so specifically different in its entry criteria from the other trials concerning verapamil, including DAVIT-I, it may be questionable to combine DAVIT-I with DAVIT-II, as done by Fischer Hansen et al,31 or to combine DAVIT-II with other verapamil trials, as done by Held and Yusuf.43 On the other hand, it seems more acceptable to combine the results of diltiazem and verapamil used in the follow-up of non–Q-wave infarcts with a 25% to 35% reduction of risk of cardiac events.44 Thus, it seems prudent to consider that not all calcium antagonists are created equal.45What Is a Prudent Clinician Expected to Do? Prudence suggests that clinicians could continue to use calcium antagonists for approved indications, avoiding hypotension as a side effect and carefully evaluating the risk/benefit ratio when using these drugs for nonlicensed indications as in the early postinfarct period. When β-blockers are contraindicated for ischemic heart disease, a heart rate–lowering calcium antagonist should be chosen rather than a dihydropyridine.46 In patients with acute-phase myocardial infarction, calcium antagonists are still used too frequently,47 when in fact they are relatively contraindicated in that setting. There is no justification for using dihydropyridines to prevent atheroma, as shown by the INTACT and the still unpublished MIDAS data (the latter on isradipine). Clearly, the use of nimodipine in cardiovascular surgery must be avoided because of the bleeding risk, which outweighs any cerebroprotective role. This drug should specifically be kept for its benefit in subarachnoid hemorrhage (but not in cerebral hemorrhage). Patients with ischemic heart disease on short-acting nifedipine should be changed to long-acting preparations. Patients with hypertension without clinical angina or ischemic heart disease already controlled by short-acting nifedipine could be left on that preparation, provided that it is combined with a β-blocker or an ACE inhibitor to offset the expected neurohumoral activation. Otherwise, short-acting nifedipine should be changed to one of the slow-release, long-acting preparations or to amlodipine because of the much more favorable trough-to-peak ratios of the latter type of compounds.48 Short-acting capsular nifedipine might have very limited indications, such as in Raynaud’s phenomenon or in selected cases when acute blood pressure reduction is required in patients who are not under threat of clinical myocardial ischemia. In the case of hypertension, the safety of nifedipine retard tablets may be ensured when the STONE study is published in full. The recent preliminary report on the STONE study49 on elderly hypertensive patients showed a highly significant (P<.001) decrease in the probability of all events by intention-to-treat analysis in the nifedipine group compared with the placebo group. Specifically, there was a decrease in the terminating events, which were stroke, heart failure, uremia, myocardial infarction, angina pectoris, severe arrhythmias, hospitalization for severe illness, and death. The decrease in events with nifedipine was significant whether “all” events or only those that were hypertension-related were considered. If substantiated in the full publication, this single-blind trial carried out on 1632 subjects 69 to 79 years old who were treated for a mean follow-up of 30 months with nifedipine tablets given in a mean dose of 20 mg twice daily would provide strong evidence against the indictment by Furberg et al, at least in the case of elderly hypertensive patients. Other data will soon be coming in on the use of the dihydropyridine calcium antagonists in hypertension (nitrendipine in the case of the SYST-EUR study,50 amlodipine in the ALLHAT study, and felodipine in the HOT study51 ). In the case of ischemic heart disease, mortality data are thus far sadly lacking for the calcium antagonists. Clearly, more outcome data are required. In the case of patients with angina pectoris, an encouraging preliminary observation is that the nifedipine retard tablet preparation gave as good an effect on hard end points as did atenolol, whereas the combination of β-blockers and calcium antagonists actually reduced hard end points.52 Likewise, this combination was beneficial in another preliminary study involving nonrandomized follow-up of patients with coronary artery disease,53 in which the use of unidentified calcium antagonists was beneficial compared with no treatment. These data conform to good clinical practice; the combination of β-blocker plus calcium antagonist has many positive arguments,554 provided that the combination is undertaken prudently, with careful attention to precautions such as the avoidance of excess hypotension. Future Work In the words of another article coauthored by Furberg, “meta-analyses seldom if ever prove specific hypotheses.”55 In other words, even though the “allegations” against short-acting nifedipine are not, in our view, based on solid fact, and despite the apparent flaws in the analysis by Furberg et al,2 nonetheless, the proof of the pudding is in the eating, and a prospective study would answer the question. Bearing in mind that at least two pharmaceutical companies have made substantial profits from the sales of short-acting nifedipine in the past, it is surely their responsibility now to invest some of those profits in proper prospective studies, with mortality as end point in patients with angina pectoris treated by dihydropyridine calcium antagonists. If such studies had already been undertaken, the indictment by Furberg et al would never have arisen. Summary It would be a serious deviation from the facts available even to hint that all calcium antagonists should be damned in all situations on the basis of a meta-analysis, the kingpin of which depends on whether a 2-week or 6-month follow-up is preferred in two of the nifedipine groups. A further lesson is that a meta-analysis is not infallible but rather should conform to certain demands, such as a very low probability value (P<.001), and be restricted to populations with similar risks. Clearly, what is needed is a more thorough prospective database and fewer meta-analyses. Table 1. Reanalysis of Mortality Data for Short-Acting Nifedipine in Ischemic Heart Disease Including Results of Long-term 6-Month Follow-up in Two Studies45Dose RangeNifedipineControl95% CIP for Comparison30-60 mg/d355/4756326/4793(0.95-1.27).2280 mg/dMuller110/9310/88Eisenberg0/250/25Gerstenblith7/685/70Lichtlen12/2142/211Total29/40017/394(0.95-2.99).102>100 mg/dGottlieb4/644/68Jaffe1/130/9Muller15/632/63Total10/1406/140(0.63-4.41).44Totals of above groups (6-month Muller data)394/5296349/5327(0.97-1.28).13Totals listed in Furberg paper (2-week Muller data)335/4171274/4183(1.05-1.43).013As above recalculated from original papers (2-week Muller data)390/5301339/5329(1.01-1.33).0462For references, see Table of Furberg et al.2P values were calculated from the χ2 test with the Yates correction. 1Studies in which 6-month follow-up was chosen by present authors; 2-week follow-up in Table of Furberg et al2 ; daily dose in both studies could be >100 mg. 2If four nonnifedipine deaths are omitted from the Lichtlen data, the probability values become P=.29 for 80 mg/d and P=.065 for final comparison. 3Calculated from listed totals in Furberg et al.2FootnotesCorrespondence to Professor L.H. Opie, Heart Research Unit of the Medical Research Council, University of Cape Town Medical School, Observatory, 7925, Cape Town, South Africa, or Franz H. Messerli, MD, Section of Hypertensive Diseases, Ochsner Clinic, 1514 Jefferson Hwy, New Orleans, LA 70121. References 1 Opie LH. Drugs for the Heart. 4th ed. Philadelphia, Pa: WB Saunders Co; 1995:iii. Google Scholar2 Furberg CD, Psaty BM, Meyer JV. Nifedipine: dose-related increase in mortality in patients with coronary heart disease. Circulation.1995; 92:1326-1331. CrossrefMedlineGoogle Scholar3 Wagenknecht LE, Furberg CD, Hammon JW, Legault C, Troost BT. Surgical bleeding: unexpected effect of a calcium antagonist. Br Med J.1995; 310:776-777. CrossrefMedlineGoogle Scholar4 Muller J, Morrison J, Stone PH, Rude RE, Rosner B, Roberts R, Pearle DL, Turi 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 comparison. Circulation.1984; 69:740-747. CrossrefMedlineGoogle Scholar5 Muller JE, Turi ZG, Pearle DL, Schneider JF, Serfas DH, Morrison J, Stone PH, Rude RE, Rosner B, Sobel BE, Tate C, Scheiner E, Roberts R, Hennekens CH, Braunwald E. Nifedipine and conventional therapy for unstable angina pectoris: a randomized, double-blind comparison. Circulation.1984; 69:728-739. CrossrefMedlineGoogle Scholar6 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 (DAVIT-II) and other recent studies. Am J Cardiol.1991; 67:1295-1297. CrossrefMedlineGoogle Scholar7 ISIS-1 (First International Study of Infarct Survival) Collaborative Group. Randomized trial of intravenous atenolol among 16,027 cases of suspected acute myocardial infarction: ISIS-I. Lancet.1986; 2:57-66. MedlineGoogle Scholar8 Lichtlen PR, Hugenholtz PG, Rafflenbeul W, Hecker H, Jost S, Nikutta P, Deckers JW, on Behalf of the INTACT Group Investigators. Retardation of coronary artery disease in humans by the calcium-channel blocker nifedipine: results of the INTACT Study (International Nifedipine Trial on Antiatherosclerotic Therapy). Cardiovasc Drugs Ther.1990; 4:1047-1068. CrossrefMedlineGoogle Scholar9 Gonser M, Vetter K, Noack F. Meta-analyses of interventional trials done in populations with different risks. Lancet.1995; 345:1304. CrossrefGoogle Scholar10 Selwyn AP, Welman E, Fox K, Horlock P, Pratt T, Klein M. The effects of nifedipine on acute experimental myocardial ischemia and infarction in dogs. Circ Res.1979; 44:16-23. CrossrefMedlineGoogle Scholar11 Israeli SPRINT Study Group. Secondary Prevention Reinfarction Israeli Nifedipine Trial (SPRINT), a randomized intervention trial on nifedipine in patients with acute myocardial infarction. Eur Heart J.1988; 9:354-364. CrossrefMedlineGoogle Scholar12 SPRINT Study Group. The Secondary Prevention Re-infarction Israeli Nifedipine Trial (SPRINT-II): design and methods, results. Eur Heart J. 1988;9(suppl):350A. Abstract. Google Scholar13 CONSENSUS Trial Study Group. Effects of enalapril on mortality in severe congestive heart failure: results of the Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS). N Engl J Med.1987; 316:1429-1435. CrossrefMedlineGoogle Scholar14 Phillips RA, Goldman ME, Ardeljan M, Eison HB, Shimabukuro S, Krakoff LR. Isolated T-wave abnormalities and evaluation of left ventricular wall motion after nifedipine for severe hypertension. Am J Hypertens.1991; 4:432-437. CrossrefMedlineGoogle Scholar15 Baumgart D, Ehring T, Heusch G. A proischemic action of nisoldipine: relationship to a decrease in perfusion pressure and comparison to dipyridamole. Cardiovasc Res.1993; 27:1254-1259. CrossrefMedlineGoogle Scholar16 Egstrup K, Andersen PE. Transient myocardial ischemia during nifedipine therapy in stable angina pectoris, and its relation to coronary collateral flow and comparison with metoprolol. Am J Cardiol.1993; 71:177-183. CrossrefMedlineGoogle Scholar17 Ruzicka M, Leenen FHH. Rele