Accepted Manuscript Anti-anginal drugs: Systematic review and clinical implications Rita Pavasini, Paolo G. Camici, Filippo Crea, Nicolas Danchin, Kim Fox, Athanasios J. Manolis, Mario Marzilli, Giuseppe M.C. Rosano, José L. Lopez-Sendon, Fausto Pinto, Cristina Balla, Roberto Ferrari PII: DOI: Reference: To appear in: Received date: Revised date: Accepted date: S0167-5273(18)36094-7 https://doi.org/10.1016/j.ijcard.2018.12.008 IJCA 27246 International Journal of Cardiology 18 October 2018 17 November 2018 3 December 2018 Please cite this article as: Rita Pavasini, Paolo G. Camici, Filippo Crea, Nicolas Danchin, Kim Fox, Athanasios J. Manolis, Mario Marzilli, Giuseppe M.C. Rosano, José L. LopezSendon, Fausto Pinto, Cristina Balla, Roberto Ferrari , Anti-anginal drugs: Systematic review and clinical implications. Ijca (2018), https://doi.org/10.1016/j.ijcard.2018.12.008 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT 1 Anti-anginal drugs: systematic review and clinical implications Rita Pavasini1, Paolo G. Camici2, Filippo Crea3, Nicolas Danchin4, Kim Fox5, Athanasios J. Manolis6, Mario Marzilli7,8, Giuseppe M. C. Rosano9,10, José L. Lopez-Sendon11, Fausto Pinto12, Cristina Balla1, Roberto Ferrari1,13 1: Centro Cardiologico Universitario, University Hospital of Ferrara, Via Aldo Moro 8, 44124 Cona, Ferrara, Italy. 2: Vita Salute University and San Raffaele Hospital, Via Olgettina Milano, 58–60, 20132 Milan, Italy. 3: Department of Cardiovascular and Thoracic Sciences, Catholic University, Largo Francesco Vito, 1, 00168 Rome, Italy. 4: Cardiology, European Hospital Georges-Pompidiou, 20 Rue Leblanc, 75015 Paris, France. 5: National Heart and Lung Institute, Imperial College and Institute of Cardiovascular MedicineandScience, RoyalBromptonHospital, SydneyStreet, London SW36NP,UK. 6: Department of Cardiology, Asklepeion General Hospital, 1, Vas. Pavlou Street, 16673 Voula, Athens, Greece. 7: Cardiothoracic Department, Lungarno Antonio Pacinotti, 43, 56126 Pisa, Italy. 8: Nottola Cardiology Division, Località Nottola, 53045 Ospedali Riuniti Valdichiana Sudest Siena, Italy. 9:Clinical Academic Group, St George’s Hospitals NHS Trust, Blackshaw Road, London SW17 0QT, UK. 10epartment of Medical Sciences IRCCS San Raffaele, Via della Pisana, 235, 00163 Rome, Italy. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 2 11: Cardiology Department, Hospital Universitario La Paz. IdiPaz, Paseo de la Castellana 261, Madrid 28036, Spain. 12: Hospital Universitário de Santa Maria/Centro Hospitalar Lisboa Norte, Centro Académico de Medicina de Lisboa, Centro Cardiovascular da Universidade de Lisboa, Faculdade de Medicina, Universidade de Lisboa, Lisboa, Portugal 13: Maria Cecilia Hospital, GVM Care & Research, Via Corriera 1, Cotignola (RA), Italy Short title: First and second line drugs: are they justified? Word count: 3493 Abstract: 214 References: 25 Keywords: beta-blockers, long acting nitrates, calcium channel blockers, nicorandil, trimetazidine, ivabradine, ranolazine, stable angina, coronary artery disease Corresponding author: Rita Pavasini Centro Cardiologico Universitario, University Hospital of Ferrara, Via Aldo Moro 8, 44124 Cona, Ferrara, Italy. [email protected] ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 3 ABSTRACT Background: The cornerstone of the treatment of patients affected by stable angina is based on drugs administration classified as first (beta-blockers, calcium channel blockers, short acting nitrates) or second line treatment (long-acting nitrates, ivabradine, nicorandil, ranolazine and trimetazidine ). However, few data on comparison between different classes of drugs justify that one class of drugs is superior to another. Methods: we performed a systematic review of the literature following PRISMA guidelines. Inclusion criteria: i) paper published in English; ii) diagnosis of stable coronary disease; iii) randomized clinical trial; iv) comparison of two anti-angina drugs; v) a sample size >100 patients; vi) a follow-up lasting at least 2 weeks; vii) paper published after 1999, when a meta-analysis of trials comparing beta-blockers, calcium antagonists, and nitrates for stable angina of Heidenreich et al. was published. Outcome: to establish whether the categorization in first and second line antianginal treatment is scientifically supported. Results: Eleven trials fulfilled inclusion criteria. The results show that there is a paucity of data comparing the efficacy of antianginal agents. The little data available show that there are not compounds superior to others in terms of improvement in exercise test duration, frequency of anginal attacks, need for sub-lingual nitroglycerin. Conclusion: The categorization of antianginal drug in first and second line is not confirmed. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 4 INTRODUCTION According to the ESC guidelines [1], drugs for symptomatic relief of angina are classified as being first line (beta blockers, calcium channel blockers, short-acting nitrates) or second line (ivabradine, nicorandil, ranolazine, long-acting nitrates, trimetazidine), with the recommendation to reserve second-line medications for patients who have contraindications to first choice agents, do not tolerate them, or remain symptomatic [1-3]. However, such categorical approach has been recently questioned [4-9]. The reasons for this criticism are multiple: 1) Second line drugs have been introduced more recently and they have been approved according to more stringent protocols, with larger sample size, longer follow-up, and safety data, compared to first line drugs which were studied in the early days with less precise description of methods, smaller sample size and shorter follow-ups [4, 9]; 2) The suggested and often-used combination of two or even three agents seems to be based on expert opinion rather than on scientific evidence [5-10]; 3) The categorical guideline recommendations do not take into account the different pathophysiological mechanisms underlying ischemia and angina (stable atherosclerotic plaque, vasospasm on the epicardial arteries, and coronary microvascular dysfunction) [4, 6]; 4) Patients with angina can have several co-morbidities requiring drugs with the appropriate auxiliary properties which are not considered in the guidelines [4, 8-9]. Recently, an expert consensus proposed a new algorithm for a more personalized medical treatment of symptomatic angina, a so called “Diamond approach” [4, Figure 1]. The authors assumed that there was no direct comparison between first-line and second-line treatments to support the superiority of one group of drugs over the other. They referred to an old meta-analysis based only on the three antianginal drugs considered first line published by Heidenreich et al. in 1999 [4, 11]. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 5 The purpose of the current systematic review is to analyze data about the more recently approved compounds which are classified as second line choice. METHODS We performed a systematic review of the literature following Preferred Reporting Items for systematic Reviews and Meta-analysis (PRISMA) [12-13]. Appropriate articles were searched in MEDLINE and in EMBASE. “Mesh” strategy was used. The terms searched were (((("angina, Stable"OR "Coronary Artery Disease") AND ("Diltiazem" OR "Verapamil" OR "Nifedipine" OR "Amlodipine" OR "Felodipine" OR "Nicardipine" OR "Nimodipine" OR "Isosorbide Dinitrate" OR "Nicorandil" OR "Ranolazine" OR "Trimetazidine" OR "Acebutolol" OR "Atenolol" OR "Bisoprolol" OR "Celiprolol" OR "Metoprolol" OR "Nadolol" OR "Propranolol" OR "carvedilol" OR "Penbutolol" OR "Nebivolol"OR "Labetalol" OR "Sotalol"OR "ivabradine") AND ("Randomized Clinical Trial"). The search was carried out between October and November 2017. The inclusion criteria were: i) paper published in English; ii) studies on patients with a diagnosis of stable coronary disease; iii) randomized clinical trial; iv) comparison of two anti-angina drugs; v) a sample size of at least 100 patients; vi) a follow-up lasting at least 2 weeks; vii) paper published after 1999 only if regarding beta-blockers, calcium antagonists and long acting nitrates (a metaanalysis of Heidenreich et al. was published in 1999 regarding trials comparing beta-blockers, calcium antagonists, and nitrates for stable angina [11])(Figure 2). Studies with the following characteristics were excluded: i) observational; ii) comparing an anti-anginal drug versus placebo; iii) comparing an anti-anginal drug versus two drug combinations; iv) comparing an anti-anginal drug versus another drug within the same class. The inclusion of the papers in the systematic review was decided after analysis of the full-text of papers selected. Divergences were solved by consensus. The outcome of interest was related to the effect of the drugs on: i) the frequency of anginal attacks; ii) the need for sub-lingual nitroglycerin; iii) results of exercise testing. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 6 The quality of the studies included was evaluated with the Cochrane Collaboration approach. In particular, the risk of analytical, selection, adjudication, and attrition bias (expressed as low, moderate, or high risk of bias, as well as incomplete reporting leading to inability to ascertain the underlying risk of bias) was assessed. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 7 RESULTS A first screening of the literature retrieved a total of 92 papers. After re-evaluation of the title and abstract, 18 studies were considered for the full-text analysis. Following the scrutiny of inclusion and exclusion criteria, 7 papers were excluded: one because it focused on hypertensive patients; one was a network meta-analysis; one a duplicate of a sample population of another study; in another one the comparator was placebo and in three studies the sample size was less than 100 patients. For all these reasons, 11 studies were included in the final analysis (Figure 1s, supplemental online material). All 11 studies were randomized trials. Figure 2s (supplemental online material) summarizes the risk of bias of studies included by Cochrane methods. Beta-blockers versus calcium channel blockers groups. The study was well planned, and diagnosis of stable angina was based on symptoms and positive exercise test. The presence of obstructive coronary artery disease was not verified by coronary angiography nor was an inclusion criterion (table 2). Another limitation is that patients received concomitant therapy with long acting nitrates (table1). In addition, it is not clear how the sample size was calculated (table 2). ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 8 Long-acting nitrates versus calcium channel blockers The report of Hall et al. [15] fulfilled our criteria (Table 1). A total of 196 patients were included in this 28 weeks randomized, double-blind parallel group study comparing amlodipine versus long acting nitrates (Table 2). The efficacy analyses were based on total exercise time, number of anginal attacks per week or the need of short acting nitrates. There was no statistically difference in terms of median number of angina attacks (surprisingly none in both groups) and in number of short acting nitrates taken. Amlodipine resulted better than long acting nitrates in terms of total exercise time achieved. No differences were found in relation to adverse events or quality of life tested with the SF36 Questionnaire. The study reported a formal a priori calculation of the sample size. One of the limitations, however, is related to the definition of stable angina. Criteria for this diagnosis were not specified and the presence of coronary artery disease was not proven by coronary angiography. It is not clear how the number of anginal attacks per week and consumption of short acting nitrates were recorded. This information is relevant to understand the reported absence of angina attacks. Nicorandil In 2007 Zhu et al. conducted a randomized double-dummy trial, comparing the effect of nicorandil versus isosorbide mononitrate in 249 patients for 4 weeks [16] (Table 2). Other anti-anginal drugs such as short acting nitrates and beta-blockers were allowed during the study period. However, the difference in the number of patients treated with these drugs at baseline was not statistically significant among study groups. The primary endpoint was exercise time to 1 mm ST segment depression whilst other exercise test parameters and number of anginal attacks were considered secondary endpoints. Sample size was calculated a priori. There were no differences in the primary endpoint between the drugs. Consumption of short acting nitrates and occurrence of angina was significantly reduced only by Nicorandil. However, the analysis on this endpoint was performed on only 41 patients receiving Nicorandil and 40 treated with isosorbide mononitrate, as the vast majority of patients did not need nitro-glycerine. Adverse events were similar between two groups, ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 9 headache being the most common (Table 1). Again, in this study the presence of obstructive coronary artery disease was not assessed (Table 2). The SWAN study, published in 1999 and not included in the meta-analysis of Heidenreich et al [11], is a randomized double-blind parallel group trial comparing Nicorandil 10 mg bid versus Amlodipine 5 mg od in 121 patients for 8 weeks (Table 2) [17]. Presence of coronary artery disease with >50% stenosis in one of the coronary arteries was one of the inclusion criteria (Table 2). The study showed that although time to onset of ST-segment depression improved only with amlodipine, time to angina attack, total exercise duration, reduction in weekly angina episodes, number of short acting nitrates used and quality of life were not different [17]. The analysis was on an intention to treat basis and the calculation of a sample size was reported (Table 2). Nicorandil seemed to be better tolerated than Amlodipine in terms of occurrence of adverse events [17]. Finally, Guermonprez et al. performed a 3 months, double-blind randomized parallel group study, in 123 patients comparing Nicorandil 20 bid versus Diltiazem 60 tid [18]. This study was published in 1993 but was not considered in the meta-analysis of Heidenreich et al. [11]. There were no statistically significant differences between groups in any of the endpoints considered. The double product at peak exercise between groups was similar. (Table1). Ranolazine Rousseau et al. compared the effect of Ranolazine versus Atenolol in 152 patients in a 3 weeks, randomized 3-period crossover, double-blind, double-dummy study [19] (Table 1-2). Even in this case, concomitant medications with short acting nitrates and calcium channel blockers were allowed. As result, 54% of the overall population was treated with calcium channel blockers at baseline, and, unfortunately, the difference between groups regarding this co-administration is not described. There was no statistically significant difference in time to angina onset or in time to onset of segment ST depression between ranolazine and atenolol. Total exercise duration was longer with ranolazine compared to atenolol 100 mg (mean difference 21.1 seconds, 95% CI 6.2-36, ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 10 p=0.006) though atenolol reduced more than ranolazine the cardiac workload and rate pressure product [19]. Both treatments had similar effect in the reduction of anginal attacks and in nitroglicerin use (table1). Trimetazidine We found two studies fulfilling our inclusion and exclusion criteria, both published before 1999 and not considered in the meta-analysis by Heidenreich et al. In the first study, Detry et al. performed a randomized double-blind trial in 149 male patients with stable angina comparing propranolol (40 mg tid) and trimetazidine (20 g tid) for 3 months [20](Table 2). The results show a similar efficacy profile in terms of angina attacks per week, exercise duration, time to 1 mm segment depression. The double product remained unchanged for patients treated with trimetazidine, but significantly decreased with propranolol, confirming that trimetazidine has anti-ischemic effect not related to the reduction in myocardial oxygen demand [20]. Koylan et al. performed a 4 weeks, multicenter randomized double-blind trial, comparing trimetazidine (20 mg tid) and diltiazem (60 mg tid) in 116 male patients. Both treatments reduced the number of anginal attacks and maximal ST-segment depression, but none changed time to ST segment depression and ST recovery on exercise test. Finally, Diltiazem prolonged PR and QRS duration [21]. Ivabradine Three studies were selected for Ivabradine, all of them with a follow-up of 3 months (Table 2). Two compared ivabradine and atenolol [22-23] and the other compared ivabradine and amlodipine [24]. Tardif et al [22], performed a double-blind parallel group trial involving 939 patients. Participants were randomized to receive ivabradine 5 mg bid, up titrated to 7.5 mg or to 10 mg bid for 12 weeks, or atenolol 50 mg od for 4 weeks increased to 100 mg for 12 weeks. Ivabradine was not inferior to atenolol in the primary outcome which was increase in total exercise duration for both the tested doses. At the peak of the activity, non-inferiority of ivabradine 5 mg compared to atenolol 50 mg ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 11 was demonstrated for all exercise test criteria (time to limiting angina, time to onset angina, time to ST segment depression). Ivabradine, as well as atenolol, reduced the number of angina attack by two thirds. An a priori sample size calculation was reported (Table 2) and evidence of CAD was documented by, at least, one of the following criteria: myocardial infarction ≥3 months before study entry, coronary angioplasty ≥ 6 months or bypass surgery ≥3 months before entry, coronary angiogram showing ≥ 1 diameter stenosis ≥ 50%, or scintigraphic/echocardiographic evidence of exercise-induced reversible myocardial ischemia. The only missing analysis is the comparison in the safety outcome, partly related to the selection process of the patients (which excluded patients with known beta-blockers intolerance excluded from the trial) [22]. These results were confirmed in a smaller study of Li et al, published in 2014 [23]. This is a randomized double-blind, double-dummy trial, on 332 patients comparing Ivabradine 5 mg bid or 7.5 mg bid, and Atenolol 12.5 or 25 mg bid for 12 weeks. Ivabradine showed to be noninferior to Atenolol in improving exercise capacity and reducing heart rate. The rate of adverse events was slightly higher with Atenolol (66 versus 73, p>0.05, respectively) [23]. Finally, Ruzyllo et al. [24] compared the effect of ivabradine 7.5 or 10 mg bid versus amlodipine 10 mg. This was a randomized, double blind, three arm parallel group trial involving 1195 patients. Also this was a well performed trial: primary and secondary endpoints as well as sample size were pre-defined (Table 2). CAD was documented by: 1) a ≥3-month history of chronic stable effort-induced angina, relieved by rest or short acting nitrates; 2) coronary artery disease (CAD) documented by occurrence of a myocardial infarction (MI) ≥3 months previously, or coronary artery bypass graft surgery (CABG) ≥3 months previously, or percutaneous coronary angioplasty (PTCA) ≥ 6 months previously, or by coronary angiography, stress echocardiography or scintigraphy; and 3) a positive bicycle exercise tolerance test (ETT) at selection and at inclusion (Table 2). The analyses were on an intention to treat basis, but also per-protocol population. The trial confirmed the non-inferiority of ivabradine compared to amlodipine in the improvement of total exercise duration, but also in all the secondary outcomes analyzed (time to angina onset, time to 1 mm ST segment depression, rate ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 12 pressure products, short acting nitrates use and angina attack frequency). In particular, the decrease in rate pressure product was significantly greater for patients treated with ivabradine and mostly at peak exercise [24]. The most frequent adverse events were visual symptoms and sinus bradycardia with ivabradine (0.8% and 0.4% withdrawals, respectively) and peripheral oedema with amlodipine (1.5% withdrawals) [24]. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 13 DISCUSSION Our overview highlights several points that warrant special attention from the Scientific Community. The first point relates to the paucity of adequate contemporary comparative data on the efficacy and tolerability of treatments with the different guideline recommended drugs for patients who have stable angina. By systematically searching Medline and Embase database and reviewing the bibliography to locate additional relevant studies published after 1999, when a meta-analysis on the comparative trials of first line antianginal drugs (beta blockers, Ca2+ blockers, and long-acting nitrates) was reported [11], we could identify only 11 randomized studies. It should be underlined that our requirements for inclusion were as simple as possible: a minimum of 100 patients, and at least one of the following outcomes: frequency of anginal attack, use of sublingual nitro-glycerine, or time on exercise test. Yet, in 18 years only 3925 patients have been randomized in comparative studies and out of these 3925 patients, 2130 in three trials related to ivabradine, vs atenolol, and/or amlodipine. The efficacy of another recently developed antianginal drugs, ranolazine, was compared to atenolol, in a trial randomizing 125 patients only [19]. Nicorandil, another antianginal was tested in 493 patients in three trials against isosorbide mono-nitrate (249 patients)[16], amlodipine (123 patients)[17] and diltiazem (123 patients)[18] (two of them published before 1999). In a recent analysis of the number of subjects required to show superiority comparing two antianginals drugs in parallel groups with 90% power and exercise testing as the primary endpoint showed that approximately 350-425 subjects would be needed [23,24]. Moreover, failure to show superiority in a clinical trial does not imply equivalence unless powered accordingly [25]. On this basis, out of the eleven selected studies only the two trials involving ivabradine were adequately powered. The second point underlines the even more critical lack of data available before 1999, when the only comparative meta-analysis was published [11]. Figure 2 shows that, in 30 years, 76 studies ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 14 were conducted comparing the drugs available at that time, which today are classified as first-line antianginal drugs. The 1999 meta-analysis included 69 trials plus some abstracts and studies related to outcome and/or adverse effects [11]. Interestingly, applying our criteria to these studies, only 7 out of 69 could be included in our analysis, mainly because of the relative small number of patients involved. Our initial intention was to conduct a second meta-analysis including all the new contemporary comparative studies. Once we realized the small number of reports available and of patients included along with the heterogeneity of the primary endpoints, we shifted our effort to a systematic review of the literature with the aim to alert the Scientific Community on the relative lack of evidence-based information about treatments for a pathology that affects millions of people worldwide and that is expected to further increase with the aging of the population. The third point relates to the typology of the available studies which, unfortunately, does not allow a proper comparison among treatments. The comparator was often tested at different dose regimen and often under-dosed. In some studies, a background therapy, mostly with long lasting nitrates and, in some cases, even with beta blockers, was allowed, thus making difficult to test the intrinsic efficacy of the study drugs. The calculation of sample size is not always available nor the diagnostic criteria of angina. As shown in Table 1, different protocols were used for exercise testing, which also might affect the results. The fourth point relates to the results of the so far available evidence which, besides all the concerns, constantly shows that the new so called “second line” drugs provide an equivalent reduction in angina severity and improvement of exercise test when compared to historical “first line beta blockers and calcium blockers. Equally resulted in similar or even reduced rate of adverse events. Earlier reviews also indicate no difference between the efficacy and occurrence of adverse events between the first and the second line drugs [4,10-11]. It should be recalled that two studies conducted with ivabradine compared to atenolol and amlodipine do not share any of the criticism of the other studies as they involved the necessary sample size calculated a priori to demonstrate noninferiority of ivabradine in comparison with a full dose of atenolol (100 mg) and amlodipine (10
