Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Deciphering the ARB and MI Data: Insight from Systematic Reviews, Meta-analyses, Meta-regression Analyses, and Non-Inferiority Trials
- Dr. Martin H. Strauss, Dr. Subodh Verma (6 October 2005)
-
ARBs and myocardial infarction - the saga continues
- James Penston (7 October 2005)
-
What are they good for?
- Kevin Pearce (15 October 2005)
-
ARB and myocardial infarction: Safety aspects in patients with history of stroke- Analysis from the MOSES study (Eprosartan compared with nitrendipine for secondary prevention of stroke)
- Stephan Lüders, Anke Kulschewski (1), Frank Hammersen (1), Kerstin Plate (2), Jürgen Berger (3), Walter Zidek (4), Peter Dominiak (5), Hans Christoph Diener(6), Joachim Schrader (1) (14 December 2005)
|
|
|||
|
Dr. Martin H. Strauss, Cardiologist North York General Hospital, Toronto, Canada M2K 1E1, Dr. Subodh Verma
Send response to journal:
|
We read with great interest the systematic review by McDonald et al(1) on ARBs and myocardial infarction, and respectfully submit that the authors may have over interpreted their findings. The authors included ARB trials with either placebo or ACEI as comparators. Two of the trials would confound the results as they had background ACEI therapy. In CHARM-Added, all of the patients, and in Val- Heft, 93% of the patients received ACEI therapy. This makes it difficult to ascertain the relative impact of an ARB alone. The review also excluded ARB trials with non-ACEI drugs as comparators, where ARBs have been demonstrated to perform inferiorly with respect to reduction in MI. We call your attention to a meta-analysis recently published by Cheung et al(2) of randomized controlled trials of ARB in hypertension that were a minimum of 2 years in duration and had greater than 100 major cardiovascular events. Three trials met these criteria; LIFE, VALUE, and SCOPE, all having non-ACEI active comparators (beta blockers or calcium channel blockers) and included 29,375 patients. ARB had a significant increase in myocardial infarction (RR ratio 1.12 95% CI: 1.01-1.26 P=.041) as compared to non-ACEI active comparators. Thus, omitting the results of LIFE and VALUE in the current paper may be considered statistically inappropriate. We are surprised that the authors, while excluding LIFE and VALUE, included SCOPE as a placebo controlled trial. It is important to point out that while SCOPE was designed as a placebo controlled trial; it was in fact switched to an active comparator trial midstream. In fact, 84% of the placebo arm in SCOPE received active treatment when the WHO hypertension guidelines changed after the HOT trial was published. If we follow the authors’ rationale for not including non-ACEI activator comparators, SCOPE should also have been excluded. The findings of McDonald et al. are reassuring in that there was no increase in MI. However, surprisingly there was no reduction in MI either. This apparent lack of coronary vascular protection, supports the recent comprehensive review of the BPLTTC(3) presented at the European Society of Hypertension in 2005 (see below). The reduction of cardiovascular events by ACEI have been shown in a recent meta-regression(4) analysis of 180,000 hypertensive patients, to be above and beyond that of blood pressure lowering alone. This meta- regression analysis(4) also suggests that a 10 mmHg reduction in systolic blood pressure should have a predictable 15% relative risk reduction in cardiovascular events. It is therefore very surprising that ARBs were not associated with a reduction in MI in the paper by McDonald et al considering the majority of those trials had BP reduction in favor of ARBs. Although MI is an important clinical endpoint, it may not always be the best measure of the vascular impact of an ARB. The OPTIMAAL trial, which drove the results of the ACEI vs. ARB arm (86.8% of weight odds ratio) in the paper by McDonald et al. is a poignant example. While losartan and captopril had similar MI rates in this post MI population with heart failure, losartan, was associated with a significant increase in cardiovascular death (RRR 95% CI: 1.17 1.01-1.34 p=.032). In RENAAL, a high risk population with diabetic nephropathy, losartan actually reduced MI, and delayed the need for dialysis, but also tended to increase mortality when dialysis was required (FDA Advisory Briefing NDA 20-386 (S- 028) www.fda.gov) The authors suggest that an apparent absence of difference in the rate of MI implies that no difference may exist. The validity of this conclusion is unclear. The vast majority of ARB trials under discussion are “superiority trials”. If one drug is not proven superior to another drug in a “superiority” trial, basic statistical principles does not suggest that the drugs under study are equivalent, even if the event rates are no different(5). Some ARB trials have had a secondary statistical analysis called “non -inferiority.” “Non-inferiority” is a relatively new statistical concept that has resulted in confusion, since many physicians have incorrectly interpreted this term to imply equivalence. “Non-inferiority” as a statistical concept, however, does not prove clinical equivalence, rather it proves that a study drug is not “substantially worse” than the gold standard(6). The landmark VALIANT trial is an excellent example of a non- inferiority trial. In VALIANT, valsartan 160 mg bid although not superior to captopril was proven non-inferior for the primary endpoint of mortality. A finding of non-inferiority for valsartan allowed a change in valsartan’s “Final Printed Labeling (FPL)” by the FDA (FDA Document NDA 21 -283/S-011 www.fda.gov) in August 2005 which states that non-inferiority makes it “unlikely that valsartan has less than about half of the estimated effect of captopril” Therefore, the product labeling of valsartan per se does not imply equivalence; rather that valsartan is an excellent 2nd line therapy for the ACEI intolerant patient. Perhaps the most compelling and comprehensive analysis of this topic is provided by the “Blood Pressure Lowering Treatment Trialists’ Collaboration” (BPLTTC). The BPLTTC are conducting prospectively planned meta-analysis of blood pressure lowering trials. Dr. Turnbull recently presented at the 15th meeting of the European Society of Hypertension, the largest and most comprehensive meta-regression analysis of ACE inhibitors and ARBs(3). In this analysis, the reduction in systolic blood pressure was plotted against the relative risk of the pre-specified end-points of stroke, heart failure, and coronary heart disease. This study consisted of 21 large-scale randomized trials [16 trials with ACE inhibitors (AASK, ABCD (H), ABCD (N), ALLHAT, ANBP2, CAPPP, DIAB-HYCAR, EUROPA, HOPE, JMIC- B, PART-2, PEACE, PROGRESS, SCAT, STOP-2, UKPDS-HDS) and 5 trials with ARBs (IDNT-placebo and CCB, LIFE, RENAAL, SCOPE, VALUE)] and involved 137,356 patients. The BPLTTC concluded that whereas there were no differences in risk reduction between ACE inhibitors and ARBs with respect to the outcomes of stroke and heart failure, a highly statistically significant (P=0.001) benefit of ACE inhibitors relative to ARBs on myocardial infarction and cardiovascular death was apparent (ACEI -9% [- 14% to -3%]; ARB +7% [-7% to +24%]). To our knowledge, the BPLTTC is the most robust data supporting the notion that ACE inhibitors may offer greater coronary vascular protection than ARBs, and that this effect is beyond what could be expected by blood- pressure lowering alone. Clinicians and patients now have important new information. For patients who are intolerant of an ACE inhibitor, an ARB appears to offer equal risk reduction in terms of stroke and heart failure, recognizing that the effects of an ACE inhibitor on coronary vascular outcomes may exceed that of an ARB. It would be appropriate to consider the BPLTTC analysis as an interim verdict of equivalence of the two classes with respect to cerebrovascular outcomes and heart failure, and superiority of ACE inhibitors for coronary vascular protection with reductions in myocardial infarction and cardiovascular death, while we await the results of ONTARGET-TRANSCEND. Dr. Martin H. Strauss, Division of Cardiology, North York General
Hospital, 4001 Leslie Street, Toronto, Canada M2K 1E1 Dr. Subodh Verma, Division of Cardiac Surgery, St. Michael’s
Hospital, University of Toronto, 30 Bond Street, 8th Floor Bond Wing,
Toronto, Canada M5B 1W8 Reference List (1) McDonald MA, Simpson SH, Ezekowitz JA, Gyenes G, Tsuyuki RT. Angiotensin receptor blockers and risk of myocardial infarction: systematic review. BMJ 2005. (2) Cheung BM, Cheung GT, Lauder IJ, Lau CP, Kumana CR. Meta- analysis of large outcome trials of angiotensin receptor blockers in hypertension. J Hum Hypertens 2005. (3) Blood pressure-independent effects for agents inhibiting the renin-angiotensin system. Program and abstracts from the Fifteenth European Meeting on Hypertension, June 17-21, 2005, Milan, Italy. Plenary session (http://www.medscape.com/viewarticle/507293). 2005. (4) Verdecchia P, Reboldi G, Angeli F, Gattobigio R, Bentivoglio M, Thijs L et al. Angiotensin-converting enzyme inhibitors and calcium channel blockers for coronary heart disease and stroke prevention. Hypertension 2005; 46(2):386-392. (5) McAlister FA, Sackett DL. Active-control equivalence trials and antihypertensive agents. Am J Med 2001; 111(7):553-558. (6) Siegel JP. Equivalence and noninferiority trials. Am Heart J 2000; 139(4):S166-S170. Competing interests: Martin Strauss has received speaking honorarium from Sanofi-Aventis Canada, Pfizer Canada, Abbot Laboratories, Bayer Canada, Roche, and Tanabe Pharma Subodh Verma has received research grants and speaking honorariums from GlaxoSmithKline Canada, Sanofi-Aventis, Pfizer Canada, Actelion Canada, AstraZeneca Canada and Merck/MerckFrosst Schering Canada |
|||
|
|
|||
|
James Penston, Consultant Physician/Gastroenterologist Scunthorpe General Hospital, Cliff Gardens, Scunthorpe, North Lincolnshire DN15 7BH
Send response to journal:
|
Sir The systematic review by McDonald et al. [1] was conducted according to the accepted standards, it involved 19 studies with a combined total of 31,569 patients and yielded results that showed that angiotensin receptor blockers were not associated with an increased risk of myocardial infarction. Why, then, is there such reticence on the part of the authors? On three occasions, they urge caution in the interpretation of the odds ratios by emphasizing that the results are compatible with either an increase of 16% or a decrease of 13-25% in the risk of myocardial infarction in patients receiving ARBs. And why do they stress the importance of awaiting further studies? If the authors are convinced about the reliability of systematic reviews, then it is surprising that they feel the need to draw attention to these issues. After all, confidence intervals hardly require such explicit elucidation whilst it goes without saying that all data are preliminary in the sense that they may be overturned by future investigations. Is it the case, then, that they harbour doubts about the reliability of systematic reviews? The authors’ qualms, though, did nothing to weaken their primary message: “Angiotensin receptor blockers do not increase the risk of myocardial infarction.” [1] But is this claim justified? If, as McDonald et al. seem to believe, we have to wait for the results of on-going large- scale randomised trials to provide the answer, then surely their assertion is premature. The study by McDonald et al. is nothing but a veiled defence of ARBs produced by authors - the majority of who have financial links to the pharmaceutical industry – intent on challenging the view that this class of drugs is associated with an increased risk of myocardial infarction. [2] But even if the muddled thinking and conflicts of interest are set aside, the underlying data – whether from meta-analyses or large-scale randomised trials – will always leave so much leeway for interpretation that the views from both courts will linger. [3] References [1] McDonald MA, Simpson SH, Ezekowitz JA, Gyenes G, Tsuyuki RT. Angiotensin receptor blockers and risk of myocardial infarction: systematic review. BMJ doi:10.1136/bmj.38595.518542.3A (29th September 2005) [2] Verma S, Strauss M. Angiotensin receptor blockers and myocardial infarction. BMJ 2004;329;1248-9 [3] Penston J. How much should patients be told about their treatment? 2nd December 2004. BMJ Rapid Response Letter to Verma & Strauss, BMJ 2004;329;1248-9. Competing interests: None declared |
|||
|
|
|||
|
Kevin Pearce, GP HA26HL
Send response to journal:
|
If these drugs have no effect one or the other on myocardial infarction ,why on earth are we continuing to prescribe them? I think if I had hypertension or existing coronary heart disease I might like a drug which worked even if it did make me cough. Competing interests: None declared |
|||
|
|
||||
|
Stephan Lüders, Medical Clinic, St.Josefs-Hospital D-49661 Cloppenburg, Germany, Anke Kulschewski (1), Frank Hammersen (1), Kerstin Plate (2), Jürgen Berger (3), Walter Zidek (4), Peter Dominiak (5), Hans Christoph Diener(6), Joachim Schrader (1)
Send response to journal:
|
McDonald et al. showed in their systematic review that there was no significant association of treatment with ARB and risk of myocardial infarction.(1) But due to large confidence interval there is still need for more data. In the review the authors analysed studies in patients with different risk factors at inclusion of the studies (diabetes, nephropathy, heart failure, myocardial infarction or coronary syndrome), but there was no study showing effects in patients with history of stroke. From the MOSES-study (Eprosartan compared with nitrendipine for secondary prevention of stroke) we can add informations to this important question in these high risk patients. (1) MOSES was the first study in 1352 patients comparing an ARB (eprosartan) with a calcium antagonist (nitrendipine) in secondary prevention of stroke, showing a significant benefit in reducing the combined end-point of total mortality, cardio-and cerebrovascular events in favour of the ARB. In this study the number of acute coronary syndrome in the eprosartan group was even less than in the nitrendipine group (39 vs. 48; not significant) showing at least the safety of treatment with eprosartan in patients with cerebrovascular disease. (1) Department of Internal Medicine, St.-Josefs-Hospital Cloppenburg,
Germany
References 1) McDonald MA, Simpson SH, Ezekowitz JA, Gyenes G, Tsuyuki RT: Angiotensin receptor blockers and risk of myocardial infarction: systematic review. BMJ doi:10.1136/bmj.38595 2) Schrader J, Luders S, Kulschewski A, et al. Morbidity and Mortality after Stroke-Eprosartan Compared With Nitrendipine for Secondary Prevention (MOSES). Stroke. 2005;36:1218-1226 Corresponding author: Dr. Stephan Lüders, Medizinische Klinik St. Josefs-Hospital, Krankenhausstr. 13 49661 Cloppenburg, Germany; e-mail: s.lueders@kh- clp.de Competing interests: None declared |
|||