Should Recommendations for Clinical use be Restricted to Patients who were Enrolled in a Pivotal Clinical Trial?

This article refers to ‘What proportion of patients with chronic heart failure are eligible for sacubitril–valsartan?’ by P. Pellicori et al., published in this issue on pages 768–778. A large-scale clinical trial in patients with chronic heart failure and a reduced ejection fraction demonstrates that a new treatment reduces the risk for cardiovascular death. The results are clinically meaningful and statistically persuasive. The drug is approved for use throughout the world. How should we select patients to receive the drug in clinical practice? Many would say that the drug should be prescribed to patients who were similar to and treated similarly to those enrolled in the pivotal trial that demonstrated its benefits. Yet, doing so precisely would represent a practical impossibility. The trial was carried out in specialized centres by research teams, who were expert in the care of patients with heart failure. The protocol specified dozens of entry criteria; many were applied only to make the trial efficient rather than to focus specifically on patients who might benefit. Patients were provided with free study medication and were followed closely by attentive study coordinators. After commercialization, should we require that all these conditions be duplicated for patients to receive treatment? We do not demand that practitioners act as clinical trialists when they manage patients. We do not mandate that treated patients be identical to those studied or be cared for under highly supervised conditions. If we did so, most patients who would benefit from the new life-prolonging drug would never receive it; the findings of the clinical trial would never be translated for the public good. To maximize the gain to society, when a regulatory agency approves a drug, the directions for its use are written to reasonably reflect the patients who might be expected to benefit. Typically, the approved indication lists almost none of the trial's inclusion or exclusion criteria, and the agency does not require that the trial's supervisory structure be replicated in clinical practice. Agencies do not demand that patients receive the medication for free. Because they seek to maximize use of a valuable medicine, regulatory bodies specify only those conditions that are necessary to ensure a drug's safe and effective use. The goal is to help, not to hinder. Do guideline committees adopt the same philosophy? As a rule, guideline committees do not issue recommendations that are more restrictive than those provided by regulatory agencies. Guideline committees are dedicated to facilitating the translation of clinical trial evidence into daily practice; requiring physicians to replicate the conditions of a clinical trial would be entirely inconsistent with their mission. Guidelines are intended to encourage the widespread adoption of beneficial practices. Whereas regulatory agencies offer options for clinical use, guideline documents make stronger recommendations, such as by indicating that the use of a drug is not only appropriate, but also mandated. Guidelines also expand the clinical use of drugs; for example, US heart failure guidelines recommend the use of spironolactone and certain angiotensin converting-enzyme (ACE) inhibitors in patients with chronic heart failure, even though these drugs are not approved for such use by the Food and Drug Administration (FDA).1 The paper by Pellicori et al.2 published in this issue highlights an example in which the recommendations of a guidelines committee differ meaningfully from those of government agencies. In 2014, the PARADIGM-HF trial demonstrated that sacubitril–valsartan prolonged life and reduced hospitalizations in patients with chronic heart failure and a reduced ejection fraction, when compared with enalapril.3 In 2015, both the US FDA and the European Medicines Agency (EMA) approved sacubitril–valsartan for the treatment of chronic heart failure and a reduced ejection fraction, in place of an ACE inhibitor or angiotensin receptor blocker.4, 5 In 2016, the US heart failure guideline issued a mandate that sacubitril–valsartan be used to replace a conventional inhibitor of the renin–angiotensin system in class II and III patients with a reduced ejection fraction.1 None of these documents (including the most recent version of the Canadian guidelines6) currently proposes that sacubitril–valsartan be used only under conditions that precisely match those in the PARADIGM-HF trial. Yet, in 2016, the European Society of Cardiology (ESC) heart failure guidelines took a different approach. Amongst patients with heart failure and a reduced ejection fraction, the ESC committee recommended the use of sacubitril–valsartan only in patients who specifically conform to the randomization criteria of the PARADIGM-HF trial.7 Specifically, the document states that patients should fit the following profile: (i) they should have an elevated plasma natriuretic peptide level (BNP ≥150 pg/mL or NT-proBNP ≥600 pg/mL or, if they had been hospitalized for heart failure within the previous 12 months, BNP ≥100 pg/mL or NT-proBNP ≥400 pg/mL); (ii) they should have an estimated glomerular filtration rate of ≥30 mL/min/1.73 m2; and (iii) they should demonstrate the ability to tolerate separate treatment periods with enalapril (10 mg b.i.d.) and sacubitril–valsartan (97/103 mg b.i.d.). In addition, the ESC guidelines recommend sacubitril–valsartan only in patients who can tolerate and remain symptomatic on a mineralocorticoid receptor antagonist, although this was not required for participation in PARADIGM-HF.3 What are the consequences of this restricted recommendation concerning the use of sacubitril–valsartan? We can answer this question by comparing the number of patients who would be treated according to the recommendations of the FDA, EMA and the US heart failure guidelines with the number who would be treated according to the recommendations of the 2016 ESC heart failure guidelines. In the European Union, approximately 4 million people have chronic heart failure (with symptoms) and a reduced ejection fraction. How many patients would receive sacubitril–valsartan if the 2016 ESC heart failure guidelines were followed? Pellicori and colleagues2 surveyed a cohort of patients referred to a community heart failure clinic. They applied a generous interpretation of the ESC recommendations in that their denominator consisted only of patients who had a contemporaneous measurement of ejection fraction and natriuretic peptides, an uncommon circumstance in clinical practice. They also did not fully apply the restrictive criteria of the ESC guidelines; if they had done so, they would have considered only patients receiving enalapril 20 mg/day as well as a mineralocorticoid receptor antagonist to be eligible. If the ESC strict criteria were applied to their patient cohort, fewer than 5% of their patients (meaningfully less than the 12% described in the paper2) would be eligible for sacubitril–valsartan. If criteria regarding the dose of a renin–angiotensin system inhibitor were to be loosened, the proportion of eligible patients would soar to 60%. If natriuretic peptides were to be eliminated as a criterion (as would be the case if the FDA, EMA and US guidelines were applied), eligibility would be extended to all patients taking some dose of an ACE inhibitor or angiotensin receptor blocker (i.e. approximately 75% of patients with heart failure and a reduced ejection fraction).8, 9 Therefore, the ESC guidelines (as written) would recommend the use of sacubitril–valsartan to fewer than 200 000 people, whereas if criteria derived from the FDA, EMA and US guidelines were to be applied, the number of eligible patients would approach 3 million. According to Pellicori et al., the 2016 ESC heart failure guidelines restrict the administration of sacubitril–valsartan to less than 10% of the population that would be treated under the recommendations of the FDA and EMA.2 It is possible that the ESC guidelines committee was concerned about cost; yet, the most cost-conscious agency in Europe, the UK National Institute for Health and Care Excellence (NICE) recommends the unrestricted use of sacubitril–valsartan in patients with class II–IV symptoms with a reduced ejection fraction.10 Hypothetical concerns about dementia (based on the inconceivable use of a neprilysin inhibitor for 15–20 years) were alleviated in the original ESC document and by recent regulatory review.11 Authors of the guideline committee document have conceded that, if they had heart failure, they would personally take sacubitril–valsartan outside the confines of the guidelines, but in a manner consistent with the recommendations of government agencies.12 It is very difficult to write guidelines, and there should be no doubt that authors of guidelines have an impossible task. Yet, we must realize that the guideline writing process has no independent access to data, and thus committees must rely on and be sensitive to the analyses carried out by authoritative government organizations. That is why the 2016 ESC heart failure guidelines committee updated its recommendations for the use of ivabradine to make the guidelines consistent with the recommendations issued by European regulatory agencies.7 When the recommendations of guidelines committees and government organizations differ, there will be confusion. The analyses presented by Pellicori et al.2 reveal that such confusion has consequences. Conflict of interest: M.P. was one of the two principal investigators in the PARADIGM-HF trial, which compared sacubitril–valsartan with enalapril in chronic heart failure. M.P. has no recent or current personal, professional or institutional financial relationship with Novartis, which manufactures sacubitril–valsartan for commercial use.