Current Guidelines in the Management of Chronic Heart Failure: Practical Issues in their Application to the Community Population

Over the last several years there have been significant advances in the management of chronic heart failure, encompassing developments in pharmacotherapy and device-based approaches to this problem. Clinical trials have demonstrated continued improvement in prognosis in this patient population as measured by reduced mortality and declining rates of hospitalisation 1 2 3. However, the anticipated improvement in outlook has not materialised in the routine community heart failure population, in part reflecting the less than optimal translation of clinical trial results 4. To help address this problem, International Societies have published guidelines to outline the appropriate application of recent proven therapies 5,6. Guideline committees assess available data, distil the information and provide consensus statements on how best to apply the knowledge gained from clinical trials and other relevant sources. However, despite the strong scientific basis underlining these guidelines, application of proven and recommended therapy remains a problem. Aside from simple failure to be aware of these guidelines, several other reasons may explain their lack of application. These reasons reflect the merging of the clinical trial data and guideline suggestions, and the physician's opinion on how they apply to the treatment of a specific patient. This underlines the fact that guidelines represent advice, but not a mandate on how to treat patients with heart failure. This article will not include a didactic review of the guidelines but will discuss some general reasons why practice may deviate from guidelines. Guidelines are based on results obtained from “clinical trial populations” which differ in many important aspects from the community patient with heart failure. In general the latter patient is often a decade older, has more extensive co-morbidities, especially renal disease, is taking more medication, both cardiac and non-cardiac, all of which can significantly alter response to therapy. In short, the community patient is more complex. In particular, these issues can significantly alter drug metabolism, heightening response to lower doses, increasing the risk of drug–drug interactions and potentially exposing the patient to drug-related complications producing significant morbidity 7,8. These issues undoubtedly contribute to the decision on whether to apply guideline suggestions in a particular circumstance. Furthermore, co-morbidities can influence management decisions in other ways. Frequent visits for gradual up-titration of medications such as angiotensin-converting enzyme inhibitors (ACEI) and beta blockers (BB) places a significant burden on elderly patients with concomitant illnesses which may limit mobility to and from clinics. This labour-intensive approach may also stretch resources. Whereas there are no data to directly implicate these factors in the proven under-utilisation of therapies such as BB, it is possible that they may influence prescription practices favouring lower doses. In this regard it is of interest that recent data suggests home titration of beta blockers can be safely performed, an approach which may facilitate more widespread uptake of this important therapy 9. A similar approach to ACEI therapy may be equally practical, encouraging prescription of higher doses and the anticipated benefits of that practice 10. Aside from differences already outlined between the clinical trial and community heart failure populations, other issues related to trial design may influence interpretation of the results and the guidelines that emanate from trials. Entry criteria for studies are routinely very restrictive leading to a high ratio of screened: enrolled patients. This raises the issue of whether therapy deemed effective in such circumstances will be similarly effective when given to patients with heart failure who would have been excluded from those clinical trials. Furthermore, on occasion, trial entry criteria in terms of background therapy may not reflect modern day practice by the time a study is completed. For example, in RALES, entry criteria did not prespecify the need for BB therapy. Consequently, only ∼10% in each group were on these agents in combination with ACEI therapy 11. As a result of the positive result of this trial, guidelines now suggest that we add spironolactone to background therapy of ACEI and BB in certain clinical circumstances. Whereas there is sound theoretical support for this advice, a careful review of the evidence would indicate that the data do not robustly defend this position. Practitioners strictly adhering to the practice of evidence-based medicine may therefore not feel justified in prescribing spironolactone in the absence of more convincing data. The recent publication by Juurlink et al may support this stance, reporting an increased incidence of spironolactone-related side-effects in patients with heart failure since the publication of the RALES data 12. The explanation for this is unclear, but it is of interest that one of the major differences noted in this report in concomitant drug therapies pre and post RALES was the increased use of BB 13. Beta blockers inhibit renin release which may reduce aldosterone production. While the recently published EPHESUS report provides further evidence for the use of aldosterone inhibitors in patients with heart failure, this study focused on a post-infarct heart failure population and therefore did not study the same group as the RALES program 14. Consequently, the argument could be made that the role of aldosterone antagonists in established Class III heart failure with reduced systolic function may require more investigation. Entry criteria may also influence the clinical interpretation of the MADIT II trial 15. Subsequent to this trial recommendations have been made to place a defibrillator in patients with ischaemic cardiomyopathy with a LVEF less than 30% who are at least 3 months remote from an infarct. However, analysis of MADIT-II demonstrates that entry criteria were very broad and did not include any specific electrophysiological index of susceptibility to arrhythmia. Including such a broad population will likely have resulted in implanting devices into many who did not require intervention. This supposition is supported by the observation that the event curves did not separate until many months after randomisation. Therefore, some practitioners may still believe that the jury is still out regarding such a widespread application of implantable cardioverter defibrillators and await further clarification of optimal use of this valuable intervention. In this regard the recent paper by Bloomfield et al. on the potential role of microvolt T-wave alternans may allow for more specific direction of this intervention 16. Heart failure clinical trials are characterised by a design that dictates that investigational drugs and devices should be assessed in the setting of standard optimal therapy. This approach has been used ever since the demonstration of benefit of ACE inhibition in this syndrome. As a result investigational agents had to prove their worth in circumstances where benefit had already been obtained by ACEI. In this setting, many trials have demonstrated the benefit of beta blockers, altering background therapy in subsequent heart failure studies to the combination of ACEI and BB. This has made it even more difficult for newer agents to demonstrate benefit. Moreover, this trial design propagates polypharmacy and increases the possibility of drug interaction and also assumes that newer agents would not prove to be superior as front-line agents compared with either ACEI or BB. The increasing number of “drug trial failures” of late may reflect this somewhat cumbersome approach, and worryingly may consign potentially effective therapies with sound theoretical basis for benefit, to the outbox of ineffective therapies. Examples of this problem include the recent trials with endothelin antagonists and the vasopeptidase inhibitor, both groups failing to show benefit in the setting of mandated background therapy 17,18. Guidelines reflect this layered approach to therapy. Accordingly, for example, it is suggested all patients with left ventricular systolic dysfunction with or without heart failure should be initially treated with an ACEI before consideration of BB, and that the ACEI be titrated to clinical trial dose. This guideline may not sit easily with practitioners all of the time. For example, initiating and titrating an ACE inhibitor before prescribing a BB may make it impossible in some circumstances to subsequently prescribe a beta blocker as a result of hypotension. In some clinical settings a practitioner may feel that a beta blocker should be considered at the outset, going against guideline recommendations. Indeed, one could develop a hypothesis that beta blockade may be as effective a first choice agent, noting that these agents have an excellent record in the post-infarct setting with reduced left ventricular systolic dysfunction. Unfortunately, this hypothesis has not been adequately challenged. While the CARMEN study may be interpreted as supporting first-line therapy with beta blockade it should be remembered that the more impressive regression of remodelling by monotherapy with BB compared with ACEI therapy may be explained by the fact that many of the patients received ACEI up until randomisation 19. The ongoing CIBIS-III trial will provide further insight into this issue by directly comparing ACEI monotherapy with beta blocker monotherapy. Furthermore, the spate of negative trials suggests that we should be looking for an alternative approach to assessing new therapies. Already, we have had some head-to-head trials, comparing monotherapy with ACEI with Angiotensin Receptor Blockade 20 and also comparing beta blocker therapy with ACEI therapy and their combination in CARMEN 19, a trial with a remodelling end-point. While small in number, these trials have at least broken the mould and will continue to provide some important insight into the relative merits of various therapies. However, an even more desirable change to pharmacotherapeutics in this syndrome would be to investigate individualisation of therapy. It is somewhat naïve to believe that the same regimen outlined by guidelines would be suitable for all patients. Indeed, it is reasonable to suggest that choice and dose of agent may vary depending on many factors including aetiology, age, gender, renal function, neuroendocrine profile, etc. Furthermore, the increasing research interest in genetic influences on metabolism and effect of pharmaceutical agents may also define what is an appropriate “cocktail” in a specific individual 21 22 23. It is also possible that such an analysis may alter the “first choice cocktail” for the same patient over time. As outlined, this approach may seem overly complex and theoretical. Nonetheless, some approach to individualisation is needed. At present, there have been preliminary attempts at this, using B-natriuretic peptide (BNP) as a biochemical marker of response to therapy 24. If proven to be of benefit one could anticipate a situation where normalisation of BNP occurs with ACE inhibitor or BB alone, obviating the need for add-on therapy. At present large-scale clinical trials are further addressing these issues. The majority of heart failure trials have assessed just one dose of the investigational drug. It is understandable that the study design will not risk under-dosing and will opt for a large dose of the investigational agent titrated upwards in the early weeks of the each patient's involvement. If the trial is positive we are left with the unanswered question of whether a smaller dose would be as effective, which if the case would possibly reduce exposure to dose-related side-effects. As a result, guidelines advocate titration to target doses, reflecting doses proven to be effective in clinical trials. Data are available suggesting that higher doses of ACEI may be more effective in reducing hospitalisation, not improving outlook 10. Retrospective analysis of data from some beta blocker trials suggests that dose may not be that relevant. Bisoprolol dose did not appear to influence outcome in CIBIS-II 25. Likewise, metoprolol dose in MERIT-HF did not affect outcome 26 and while data from the US Carvedilol Heart Failure Program did indicate an impact of dose on ventricular function, it should be noted that this information comes from a relatively small group of patients 27. Achieving a BB effect as measured by heart rate and blood pressure may be more meaningful than dose achieved. As mentioned above, the trial design in heart failure, while productive in many respects, has not provided any insight into how to individualise therapy. This issue is coming to the forefront as we realise that there is a limit to the number of agents that patients with heart failure can tolerate. Without information on how to individualise therapy we are presently left with no clear knowledge as what should be added to the combination of ACEI and BB to reduce morbidity in those with persistent symptoms but no overt congestion. The present data and guidelines cannot be precise in this regard as there are no trials assessing the relative merits of spironolactone, angiotensin receptor blockers or other therapies such as the combination of nitrates and hydralazine. As mentioned RALES does not help us accurately place spironolactone in modern day therapy of combination ACEI and BB. We now have data from the CHARM study demonstrating morbidity benefit from the addition of an angiotensin receptor blocker 28. Less conclusive data on this issue come from Val-HeFT where retrospective analysis of those on background ACEI and BB suggested potential harm from “triple therapy”, though it must be stressed that that this was a post hoc analysis and therefore of questionable significance 29. Nonetheless, it does raise the issue that multiple vasoactive therapies may have drawbacks in certain circumstances. Our own experience in monitoring 24-h ambulatory blood pressure recordings in patients on escalating numbers of vasodilator agents does highlight a concerning degree of systemic hypotension which could be of potential harm, especially where coronary perfusion is important 30 (fig. Fig. 1). Therefore, there may be some doubt about how and when to apply a third-line agent and which agent to chose. While the few data available indicate potential for further symptomatic improvement, and possibly improvement in prognosis in African-Americans with the addition of hydralazine and nitrates, the information to hand at present does not allow for the development of precise guidelines 31. Furthermore, it is not clear whether cardiac resynchronisation therapy in patients with proven dyssynchrony is a better approach than a third-line pharmaceutical agent. This decision on additional strategies underlines the importance at looking critically at individualisation of therapy. One of the main purposes for clinical trials is to prove benefit of agents using certain established parameters. Consequently the primary end-points in heart failure studies are mortality and heart failure related morbidity, the latter usually measured by hospitalisation. These end-points are also generally accepted in clinical practice as being the most meaningful. However, in certain circumstances, these end-points may not be that important. For example prognosis is not a major concern in certain subsets of the heart failure population such as the very elderly, those with refractory symptoms from heart failure for whom transplantation or device implantation is not an option, or in individuals with terminal malignancy. Indeed, many such patients would often accept heart failure therapy which would sacrifice duration of life-span for adequate symptom relief 32. Unfortunately, while agents with such a profile have been defined, their development has been stopped primarily as a result of commercial concerns. A potential niche role for their use is not alluded to in guidelines, often leaving practitioners with fewer options to treat this small percentage of heart failure patients who need terminal care. The major cost associated with heart failure care is hospitalisation 33. Consequently reduction in hospital admission for heart failure remains a critical assessment of any new therapy. Indeed, this end-point could influence choice of agent. However, the relative efficacy of various interventions in this regard has not been compared, predominantly because every new therapy is assessed in the context of additional effectiveness it may provide to standard background therapy. Therefore, it is difficult to comment on relative cost-effectiveness of therapeutic interventions. In terms of cost savings it is likely that the structure of care will be of greater importance than the actual agents chosen, as evidenced by the significant reduction in hospital readmission achieved with multidisciplinary care 34. Pharmacotherapy and device-based therapy is becoming increasingly complex, producing continued incremental benefit in clinical trial populations. Translating these benefits into the routine community heart failure population has proven difficult. Development of guidelines has helped address this discrepancy but reasons as outlined above, other than lack of awareness of benefit, can lead physicians to deviate from suggested protocols in an attempt to apply the outcome of clinical trials in the most suitable manner to the individual patient.