Despite recent advances in therapy, heart failure (HF) remains a major cardiovascular epidemic worldwide. The increasing prevalence of the syndrome demands further improvement in practice and management to curb the economic and morbidity burden.
Current Dilemmas in Heart Failure Diagnosis and Care
Heart failure is a condition characterised by non-specific symptoms and unremarkable physical examination, which is why the medical need exists for a widely available objective marker of HF status. Such a marker could be of particular use in several settings, including defining those at risk for HF and providing objective analysis of status in people with established HF presenting with possible clinical deterioration. Natriuretic peptides (NPs) might be such a marker. This article will review how NPs may aid in addressing these dilemmas in present-day HF management.
At-risk Staging for Heart Failure
A workable prevention strategy for HF would have a dramatic impact on the natural history of this syndrome, but the evolution of this approach has been slow, reflecting concerns about how to proceed. In general terms, an effective screening programme for classifying at-risk status for HF should begin by documenting, if asymptomatic left ventricular systolic dysfunction (ALVSD) is present. ALVSD is a clearly defined at-risk stage for HF, and it has a recognised treatment that has been shown to be effective. NPs comprise an easy-to-use screening tool to curtail excessive and sometimes inappropriate use of Doppler-echocardiography. Recent work on focused screening, based on signs and symptoms with NP assessment in those with at-risk characteristics for HF (such as ischaemic heart disease, hypertension and diabetes), would appear to be a more effective strategy. While false positive rates remain with this approach, further work is needed to define the true meaning of these observations, such as epidemiological studies1 that have highlighted the prognostic impact of a modest elevation in NP blood levels. Many people with established ALVSD are not on disease-modifying therapies; therefore, focused use of NPs may be the most effective way of developing a HF prevention strategy. Figure 1 illustrates a proposed algorithm for more accurate diagnoses of new-onset HF based on elevated BNP blood values, which could also potentially reduce referral for more costly echocardiography.2
Clinical Deterioration in Established Heart Failure
One of the classical features of established HF are the inevitable episodes of clinical deterioration that can worsen to the extent that hospitalisation or emergency room visit is required. While the prevalence of this condition is uncertain, our own experience from survivors of a hospital admission for HF indicated that 15% of people will experience one such episode every three months after discharge, as displayed in Figure 2.
Indeed, these data also demonstrate that the major morbidity for people with HF, post discharge from hospital, is readmission to hospital for non-HF causes, reflecting the co-morbidities characteristic of an HF population. The essence of good clinical management of this problem is early identification of emerging clinical deterioration, distinguishing it from other non-HF causes of declining wellbeing, and to allow for prompt initiation of therapies to stabilise the person’s condition and prevent hospitalisation.
At present, early identification of emerging clinical deterioration is a difficult task. Body weight measurement, the standard early warning system in use, has been shown to be insensitive in this regard.3 While the same study indicates that NPs may also lack sensitivity, it should be noted that in contrast to the assessment of weight change data, change in NP plasma concentration was often calculated from a remote stable NP measurement, therefore potentially compromising its assessment. Additional precision/information is needed in this important area. More frequent assessment of NP levels may provide a critical early warning signal for emerging deterioration.
Natriuretic Peptide Use in the Community
While the major focus for use of NPs has been in the ER and in people hospitalised with acute decompensated HF, there are established and emerging uses for these peptides in the management of HF in the community. These include the role of NPs at the time of possible new presentation of HF, its role in prognostication and, finally, the increasing interest in using NPs to guide therapy in the outpatient setting.
New Diagnosis of Heart Failure
The initial presentation of HF is often non-specific and can be difficult to diagnose. Subtle dyspnoea or fatigue can be due to a variety of illnesses. What is most needed to aid diagnosis in such circumstances are tests which can either effectively rule out HF, allowing focus to be put on other potential diagnoses, or alternatively, encourage further investigation of possible HF.
It is this setting where BNP is most effective. Normal values for BNP, less than 50–100pg/ml, have excellent negative predictive value (NPV) in excluding HF as a diagnosis, especially when combined with a normal ECG. BNP values that are significantly elevated, such as greater than 500pg/ml, make the diagnosis of HF more likely and should initiate on-going procedures such as Doppler-echocardiography. Values in the ‘grey zone’ need further work up. Caution should be exercised in interpreting BNP levels owing to the fact that values can be influenced by factors such as age, renal function, body mass index (BMI) and pharmacological therapies. Use of BNP in this new diagnostic setting can be very effective, avoiding unnecessary more expensive tests such as echocardiography.
Using NPs to aid in formulating prognosis in people with HF in the community, is presently a less common use. Nonetheless, its value in this situation has been well established, both in those with stable HF and in those discharged from the hospital. In these reports, NPs were shown to be an independent predictor of outcome, including likelihood of hospitalisation. More widespread use of this observation could have significant resource implication. For example, on discharge from the hospital, the likelihood of a recurring clinical event is significantly elevated when BNP concentration is >250–300pg/ml. This should encourage triage of this group to more close clinical follow-up under specialist cardiology review, whereas a less intensive follow-up strategy could be advocated for those whose BNP concentration at discharge is <100–150pg/ml. Similarly, persistent elevated BNP in a person with stable HF points to a more complicated course and may encourage change in therapy despite clinical stability and continued close monitoring by specialist services.
Guided Therapy in Chronic Heart Failure
Studies have shown that increases in LV filling pressure start before people develop symptoms or signs of HF. These results indicate that people have already had haemodynamic characteristics of depressed cardiac function for a period before medical treatment is started.4
Based on these observations, it can be argued that clinical symptoms and signs on their own are not the optimal way to guide treatment. NPs (BNP or NT-proBNP) are highly correlated to the LV filling pressure4 and their measurement might be a useful tool to guide the medical treatment of people with HF.
NPs have been shown to provide important prognostic information for cardiovascular mortality.5,6 Additionally, the plasma concentration of the NPs is lowered with evidence-based pharmacological treatment of HF.7 A logical conclusion would be to use the plasma concentrations of NPs as a tool to tailor optimal treatment for people with HF. This approach was first demonstrated by Troughton, et al.8 several years ago in a small study showing borderline beneficial results. This concept has been tested in several larger studies and in different patient populations.
Two meta-analyses have been published in order to evaluate the impact of biomarker-guided therapy on mortality in people with HF.14,15 Most of these studies have been performed in people with systolic heart failure and with a dominating ischemic aetiology (around 60 %). Only two studies have included people with HF with preserved systolic function (Lainchbury11 and Eurlings12) and in those two studies the mean ejection fraction (EF) was 38% in contrast to the other trials where it was mostly <30%. However, the results from this group of people have not been separately presented. In most of the studies, all included people with HF have been well-treated and in most of the people in the group guided by NPs, the treatment has been significantly improved during the trial. The studies have had different primary endpoints and only two of them (Troughton8 and Jourdain9) have shown significant improvement of their composite primary endpoint (p<0.05). Two other studies (Pfisterer10 and Lainchbury11) had a stratified randomisation according to age (<75 years versus >75 years) and also pre-specified subgroup analyses based on those age groups. This subgroup analysis in both studies demonstrated beneficial effects in terms of reduction in mortality and morbidity in the younger people but not in the older. This has raised the question whether older people (the majority of people with HF) are receiving any beneficial effects from NP-guided treatment. One explanation might be that there is more structural heart disease and stiffer myocardium in older people. Only one trial has included older people in the community (Persson13) and the results from this trial are not different from the others.
However, in three of the six trials presented at major cardiovascular meetings, people have also been divided into responders (>30% reduction of the NP concentration during the study) and non-responders (<30% reduction of the NP concentration). Based on this pre-specified subgroup analysis in all three studies, the responders demonstrated a lower risk for the primary endpoint, a composite of mortality and morbidity.
This result is in line with studies showing that people hospitalised with HF with a reduction in plasma concentration of NPs had a better outcome compared with those without a corresponding response even if the two groups were treated similarly. Thus the mechanisms of response by lowered NP concentration seem to be a property that is central in those people. Similar results were shown in a post-hoc analysis of the Cohen-Solal trial.16 Beyond differences in age, more work is needed to clarify why people respond to NP-guided therapy versus those who do not. Based on these differences in people’s responses, perhaps different therapies may be required for different people.
Currently, we do not have firm evidence from prospectively-defined published trials that the addition of NP-guided therapy is beneficial. Both meta-analyses have concluded that NP-guided therapy is associated with a significant reduction of all-cause mortality compared to usual care in people with HF and, in the latest one,15 the investigators found that this significant outcome is especially true in people younger than 75 years. However, a large study powered to evaluate mortality is still needed before incorporating NP-guided therapy as routine treatment. Furthermore, in most studies, people with HF and normal EF have not been included so we do not know if this group would benefit from guided therapy. Moreover, the initial findings that responders might have a better outcome compared to non-responders needs to be further investigated in future studies.
Six published studies have evaluated the impact of biomarker-guided therapy on mortality outcome8–13 and an additional three have only been presented at major cardiovascular meetings (the Strategies for tailoring advanced heart failure regimens in the outpatient setting: brain natriuretic peptide levels versus the clinical congestion score [STARBRITE] study, the Pro-B type natriuretic peptide outpatient tailored chronic heart failure therapy [PROTECT-AHA 2010] study and the [Use of peptides in tailoring effective heart failure [UPSTEP-AHA 2009] project). In each of the six studies, the point estimate for all-cause mortality favoured the biomarker-guided strategy over control. The overall hazard ratio for mortality was 0.69 (95% CI 0.55–0.86). The risk ratio showed a statistically significant reduction in mortality with biomarker-guided therapy,15 ranging from 0.62 to 0.71. Figure 3 displays the relative risk of all-cause mortality for people with CHF randomised to biomarker-guided therapy versus control14 in each of the six studies individually and when combined. These trials have been small, including between 69 and 499 people, and are therefore underpowered to provide definitive evidence of benefit in hard endpoints as mortality.
Potential Uses of Natriuretic Peptides
Several studies have demonstrated the powerful role NPs have in the new diagnosis of HF and also in assessing the prognosis of people with HF.17 Despite this strong evidence, it is uncertain if NPs are widely used in the routine management of people with HF worldwide. As outlined above, risk-stratification based on the use of NPs would allow focused and more aggressive treatment of people with a poor prognosis, and less intensive treatment in those with low levels of NPs, assuming they have a more favourable outcome.
Another interesting area is HF treatment guided by NPs. As already described, several studies have been performed in this area, with conflicting results. A large prospective controlled study, statistically powered to look at hard endpoints such as mortality and morbidity, is required before we reach any consensus on this approach.
Our goal in the management of people with HF is to reduce symptoms, reduce hospitalisation, increase survival and finally, if possible, prevent the development of HF. Screening people for ALVSD by the use of NPs may well be an effective approach to help prevent HF. It would be a great success if the use of moderately elevated NP levels could accurately and cost-effectively predict future cardiovascular events in those people. If this is true, it would expand the field enormously in how to use NPs alone or in combination with other biomarkers in order to predict future cardiovascular complications.
The new area of distributed healthcare has been discussed, during the last few years, moving the management of people with HF to their homes by use of technical solutions with sensoring techniques. Telemonitoring, one such technical solution, has been evaluated in many studies and again with conflicting results.18 However, adding the value of serial measurement of NPs to information of clinical parameters, such as changes of weight, blood pressure, heart rate and symptoms, might help the clinician to monitor people in a more effective way; and may, in time, detect features of worsening HF in need of intensification of medication. Most of this work can be performed at a distance by transmitting the data by Bluetooth technology, from the person to the heart failure clinic, thereby reducing unnecessary visits to the clinic. This is one way to move the heart failure clinic to people’s homes. The concept, however, must be carefully evaluated in future prospective studies. Ongoing work assessing the role of frequent home measurement of NP in people who are stable will provide us with further insight into whether this improves our ability to manage people in their homes.
NP assessment in the management of people with, or suspected of, HF is now an accepted, necessary tool in medical and clinical practice. Based on present research and utility of NPs, the possibilities for future use of this tool should continue to expand.