Biomarkers to Diagnose and Determine Risk

General description of procedure, equipment, technique

What is a biomarker?

The NIH defines a biomarker as “a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or the response to a therapeutic intervention.” Based on this definition, many commonly used clinical assessments (e.g., blood pressure, heart rate) would qualify as biomarkers.

For the purposes of this chapter, we will assume that the term “biomarker” refers to a biologic characteristic that can be measured in biologic samples (e.g., blood, plasma, urine) and used in the clinical care of patients.

Attributes of an ideal biomarker

Morrow and de Lemos have proposed a set of criteria by which biomarkers should be evaluated, which are shown below:

Criteria for Evaluation of New Biomarkers

• Can the clinician measure it?

• Accurate and reproducible methods

• Rapid turn around

• Reasonable costs

• Does it add new information?

• Strong and consistent association between marker and outcome or disease of interest in multiple studies

• Decision limits are validated in generalizable populations

• Will it help with management?

• Superior performance to existing tests

• Evidence that it enhances outcomes or process of care

Indications and patient selection

Potential biomarkers in heart failure

A large number of potential biomarkers have been identified or proposed to be useful in heart failure, a comprehensive listing and review of which would be beyond the scope of this chapter. Below is a brief summary of the rationale and biology behind some of the most commonly used and emerging biomarkers in heart failure.

Natriuretic Peptides

The natriuretic peptide family is composed of three molecules that share a common 17 amino acid ring and similar biology: A-type natriuretic peptide (ANP) is secreted primarily by the atria, B-type natriuretic peptide (BNP) is secreted primarily by the ventricles, and C-type natriuretic peptide (CNP) is secreted by endothelial cells. All three molecules share similar physiologic effects, including balanced (arterial and venous) vasodilation, lusitropy, antifibrotic, and pronatriuretic effects.

In general, the physiologic effects of the natriuretic peptides can be seen as counteracting the effects of chronic neurohormonal stimulation that is a central component of the heart failure syndrome. By far the most well studied and characterized natriuretic peptides in heart failure are BNP and its amino terminal fragment NT-proBNP.

BNP is secreted from ventricular myocytes in response to myocyte stretch, typically from pressure or volume overload. BNP is secreted from myocytes as a prohormone (proBNP), which is then cleaved in the plasma into an active peptide (BNP) and a biologically inert N-terminal fragment (NT-proBNP) (Figure 1). Assays of both BNP and NT-proBNP are clinically available, and include both laboratory based and point of care assays.

Figure 1.

These molecules differ with regard to their half-life in the plasma, their mode of clearance, and the effect of age, gender, renal function, and concomitant medications, their “normal” values, and the effect of some medications as discussed further below. Although these assays appear to provide similar clinical information, the differences between the assays make it critical to specify the assay being used when using these measurements clinically or when interpreting published studies. Differences between BNP and NT-proBNP are summarized in Table I.

Table I.n

Levels of both molecules are elevated with aging, and are higher in women than in men. Obesity has also recently emerged as another possible confounder of natriuretic peptides measurements, with lower natriuretic peptide levels observed with a progressively higher body mass index. The recent development and approval of valsartan/sacubitril (Entresto) as a chronic therapy has introduced additional complexity into the interpretation of natriuretic peptide values. This therapy increases circulating levels of BNP (presumably through inhibition of the enzyme neprilysin, which is responsible for the degradation of BNP) but does not affect levels of NT-proBNP. The extent to which this treatment may require a reassessment of current decision thresholds for BNP is an area of active research.

A common clinical question without a simple answer is “how can I convert between BNP and NT-proBNP results?” In general, NT-proBNP values are between 3- and 10-fold higher in a given patient than BNP values, although this relationship is altered by a variety of factors such as age, gender, and renal function. Broadly speaking, the multiple between BNP and NT-proBNP is increased in older patients and in patients with more advanced disease.

Contraindications

Troponins

Death of cardiomyocytes represents a common “final common pathway” for progression of heart failure. Although classically associated with acute coronary syndromes, it is now clear that troponin elevation occurs in heart failure, even in the absence of clinically overt ischemia or epicardial coronary artery disease.

Troponin elevation is now recognized to be frequent in patients with heart failure, and is more common in patients with advanced heart failure or acute heart failure. The mechanisms of troponin elevation in HF include both traditional ischemic injury as well as nonischemic mechanisms, such as alterations in calcium handling, loss of cell membrane integrity, apoptosis, and excessive wall tension. The prevalence of troponin elevation in HF patients depends on the patient population and the assay used, but data with contemporary assays suggest that circulating cardiac troponins are detectable in the vast majority of patients with acute heart failure and a substantial portion of patients with chronic HF. As the sensitivity of available assays increases, it appears likely that troponin will increasingly be viewed as a continuous variable that is predictive of risk (like BNP) rather than a dichotomous variable (i.e., “positive” or “negative”).

ST2

ST2 is a member of the interleukin 1 (IL)-1 receptor family. Research in animal models has shown that the ligand for ST2 is IL-33, which is also released by cardiac fibroblasts during myocardial stretch. IL-33/ST2 signaling is a mechanically activated cardioprotective signaling system that acts to shield the myocardium against the adverse effects of volume and/or pressure overload.

Circulating ST2 appears to act as a decoy receptor to modulate IL-33/ST2 signaling. Elevated plasma levels of ST2 are strongly associated with HF severity and increased risk of death or transplantation in patients with chronic heart failure. ST2 has also been shown to be a potent marker of risk in acute heart failure.

A recently developed high-sensitivity ST2 assay will allow for detection across lower serum concentrations, which may potentially expand its use to less severe forms of heart failure. ST2 may also have potential uses in targeting therapies to patients most likely to benefit. Data from Weir et al in postmyocardial infarction patients has shown that elevations of ST2 identify a group more likely to respond to the beneficial anti-remodeling effects of eplerenone.

Galectin-3

Galectin-3 is a member of a family of soluble β-galactoside-binding lectins that are expressed in many tissues and play a fundamental role modulating fibrosis. In chronic HF, galectin-3 augments fibrosis and modulates immune response, both pivotal processes in maladaptive cardiac remodeling.

In animal models, galectin-3 appears to be a mediator (rather than just a marker) of HF progression. Galectin-3 levels are elevated in patients with chronic HF and provide important independent prognostic information in some studies.

Unlike the natriuretic peptides or ST2, galectin-3 levels do not change significantly in response to acute decompensation. Some preliminary data have suggested that elevated galectin-3 may identify a specific “sub-phenotype” of heart failure patients that may be less responsive to traditional therapies, although these observations will need to be confirmed in additional datasets before they can influence clinical care.

Adrenomedullin

Adrenomedullin (ADM) is a novel biomarker that is released in response to hemodynamic stress. ADM is a vasoactive peptide that has complex biologic properties, but appears to be related to endothelial and vascular function. Due to the short half-life of biologically active ADM in the circulation, measurement of ADM activity is accomplished through the measurement of mid-regional proADM (MR-proADM), a by-product of ADM posttranslational modification.

MR-proADM has been shown in at least one large registry study (the Biomarkers in the Assessment of Congestive Heart (BACH) failure study) to be superior to natriuretic peptide measurement in the prediction of 90 day outcomes in patients presenting with dyspnea. Of note, the prognostic utility of MR-proADM also extended to patients who did not ultimately have heart failure as the primary diagnosis. Ongoing research will examine the optimal place for MR-proADM in heart failure management.

Growth-Differentiation Factor-15 (GDF-15)

GDF-15 is a member of the transforming growth factor cytokine superfamily. It is expressed in response to inflammation and tissue injury. Unlike the natriuretic peptides, GDF-15 is not cardiac specific and is produced by other cells including vascular smooth muscle and endothelial cells. Some data suggest that GDF-15 levels in heart failure patients are strong predictors of adverse outcome (death or hospitalization) even after adjustment for other known prognostic factors. As with other biomarkers, the precise role that GDF-15 may play in clinical practice remains to be clearly defined.

Interpretation of results

Use of biomarkers for diagnosing heart failure

Heart failure is a clinical syndrome rather than a specific disease, and as such there is no “gold standard” diagnostic test for the diagnosis of heart failure. Although in many cases the diagnosis of heart failure is clear, in other circumstances there can be substantial diagnostic uncertainty, since the cardinal signs and symptoms of heart failure (dyspnea, fluid retention, fatigue, and edema) can overlap with many other cardiovascular and noncardiovascular conditions. In the acute setting in particular, accurate diagnosis of heart failure is critical in order to initiate early treatment and to avoid unnecessary testing and interventions

By far the most extensively studied biomarkers for making the diagnosis of heart failure are the natriuretic peptides BNP and NT-proBNP. Current AHA-ACC guidelines give a class 1 recommendation (“should be performed”) for the use of NP measurement in the diagnosis of patients who present with unexplained dyspnea to acute care settings (such as the emergency department) as follows:

“Measurement of BNP or NT-proBNP is useful to support clinical judgment for the diagnosis of
acutely decompensated HF, especially in the setting of uncertainty for the diagnosis (Level of Evidence A)”

The rationale for use of natriuretic peptide testing in patients with acute dyspnea is that the vast majority of presentations for acute decompensated heart failure are related to elevated filling pressure, and there is a reasonably close relationship between elevated filling pressure and elevated natriuretic peptide levels in most clinical scenarios

For BNP, the most commonly proposed diagnostic cut point for BNP in the diagnosis of heart failure is 100 pg/ml. For NT-proBNP, distinct cut points for different age ranges are typically proposed (125 pg/ml for patients <75 years old, and 450 pg/ml for patients greater than 75 years old). The best choice of specific cut points for diagnosis may vary in different clinical scenarios.

In acutely symptomatic patients, sensitivity is more important, and false positives are less of a concern since further evaluation is likely to be performed. Conversely, in screening asymptomatic persons, specificity is most important, since false positives can generate a substantial burden of confirmatory testing in patients who do not have the disease.

A generally important feature of the use of natriuretic peptides for the diagnosis of heart failure is their high negative predictive value. A BNP or NT-proBNP value below the diagnostic cut point in a symptomatic patient rules out HF as a cause of acute symptoms with a high degree of accuracy (90% negative predictive value).

High values are suggestive of heart failure, but positive predictive value is generally lower since many other acute cardiovascular diagnoses (e.g., myocardial infarction (MI), pulmonary embolus) may lead to elevations of BNP. Very high levels of natriuretic peptides are highly predictive of heart failure as a cause of symptoms, since alternative diagnoses such as MI or PE do not generally tend to produce very high levels. Thus as a practical approach to interpreting natriuretic peptide values in acute settings is shown below:

• Very low values (BNP <100 pg/ml) strongly suggest HF is not an explanation for symptoms.

• Very high levels (BNP >500 pg/ml) strongly suggest HF as an explanation for acute symptoms.

• Intermediate levels (BNP between 100 and 500, the so called “grey zone”) are consistent with HF but also with alternate diagnosis.

Performance characteristics of the procedure

Important caveats (false positives and false negatives)

As with all diagnostic tests, false positives and false negatives are possible when using natriuretic peptides for the diagnosis of heart failure. Due to the high negative predictive value, false negatives are relatively rare, but one important source of false negatives is obesity.

Multiple studies have shown an inverse association between body mass index (BMI) and natriuretic peptide levels. Rarely, patients with sudden onset of increased filling pressures (such as acute mitral regurgitation) may present before natriuretic peptides become elevated, resulting in a false negative value.

False positives are generally more common, since many of the diagnoses that can be confused with heart failure also lead to elevations (generally of a lesser magnitude than in heart failure) or natriuretic peptide levels. Important examples include acute coronary syndrome and pulmonary embolus, both of which can be associated with elevations of natriuretic peptide levels.

As noted above, the recent introduction of valsartan/sacubitril (Entresto) into clinical clinical practice may require a reappraisal of the appropriate diagnostic thresholds for BNP (which is increased by this therapy) in patients taking this agent.

Alternative and/or additional procedures to consider

Using biomarkers for estimating prognosis in heart failure

There is a large body of literature linking levels of biomarkers to outcome in all forms of heart failure. In particular, the association between elevated natriuretic peptide levels and adverse outcomes is consistent across the disease spectrum of heart failure, including asymptomatic LV dysfunction, ambulatory heart failure patients, acutely decompensated patients in the hospital, and end-stage (stage D) patients.

In many predictive models, natriuretic peptides are the most powerful single predictor of outcome. For prognosis, the relationship between natriuretic peptide levels and outcomes is generally linear, and no specific “cut point” is used. In acutely hospitalized patients with heart failure, a fall of 30% or greater in natriuretic peptide levels is associated with a low post-discharge event rate, whereas failure of natriuretic peptide levels to decrease with acute therapy is generally associated with a high risk of rehospitalization or death pos-tdischarge.

The concept of using natriuretic peptide measurements to guide decisions about duration and intensity of therapy in hospitalized patients is attractive, but has not been proven in prospective studies. A common sense approach to using natriuretic peptides in patients hospitalized with heart failure would be to measure a value on admission and again when the patient appears to be approaching readiness for discharge. A failure of natriuretic peptide levels to decrease significantly with acute therapy should suggest the need for careful assessment of volume status with an eye towards potentially increasing the intensity of medical therapy prior to discharge.

As described above, a variety of other biomarkers, including troponins, ST2, galectin-3, and GDF-15 (among many others) have been associated with adverse outcomes in heart failure. Although some studies have suggested that these markers can add significant prognostic value to that obtained from clinical evaluation and natriuretic peptides, further research will be required to clarify the optimal use of these additional markers in heart failure.

Complications and their management

Biomarker guided therapy

Changes in natriuretic peptide levels over time are strongly associated with outcomes—patients in whom natriuretic peptide levels decrease in response to ongoing medical management have a better prognosis than patients in whom levels do not change or increase.

Additionally, heart failure therapies proven to have beneficial long-term effects on morbidity and mortality generally decrease natriuretic peptide levels over time. These observational data have suggested the hypothesis that serial measurements of natriuretic peptides may serve as a guide to the titration of medical therapy in HF-“biomarker guided therapy.”

Biomarker-guided therapy involves titration of chronic heart failure therapy to achieve a given level of suppression of natriuretic peptides (rather than simply targeting the doses studied in specific clinical trials). Thus, this is a form of “personalized medicine,” and can be considered analogous to the way we manage other chronic disease such as diabetes mellitus (using glycosylated hemoglobin, for example).

This concept has been tested over the last decade in multiple small randomized controlled studies. Two recent systematic meta-analyses both suggested that biomarker-guided therapy using natriuretic peptides in ambulatory heart failure patients was associated with a substantial improvement in long-term outcomes (particularly mortality).

Although provocative, these findings would need to be confirmed in larger randomized studies (currently being planned) before they could impact clinical practice guidelines and clinical care. At present, significant elevations in BNP and or NT-proBNP in apparently stable ambulatory heart failure patients should warrant careful evaluation for subclinical volume overload, as well as attention to appropriate up-titration of chronic neuro-hormonal antagonists (beta-blockers, ACE-inhibitors, and aldosterone-blockers.

Conclusions and take home messages

Biomarkers are useful clinical tools, which can provide objective, quantitative, reproducible physiologic information. Despite these obvious attractions, it is imperative that biomarkers (like all laboratory data) be interpreted in the clinical context.

Prior data have shown that the best result (e.g., the most diagnostic accuracy) is obtained by combining clinical evaluation and biomarker data, rather looking at either in isolation. Natriuretic peptides remain the best-studied and most proven biomarkers for making the diagnosis and assessing the prognosis of patients with heart failure.

A variety of other markers are emerging as potentially important additional markers in heart failure. “Multi-marker” approaches that combine information from multiple biomarkers (especially to the extent they provide non-overlapping information) will be increasingly employed as decision tools, especially as the migration to electronic medical records creates opportunities for “decision support tools” based on physiologic data to be provided at the point of care.

What's the evidence?

General Reviews

Braunwald, E. ” Biomarkers in heart failure”. N Engl J Med. vol. 358. 2008. pp. 2148-59.

Morrow, DA, de Lemos, JA. “Benchmarks for the assessment of novel cardiovascular biomarkers”. Circulation. vol. 115. 2007. pp. 949-52.

Ahmad, T, Fiuzat, M, Felker, GM, O’Connor, CM. “Novel biomarkers in chronic heart failure”. Nat Rev Cardiol. vol. 9. 2012 Mar 27. pp. 347-59.

Specific Biomarkers

Maisel, AS, Krishnaswamy, P, Nowak, RM. ” Rapid measurement of B-type natriuretic peptide in the emergency diagnosis of heart failure”. N Engl J Med. vol. 347. 2002. pp. 161-7.

Januzzi, JL, Peacock, WF, Maisel, AS. ” Measurement of the interleukin family member ST2 in patients with acute dyspnea: Results from the PRIDE (Pro-Brain Natriuretic Peptide Investigation of Dyspnea in the Emergency Department) Study”. J Am Coll Cardiol. vol. 50. 2007. pp. 607-13.

Januzzi, JL, Sakhuja, R, O’Donoghue, M. ” Utility of amino-terminal pro-brain natriuretic peptide testing for prediction of 1-year mortality in patients with dyspnea treated in the emergency department”. Arch Intern Med. vol. 166. 2006. pp. 315-20.

Felker, GM, Fiuza, M, Shaw, LK. “Galectin-3 in ambulatory patients with heart failure: Results from the HF-ACTION study”. Circ Heart Fail. vol. 5. 2012. pp. 72

Maisel, A, Mueller, C, Nowak, RM. “Midregion prohormone adrenomedullin and prognosis in patients presenting with acute dyspnea: results from the BACH (Biomarkers in Acute Heart Failure) trial”. J Am Coll Cardiol. vol. 58. 2011. pp. 1057-67.

Kociol, RD, Pang, PS, Gheorghiade, M. ” Troponin elevation in heart failure prevalence, mechanisms, and clinical implications”. J Am Coll Cardiol. vol. 56. 2010. pp. 1071-8.

Anand, IS, Kempf, T, Rector, TS. “Serial measurement of growth-differentiation factor-15 in heart failure: relation to disease severity and prognosis in the Valsartan Heart Failure Trial”. Circulation. vol. 122. 2010 Oct 5. pp. 1387-95.

Felker, GM, Fiuzat, M, Thompson, V. “Soluble ST2 in ambulatory patients with heart failure: Association with functional capacity and long-term outcomes”. Circ Heart Fail. vol. 6. 2013 Nov. pp. 1172-9.

Packer, M, McMurray, JJ, Desai, AS. ” Angiotensin receptor neprilysin inhibition compared with enalapril on the risk of clinical progression in surviving patients with heart failure”. Circulation. vol. 131. 2015 Jan 6. pp. 54-61.

Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.

No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.