Sudden Cardiac Death Risk Stratification in Heart Failure -The Potential Role of Biomarkers Part 2

Serum Biomarkers to Predict Sudden Cardiac Death in Heart Failure

The value of serum biomarkers to predict SCD in heart failure has been specifically evaluated in two prospective studies (Table 1) [83,84]. One enrolled patients with chronic heart failure of ischaemic and non-ischaemic aetiology [83], and the other post-MI patients [84]. Both demonstrated a significant association between a single serum biomarker measurement and subsequent SCD risk.

Berger et al examined the association of 4 serum biomarkers - BNP, N-terminal BNP (NT-BNP), N-terminal atrial natriuretic peptide (NT-ANP), and big endothelin – with SCD in 452 ambulatory patients with heart failure and LVEF <35% [83]. The aetiology of heart failure in the majority of these patients (65%) was non-ischaemic. During follow-up (592+/-387 days) there were 89 deaths of which 44 were sudden. Using univariate analyses the only significant predictors of sudden death were log BNP (p=0.0006), log N-ANP (p=0.0028), LVEF (p=0.0054), log N-BNP (p=0.0057), systolic blood pressure (p=0.0138), big endothelin (p=0.0326), and NYHA class (p=0.0375). However in multivariate analysis only log BNP (p=0.0006) was still significantly associated with SCD. The use of specific cardiac medication including beta-blockers, ACE-I and amiodarone, as well as the presence of IHD and diabetes, were not predictive of SCD.

Tapanainen et al prospectively evaluated the accuracy of plasma ANP, N-ANP, BNP and depressed LVEF in predicting SCD in 521 survivors of acute MI [84]. During a mean follow-up of 43 +/-13 months there were 33 deaths of which 16 were due to SCD. In univariate analysis, BNP (relative risk 4.4, p=0.011), ANP (RR 4.1, p=0.014) and N-ANP (RR 3.4, p=0.018) had similar accuracy as LVEF (RR 4.9, p=0.013) in predicting SCD. In multivariate analysis, after adjusting for clinical variables, only elevated BNP (p = 0.02) and low LVEF (<40%) (p = 0.03) remained as significant predictors of SCD. It should be noted that there was a high use of contemporary post-MI medical therapy in the cohort, including 97% beta-blockade.

Serum Biomarkers to Predict ICD Discharges

Implantable cardioverter defibrillators are extremely effective in terminating episodes of ventricular fibrillation (VF) and ventricular tachycardia (VT) that may otherwise have led to SCD. Therefore evaluating the relationship of biomarkers to SCD in patients with ICDs is potentially difficult. In addition to this therapeutic role however, ICDs also accurately record the occurrence of these malignant arrhythmias and the treatment given by the device, termed anti-tachycardic therapy. Thus the incidence of potentially life-threatening arrhythmias, as determined by device interrogation, may be used as a surrogate marker of SCD in these patients.

Table 2. Studies evaluating the association of serum biomarkers with malignant ventricular arrhythmias in ICD recipients.

IHD, ischaemic heart disease; NICM, non-ischaemic cardiomyopathy; BNP, brain natriuretic peptide; CRP, C-reactive protein; NT-proBNP, N-terminal pro-brain natriuretic peptide; PINP procollagen type I aminoterminal peptide; PIIINP, procollagen type III aminoterminal peptide, TIMP1, membrane metalloproteinase I; IL-6, interleukin 6; TNF-a, tumour necrosis factor alpha; VT, ventricular tachycardia; VF, ventricular fibrillation.

The ability of a range of serum biomarkers to predict malignant arrhythmias in ICD recipients has been assessed in 8 studies, enrolled 890 patients (Table 2) [85-92]. These studies, 6 of which were prospective, enrolled patients with both ischaemic and non-ischaemic heart failure. All but one study found biomarkers were able to predict the occurrence of malignant ventricular arrhythmias. The only study with negative findings was small and examined only 50 patients over 6 months [92].

Six studies have investigated BNP (or N-terminal pro-BNP) and demonstrated it independently predicts malignant arrhythmias in patients with ICDs [85-90]. Three of the larger studies reported patients with BNP levels over the 50th centile had significantly more malignant arrhythmias (risk ratios between 2.2 and 3.8) [86,88,89]. Multivariate regression analyses in these studies examining traditional clinical and echocardiographic risk factors for SCD, found BNP most strongly predicted malignant arrhythmias and performed better than reduced LVEF.

Two studies investigated a broader range of serum biomarkers. Blangy et al prospectively evaluated markers of cardiac fibrosis [procollagen type I aminoterminal peptide (PINP), procollagen type III aminoterminal peptide (PIIINP), membrane metalloproteinase I (TIMP1)], myocardial pressure overload [brain natriuretic peptide (BNP)] and inflammation [high sensitivity (hs)-C-reactive protein] [88]. They observed 121 patients with IHD over 12 months. During this time 38 patients had appropriate device therapy for VT. In a multivariate analysis, LVEF <0.35 (OR = 2.19, P = 0.049), an increased serum BNP (OR = 3.75, P = 0.014), an increased hs-C-reactive protein (OR = 3.2, P = 0.006), an increased PINP (OR = 3.71, P = 0.009), and a decreased PIIINP (OR = 0.21, P = 0.003) were associated with a higher VT incidence. Biasucci et al studied 65 patients and confirmed the association with hsCRP [91].

One study has compared the predictive value of N-terminal pro-BNP (NT-pro-BNP) to the gold-standard of EPS [89]. Yu et al prospectively studied 99 patients with ICDs for prevention of SCD following MI. EPS and measurement of NT-pro-BNP were performed at study entry. During a mean follow-up of 556 (+/-122) days 23 patients received appropriate device therapy for VF/VT. On multivariate Cox regression analysis, only NT-pro-BNP level at or greater than median (497 ng/L) was a significant predictor for VT/VF occurrence (p=0.047). Neither univariate or multivariate analysis demonstrated any relationship between inducibility at EPS and the study end-points.

Serum Biomarkers to Guide ICD use?

Multiple studies have demonstrated that serum biomarkers can accurately predict adverse outcomes in patients with heart failure and asymptomatic left ventricular dysfunction of both ischaemic and non-ischaemic aetiology. A range of biomarkers predict both the occurrence of SCD in patients without ICDs and the occurrence of malignant arrhythmias in patients with devices (Table 3). In these studies individual biomarkers are at least as good as the current best marker of SCD risk, depressed LVEF. In the only trial to compare biomarkers to electrophysiological testing, serum NT-BNP was considerably more accurate than EPS in predicting malignant arrhythmias [89].

As predictive tests, biomarkers have significant advantages over current tools. Assessment of LVEF can be expensive, if performed by the gold-standard magnetic resonance imaging, and inaccurate, if performed using two-dimensional transthoracic echocardiography. EPS is expensive, invasive, associated with small but important risks to the patient, and often only available in larger cardiac centres. Ambulatory monitoring, to look for spontaneous ventricular arrhythmias, is not particularly reproducible [93]. In contrast, biomarker measurement is simple, relatively inexpensive, reproducible, and without direct patient risk.

Table 3. Biomarkers demonstrated to predict the occurrence of sudden cardiac death or ventricular arrhythmias in patients with heart failure. 

IHD, ischaemic heart disease; NICM, non-ischaemic cardiomyopathy.

The genesis of ventricular arrhythmias that lead to SCD is a complex process requiring the presence of both an abnormal myocardial substrate, needed to initiate and sustain an arrhythmia, and pro-arrhythmic triggers [11]. A range of electrophysiological and molecular alterations contribute to arrhythmogenesis in the failing heart, including changes in ion channel expression and neurohormonal modulation, and serum biomarkers may provide an assessment of these various processes. It is likely that optimal SCD risk stratification will require the combination of multiple tests that reflect these diverse upstream processes. As such the greatest potential benefit of biomarkers may be in measuring multiple complementary markers that assess distinct aspects of arrhythmic risk, or in combining biomarkers with traditional risk stratification tools. Currently there have been no studies evaluating this.

Conclusion

Despite the availability of a number of well characterised tests, risk stratification of SCD in patients with heart failure is currently sub-optimal. The value of serum biomarkers in cardiovascular disease is well established. There is increasing data to suggest that individual serum biomarkers predict SCD at least as well as established risk stratification tools in heart failure patients. Biomarkers are available that provide an assessment of the diverse pathophysiological processes that are central to ventricular arrhythmogenesis, including myocardial stretch, inflammation, and neurohormonal activation. There is therefore significant need for further studies to evaluate the potential role of biomarkers, individually or in combination, in patient selection for ICDs.