Prediction of individual response to heart failure therapy.

Heart failure is associated with an adverse prognosis, and aggressive therapy is needed. In most patients, medical therapy is the cornerstone of treatment. Although the response to medical therapy is good in heart failure patients, benefit varies significantly among individual patients. This has been particularly demonstrated for beta-blocker therapy. Packer and colleagues evaluated .2000 heart failure patients, who were randomized to placebo or carvedilol. The authors reported a 35% decrease in the risk of death with carvedilol as compared with placebo; moreover, a 24% decrease in the combined risk of death or hospitalization was observed. Other studies demonstrated that beta-blocker therapy in heart failure patients was associated with an improvement in left ventricular ejection fraction (LVEF), although a substantial percentage of individual patients did not exhibit a significant improvement in LVEF (defined as an increase .5%). Data from the Carvedilol Hibernation Reversible Ischemia Trial: Marker of Success (CHRISTMAS) trial showed that the magnitude of improvement in LVEF after carvedilol therapy was related to the extent of viable tissue (on single photon emission computed tomography) in the left ventricle. Similarly, Bello et al. evaluated 45 heart failure patients who received beta-blocker therapy, with contrast-enhanced cardiovascular magnetic resonance (CMR) to assess the extent of scar tissue in the left ventricle. The authors demonstrated that significant improvement in LVEF (defined as .5%) after 6 months of beta-blocker therapy was not observed in 43% of the patients. Moreover, the extent of scar tissue on CMR was inversely related to the likelihood of improvement in LVEF. Another study in 43 heart failure patients demonstrated that the presence of contractile reserve on dobutamine echocardiography was associated with good response to beta-blocker therapy, and that the improvement in function occurred earlier after initiation of beta-blocker therapy in patients with idiopathic dilated cardiomyopathy as compared with patients with ischaemic cardiomyopathy. These findings provide some preliminary evidence for the possibility of identifying heart failure patients that may respond better or worse to beta-blocker therapy. The varying response in heart failure patients has also been observed in device therapy, in particular cardiac resynchronization therapy (CRT). In the Cardiac Resynchronization-Heart Failure (CARE-HF) study, 813 heart failure patients [New York Heart Association (NYHA) class III– IV] were randomized to optimized medical therapy or CRT. Over a follow-up period of 29.4 months, mortality was significantly lower with CRT as compared with medical therapy (20% vs. 30%, P , 0.002). Moreover, the combined endpoint of all-cause mortality or unplanned cardiovascular hospitalization was reached in 39% of patients undergoing CRT as compared with 55% of patients receiving medical therapy (P , 0.001). Also, patients with CRT exhibited an improvement in LVEF, with a reduction in left ventricular (LV) endsystolic volume and mitral regurgitation, as well as a reduction in symptoms and an improvement in quality of life. Based on the CARE-HF trial and other studies, severe heart failure patients with wide QRS complex (≥120 ms) and depressed LVEF (≤35%) have a class I indication for CRT; more recently, these guidelines have been updated based on new evidence, now also including patients with less severe heart failure as candidates for CRT. Despite these selection criteria, the individual response to CRT varies significantly, with a relatively high percentage of nonresponders, depending on the definition of non-response. In general, absence of reverse LV remodelling is observed more often than absence of improvement in heart failure symptoms. This has prompted an extensive search for prediction of response to CRT. Various variables have been proposed, e.g. the QRS duration, cardiac dyssynchrony, and the presence of extensive scar tissue. For example, patients with a QRS duration .150 ms tend to respond better to CRT as compared with patients with a QRS duration between 120 and 150 ms. Many studies have used cardiac dyssynchrony (predominantly assessed by