There are no parameters predicting the individual probability of “full response” to cardiac resynchronization therapy (CRT). The aim of this work was to find prognostic factors of full clinical and echocardiographic responses (i.e., ≥50% left ventricular ejection fraction [LVEF] and New York Heart Association class I) after 1 year of CRT. This was a prospective follow-up study that involved 2 hospitals. Patients (n = 75) with advanced heart failure (64 ± 9 years of age, 87% men, LVEF 24 ± 7%) who received CRT were followed for 17 ± 9 months. Univariate and multivariate regression analyses were used to identify predictors of full CRT response. A nomogram predicting the individual probability of full CRT response during follow-up was calculated. There were 13 patients with restoration of normal LVEF versus 62 without (mean LVEF 56% ± 5% vs 31% ± 8%, respectively, p <0.001). Predictors of full response included cause of heart disease, baseline QRS width, and degree of QRS shortening in response to CRT. Patients with nonischemic heart disease, baseline QRS width ≤150 ms, and QRS shortening ≥40 ms in response to CRT had a >75% probability of restoration of normal LVEF. In conclusion, our nomogram using a combination of cause, baseline QRS width, and degree of QRS shortening in response to CRT allows assessment of individual probability of full response. This observation awaits further confirmation from larger series.
Cardiac resynchronization therapy (CRT) is an established therapy for patients with heart failure and left ventricular (LV) dysfunction ; predicting the response of an individual patient to CRT is still a debated issue. The primary goal of this study was to create a nomogram allowing, at time of implantation, determination of an individual probability of being a “full CRT responder” (left ventricular ejection fraction [LVEF] ≥50%, New York Heart Association [NYHA] class I). Restoration of normal LV function with complete reverse remodeling was chosen to limit the confounding influence represented by different definitions of responders to CRT in the literature.
This is a prospective follow-up study involving 2 hospitals. The study population included 75 consecutive patients with symptomatic heart failure despite optimal medical treatment (mean age 64 ± 8.6 years, 87% men) who underwent CRT according to current indications. All patients were in NYHA functional class III or IV caused by ischemic cardiomyopathy or nonischemic dilated cardiomyopathy (NIDC) diagnosed ≥1 year before implantation. Diagnosis of IDC was made according to World Health Organization criteria. An implantable cardioverter–defibrillator (ICD) was added to CRT in patients at risk of sudden cardiac death according to current indications. Patients with previous pacemakers or ICDs were included in the study if the burden of right ventricular pacing did not exceed 5% of their telemetry-monitored rhythm.
The study was approved and conducted in compliance with regulations of the local ethics committees of the 2 institutions and with the Declaration of Helsinki. All patients provided oral and written informed consent to device implantation and agreed to data retrieval and analysis.
Biventricular pacemakers or ICDs were implanted as described in the current literature. In brief, the LV lead was inserted using a transvenous approach through the coronary sinus into the cardiac veins. Location of the LV lead in the lateral or posterolateral cardiac vein was preferred whenever possible.
Standard echocardiography at baseline and during follow-up was performed using commercially available equipment (Philips Sonos 5500, Philips IE33, Andover, Massachusetts) by cardiologists expert in echocardiography and blinded to a patient’s clinical response. LVEF was assessed by biplane Simpson equation using apical 4-chamber and 2-chamber views. Interobserver and intraobserver variabilities for volumetric assessment were 7% ± 3.3% and 2% ± 4%, respectively. LV diastolic function was assessed by transmitral and pulmonary veins pulse-wave Doppler. LV filling pattern at transmitral Doppler was defined as restrictive when the E wave-deceleration time was ≤120 ms or E/A wave peak velocity ratio ≥2 plus E wave-deceleration time ≤150 ms. Severity of mitral regurgitation was assessed by percent jet area relative to left atrial size in the apical 4-chamber view and graded in 4 degrees. Right ventricular function was assessed as shortening fraction area from an apical approach.
Patients were followed with regular clinical evaluation, echocardiographic and electrocardiographic assessments, and device interrogation; evaluation intervals were based on clinical judgment, but there was ≥1 assessment in the first 6 months and thereafter ≥1 every year. The following variables were recorded and collected in a database: cardiac rhythm, NYHA functional class, QRS duration at baseline (i.e., before implantation) and QRS shortening in response to CRT at time of implantation, echocardiographic left atrial and LV dimensions and volumes, LVEF, right ventricular systolic function, mitral regurgitation, transmitral flow velocity, rehospitalization for worsening heart failure, vital status (death vs heart transplantation), and medications before and after device implantation.
Patients were classified as full responders if at 1-year follow-up they showed NYHA functional class I and LVEF >50% in the absence of significant mitral regurgitation. As a study subanalysis, patients were also categorized as responders or nonresponders depending on whether they had an increase in LVEF ≥10 percentage points.
Results for continuous variables are presented as mean ± SD and for discrete variables as frequency (percentage). A p value <0.05 was considered statistically significant and all tests were 2-sided. Differences between baseline and follow-up characteristics in the same patients were investigated by paired t tests for continuous variables and McNemar test for binary variables. Differences between groups were tested by chi-square test for comparison of discrete variables and by analysis of variance for comparison of continuous variables. Univariate odds ratios and multivariate stepwise logistic regression analyses were used for identifying variables predictive of restoration of normal LV function in response to CRT. Parameters investigated for univariate analysis are listed in Table 1 . In the stepwise procedure all variables significant at univariate analysis were entered; because of the limited number of events, we adopted the stepwise backward-conditional procedure with removal (p = 0.1). Solutions to regression analyses provided the equations used to generate a nomogram calculating the individual probability of full CRT response. Simulation techniques were used to ascertain whether predicted values from the model were likely to accurately predict responses in patients with clinical characteristics similar to those used to generate the model. Calibration and reliability were assessed to test the predictive accuracy of the model. A bootstrap technique was used to obtain estimates of model performance. The set of indexes computed for summarizing the model validation were maximum error in predicted probabilities, intercept/slope calibration indexes (which have to be close to 0 and 1, respectively), and the Somer concordance index (concordance between predicted and observed outcomes). SPSS 14.0 (SPSS, Inc., Chicago, Illinois) and R 2.6.1 (available at: www.r-project.org/ ) were used for analysis.
|Parameter||Baseline||At Last Follow-Up||p Value|
|Age (years)||64 ± 8.6|
|Age ≤65 years||34 (45%)|
|Nonischemic cause||39 (52%)|
|Duration of disease (months), median (interquartile range)||54 (20–113)|
|Atrial fibrillation||10 (13%)|
|Baseline QRS duration (ms)||165 ± 29|
|Enalapril equivalent dose (mg/day)||23 ± 14|
|Metoprolol equivalent dose (mg/day)||67 ± 52|
|Furosemide dose (mg/day)||57 ± 55|
|Stimulation in lateral vein||29 (39%)|
|Baseline QRS width (ms)||165 ± 30|
|QRS duration of cardiac resynchronization therapy paced beats (ms)||149 ± 24|
|New York Heart Association class IV||16 (21%)||1 (2%)||0.002|
|Left ventricular ejection fraction (%)||24 ± 7||35 ± 13||<0.001|
|End-diastolic diameter (mm)||73 ± 11||69 ± 12||0.02|
|End-systolic diameter (mm)||62 ± 11||54 ± 14||<0.001|
|End-diastolic volume (ml)||256 ± 99||220 ± 121||0.01|
|End-systolic volume (ml)||195 ± 89||157 ± 108||0.004|
|Left atrial diameter (mm)||49 ± 8||48 ± 9||NS|
|Left atrial area (cm 2 )||33 ± 9||30 ± 10||NS|
|Mitral regurgitation grade ≥2||48 (64%)||25 (33%)||NS|
|Mitral regurgitation area (cm 2 )||10 ± 6||9 ± 5||NS|
|Restrictive pattern||17 (23%)||10 (13%)||NS|