Prognostic Value of Changes in N-Terminal Pro-Brain Natriuretic Peptide in Val-HeFT (Valsartan Heart Failure Trial)
Serge Masson, PhD*,*,
Roberto Latini, MD*, ,
Inder S. Anand, MD, FACC, FRCP, DPhil (Oxon) ,
Simona Barlera, MS*,
Laura Angelici, MS*,
Tarcisio Vago, BiolD ,
Gianni Tognoni, MD||,
Jay N. Cohn, MD, FACC¶ for the Val-HeFT Investigators
* Department of Cardiovascular Reearch, Istituto Mario Negri, Milan, Italy
Department of Medicine, New York Medical College, Valhalla, New York
Cardiology Section, Veterans Affairs Medical Center, Minneapolis, Minnesota
Laboratorio di Endocrinologia, Ospedale Luigi Sacco, Milan, Italy
|| Consorzio Mario Negri Sud, S. Maria Imbaro, Chieti, Italy
¶ Cardiovascular Division, Department of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota
View larger version (23K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1 Rate of Death by Quartiles of NT-proBNP Changes
Patients were divided according to quartiles of absolute (A) or percent relative (B) changes of NT-proBNP over 4 months and rates of death at end point are shown. Absolute changes were calculated as NT-proBNP4-month – NT-proBNPbaseline. Relative changes were calculated as 100 x ([NT-proBNP4-month – NT-proBNPbaseline]/NT-proBNPbaseline). Number of events and total number of patients for each quartile are shown above the bars. NT-proBNP = N-terminal pro-brain natriuretic peptide.
|
|
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2 Time-Dependent Receiver-Operator Characteristic Curves for All-Cause Mortality
Curves are based on univariate models of the prediction of risk at 24 months with single determinations of NT-proBNP at baseline or 4 months or as percentage relative or absolute changes from baseline. Areas under the curves were compared with the C-index. NT-proBNP = N-terminal pro-brain natriuretic peptide.
|
|
View larger version (13K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3 Kaplan-Meier Curves for All-Cause Mortality in the 4 Categories of Patients
Patients were divided into 4 categories according to NT-proBNP concentrations at baseline and 4 months relative to a threshold (1,078 pg/ml) determined by ROC curves. Low low (patients with NT-proBNP below threshold at baseline and 4 months, n = 904 patients), highhigh (patients with NT-proBNP above threshold at baseline and 4 months, n = 599), highlow (NT-proBNP above to below threshold, n = 125), and lowhigh (NT-proBNP below to above threshold, n = 114). Log-rank test, chi-square = 96.5, p < 0.0001. NT-proBNP = N-terminal pro-brain natriuretic peptide; ROC = receiver-operator characteristic.
|
|
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4 Risk of Death by Baseline NT-proBNP and Categories of Changes Over 4 Months
Patients were divided into 4 categories according to NT-proBNP concentration at baseline and 4 months relative to a threshold of 1,078 pg/ml. A multivariate Cox proportional hazards model was used to compare the risk of death among the 4 categories of patients and by baseline NT-proBNP concentration (log-transformed) using the lowlow group as a referent. The model was adjusted for clinical variables with statistically significant association to outcome (age, body mass index, serum creatinine at baseline, ischemic etiology of heart failure, New York Heart Association functional class, left ventricular ejection fraction and diameter, and prescription of digoxin and diuretics). Data are presented as hazard ratio (HR) and 95% confidence interval (CI). NT-proBNP = N-terminal pro-brain natriuretic peptide.
|
|
|
