HEART FAILURE
Predictors of mortality in patients with heart failure and preserved systolic function in the Digitalis Investigation Group trial
R.Christopher Jones, MD,
Gary S. Francis, MD, FACC and
Michael S. Lauer, MD, FACC*
Department of Cardiovascular Medicine, Cleveland Clinic Foundation, Cleveland, Ohio, USA
Manuscript received March 23, 2004;
revised manuscript received April 19, 2004,
accepted May 3, 2004.
* Reprint requests and correspondence: Dr. Michael S. Lauer, Desk F25, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195 (Email: Lauerm{at}ccf.org).
 |
Abstract
|
|---|
OBJECTIVES: We identified predictors of mortality in patients with preserved ejection fraction (EF) and clinical heart failure (HF).
BACKGROUND: Although diastolic HF is common, the factors that predict mortality have not been clearly defined.
METHODS: We studied 988 patients with HF and preserved EF enrolled in the Digitalis Investigation Group (DIG) trial. Survival analyses were employed to identify variables associated with mortality.
RESULTS: During 3.1 years of follow-up, 231 (23%) patients died. Among 18 variables considered, the strongest independent predictors of death were glomerular filtration rate (adjusted hazard ratio for one standard deviation decrease 1.50, 95% confidence interval [CI] 1.35 to 1.67, p < 0.0001), New York Heart Association functional class III or IV (adjusted hazard ratio 1.64, 95% CI 1.20 to 2.18, p = 0.0011), male gender (adjusted hazard ratio 1.71, 95% CI 1.26 to 2.32, p = 0.0005), and older age (adjusted hazard ratio for one standard deviation increase of age2 1.28, 95% CI 1.08 to 1.50, p = 0.0019). A risk score was developed to estimate long-term mortality.
CONCLUSIONS: Diastolic HF is associated with a high death rate. Important predictors of death include impaired renal function, worse functional class, male gender, and older age.
|
Abbreviations and Acronyms
| | BMI = body mass index | | CI = confidence interval | | CT = cardiothoracic | | DIG = Digitalis Investigation Group | | EF = ejection fraction | | GFR = glomerular filtration rate | | HF = heart failure | | MDRD = Modification of Diet in Renal Disease | | NYHA = New York Heart Association |
|
The Digitalis Investigation Group (DIG) enrolled 6,800 patients between 1991 and 1993 with clinical heart failure (HF) and depressed ejection fraction (EF) in a randomized, double-blind, placebo-controlled trial of digoxin therapy (1). Upon the advice of the late Dr. Richard Gorlin, an ancillary study was concomitantly conducted consisting of 988 patients with clinical HF and EF more than 0.45. While the overall mortality in this group (23%) was lower than that in the main trial (35%), the absolute mortality was still substantial. Although it was determined that digoxin therapy was not associated with altered mortality in either population, other predictors of mortality for the patients in the ancillary trial were not reported. Accordingly, we undertook this analysis to better define prognostic indicators in patients with clinical HF and preserved systolic function.
|
Methods
|
|---|
Study design.
The details of the trial design and rationale have been previously published (1). Patients were enrolled from 302 North American centers, and were eligible for the ancillary trial if they had stable HF, EF >0.45, and sinus rhythm. Heart failure was defined by predefined symptoms, signs, and chest radiograph criteria. Cardiothoracic (CT) ratio was estimated from a chest radiograph. Glomerular filtration rate (GFR) was estimated from the serum creatinine based on the Modification of Diet in Renal Disease (MDRD) equation (2). Informed consent was obtained from all patients. The primary outcome used for the ancillary trial was combined overall mortality and hospitalization from worsening HF.
Statistical analysis.
Associations of candidate variables with mortality were assessed by Kaplan-Meier plots and log-rank chi-square statistics. All variables were considered in multivariable Cox regression analyses. Bootstrap resampling with replacement (3) was performed to create 1,000 new data sets, each with a size equal to the original data set. A resampling analysis with 1,000 iterations was performed to identify the variables that entered into 50% of the Cox regression models, with p < 0.05 for retention of variables (4). A second series of 1,000 iterations was performed with only the variables that were retained in the first iteration. This second analysis was used to estimate hazard ratios and confidence intervals (CI).
Analyses were performed using SAS version 8.2 software (SAS Institute, Cary, North Carolina). Values of p < 0.05 were considered significant.
|
Results
|
|---|
Baseline characteristics.
Baseline characteristics of the 988 patients in the ancillary trial are listed in Table 1. There were 285 patients (29%) with diabetes, which was very similar to the proportion of patients with diabetes among patients with systolic dysfunction (1,933 of 6,800 [28%], p = 0.78).
Mortality.
There were 231 total deaths in a mean follow-up period of 3.5 ± 0.7 years. Univariate predictors of mortality included increasing age, decreasing GFR, decreasing body mass index (BMI), worsening functional status, increasing CT ratio, presence of diabetes, use of vasodilators, and use of diuretics (Figs. 1A and 1B). Etiology of HF, gender, and EF did not predict death (Fig. 1C).
View larger version (39K):
[in this window]
[in a new window]
|
Figure 1 (A) Univariable predictors of mortality. Age, glomerular filtration rate (GFR), and body mass index (BMI) are presented in quartiles. (B) Univariable predictors of mortality. Cardiothoracic (CT) ratio is presented in quartiles. (C) Univariable predictors of mortality. Ejection fraction (EF) is presented in quartiles. HTN = hypertension; NYHA = New York Heart Association.
|
|
Multivariable Cox regression analyses.
The only variables to enter more than 90% of the bootstrap-generated models were GFR and New York Heart Association (NYHA) functional class III or IV. Male gender, increasing age, use of diuretics, decreasing BMI, increasing CT ratio, diabetes, and use of vasodilators entered more than 50% of the models (Table 2).
View this table:
[in this window]
[in a new window]
|
Table 2. Potential Variables Associated With Mortality in Bootstrap Models, Presented in Descending Order of Strength; 1,000 Bootstrap Models Performed in Each Iteration
|
|
Risk score.
A risk score for mortality was developed using each of the 10 variables from the multivariable bootstrap model entering more than 50% of the models (Appendix). A Kaplan-Meier curve demonstrating worsening survival for the high-risk score quartiles is presented in Figure 2. The fourth quartile of risk score was a high-risk group, with a four-year mortality of nearly 50%.
Treatment with digoxin or placebo.
There was no association between treatment status and mortality, as previously reported (1). We found no interaction between renal function and treatment status for prediction of death (p for interaction 0.17).
Blood levels for digoxin that were obtained within one month of enrollment were available in 289 patients, among whom 75 died. We found no association between digoxin level and mortality risk (hazard ratio for 0.2 mg/dl increase in digoxin level 1.03, 95% CI 0.96 to 1.10, p = 0.49). This lack of association persisted even after adjusting for renal function (adjusted hazard ratio 0.99, 95% CI 0.92 to 1.07, p = 0.81).
|
Discussion
|
|---|
The DIG ancillary study enrolled 988 patients with clinical HF and preserved systolic function, thus representing one of the largest such prospectively gathered populations assembled to date. A unique feature of this cohort is that it was largely derived from an ambulatory population. While the primary outcome in the original ancillary trial was the combined end points of death or hospitalization from HF, the investigators demonstrated that digoxin was not associated with any difference in mortality in the ancillary trial (1). We have shown that important predictors of mortality in these patients with clinical HF and preserved systolic function include worsening renal function, NYHA functional class III or IV, and male gender.
The findings that poor functional class, worsening renal function, and male gender are associated with an adverse effect in a general HF patient population are not novel (5–9). Our group recently published a post-hoc analysis of the DIG trial exercise substudy demonstrating increasing mortality with decreasing quartile of creatinine clearance and 6-min walk distance <262 m (10). Similarly, a post-hoc analysis of the Studies Of Left Ventricular Dysfunction (SOLVD) registry demonstrated worse combined outcome of hospitalization or death for the subset of patients with creatinine clearance <60 ml/min (11). Explanations for why renal insufficiency may worsen prognosis in patients with diastolic HF are provisional. Two groups, however, found improvement in echocardiographic measures of diastolic function in end-stage renal disease patients undergoing hemodialysis (12,13). One hypothesis for the improvement was attributed to a possible decrease in interstitial edema.
The risk score has several potential uses. With current hand-held computer capability, a patient's risk score may be readily determined during the initial encounter. This score may then be used to estimate prognosis, thus potentially facilitating discussions between clinician and patient regarding therapies and lifestyle changes. We also foresee our risk score being used to identify higher risk populations for entry into clinical trials of therapeutics. Two large, randomized, controlled trials have now been performed attempting to identify appropriate therapy for patients with clinical HF and preserved EF without significant success (1,14). Our risk score would allow for sicker patients to be identified a priori, thus promoting the entry of patients with larger expected event rates.
Several limitations are worthy of mention. Although the DIG trial data set is well validated, it is aging. Records regarding beta-blocker therapy were not collected, as beta-blockade was not considered to be part of the therapeutic armamentarium for congestive HF at the time of the study design and enrollment. Similarly, no information specifically regarding aldosterone antagonists or angiotensin-receptor blockers was collected. Echocardiography was not performed as a systematic component in the DIG trial; this is an important acknowledged limitation. It is indeed quite likely that future work in other cohorts may demonstrate that certain echocardiographic measures, like left atrial size, left ventricular mass, or left ventricular wall thickness, may well emerge as important prognostic markers in diastolic HF. The same applies to other newer measures, including neurohormonal perturbations. We cannot contend that our patients had diastolic dysfunction as we do not have echocardiographic parameters indicating abnormal filling patterns (15–18). The estimation of renal function was not performed by the more commonly used Cockcroft-Gault equation as we had only BMI, and not weight, information (19). Nevertheless, some argue that the MDRD equation may be more accurate (20). We did not have data on the extent or severity of coronary disease, which may well have prognostic importance. Lastly, the risk score needs validation in a separate cohort.
Conclusions.
Renal insufficiency and poor exercise capacity add substantially to the annualized mortality in patients with HF and preserved EF, as they do for patients with HF and impaired systolic function. In planning future clinical trials, which are desperately needed for patients with HF and preserved EF (21), one might be able to enrich the population by incorporating such entry criteria, thus ensuring that primary end points will be met with a smaller sample size than what otherwise might be deemed necessary.
|
Appendix
|
|---|
where, NYHA functional class III/IV = 1, if NYHA functional class III/IV is present, female gender = 1, if female gender is present, diuretics = 1, if diuretics are used, diabetes = 1, if diabetes is present, vasodilators = 1, if vasodilators are used, and K-sparing diuretics = 1, if potassium-sparing diuretics are used.
|
References
|
|---|
- The Digitalis Investigation Group The effect of digoxin on mortality and morbidity in patients with heart failure N Engl J Med 1997;336:525-533.[Abstract/Free Full Text]
- Levey AS. Clinical practice: nondiabetic kidney disease N Engl J Med 2002;347:1505-1511.[Free Full Text]
- Efron B, Tibshirani RJ. An Introduction to the Bootstrap. New York, NY: Chapman and Hall, 1993..
- Chen CH, Georg SL. The bootstrap and identification of prognostic factors via Cox's proportional hazards regression model Stat Med 1985;4:39-46.[Medline]
- Adams Jr. KF, Sueta CA, Gheorghiade M, et al. Gender differences in survival in advanced heart failure: insights from the FIRST study Circulation 1999;99:1816-1821.[Abstract/Free Full Text]
- Bittner V, Weiner DH, Yusuf S, et al. Prediction of mortality and morbidity with a 6-minute walk test in patients with left ventricular dysfunction: SOLVD investigators JAMA 1993;270:1702-1707.[Abstract]
- Cowie MR, Wood DA, Coats AJ, et al. Survival of patients with a new diagnosis of heart failure: a population based study Heart 2000;83:505-510.[Abstract/Free Full Text]
- Martinez-Selles M, Robles JA, Prieto L, et al. Systolic dysfunction is a predictor of long-term mortality in men but not in women with heart failure Eur Heart J 2003;24:2046-2053.[Abstract/Free Full Text]
- McAlister FA, Teo KK, Taher M, et al. Insights into the contemporary epidemiology and outpatient management of congestive heart failure Am Heart J 1999;138:87-94.[CrossRef][Medline]
- Mahon NG, Blackstone EH, Francis GS, Starling 3rd RC, Young JB, Lauer MS. The prognostic value of estimated creatinine clearance alongside functional capacity in ambulatory patients with chronic congestive heart failure J Am Coll Cardiol 2002;40:1106-1113.[Abstract/Free Full Text]
- Dries DL, Exner DV, Domanski MJ, Greenberg B, Stevenson LW. The prognostic implications of renal insufficiency in asymptomatic and symptomatic patients with left ventricular systolic dysfunction J Am Coll Cardiol 2000;35:681-689.[Abstract/Free Full Text]
- Fatema K, Hirono O, Takeishi Y, et al. Hemodialysis improves myocardial interstitial edema and left ventricular diastolic function in patients with end-stage renal disease: noninvasive assessment by ultrasonic tissue characterization Heart Vessels 2002;16:227-231.[CrossRef][Medline]
- Gupta S, Dev V, Kumar MV, Dash SC. Left ventricular diastolic function in end-stage renal disease and the impact of hemodialysis Am J Cardiol 1993;71:1427-1430.[CrossRef][Medline]
- Yusuf S, Pfeffer MA, Swedberg K, et al. Effects of candesartan in patients with chronic heart failure and preserved left ventricular ejection fraction: the CHARM-Preserved trial Lancet 2003;362:777-781.[CrossRef][Medline]
- Chen HH, Lainchbury JG, Senni M, Bailey KR, Redfield MM. Diastolic heart failure in the community: clinical profile, natural history, therapy, and impact of proposed diagnostic criteria J Card Fail 2002;8:279-287.[CrossRef][Medline]
- Brutsaert DL, Sys SU. Diastolic dysfunction in heart failure J Card Fail 1997;3:225-242.[CrossRef][Medline]
- Zile MR, Brutsaert DL. New concepts in diastolic dysfunction and diastolic heart failure: part I: diagnosis, prognosis, and measurements of diastolic function Circulation 2002;105:1387-1393.[Free Full Text]
- European Study Group on Diastolic Heart Failure How to diagnose diastolic heart failure Eur Heart J 1998;19:990-991003.[Free Full Text]
- Gault MH, Longerich LL, Harnett JD, Wesolowski C. Predicting glomerular function from adjusted serum creatinine Nephron 1992;62:249-256.[Medline]
- Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation: Modification of Diet in Renal Disease study group Ann Intern Med 1999;130:461-470.[Abstract/Free Full Text]
- Zile MR, Brutsaert DL. New concepts in diastolic dysfunction and diastolic heart failure: part II: causal mechanisms and treatment Circulation 2002;105:1503-1508.[Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
D. M. Henkel, M. M. Redfield, S. A. Weston, Y. Gerber, and V. L. Roger
Death in Heart Failure: A Community Perspective
Circ Heart Fail,
July 1, 2008;
1(2):
91 - 97.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. C. Fonarow, W. G. Stough, W. T. Abraham, N. M. Albert, M. Gheorghiade, B. H. Greenberg, C. M. O'Connor, J. L. Sun, C. W. Yancy, J. B. Young, et al.
Characteristics, Treatments, and Outcomes of Patients With Preserved Systolic Function Hospitalized for Heart Failure: A Report From the OPTIMIZE-HF Registry
J. Am. Coll. Cardiol.,
August 21, 2007;
50(8):
768 - 777.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. A. Fiack, H. W. Farber, M. A. Arias, A. Alonso-Fernandez, F. Garcia-Rio, K. M. Kessler, A. Ahmed, J. L. Fleg, M. Gheorghiade, T. E. Owan, et al.
Heart failure with preserved ejection fraction.
N. Engl. J. Med.,
October 26, 2006;
355(17):
1828 - 1828.
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. C. Huynh, A. Rovner, and M. W. Rich
Long-term Survival in Elderly Patients Hospitalized for Heart Failure: 14-Year Follow-up From a Prospective Randomized Trial.
Arch Intern Med,
September 25, 2006;
166(17):
1892 - 1898.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. S. Bhatia, J. V. Tu, D. S. Lee, P. C. Austin, J. Fang, A. Haouzi, Y. Gong, and P. P. Liu
Outcome of heart failure with preserved ejection fraction in a population-based study.
N. Engl. J. Med.,
July 20, 2006;
355(3):
260 - 269.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. L. Hillege, D. Nitsch, M. A. Pfeffer, K. Swedberg, J. J.V. McMurray, S. Yusuf, C. B. Granger, E. L. Michelson, J. Ostergren, J. H. Cornel, et al.
Renal Function as a Predictor of Outcome in a Broad Spectrum of Patients With Heart Failure
Circulation,
February 7, 2006;
113(5):
671 - 678.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. J. Pocock, D. Wang, M. A. Pfeffer, S. Yusuf, J. J.V. McMurray, K. B. Swedberg, J. Ostergren, E. L. Michelson, K. S. Pieper, C. B. Granger, et al.
Predictors of mortality and morbidity in patients with chronic heart failure
Eur. Heart J.,
January 1, 2006;
27(1):
65 - 75.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Q. Yu, R. R. Watson, J. J. Marchalonis, and D. F. Larson
A role for T lymphocytes in mediating cardiac diastolic function
Am J Physiol Heart Circ Physiol,
August 1, 2005;
289(2):
H643 - H651.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. McMurray
Making sense of SENIORS
Eur. Heart J.,
February 1, 2005;
26(3):
203 - 206.
[Full Text]
[PDF]
|
|
|
|
|
|
|
Predicting Death in Heart-Failure Patients with Preserved LVEFs
Journal Watch Cardiology,
November 12, 2004;
2004(1112):
2 - 2.
[Full Text]
|
|
|
|
Top
Abstract
Methods
