- Email: [email protected]
Relation between renal function and response to cardiac resynchronization therapy in Multicenter Automatic Defibrillator Implantation Trial—Cardiac Resynchronization Therapy (MADIT-CRT)
Relation between renal function and response to cardiac resynchronization therapy in Multicenter Automatic Defibrillator Implantation Trial—Cardiac Resynchronization Therapy (MADIT-CRT) Ilan Goldenberg, MD, MA,* Arthur J. Moss, MD,* Scott McNitt, MS,* Alon Barsheshet, MD,* Daniel Gray, MD,† Mark L. Andrews, BBA,* Mary W. Brown, MA,* Wojciech Zareba, MD, PhD,* Edward Sze, MD,* Scott D. Solomon, MD,‡ Marc A. Pfeffer, MD,‡ for the Multicenter Automatic Defibrillator Implantation Trial—Cardiac Resynchronization Therapy (MADIT-CRT) Investigators From the *Cardiology and †Nephrology (DG) Units of the Department of Medicine, University of Rochester Medical Center, Rochester, New York, and the ‡Cardiovascular Division of the Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts. BACKGROUND Cardiorenal interactions have been shown to affect outcome in heart failure patients but were not related to response to cardiac resynchronization therapy (CRT). OBJECTIVE The purpose of this study was to test our hypothesis that assessment of markers of prerenal failure may help identify mildly symptomatic HF patients with diminished effective circulating blood volume who will benefit from CRT. METHODS The benefit of CRT with a defibrillator (CRT-D) as compared with defibrillator-only therapy in reducing the risk of HF or death was assessed by renal function parameters (including serum creatinine [SCr], blood urea nitrogen [BUN], and the ratio of BUN to SCr [BUN:SCr], dichotomized at median values and into approximate quartiles) among 1,803 patients enrolled in the Multicenter Automatic Defibrillator Implantation Trial—Cardiac Resynchronization Therapy. RESULTS Multivariate analysis showed that the benefit of CRT-D was inversely related to SCr levels and directly related to BUN levels. Combined assessment of the two renal function parameters showed a significant difference in the benefit of CRT-D between patients with low (ⱕ18 mg/dL, HR ⫽ 0.85, P ⫽ .28) and elevated (⬎18 mg/dL, HR ⫽ 0.46, P⬍.001) BUN:SCr (P-value for interaction ⫽ .005). Consistently, the benefit of CRT-D was significantly
increased with increasing quartiles of BUN:SCr (Q1: HR ⫽ 1.06 [P ⫽ .79], Q2: HR ⫽ 0.64 [P ⫽ .04], Q3: HR ⫽ 0.47 [P⬍.001], Q4: HR ⫽ 0.44 [P⬍.001]; P-value for trend ⫽ .005). CONCLUSIONS In MADIT-CRT, patients with an elevated ratio of BUN to SCr experienced a significantly greater reduction in the risk of HF or death with CRT-D therapy as compared with patients with a low ratio. These findings suggest an association between prerenal function and response to CRT. KEYWORDS Cardiac resynchronization therapy; Implantable cardioverter defibrillator; Heart failure; Renal function; Mortality ABBREVIATIONS: BUN ⫽ blood urea nitrogen; CI ⫽ confidence interval; CRT ⫽ cardiac resynchronization therapy; CRT-D ⫽ cardiac resynchronization therapy with defibrillator; EF ⫽ ejection fraction; eGFR ⫽ estimated glomerular filtration rate; HF ⫽ heart failure; ICD ⫽ implantable cardioverter-defibrillator; LV ⫽ left ventricular; MADIT-CRT ⫽ Multicenter Automatic Defibrillator Implantation Trial—Cardiac resynchronization therapy; MADIT-II ⫽ Multicenter Automatic Defibrillator Implantation Trial-II; NYHA ⫽ New York Heart Association; SCr ⫽ serum creatinine; SD ⫽ standard deviation
Impaired renal function is an important independent risk factor for morbidity and mortality in patients with heart failure (HF).1– 6 Even minor abnormalities in renal function have been shown to impair the normal regulation of fluid volume and increase the frequency of acute HF exacerbations, underscoring the link between the heart and the kid-
ney.7 Furthermore, a state of prerenal azotemia may exist among patients with systolic left ventricular (LV) dysfunction due to redistributed excess total body salt and water and a diminished effective circulating blood volume. This clinical status may be manifested in renal function tests as disproportionately elevated blood urea nitrogen (BUN) levels relative to serum creatinine (SCr) levels.8 –10 We have recently shown that an elevated BUN-to-SCr ratio was the most powerful predictor of recurrent HF events in the Multicenter Automatic Defibrillator Implantation Trial-II (MADIT-II).11 Thus, cardiorenal connections may affect outcomes in patients with LV dysfunction with or without overt kidney disease. The Multicenter Automatic Defibrillator Implantation Trial—Cardiac Resynchronization Therapy (MADIT-CRT) demonstrated that CRT with a defibrillator (CRT-D) was associated with a significant reduction in HF or death as
The MADIT-CRT study was supported by a research grant from Boston Scientific, St. Paul, Minnesota, to the University of Rochester School of Medicine and Dentistry. Drs. Moss, Solomon, and Zareba have received research support for the conduct of the MADIT-CRT trial from Boston Scientific through a grant to the University of Rochester. Address reprint requests and correspondence: Ilan Goldenberg, M.D., Heart Research Follow-up Program, Box 653, University of Rochester Medical Center, Rochester, New York 14642. E-mail address: Ilan. [email protected]. (Received July 28, 2010; accepted September 2, 2010.)
(Heart Rhythm 2010;7:1777–1782) © 2010 Heart Rhythm Society. All rights reserved.
1547-5271/$ -see front matter © 2010 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2010.09.005
1778
Heart Rhythm, Vol 7, No 12, December 2010
compared with defibrillator-only therapy among asymptomatic or mildly symptomatic patients with reduced ejection fraction (EF) and a wide QRS duration.12 However, the benefit of CRT-D in this population is not uniform.12 It is possible that renal regulation may affect the response to this mode of HF therapy. We hypothesized that renal markers of effective circulating blood volume relate to the benefit of CRT.
the classification of deaths15); and (3) the incremental echocardiographic response to CRT-D versus ICD therapy, defined as the percent reduction in cardiac volumes (LV end-systolic and end-diastolic volumes [LVESV and LVEDV, respectively]) between enrollment and 1 year (calculated as the difference between 1-year cardiac volumes and baseline cardiac volumes, divided by baseline cardiac volumes).
Methods
Statistical analysis
MADIT-CRT
The clinical characteristics of study patients were compared by BUN:SCr categories using the 2-test for categorical variables and the t-test for continuous variables. KaplanMeier estimates for HF or all-cause mortality in each treatment group, stratified according to the renal function categories, were determined and statistically evaluated with the log-rank test. The Cox proportional-hazards regression model was used to evaluate the independent contribution of baseline clinical factors to the development of the endpoint. All models included the following prespecified covariates: age ⬎65 years; gender; ischemic status; QRS ⬎150 ms; EF ⬍25%; diabetes mellitus; left bundle branch block; systolic blood pressure ⱖ130 mmHg; heart rate ⱖ80 bpm; and medical therapies with beta-blockers, angiotensin-converting enzyme inhibitors, and diuretics. The benefit of CRT-D compared with ICD-only therapy in reducing the primary endpoint in each renal function category was assessed by including a renal function-by-treatment interaction term in the multivariate models. In the quartile analyses, the probability value for trend was determined by using the four quartiles of each renal function parameter as an ordinal categorical variable in the multivariate models. Multivariate regression analysis was used to evaluate the incremental echocardiographic response to CRT-D versus ICD-only therapy (assessed as percent reductions in LVESV and LVEDV between baseline and 1-year follow-up echocardiograms). Regression models included the covariates described above. The benefit of CRT-D compared with ICD-only therapy in reducing cardiac volumes was assessed in each BUN:SCr category by including a treatment-by-BUN:SCr interaction term in the multivariate regression models. The Proportional hazards assumption for each renal parameter was tested by including the treatment-bytime interaction term in the multivariate Cox models. All P-values were two-sided, and a P-value ⬍.05 was considered statistically significant. Analyses were performed with the use of SAS software (ver. 9.20, SAS Institute, Cary, North Carolina).
The design and results of MADIT-CRT have been reported elsewhere.12 Briefly, 1,820 patients with ischemic or nonischemic cardiomyopathy, an EF of 30% or less, a QRS duration of 130 ms or more, and New York Heart Association (NYHA) class I or II symptoms were randomly assigned in a 3:2 ratio to receive CRT-D or an implantable cardioverter-defibrillator (ICD) alone. Exclusion criteria included coronary revascularization or myocardial infarction within the previous 3 months and the presence of any other disease associated with a reduced likelihood of survival for the duration of the trial, including advanced kidney disease (BUN ⬎70 mg/dL or SCr ⬎3.0 mg/dL). After an average follow-up of 2.4 years, treatment with CRT-D was associated with a significant 34% (P ⬍.001) reduction in the risk of the primary endpoint of the study, comprising death from any cause or a nonfatal HF event. Two-dimensional echocardiography was obtained at baseline and at 1-year follow-up to assess change in LV volumes and EF during this time period in the two treatment arms (CRT-D with CRT on at 1 year, n ⫽ 766; ICD only, n ⫽ 628).
Definitions and outcomes Renal function tests at enrollment were available for 1,803 (99%) patients who made up the present study population. The renal function parameters that were evaluated included BUN, SCr, and the estimated glomerular filtration rate (eGFR). Combined assessment of BUN and SCR (prespecified as BUN:SCr) was used as a marker of prerenal function.8,9 The individual renal function variables (including BUN, SCr, and eGFR) were categorized into approximate quartiles, and the ratio of BUN to SCr was analyzed both by dichotomizing values at the median level and by categorization into approximate quartiles. Estimated GFR was assessed using the four-component Modification of Diet in Renal Disease equation incorporating age, race, sex, and SCr level: eGFR (mL/min/1.73 m2) ⫽ 186 ⫻ (SCr level [in mg/dL])⫺1.154 ⫻ (age [in years])⫺0.203. For women and blacks, the product of this equation was multiplied by a correction factor of 0.742 and 1.21, respectively.13 The presence of kidney disease was defined as eGFR ⬍60 mL/min/1.73 m2, according to the guidelines of the National Kidney Foundation.14 We evaluated the relationship between each of the prespecified renal function parameters and (1) the benefit of CRT-D versus ICD therapy in reducing the primary endpoint of death or a nonfatal first HF event; (2) the benefit of CRT-D versus ICD therapy in reducing the separate endpoints of nonfatal HF, all-cause mortality, and nonfatal HF or HF-related death (using a modified Hinkle-Thaler system for
Results Among the 1,803 study patients, mean ⫾ standard deviation (SD) values of the prespecified renal parameters included BUN levels of 21.5 ⫾ 8.9 mg/dL (median 20.0 mg/dL; interquartile range 15.8 –25.0 mg/dL), SCr levels of 1.2 ⫾ 0.4 mg/dL (median 1.1 mg/dL; interquartile range 0.9 –1.3 mg/dL), and eGFR levels of 66 ⫾ 27 mL/min/1.73 m2 (median 66 mL/min/ 1.73 m2; interquartile range 46 – 85 ml/min/1.73 m2). Values of the ratio BUN to SCr were normally distributed in the study population, with a mean ⫾ SD of 18.6 ⫾ 5.8 (median 17.8; interquartile range 14.8 –22.0).
Goldenberg et al
CRT and Renal Function
1779
Table 1 Baseline clinical characteristics of study patients by BUN to creatinine ratio BUN to creatinine ratio (median ⫽ 18)
Characteristic BUN, mg/dL Creatinine, mg/dL eGFR, mL/min/1.73 m2 CRT-D Ischemic cardiomyopathy Age, years Age ⬎65 years Female gender NYHA class II Hypertension Diabetes mellitus Past coronary artery bypass surgery Past percutaneous coronary revascularization QRS duration ⱖ150 ms Left bundle branch block EF, mean ⫾ SD EF ⬍25% Atrial fibrillation ⬎1 month before enrollment Cigarette smoking anytime Systolic blood pressure ⱖ130 mmHg Diastolic blood pressure ⱕ80 mmHg Heart rate per ⱖ80 bpm Body mass index ⱖ30 kg/ m2 Apical lead position† Baseline echocardiogram LVESV, mL LVEDV, mL Medications Angiotensin-converting enzyme inhibitors Angiotensin receptor blockers Class I antiarrhythmics Digitalis -blockers Amiodarone Statins Diuretics Furesemide dose
were similar between CRT-D and ICD patients within the two BUN:SCr subgroups (not shown).
Relation of individual renal function parameters to clinical benefit of CRT-D
ⱕMedian (n ⫽ 936)
⬎Median (n ⫽ 867)
17 ⫾ 6 1.2 ⫾ 0.4 71 ⫾ 25 61 56 63 ⫾ 11 44 20 85 63 25 29
26 ⫾ 9* 1.1 ⫾ 0.3* 64 ⫾ 28* 58 54 66 ⫾ 10* 55* 30* 86 64 35* 29
28
26
63 68 24 ⫾ 5 60 11
66 73* 24 ⫾ 5 59 12
15 38
9* 33*
83
88*
36
37
15
12
177 ⫾ 51 248 ⫾ 63
177 ⫾ 49 248 ⫾ 60
76
77
19
22
1 26 93 8 68 69 48 ⫾ 45
0.6 26 93 6 67 81* 52 ⫾ 40
Values are percentages or mean ⫹/- SD. *P⬍.05 for the difference between BUN:SCr categories. †Data regarding LV lead position were available for 799 of the 1,089 patients who were allocated to CRT-D at enrollment (74%).
The baseline characteristics of study patients by BUN:SCr are shown in Table 1. Patients with BUN:SCr values in the abovemedian range (⬎18) comprised a higher proportion of females and exhibited higher frequency risk factors, including an older age, diabetes mellitus, and left bundle branch block, as compared with study subjects who had a lower BUN:SCr. Baseline medications were administered to a similar proportion of patients in the two BUN:SCr subgroups with the exception of diuretics, which were administered more frequently to patients with a higher BUN: SCr (Table 1). Baseline clinical characteristics and medications
Multivariate analysis showed that increased levels of each of the individual renal function parameters was independently associated with a significant increase in the risk of HF or death in the total study population (BUN ⱖ20 mg/dL, HR ⫽ 1.38 [95% confidence interval (CI) 1.07–1.78], P ⫽ .01; SCr ⬎1.1: HR ⫽ 1.46 [95% CI 1.13–1.86], P ⫽ .003; eGFR ⬍60 mL/ min/1.73 m2; HR ⫽ 1.49 [95% CI 1.15–1.93], P ⫽ .002). The benefit of treatment with CRT-D as compared with ICD-only therapy was shown to be directly related to BUN and inversely related to SCr (Table 2). Thus, multivariate analysis showed that risk reduction associated with CRT-D therapy was highest among patients with upper quartile BUN (57% reduction in the risk of HF or death compared with ICD-only therapy [P⬍.001]) and lower quartile SCr (60% reduction in the risk of HF or death compared with ICD-only therapy [P ⫽ .001]). In contrast, among patients with lower quartile BUN and upper quartile SCr there was no statistically significant difference in the risk of HF or death between the two treatment arms (Table 2). The benefit of CRT-D was similar between patients with and without kidney disease (eGFR ⬍60 mL/min/1.73 m2; HR ⫽ 0.67 [95% CI 0.50 – 0.89]; eGFR ⱖ60 mL/min/1.73 m2; HR ⫽ 0.61 [95% CI 0.50 – 0.89]; P-value for interaction ⫽ .76). However, consistent with the SCr results, patients with eGFR levels in the upper quartile range experienced the most pronounced risk reduction with CRT-D as compared with ICDonly therapy (HR ⫽ 0.42 [95% CI 0.22–0.82]; P ⫽ .01).
Relation of the ratio of BUN to SCr to the clinical benefit of CRT-D Combined assessment of BUN and SCr levels demonstrated a significant difference in the benefit of CRT-D between patients who exhibited BUN:SCr levels above and below median values (Table 3). Accordingly, multivariate analysis showed that the overall 37% (P ⬍.001) reduction in the risk of HF or death with CRT-D in the total study population was related to a pronounced 54% (P ⬍.001) risk reduction among patients with BUN:SCr ⬎18, whereas among patients with lower BUN:SCr values, treatment with CRT-D was not associated with a statistically significant benefit (HR ⫽ 0.85 [P ⫽ .28]; P-value for BUN:SCr-by-treatment interaction ⫽ .005; Table 3). Consistent with these findings, patients with BUN:SCr ⬎18 experienced pronounced and significant reductions in the rates of HF or death with CRT-D as compared with ICD-only therapy (Figure 1A), whereas patients with BUN:Cr ⱕ18 exhibited virtually identical rates of HF or death in the two treatment groups at 2 and 3 years of follow-up (Figure 1B). Notably, the clinical benefit of CRT-D therapy among patients with BUN:SCr ⬎18 was more pronounced at ⬎1 year of follow-up than during the first year after enrollment (Figure 1A). However, this time-related difference in the benefit of CRT-D was not
1780 Table 2
Heart Rhythm, Vol 7, No 12, December 2010 Multivariate analysis: CRT-D versus ICD-only risk of HF or death by individual renal function parameters
BUN, mg/dL: Q1: ⬍15 (n ⫽ 389) Q2: 15–19 (n ⫽ 499) Q3: 20–24 (n ⫽ 430) Q4: ⱖ25 (n ⫽ 485) SCr, mg/dL: Q1: ⬍0.9 (n ⫽ 446) Q2: 0.9–1.1 (n ⫽ 511) Q3: ⬎1.1–1.3 (n ⫽ 432) Q4: ⬎1.3 (n ⫽ 413)
Adjusted hazard ratio
95% Confidence interval
P
P-value for trend
0.81 0.69 0.64 0.43
0.50–1.28 0.46–0.99 0.43–0.97 0.28–0.68
.37 .04 .04 ⬍.001
.08
0.40 0.52
0.24–0.71 0.35–0.79
.001 .001
.09
0.83
0.61–1.32
.67
0.78
0.48–1.09
.16
Note: All models were adjusted for the following additional covariates: age ⬎65 years; gender; ischemic status; QRS ⬎150 ms; EF ⬍25%; diabetes mellitus; left bundle branch block; systolic blood pressure ⱖ130 mmHg; heart rate ⱖ80 bpm; BUN ⱖ20 mg/dL (in the SCr model); SCr ⬎1.1 mg/dL (in the BUN model); and medical therapies with beta-blockers, angiotensin-converting enzyme inhibitors, and diuretics. CRT-D versus ICD efficacy in each renal parameter subset was assessed using by adding renal function-by-treatment interaction terms to the multivariate models.
statistically significant (P-value for treatment-by-time interaction in the multivariate Cox model ⫽ .61). Similar to the results that related CRT-D benefit to median BUN:SCr levels, quartile analysis showed a significant increase in the benefit of CRT with increasing BUN:SCr quartiles (P-value for trend ⫽ .005; Table 3) that corresponded to a direct correlation between increasing BUN: SCr quartiles and the cumulative event rate reduction at 2.5 years of follow-up associated with CRT-D (Figure 2). Notably, the significant 56% (P ⬍.001) reduction in the risk of HF or death associated with CRT-D among patients in the highest BUN:SCr quartile was related to a 52% (P ⫽ .06) reduction in all-cause mortality, a 59% (P ⫽ .001) reduction in the risk of nonfatal HF, and a 61% (P ⬍.001) reduction in the risk of nonfatal HF or HF-related death.
Relation of BUN-to-SCr ratio to echocardiographic response to CRT-D There was no significant difference in baseline LVESV and LVEDV between patients with low or elevated BUN:SCr Table 3
(Table 1). Despite this, however, percent reduction in both LVESD and LVEDV between baseline and 1-year values among CRT-D-treated patients displayed a direct correlation with increasing BUN:SCr quartiles (Figure 3A and 3B, respectively), whereas reductions in cardiac volumes in the ICD-only arm of the trial were significantly lower and unrelated to BUN:SCr (not shown). Consistently, multivariate regression analysis showed that the incremental reductions in cardiac volumes with CRT-D as compared with ICD-only therapy displayed a direct correlation with increasing BUN:SCr quartiles (LVESV: P-value for trend ⫽ .004; LVEDV: P-value for trend ⫽ .005).
Discussion To our knowledge, the current study is the first to evaluate the relation between markers of renal function and response to CRT. Assessment of individual renal function parameters showed that the benefit of treatment with CRT-D was directly correlated with BUN levels and inversely related to SCr. Furthermore, combined assessment of the two renal markers was shown to identify a significant differential
Multivariate analysis: CRT-D versus ICD-only risk of HF or death by BUN:SCr
BUN:SCr All patients By BUN:SCr: Median: Lower: ⱕ18 (n ⫽ 936) Upper: ⬎18 (n ⫽ 837) Quartiles: Q1: ⬍15 (n ⫽ 459) Q2: 15-18 (n ⫽ 477) Q3: ⬎18-22 (n ⫽ 426) Q4: ⬎22 (n ⫽ 441)
Adjusted hazard ratio
95% Confidence interval
P
0.63
0.51–0.77
⬍.001
0.85
0.61–1.14
.28
0.46
0.34–0.62
⬍.001
1.06 0.64
0.70–1.60 0.41–0.99
.79 .04
0.47
0.30–0.74
⬍.001
0.44
0.29–0.66
⬍.001
P-value for interaction .005
P-value for trend .005
Note: All models were adjusted for the following additional covariates: age ⬎65 years; gender; ischemic status; QRS ⬎150 ms; EF ⬍25%; diabetes mellitus; left bundle branch block; systolic blood pressure ⱖ130 mmHg; heart rate ⱖ80 bpm; SCr ⬎1.1 mg/dL; BUN ⱖ20 mg/dL; and medical therapies with beta-blockers, angiotensin-converting enzyme inhibitors, and diuretics. CRT-D versus ICD efficacy in BUN:SCr subset was assessed by adding a BUN:SCr-by-treatment interaction term to the multivariate models. Virtually identical results were obtained in multivariate models that included further adjustment for time-dependent furosemide dosage.
Goldenberg et al
CRT and Renal Function
1781
Figure 3 Reduction in percent of (A) LVESVs and (B) LVEDVs among patients in the CRT arm by BUN:SCr quartiles. Percent reduction was calculated as (1-year volume ⫺ baseline volume)/baseline volume.
Figure 1 Kaplan-Meier estimates of the probability of heart failure or death by treatment in patients with (A) BUN:SCr ⬎18 mg/dL and (B) BUN:SCr ⱕ18 mg/dL.
clinical and echocardiographic response to CRT-D in the MADIT-CRT population. These findings suggest that cardiorenal interactions relate to response to CRT. In patients with LV dysfunction, an elevated BUN in the setting of relatively normal SCr may reflect a physiological response to a relative decrease of blood flow to the kidney due to diminished effective circulating blood volume.8 Thus, an increase in the ratio of BUN to SCr may identify HF patients who have not yet developed intrinsic renal impairment despite a reduction in renal blood flow, possibly due to adaptive neurohormonal mechanisms to preserve glomerular filtration rate.16 Consistent with this mechanism,
Figure 2 Two-and-a-half-year Kaplan-Meier rates of HF or death in the two treatment groups by BUN:SCr quartiles. The time point of 2.5 years for assessment of event rates was selected since this was the average follow-up among study patients enrolled in the trial.
study patients with increased BUN:SCr exhibited high BUN levels in the presence of normal SCr (Table 1). CRT has been demonstrated to be an efficacious therapy for patients with reduced LV function and ventricular dyssynchrony, improving LV function, symptomatic and functional status, and survival.17–22 The primary mechanism by which this is achieved is improved ventricular pumping efficiency through better temporal coordination of LV activation and contraction.23,24 MADIT-CRT was the first clinical trial to show a significant reduction in the risk of HF or death with CRT-D in a population of patients with LV dysfunction and asymptomatic or mildly symptomatic HF.12 Data from the current study suggest that the benefit of CRT-D in the trial may be related to prerenal function. We have shown that CRT-D therapy was associated with a significantly greater benefit in patients with an elevated ratio of BUN to SCr as compared with those who had a lower ratio. These findings extend recent data from the MADIT-II trial, in which we have shown that an elevated ratio of BUN to SCr was the most powerful predictor for the development of recurrent HF events after ICD implantation.11 Thus, markers of prerenal failure may identify patients with diminished effective circulating blood volume and increased risk for HF events who exhibit a favorable echocardiographic and clinical response to CRT. Notably, the clinical benefit of CRT-D among patients with elevated BUN:SCr was more pronounced after the first year follow-up in the trial, possibly due to the fact that a longer follow-up is required for the echocardiographic benefit of resynchronization therapy to translate into a clinical benefit. In contrast, a lower BUN-to-SCr ratio in the present study occurred in the setting of relatively normal BUN and SCr levels (Table 1), possibly among lower risk patients, who have not yet developed a marked reduction in effective circulating blood volume
1782 and prerenal azotemia, in whom early intervention with CRT-D may not result in a pronounced benefit.
Study limitations Renal function was not a randomization factor in MADITCRT-D, and differences in CRT-D benefit among the prespecified renal function categories may be related to unmeasured risk factors. It should also be noted that a significantly greater proportion of patients with elevated BUN:SCr received diuretic therapy. Our findings, however, persisted after multivariate adjustment for medication usage, including diuretic therapy and furosemide dosage. Furthermore, subgroup analyses demonstrated that the relation between BUN:SCr and CRT-D efficacy was consistent among patients who did and did not receive diuretic therapy (P-value for trend for CRT-D benefit by increasing BUN:SCr quartiles ⬍.05 in both subgroups). Thus, it is unlikely that medication usage affected the present results significantly. Follow-up data regarding renal function parameters were not collected in MADIT-CRT. Therefore, the ratio of BUN to SCr was not assessed as a time-dependent covariate in the multivariate models. Furthermore, data regarding sodium and serum osmolality, as additional markers of intravascular status, were not collected in the trial. Thus, the implications of the current findings are limited to the association between baseline renal function at enrollment and subsequent outcome during an average follow-up period of 2.4 years. MADIT-CRT enrolled asymptomatic or mildly symptomatic HF patients with reduced EF. Therefore, it is unclear whether the findings from the present study are applicable to patients who are currently being referred for treatment with CRT in the community, who may be sicker and have more advanced HF symptoms. Furthermore, patients with advanced renal dysfunction were excluded from MADIT-CRT. Only 89 study patients (5%) had severe kidney disease (defined as eGFR ⬍30 mL/min/1.73 m2).14 Thus, our results regarding the relation between the benefit of CRT-D and levels of BUN and SCr may not reflect the outcome with CRT-D among patients with severe kidney disease.
Heart Rhythm, Vol 7, No 12, December 2010
2. 3.
4.
5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
Conclusions and clinical implications Renal dysfunction is frequently observed in HF patients and is strongly correlated with the severity and prognosis of this condition.1–7 However, evidence for therapeutic strategies to actually reduce mortality and morbidity in HF patients with renal impairment is still lacking. Our findings suggest that CRT-D is a highly effective therapy for mildly symptomatic low-EF patients who exhibit markers of prerenal azotemia. In this population, we have shown that CRT-D was associated with a ⬎50% reduction in the risk of HF or death and a pronounced echocardiographic response, whereas among patients with a low BUN:SCr, CRT-D therapy was associated with a significantly lower clinical and echocardiographic response. These findings suggest that cardiorenal interactions should be assessed among candidates for CRT.
References 1. Dries DL, Exner DV, Domanski MJ, Greenberg B, Stevenson LW. The prognostic
19.
20.
21.
22.
23.
24.
implications of renal insufficiency in asymptomatic and symptomatic patients with left ventricular systolic dysfunction. J Am Coll Cardiol 2000;35:681–689. Hillege HL, Girbes AR, de Kam PJ, et al. Renal function, neurohormonal activation, and survival in patients with chronic heart failure. Circulation 2000;102:203–210. Al-Ahmad A, Rand WM, Manjunath G, et al. Reduced kidney function and anemia as risk factors for mortality in patients with left ventricular dysfunction. J Am Coll Cardiol 2001;38:955–962. Hillege HL, Nitsch D, Pfeffer MA, et al. Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity (CHARM) Investigators. Renal function as a predictor of outcome in a broad spectrum of patients with heart failure. Circulation 2006;113:671– 678. Anavekar NS, McMurray JJ, Velazquez EJ, et al. Relation between renal dysfunction and cardiovascular outcomes after myocardial infarction. N Engl J Med 2004;351:1285–1295. Forman DE, Butler J, Wang Y, et al. Incidence, predictors at admission, and impact of worsening renal function among patients hospitalized with heart failure. J Am Coll Cardiol 2004;43:61– 67. Krumholz HM, Chen YT, Vaccarino V, et al. Correlates and impact on outcomes of worsening renal function in patients ⬎ or ⫽ 65 years of age with heart failure. Am J Cardiol 2000;85:1110 –1113. Morgan DB, Carver ME, Payne RB. Plasma creatinine and urea: creatinine ratio in patients with raised plasma urea. BMJ 1977;2:929 –932. Dossetor JB. Creatininemia versus uremia. The relative significance of blood urea nitrogen and serum creatinine concentrations in azotemia. Ann Intern Med 1966;65: 1287–1299. Lin HJ, Chao CL, Chien KL, et al. Elevated blood urea nitrogen-to-creatinine ratio increased the risk of hospitalization and all-cause death in patients with chronic heart failure. Clin Res Cardiol 2009;98:487– 492. Sze E, Moss AJ, McNitt S, Barsheshet A, Andrews ML, Zareba W, Goldenberg I, for the Multicenter Automatic Defibrillator Implantation Trial-II Investigators. Risk Factors for Recurrent Heart Failure Events in the Multicenter Automatic Defibrillator Implantation Trial II (MADIT-II). J Cardiovasc Electrophysiol 2010. Moss AJ, Hall WJ, Cannom DS, et al. MADIT-CRT Trial Investigators. Cardiac-resynchronization therapy for the prevention of heart-failure events. N Engl J Med 2009;361:1329 –1338. Levey AS, Bosch JP, Lewis JB, Greene T, Rodgers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130:461– 470. National Kidney Foundation. Kidney Disease Outcome Quality Initiative (K/ DOQI) clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39(Suppl 1):S1–S266. Greenberg H, Case RB, Moss AJ, Brown MW, Carroll ER, Andrews ML. Analysis of mortality events in the Multicenter Automatic Defibrillator Implantation Trial (MADIT-II). J Am Coll Cardiol 2004;43:1459 –1465. Givertz MM, Braunwald E. Neurohormones in heart failure: predicting outcomes, optimizing care. Eur Heart J 2004;25:281–282. Cazeau S, Leclercq C, Lavergne T, et al. Multisite Stimulation in Cardiomyopathies (MUSTIC) Study Investigators. Effects of multisite biventricular pacing in patients with heart failure and intraventricular conduction delay. N Engl J Med 2001;344:873– 880. Linde C, Leclercq C, Rex S, et al. Long-term benefits of biventricular pacing in congestive heart failure: results from the MUltisite STimulation in cardiomyopathy (MUSTIC) study. J Am Coll Cardiol 2002;40:111–118. Auricchio A, Stellbrink C, Sack S, et al. Pacing Therapies in Congestive Heart Failure (PATH-CHF) Study Group. Long-term clinical effect of hemodynamically optimized cardiac resynchronization therapy in patients with heart failure and ventricular conduction delay. J Am Coll Cardiol 2002;39:2026–2033. Abraham WT, Fisher WG, Smith AL, et al. MIRACLE Study Group. Multicenter InSync Randomized Clinical Evaluation. Cardiac resynchronization in chronic heart failure. N Engl J Med 2002;346:1845–1853. Cleland JG, Daubert JC, Erdmann E, et al. Cardiac Resynchronization-Heart Failure (CARE-HF) Study Investigators. The effect of cardiac resynchronization on morbidity and mortality in heart failure. N Engl J Med 2005;352: 1539 –1549. Cleland JG, Daubert JC, Erdmann E, et al. Longer-term effects of cardiac resynchronization therapy on mortality in heart failure [the CArdiac REsynchronization-Heart Failure (CARE-HF) trial extension phase]. Eur Heart J 2006;27:1928–1932. St. John Sutton MG, Plappert T, Abraham WT, et al. Multicenter InSync Randomized Clinical Evaluation (MIRACLE) Study Group. Effect of cardiac resynchronization therapy on left ventricular size and function in chronic heart failure. Circulation 2003;107:1985–1990. Saxon LA, De Marco T, Schafer J, Chatterjee K, Kumar UN, Foster E, VIGOR Congestive Heart Failure Investigators. Effects of long-term biventricular stimulation for resynchronization on echocardiographic measures of remodeling. Circulation 2002;105:1304 –1310.
increased with increasing quartiles of BUN:SCr (Q1: HR ⫽ 1.06 [P ⫽ .79], Q2: HR ⫽ 0.64 [P ⫽ .04], Q3: HR ⫽ 0.47 [P⬍.001], Q4: HR ⫽ 0.44 [P⬍.001]; P-value for trend ⫽ .005). CONCLUSIONS In MADIT-CRT, patients with an elevated ratio of BUN to SCr experienced a significantly greater reduction in the risk of HF or death with CRT-D therapy as compared with patients with a low ratio. These findings suggest an association between prerenal function and response to CRT. KEYWORDS Cardiac resynchronization therapy; Implantable cardioverter defibrillator; Heart failure; Renal function; Mortality ABBREVIATIONS: BUN ⫽ blood urea nitrogen; CI ⫽ confidence interval; CRT ⫽ cardiac resynchronization therapy; CRT-D ⫽ cardiac resynchronization therapy with defibrillator; EF ⫽ ejection fraction; eGFR ⫽ estimated glomerular filtration rate; HF ⫽ heart failure; ICD ⫽ implantable cardioverter-defibrillator; LV ⫽ left ventricular; MADIT-CRT ⫽ Multicenter Automatic Defibrillator Implantation Trial—Cardiac resynchronization therapy; MADIT-II ⫽ Multicenter Automatic Defibrillator Implantation Trial-II; NYHA ⫽ New York Heart Association; SCr ⫽ serum creatinine; SD ⫽ standard deviation
Impaired renal function is an important independent risk factor for morbidity and mortality in patients with heart failure (HF).1– 6 Even minor abnormalities in renal function have been shown to impair the normal regulation of fluid volume and increase the frequency of acute HF exacerbations, underscoring the link between the heart and the kid-
ney.7 Furthermore, a state of prerenal azotemia may exist among patients with systolic left ventricular (LV) dysfunction due to redistributed excess total body salt and water and a diminished effective circulating blood volume. This clinical status may be manifested in renal function tests as disproportionately elevated blood urea nitrogen (BUN) levels relative to serum creatinine (SCr) levels.8 –10 We have recently shown that an elevated BUN-to-SCr ratio was the most powerful predictor of recurrent HF events in the Multicenter Automatic Defibrillator Implantation Trial-II (MADIT-II).11 Thus, cardiorenal connections may affect outcomes in patients with LV dysfunction with or without overt kidney disease. The Multicenter Automatic Defibrillator Implantation Trial—Cardiac Resynchronization Therapy (MADIT-CRT) demonstrated that CRT with a defibrillator (CRT-D) was associated with a significant reduction in HF or death as
The MADIT-CRT study was supported by a research grant from Boston Scientific, St. Paul, Minnesota, to the University of Rochester School of Medicine and Dentistry. Drs. Moss, Solomon, and Zareba have received research support for the conduct of the MADIT-CRT trial from Boston Scientific through a grant to the University of Rochester. Address reprint requests and correspondence: Ilan Goldenberg, M.D., Heart Research Follow-up Program, Box 653, University of Rochester Medical Center, Rochester, New York 14642. E-mail address: Ilan. [email protected]. (Received July 28, 2010; accepted September 2, 2010.)
(Heart Rhythm 2010;7:1777–1782) © 2010 Heart Rhythm Society. All rights reserved.
1547-5271/$ -see front matter © 2010 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2010.09.005
1778
Heart Rhythm, Vol 7, No 12, December 2010
compared with defibrillator-only therapy among asymptomatic or mildly symptomatic patients with reduced ejection fraction (EF) and a wide QRS duration.12 However, the benefit of CRT-D in this population is not uniform.12 It is possible that renal regulation may affect the response to this mode of HF therapy. We hypothesized that renal markers of effective circulating blood volume relate to the benefit of CRT.
the classification of deaths15); and (3) the incremental echocardiographic response to CRT-D versus ICD therapy, defined as the percent reduction in cardiac volumes (LV end-systolic and end-diastolic volumes [LVESV and LVEDV, respectively]) between enrollment and 1 year (calculated as the difference between 1-year cardiac volumes and baseline cardiac volumes, divided by baseline cardiac volumes).
Methods
Statistical analysis
MADIT-CRT
The clinical characteristics of study patients were compared by BUN:SCr categories using the 2-test for categorical variables and the t-test for continuous variables. KaplanMeier estimates for HF or all-cause mortality in each treatment group, stratified according to the renal function categories, were determined and statistically evaluated with the log-rank test. The Cox proportional-hazards regression model was used to evaluate the independent contribution of baseline clinical factors to the development of the endpoint. All models included the following prespecified covariates: age ⬎65 years; gender; ischemic status; QRS ⬎150 ms; EF ⬍25%; diabetes mellitus; left bundle branch block; systolic blood pressure ⱖ130 mmHg; heart rate ⱖ80 bpm; and medical therapies with beta-blockers, angiotensin-converting enzyme inhibitors, and diuretics. The benefit of CRT-D compared with ICD-only therapy in reducing the primary endpoint in each renal function category was assessed by including a renal function-by-treatment interaction term in the multivariate models. In the quartile analyses, the probability value for trend was determined by using the four quartiles of each renal function parameter as an ordinal categorical variable in the multivariate models. Multivariate regression analysis was used to evaluate the incremental echocardiographic response to CRT-D versus ICD-only therapy (assessed as percent reductions in LVESV and LVEDV between baseline and 1-year follow-up echocardiograms). Regression models included the covariates described above. The benefit of CRT-D compared with ICD-only therapy in reducing cardiac volumes was assessed in each BUN:SCr category by including a treatment-by-BUN:SCr interaction term in the multivariate regression models. The Proportional hazards assumption for each renal parameter was tested by including the treatment-bytime interaction term in the multivariate Cox models. All P-values were two-sided, and a P-value ⬍.05 was considered statistically significant. Analyses were performed with the use of SAS software (ver. 9.20, SAS Institute, Cary, North Carolina).
The design and results of MADIT-CRT have been reported elsewhere.12 Briefly, 1,820 patients with ischemic or nonischemic cardiomyopathy, an EF of 30% or less, a QRS duration of 130 ms or more, and New York Heart Association (NYHA) class I or II symptoms were randomly assigned in a 3:2 ratio to receive CRT-D or an implantable cardioverter-defibrillator (ICD) alone. Exclusion criteria included coronary revascularization or myocardial infarction within the previous 3 months and the presence of any other disease associated with a reduced likelihood of survival for the duration of the trial, including advanced kidney disease (BUN ⬎70 mg/dL or SCr ⬎3.0 mg/dL). After an average follow-up of 2.4 years, treatment with CRT-D was associated with a significant 34% (P ⬍.001) reduction in the risk of the primary endpoint of the study, comprising death from any cause or a nonfatal HF event. Two-dimensional echocardiography was obtained at baseline and at 1-year follow-up to assess change in LV volumes and EF during this time period in the two treatment arms (CRT-D with CRT on at 1 year, n ⫽ 766; ICD only, n ⫽ 628).
Definitions and outcomes Renal function tests at enrollment were available for 1,803 (99%) patients who made up the present study population. The renal function parameters that were evaluated included BUN, SCr, and the estimated glomerular filtration rate (eGFR). Combined assessment of BUN and SCR (prespecified as BUN:SCr) was used as a marker of prerenal function.8,9 The individual renal function variables (including BUN, SCr, and eGFR) were categorized into approximate quartiles, and the ratio of BUN to SCr was analyzed both by dichotomizing values at the median level and by categorization into approximate quartiles. Estimated GFR was assessed using the four-component Modification of Diet in Renal Disease equation incorporating age, race, sex, and SCr level: eGFR (mL/min/1.73 m2) ⫽ 186 ⫻ (SCr level [in mg/dL])⫺1.154 ⫻ (age [in years])⫺0.203. For women and blacks, the product of this equation was multiplied by a correction factor of 0.742 and 1.21, respectively.13 The presence of kidney disease was defined as eGFR ⬍60 mL/min/1.73 m2, according to the guidelines of the National Kidney Foundation.14 We evaluated the relationship between each of the prespecified renal function parameters and (1) the benefit of CRT-D versus ICD therapy in reducing the primary endpoint of death or a nonfatal first HF event; (2) the benefit of CRT-D versus ICD therapy in reducing the separate endpoints of nonfatal HF, all-cause mortality, and nonfatal HF or HF-related death (using a modified Hinkle-Thaler system for
Results Among the 1,803 study patients, mean ⫾ standard deviation (SD) values of the prespecified renal parameters included BUN levels of 21.5 ⫾ 8.9 mg/dL (median 20.0 mg/dL; interquartile range 15.8 –25.0 mg/dL), SCr levels of 1.2 ⫾ 0.4 mg/dL (median 1.1 mg/dL; interquartile range 0.9 –1.3 mg/dL), and eGFR levels of 66 ⫾ 27 mL/min/1.73 m2 (median 66 mL/min/ 1.73 m2; interquartile range 46 – 85 ml/min/1.73 m2). Values of the ratio BUN to SCr were normally distributed in the study population, with a mean ⫾ SD of 18.6 ⫾ 5.8 (median 17.8; interquartile range 14.8 –22.0).
Goldenberg et al
CRT and Renal Function
1779
Table 1 Baseline clinical characteristics of study patients by BUN to creatinine ratio BUN to creatinine ratio (median ⫽ 18)
Characteristic BUN, mg/dL Creatinine, mg/dL eGFR, mL/min/1.73 m2 CRT-D Ischemic cardiomyopathy Age, years Age ⬎65 years Female gender NYHA class II Hypertension Diabetes mellitus Past coronary artery bypass surgery Past percutaneous coronary revascularization QRS duration ⱖ150 ms Left bundle branch block EF, mean ⫾ SD EF ⬍25% Atrial fibrillation ⬎1 month before enrollment Cigarette smoking anytime Systolic blood pressure ⱖ130 mmHg Diastolic blood pressure ⱕ80 mmHg Heart rate per ⱖ80 bpm Body mass index ⱖ30 kg/ m2 Apical lead position† Baseline echocardiogram LVESV, mL LVEDV, mL Medications Angiotensin-converting enzyme inhibitors Angiotensin receptor blockers Class I antiarrhythmics Digitalis -blockers Amiodarone Statins Diuretics Furesemide dose
were similar between CRT-D and ICD patients within the two BUN:SCr subgroups (not shown).
Relation of individual renal function parameters to clinical benefit of CRT-D
ⱕMedian (n ⫽ 936)
⬎Median (n ⫽ 867)
17 ⫾ 6 1.2 ⫾ 0.4 71 ⫾ 25 61 56 63 ⫾ 11 44 20 85 63 25 29
26 ⫾ 9* 1.1 ⫾ 0.3* 64 ⫾ 28* 58 54 66 ⫾ 10* 55* 30* 86 64 35* 29
28
26
63 68 24 ⫾ 5 60 11
66 73* 24 ⫾ 5 59 12
15 38
9* 33*
83
88*
36
37
15
12
177 ⫾ 51 248 ⫾ 63
177 ⫾ 49 248 ⫾ 60
76
77
19
22
1 26 93 8 68 69 48 ⫾ 45
0.6 26 93 6 67 81* 52 ⫾ 40
Values are percentages or mean ⫹/- SD. *P⬍.05 for the difference between BUN:SCr categories. †Data regarding LV lead position were available for 799 of the 1,089 patients who were allocated to CRT-D at enrollment (74%).
The baseline characteristics of study patients by BUN:SCr are shown in Table 1. Patients with BUN:SCr values in the abovemedian range (⬎18) comprised a higher proportion of females and exhibited higher frequency risk factors, including an older age, diabetes mellitus, and left bundle branch block, as compared with study subjects who had a lower BUN:SCr. Baseline medications were administered to a similar proportion of patients in the two BUN:SCr subgroups with the exception of diuretics, which were administered more frequently to patients with a higher BUN: SCr (Table 1). Baseline clinical characteristics and medications
Multivariate analysis showed that increased levels of each of the individual renal function parameters was independently associated with a significant increase in the risk of HF or death in the total study population (BUN ⱖ20 mg/dL, HR ⫽ 1.38 [95% confidence interval (CI) 1.07–1.78], P ⫽ .01; SCr ⬎1.1: HR ⫽ 1.46 [95% CI 1.13–1.86], P ⫽ .003; eGFR ⬍60 mL/ min/1.73 m2; HR ⫽ 1.49 [95% CI 1.15–1.93], P ⫽ .002). The benefit of treatment with CRT-D as compared with ICD-only therapy was shown to be directly related to BUN and inversely related to SCr (Table 2). Thus, multivariate analysis showed that risk reduction associated with CRT-D therapy was highest among patients with upper quartile BUN (57% reduction in the risk of HF or death compared with ICD-only therapy [P⬍.001]) and lower quartile SCr (60% reduction in the risk of HF or death compared with ICD-only therapy [P ⫽ .001]). In contrast, among patients with lower quartile BUN and upper quartile SCr there was no statistically significant difference in the risk of HF or death between the two treatment arms (Table 2). The benefit of CRT-D was similar between patients with and without kidney disease (eGFR ⬍60 mL/min/1.73 m2; HR ⫽ 0.67 [95% CI 0.50 – 0.89]; eGFR ⱖ60 mL/min/1.73 m2; HR ⫽ 0.61 [95% CI 0.50 – 0.89]; P-value for interaction ⫽ .76). However, consistent with the SCr results, patients with eGFR levels in the upper quartile range experienced the most pronounced risk reduction with CRT-D as compared with ICDonly therapy (HR ⫽ 0.42 [95% CI 0.22–0.82]; P ⫽ .01).
Relation of the ratio of BUN to SCr to the clinical benefit of CRT-D Combined assessment of BUN and SCr levels demonstrated a significant difference in the benefit of CRT-D between patients who exhibited BUN:SCr levels above and below median values (Table 3). Accordingly, multivariate analysis showed that the overall 37% (P ⬍.001) reduction in the risk of HF or death with CRT-D in the total study population was related to a pronounced 54% (P ⬍.001) risk reduction among patients with BUN:SCr ⬎18, whereas among patients with lower BUN:SCr values, treatment with CRT-D was not associated with a statistically significant benefit (HR ⫽ 0.85 [P ⫽ .28]; P-value for BUN:SCr-by-treatment interaction ⫽ .005; Table 3). Consistent with these findings, patients with BUN:SCr ⬎18 experienced pronounced and significant reductions in the rates of HF or death with CRT-D as compared with ICD-only therapy (Figure 1A), whereas patients with BUN:Cr ⱕ18 exhibited virtually identical rates of HF or death in the two treatment groups at 2 and 3 years of follow-up (Figure 1B). Notably, the clinical benefit of CRT-D therapy among patients with BUN:SCr ⬎18 was more pronounced at ⬎1 year of follow-up than during the first year after enrollment (Figure 1A). However, this time-related difference in the benefit of CRT-D was not
1780 Table 2
Heart Rhythm, Vol 7, No 12, December 2010 Multivariate analysis: CRT-D versus ICD-only risk of HF or death by individual renal function parameters
BUN, mg/dL: Q1: ⬍15 (n ⫽ 389) Q2: 15–19 (n ⫽ 499) Q3: 20–24 (n ⫽ 430) Q4: ⱖ25 (n ⫽ 485) SCr, mg/dL: Q1: ⬍0.9 (n ⫽ 446) Q2: 0.9–1.1 (n ⫽ 511) Q3: ⬎1.1–1.3 (n ⫽ 432) Q4: ⬎1.3 (n ⫽ 413)
Adjusted hazard ratio
95% Confidence interval
P
P-value for trend
0.81 0.69 0.64 0.43
0.50–1.28 0.46–0.99 0.43–0.97 0.28–0.68
.37 .04 .04 ⬍.001
.08
0.40 0.52
0.24–0.71 0.35–0.79
.001 .001
.09
0.83
0.61–1.32
.67
0.78
0.48–1.09
.16
Note: All models were adjusted for the following additional covariates: age ⬎65 years; gender; ischemic status; QRS ⬎150 ms; EF ⬍25%; diabetes mellitus; left bundle branch block; systolic blood pressure ⱖ130 mmHg; heart rate ⱖ80 bpm; BUN ⱖ20 mg/dL (in the SCr model); SCr ⬎1.1 mg/dL (in the BUN model); and medical therapies with beta-blockers, angiotensin-converting enzyme inhibitors, and diuretics. CRT-D versus ICD efficacy in each renal parameter subset was assessed using by adding renal function-by-treatment interaction terms to the multivariate models.
statistically significant (P-value for treatment-by-time interaction in the multivariate Cox model ⫽ .61). Similar to the results that related CRT-D benefit to median BUN:SCr levels, quartile analysis showed a significant increase in the benefit of CRT with increasing BUN:SCr quartiles (P-value for trend ⫽ .005; Table 3) that corresponded to a direct correlation between increasing BUN: SCr quartiles and the cumulative event rate reduction at 2.5 years of follow-up associated with CRT-D (Figure 2). Notably, the significant 56% (P ⬍.001) reduction in the risk of HF or death associated with CRT-D among patients in the highest BUN:SCr quartile was related to a 52% (P ⫽ .06) reduction in all-cause mortality, a 59% (P ⫽ .001) reduction in the risk of nonfatal HF, and a 61% (P ⬍.001) reduction in the risk of nonfatal HF or HF-related death.
Relation of BUN-to-SCr ratio to echocardiographic response to CRT-D There was no significant difference in baseline LVESV and LVEDV between patients with low or elevated BUN:SCr Table 3
(Table 1). Despite this, however, percent reduction in both LVESD and LVEDV between baseline and 1-year values among CRT-D-treated patients displayed a direct correlation with increasing BUN:SCr quartiles (Figure 3A and 3B, respectively), whereas reductions in cardiac volumes in the ICD-only arm of the trial were significantly lower and unrelated to BUN:SCr (not shown). Consistently, multivariate regression analysis showed that the incremental reductions in cardiac volumes with CRT-D as compared with ICD-only therapy displayed a direct correlation with increasing BUN:SCr quartiles (LVESV: P-value for trend ⫽ .004; LVEDV: P-value for trend ⫽ .005).
Discussion To our knowledge, the current study is the first to evaluate the relation between markers of renal function and response to CRT. Assessment of individual renal function parameters showed that the benefit of treatment with CRT-D was directly correlated with BUN levels and inversely related to SCr. Furthermore, combined assessment of the two renal markers was shown to identify a significant differential
Multivariate analysis: CRT-D versus ICD-only risk of HF or death by BUN:SCr
BUN:SCr All patients By BUN:SCr: Median: Lower: ⱕ18 (n ⫽ 936) Upper: ⬎18 (n ⫽ 837) Quartiles: Q1: ⬍15 (n ⫽ 459) Q2: 15-18 (n ⫽ 477) Q3: ⬎18-22 (n ⫽ 426) Q4: ⬎22 (n ⫽ 441)
Adjusted hazard ratio
95% Confidence interval
P
0.63
0.51–0.77
⬍.001
0.85
0.61–1.14
.28
0.46
0.34–0.62
⬍.001
1.06 0.64
0.70–1.60 0.41–0.99
.79 .04
0.47
0.30–0.74
⬍.001
0.44
0.29–0.66
⬍.001
P-value for interaction .005
P-value for trend .005
Note: All models were adjusted for the following additional covariates: age ⬎65 years; gender; ischemic status; QRS ⬎150 ms; EF ⬍25%; diabetes mellitus; left bundle branch block; systolic blood pressure ⱖ130 mmHg; heart rate ⱖ80 bpm; SCr ⬎1.1 mg/dL; BUN ⱖ20 mg/dL; and medical therapies with beta-blockers, angiotensin-converting enzyme inhibitors, and diuretics. CRT-D versus ICD efficacy in BUN:SCr subset was assessed by adding a BUN:SCr-by-treatment interaction term to the multivariate models. Virtually identical results were obtained in multivariate models that included further adjustment for time-dependent furosemide dosage.
Goldenberg et al
CRT and Renal Function
1781
Figure 3 Reduction in percent of (A) LVESVs and (B) LVEDVs among patients in the CRT arm by BUN:SCr quartiles. Percent reduction was calculated as (1-year volume ⫺ baseline volume)/baseline volume.
Figure 1 Kaplan-Meier estimates of the probability of heart failure or death by treatment in patients with (A) BUN:SCr ⬎18 mg/dL and (B) BUN:SCr ⱕ18 mg/dL.
clinical and echocardiographic response to CRT-D in the MADIT-CRT population. These findings suggest that cardiorenal interactions relate to response to CRT. In patients with LV dysfunction, an elevated BUN in the setting of relatively normal SCr may reflect a physiological response to a relative decrease of blood flow to the kidney due to diminished effective circulating blood volume.8 Thus, an increase in the ratio of BUN to SCr may identify HF patients who have not yet developed intrinsic renal impairment despite a reduction in renal blood flow, possibly due to adaptive neurohormonal mechanisms to preserve glomerular filtration rate.16 Consistent with this mechanism,
Figure 2 Two-and-a-half-year Kaplan-Meier rates of HF or death in the two treatment groups by BUN:SCr quartiles. The time point of 2.5 years for assessment of event rates was selected since this was the average follow-up among study patients enrolled in the trial.
study patients with increased BUN:SCr exhibited high BUN levels in the presence of normal SCr (Table 1). CRT has been demonstrated to be an efficacious therapy for patients with reduced LV function and ventricular dyssynchrony, improving LV function, symptomatic and functional status, and survival.17–22 The primary mechanism by which this is achieved is improved ventricular pumping efficiency through better temporal coordination of LV activation and contraction.23,24 MADIT-CRT was the first clinical trial to show a significant reduction in the risk of HF or death with CRT-D in a population of patients with LV dysfunction and asymptomatic or mildly symptomatic HF.12 Data from the current study suggest that the benefit of CRT-D in the trial may be related to prerenal function. We have shown that CRT-D therapy was associated with a significantly greater benefit in patients with an elevated ratio of BUN to SCr as compared with those who had a lower ratio. These findings extend recent data from the MADIT-II trial, in which we have shown that an elevated ratio of BUN to SCr was the most powerful predictor for the development of recurrent HF events after ICD implantation.11 Thus, markers of prerenal failure may identify patients with diminished effective circulating blood volume and increased risk for HF events who exhibit a favorable echocardiographic and clinical response to CRT. Notably, the clinical benefit of CRT-D among patients with elevated BUN:SCr was more pronounced after the first year follow-up in the trial, possibly due to the fact that a longer follow-up is required for the echocardiographic benefit of resynchronization therapy to translate into a clinical benefit. In contrast, a lower BUN-to-SCr ratio in the present study occurred in the setting of relatively normal BUN and SCr levels (Table 1), possibly among lower risk patients, who have not yet developed a marked reduction in effective circulating blood volume
1782 and prerenal azotemia, in whom early intervention with CRT-D may not result in a pronounced benefit.
Study limitations Renal function was not a randomization factor in MADITCRT-D, and differences in CRT-D benefit among the prespecified renal function categories may be related to unmeasured risk factors. It should also be noted that a significantly greater proportion of patients with elevated BUN:SCr received diuretic therapy. Our findings, however, persisted after multivariate adjustment for medication usage, including diuretic therapy and furosemide dosage. Furthermore, subgroup analyses demonstrated that the relation between BUN:SCr and CRT-D efficacy was consistent among patients who did and did not receive diuretic therapy (P-value for trend for CRT-D benefit by increasing BUN:SCr quartiles ⬍.05 in both subgroups). Thus, it is unlikely that medication usage affected the present results significantly. Follow-up data regarding renal function parameters were not collected in MADIT-CRT. Therefore, the ratio of BUN to SCr was not assessed as a time-dependent covariate in the multivariate models. Furthermore, data regarding sodium and serum osmolality, as additional markers of intravascular status, were not collected in the trial. Thus, the implications of the current findings are limited to the association between baseline renal function at enrollment and subsequent outcome during an average follow-up period of 2.4 years. MADIT-CRT enrolled asymptomatic or mildly symptomatic HF patients with reduced EF. Therefore, it is unclear whether the findings from the present study are applicable to patients who are currently being referred for treatment with CRT in the community, who may be sicker and have more advanced HF symptoms. Furthermore, patients with advanced renal dysfunction were excluded from MADIT-CRT. Only 89 study patients (5%) had severe kidney disease (defined as eGFR ⬍30 mL/min/1.73 m2).14 Thus, our results regarding the relation between the benefit of CRT-D and levels of BUN and SCr may not reflect the outcome with CRT-D among patients with severe kidney disease.
Heart Rhythm, Vol 7, No 12, December 2010
2. 3.
4.
5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
Conclusions and clinical implications Renal dysfunction is frequently observed in HF patients and is strongly correlated with the severity and prognosis of this condition.1–7 However, evidence for therapeutic strategies to actually reduce mortality and morbidity in HF patients with renal impairment is still lacking. Our findings suggest that CRT-D is a highly effective therapy for mildly symptomatic low-EF patients who exhibit markers of prerenal azotemia. In this population, we have shown that CRT-D was associated with a ⬎50% reduction in the risk of HF or death and a pronounced echocardiographic response, whereas among patients with a low BUN:SCr, CRT-D therapy was associated with a significantly lower clinical and echocardiographic response. These findings suggest that cardiorenal interactions should be assessed among candidates for CRT.
References 1. Dries DL, Exner DV, Domanski MJ, Greenberg B, Stevenson LW. The prognostic
19.
20.
21.
22.
23.
24.
implications of renal insufficiency in asymptomatic and symptomatic patients with left ventricular systolic dysfunction. J Am Coll Cardiol 2000;35:681–689. Hillege HL, Girbes AR, de Kam PJ, et al. Renal function, neurohormonal activation, and survival in patients with chronic heart failure. Circulation 2000;102:203–210. Al-Ahmad A, Rand WM, Manjunath G, et al. Reduced kidney function and anemia as risk factors for mortality in patients with left ventricular dysfunction. J Am Coll Cardiol 2001;38:955–962. Hillege HL, Nitsch D, Pfeffer MA, et al. Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity (CHARM) Investigators. Renal function as a predictor of outcome in a broad spectrum of patients with heart failure. Circulation 2006;113:671– 678. Anavekar NS, McMurray JJ, Velazquez EJ, et al. Relation between renal dysfunction and cardiovascular outcomes after myocardial infarction. N Engl J Med 2004;351:1285–1295. Forman DE, Butler J, Wang Y, et al. Incidence, predictors at admission, and impact of worsening renal function among patients hospitalized with heart failure. J Am Coll Cardiol 2004;43:61– 67. Krumholz HM, Chen YT, Vaccarino V, et al. Correlates and impact on outcomes of worsening renal function in patients ⬎ or ⫽ 65 years of age with heart failure. Am J Cardiol 2000;85:1110 –1113. Morgan DB, Carver ME, Payne RB. Plasma creatinine and urea: creatinine ratio in patients with raised plasma urea. BMJ 1977;2:929 –932. Dossetor JB. Creatininemia versus uremia. The relative significance of blood urea nitrogen and serum creatinine concentrations in azotemia. Ann Intern Med 1966;65: 1287–1299. Lin HJ, Chao CL, Chien KL, et al. Elevated blood urea nitrogen-to-creatinine ratio increased the risk of hospitalization and all-cause death in patients with chronic heart failure. Clin Res Cardiol 2009;98:487– 492. Sze E, Moss AJ, McNitt S, Barsheshet A, Andrews ML, Zareba W, Goldenberg I, for the Multicenter Automatic Defibrillator Implantation Trial-II Investigators. Risk Factors for Recurrent Heart Failure Events in the Multicenter Automatic Defibrillator Implantation Trial II (MADIT-II). J Cardiovasc Electrophysiol 2010. Moss AJ, Hall WJ, Cannom DS, et al. MADIT-CRT Trial Investigators. Cardiac-resynchronization therapy for the prevention of heart-failure events. N Engl J Med 2009;361:1329 –1338. Levey AS, Bosch JP, Lewis JB, Greene T, Rodgers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130:461– 470. National Kidney Foundation. Kidney Disease Outcome Quality Initiative (K/ DOQI) clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39(Suppl 1):S1–S266. Greenberg H, Case RB, Moss AJ, Brown MW, Carroll ER, Andrews ML. Analysis of mortality events in the Multicenter Automatic Defibrillator Implantation Trial (MADIT-II). J Am Coll Cardiol 2004;43:1459 –1465. Givertz MM, Braunwald E. Neurohormones in heart failure: predicting outcomes, optimizing care. Eur Heart J 2004;25:281–282. Cazeau S, Leclercq C, Lavergne T, et al. Multisite Stimulation in Cardiomyopathies (MUSTIC) Study Investigators. Effects of multisite biventricular pacing in patients with heart failure and intraventricular conduction delay. N Engl J Med 2001;344:873– 880. Linde C, Leclercq C, Rex S, et al. Long-term benefits of biventricular pacing in congestive heart failure: results from the MUltisite STimulation in cardiomyopathy (MUSTIC) study. J Am Coll Cardiol 2002;40:111–118. Auricchio A, Stellbrink C, Sack S, et al. Pacing Therapies in Congestive Heart Failure (PATH-CHF) Study Group. Long-term clinical effect of hemodynamically optimized cardiac resynchronization therapy in patients with heart failure and ventricular conduction delay. J Am Coll Cardiol 2002;39:2026–2033. Abraham WT, Fisher WG, Smith AL, et al. MIRACLE Study Group. Multicenter InSync Randomized Clinical Evaluation. Cardiac resynchronization in chronic heart failure. N Engl J Med 2002;346:1845–1853. Cleland JG, Daubert JC, Erdmann E, et al. Cardiac Resynchronization-Heart Failure (CARE-HF) Study Investigators. The effect of cardiac resynchronization on morbidity and mortality in heart failure. N Engl J Med 2005;352: 1539 –1549. Cleland JG, Daubert JC, Erdmann E, et al. Longer-term effects of cardiac resynchronization therapy on mortality in heart failure [the CArdiac REsynchronization-Heart Failure (CARE-HF) trial extension phase]. Eur Heart J 2006;27:1928–1932. St. John Sutton MG, Plappert T, Abraham WT, et al. Multicenter InSync Randomized Clinical Evaluation (MIRACLE) Study Group. Effect of cardiac resynchronization therapy on left ventricular size and function in chronic heart failure. Circulation 2003;107:1985–1990. Saxon LA, De Marco T, Schafer J, Chatterjee K, Kumar UN, Foster E, VIGOR Congestive Heart Failure Investigators. Effects of long-term biventricular stimulation for resynchronization on echocardiographic measures of remodeling. Circulation 2002;105:1304 –1310.