Abstract
Aims
Recent observations in chronic stable heart failure suggest that high-dose loop diuretics (HDLDs) have detrimental prognostic effects in patients with high blood urea nitrogen (BUN), but recent findings have also indicated that diuretics may improve renal function. Carbohydrate antigen 125 (CA125) has been shown to be a surrogate of systemic congestion. We sought to explore whether BUN and CA125 modulate the mortality risk associated with HDLDs following a hospitalization for acute heart failure (AHF).
Methods and results
We analysed 1389 consecutive patients discharged for AHF. CA125 and BUN were measured at a mean of 72 ± 12 h after admission. HDLDs (≥120 mg/day in furosemide equivalent dose) were interacted to a four-level variable according to CA125 (>35 U/mL) and BUN (above the median), and related to all-cause mortality. At a median follow-up of 21 months, 561 (40.4%) patients died. The use of HDLDs was independently associated with increased mortality [hazard ratio (HR) 1.23, 95% confidence interval (CI) 1.01–1.50], but this association was not homogeneous across CA125–BUN categories (P for interaction <0.001). In patients with normal CA125, use of HDLDs was associated with high mortality if BUN was above the median (HR 2.29, 95% 1.51–3.46), but not in those with BUN below the median (HR 1.22, 95% CI 0.73–2.04). Conversely, in patients with high CA125, HDLDs showed an association with increased survival if BUN was above the median (HR 0.73, 95% CI 0.55–0.98) but was associated with increased mortality in those with BUN below the median (HR 1.94, 95% CI 1.36–2.76).
Conclusion
The risk associated with HDLDs in patients after hospitalization for AHF was dependent on the levels of BUN and CA125. The information provided by these two biomarkers may be helpful in tailoring the dose of loop diuretics at discharge for AHF.
Keywords: Loop diuretics, Mortality, Acute heart failure, Carbohydrate antigen 125, Blood urea nitrogen
Introduction
Loop diuretics are nearly universally used for relieving symptoms of systemic congestion in patients with heart failure (HF), especially during episodes of clinical decompensation.1,2 Paradoxically, a number of studies3−6 have reported an increased risk for adverse outcomes associated with higher doses of loop diuretic treatment. Furthermore, the optimal use of loop diuretics remains a real clinical challenge, and their dose titration is largely intuitive.
Blood urea nitrogen (BUN) is a well-known marker of renal function, and hence its serum concentrations vary according to changes in glomerular filtration rate (GFR). Recent studies7−9 have also shown that BUN correlates with neurohormonal activation parameters. Testani et al.9 reported a significant interaction between high-dose loop diuretics (HDLDs) and BUN in a selected cohort of 2456 compensated HF patients with left ventricular ejection fraction (LVEF) ≤35%. These authors found that the risk associated with HDLD use was strongly dependent on BUN concentrations, with reduced survival when BUN was above the median.
On the other hand, recent studies10–13 have also highlighted the importance of venous congestion in the pathophysiology of renal dysfunction in HF. For instance, a recent study by Damman et al.13 suggested that furosemide may prevent tubular renal injury in a small group of patients with HF. Nevertheless, it is recognized that the accuracy of symptoms and signs for quantifying systemic congestion in HF is limited.14 In this regard, various studies15−18 have suggested that the serum tumour marker carbohydrate antigen 125 (CA125) may be a reliable surrogate for systemic congestion, and associated with adverse outcomes in acute and chronic HF. Thus, we hypothesize that the prognostic effect of HDLDs is modulated by the balance between the beneficial decongestion vs. the negative neurohormonal effect.
We sought to explore, in a cohort of patients hospitalized with acute heart failure (AHF), the relationship between discharge HDLD and all-cause mortality, and whether the association is modulated by surrogate markers of systemic congestion (CA125) and renal dysfunction/neurohormonal activation (BUN).
Methods
Study group and protocol
We prospectively studied a cohort of 1538 patients consecutively admitted to the cardiology department from a tertiary hospital (Hospital Clínico Universitario de Valencia) with the diagnosis of AHF. AHF was defined according to current guidelines.1,2,19 Patients were followed-up from hospital discharge occurring between 1 January 2004 and 9 March 2011. By design, patients who died (n = 80) or received a heart valve replacement during the index hospitalization were excluded (n = 69), leaving 1389 patients as the study sample. In addition, patients with a final diagnosis of acute coronary syndrome, active sepsis/pneumonia, terminal cancer, or end-stage renal disease on dialysis were excluded from the study. Demographic information, medical history, vital signs, 12-lead electrocardiogram, and laboratory and drug utilization data were routinely determined on admission and throughout the hospital course, using pre-established registry questionnaires. All patients received intravenous treatment with furosemide for at least the first 48 h. LVEF was assessed with echo (Agilent Sonos 5500-Phillips) during the index hospitalization. Treatment with angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta-blockers, aldosterone antagonists, anticoagulants, diuretics, and other therapeutic strategies were individualized following established guidelines operating at the time the patient was recruited in the registry.1,2,19
Outcomes
Patients were followed-up until death, lost to follow-up, valve replacement, or cardiac transplantation. All-cause mortality was selected as the main endpoint, and cardiovascular (CV) mortality as a secondary endpoint. The information regarding the cause of death was extracted from the patients' clinical chart, and adjudicated by an investigator who was blinded to the hypothesis of the study. Deaths related to CV aetiology included sudden death, progressive HF death, and other CV-related deaths. Moreover, those patients who died outside the hospital (n = 157), in which the circumstances concerning the death were unknown, were assumed to be CV in origin for the purpose of analysis. This study conforms to the principles outlined in the Declaration of Helsinki and was approved by an institutional review committee. All patients gave informed consent.
Blood urea nitrogen and carbohydrate antigen 125 measurements
Blood urea nitrogen and serum CA125 were obtained simultaneously during patients' hospitalization (at a mean of 72 ± 12 h after admission). CA125 was measured using commercially available immunoassay kits (Elecsys CA125 II assay-Roche Diagnostics) and BUN was measured using a kinetic test with urease (Roche-Hitachi systems cobas c). A derived variable was constructed dichotomizing BUN by its median ( < 24.8 mg/dL or ≥24.8 mg/dL) and serum CA125 by its upper limit of normality (<35 U/mL or ≥35 U/mL). Hence, the study sample was stratified on the following four categories: C1 (n = 239), with low CA125 and high BUN; C2 (n = 269), with low CA125 and low BUN; C3 (n = 437), with high CA125 and low BUN; and C4 (n = 444), with high CA125 and high BUN.
Loop diuretic treatment
Overall, patients' treatment decisions were left at the discretion of the cardiologist in charge of the patient. No specific recommendations regarding prescription of diuretics were followed according to the levels of any marker. All patients were discharged on diuretics (furosemide = 69.1%, torasemide = 24.9%, furosemide + hydrochlorothiazide (HCTZ) = 4.2%, torasemide + HCTZ = 1.44%, and HCTZ alone = 0.36%). Total loop diuretic dose (mg/day) was converted to furosemide equivalent dose (FED) following the equation used by Levy et al.20 The conversion used was furosemide 80 mg = torasemide 40 mg = HCTZ 25 mg. HCTZ contributed only when added to loop diuretics. Thus, five patients were included in the analysis with FED = 0. For the association with mortality risk, FED was explored as continuous and dichotomized according to a pre-specified cut-off point used to define HDLD ( ≥120 mg/day).
Statistical analysis
Continuous variables with and without symmetrical distributions were expressed as mean ±SD and median [interquartile range (IQR)], respectively. For their comparison, Student's t-test, analysis of variance (ANOVA), or Kruskal–Wallis rank test was used as appropriate. Discrete variables were presented as percentages and compared with χ2 test. HDLD mortality rates were depicted among BUN–CA125 categories using the Kaplan–Meier method, and their differences tested by the Peto–Peto Prentice test. As a pre-specified hypothesis, we intentionally tested for homogeneity of the effect of continuous and dichotomized FED (into HDLD ≥120 mg/day) on mortality among the CA125–BUN categories. Multivariable analysis for all-cause mortality was performed by using a flexible parametric survival analysis described by Royston et al.21 Baseline hazard function was modelled with three degrees of freedom (df) restrictive cubic splines (RCS). Candidate covariates for the initial multivariable model were chosen based on previous medical knowledge, and regardless of their P-value. Then, a reduced, although highly predictive model, was derived by backward elimination using a ‘multivariable fractional polynomial’ algorithm.22The final model included as covariates age, gender, obesity, prior admission for AHF, last known New York Heart Association (NYHA) class before admission (under stable conditions), hypertension, diabetes mellitus, history of myocardial infarction, dementia, systolic blood pressure, LVEF <50%, heart rate, atrial fibrillation, serum creatinine, hyponatraemia (sodium <135 mEq/L), anaemia (haemoglobin ≤12 g/L for women and ≤13 g/L for men), brain natriuretic peptide, high sensitivity C-reactive protein, and treatment with beta-blockers, oral anticoagulants, statins, and mineralcorticoid receptor inhibitors. For CV mortality, a multivariable competing risks analysis23 was used, and the estimates are presented as the subdistribution hazard rate (SHR) with 95% confidence intervals (CIs). The CV mortality final model included a similar set of covariates to that of the main model. The proportionality assumption for the hazard function over time was tested by interacting the variables retained in the final model with time. Anaemia, use of angiotensin-converting enzyme inhibitors, and hypertension were included in the final models with time-dependent effects. HDLD mortality rates [expressed as per 10 person-years (PYs)] were estimated from the multivariable regression model. The performance of the survival models was assessed by the Harrell's C-statistic. A two-sided P-value of <0.05 was considered to be statistically significant for all analyses. All analyses were performed using Stata 12 (StataCorp, 2011, Stata Statistical Software: Release 12. College Station, TX, USA).
Results
The mean age in our sample was 72.7 ± 11.5 years; 51% were female, 46.6% exhibited LVEF >50%, 37.9% had prior history of ischaemic heart disease, and median length of stay was 7 days (5–11). The medians (IQR) for FED, BUN, and serum CA125 were 80 (40–100) mg/day, 24.8 (19.2–33.6) mg/dL, and 54 (24–125) U/mL, respectively. All-cause mortality rates are depicted through Kaplan–Meier curves in Supplementary material, Figure S1. Overall, CA125 and BUN markers identified four subpopulations that differ in most of the indicators of disease severity (Table 1). Indeed, patients in C1 and C4 showed the worst baseline risk profile, including higher dose of HDLDs (24% and 33%, respectively, as compared with 17% and 21% for C2 and C3) and the higher mortality risk (Supplementary material, Figure S1). Moreover, patients receiving HDLDs (24.4% of our population) were shown to be sicker and exhibited higher mortality rates (Table 2). The adjusted interaction between CA125 (>35 U/mL/ ≤35 U/mL) and BUN (above/below the median) was not significant (P = 0.930), indicating that the prognostic value of high CA125 did not differ substantially according to BUN status.
Table 1.
Baseline characteristics of the population stratified by carbohydrate antigen 125 and blood urea nitrogen status
| Variables | C1 (n = 239) | C2 (n = 269) | C3 (n = 437) | C4 (n = 444) | Omnibus P-value |
|---|---|---|---|---|---|
| Demographic and medical history | |||||
| Age, years | 76 ± 9 | 70 ± 12 | 69 ± 13 | 76 ± 9 | <0.001 |
| Male, n (%) | 109 (45.6) | 117 (43.5) | 229 (52.4) | 225 (50.7) | 0.076 |
| Previous admission for AHF, n (%) | 142 (59.4) | 181 (67.3) | 321 (73.5) | 267 (60.1) | <0.001 |
| LOS, daysa | 7 (5–10) | 7 (5–9) | 7 (5–11) | 8 (6–12.5) | 0.001 |
| Hypertension, n (%) | 215 (90) | 214 (79.6) | 302 (69.1) | 358 (80.6) | <0.001 |
| Diabetes mellitus, n (%) | 98 (41.0) | 105 (39.0) | 165 (37.8) | 211 (47.5) | 0.020 |
| Dyslipidaemia, n (%) | 119 (49.8) | 130 (48.3) | 162 (37.1) | 221 (49.8) | <0.001 |
| Current smoker, n (%) | 12 (5.0) | 40 (14.9) | 62 (14.2) | 50 (11.3) | 0.001 |
| Previous smoker, n (%) | 51 (21.3) | 47 (17.5) | 95 (21.7) | 93 (20.9) | 0.558 |
| Ischaemic heart disease, n (%) | 100 (41.8) | 91 (33.8) | 145 (33.2) | 191 (43) | 0.006 |
| Valvular heart disease, n (%) | 52 (21.8) | 60 (22.3) | 109 (24.9) | 114 (25.7) | 0.580 |
| ADHF, n (%) | 142 (59.4) | 156 (58) | 336 (76.9) | 327 (73.6) | <0.001 |
| Acute pulmonary oedema, n (%) | 61 (25.5) | 78 (29.0) | 68 (15.6) | 84 (18.9) | <0.001 |
| Hypertensive AHF, n (%) | 32 (13.4) | 34 (12.6) | 31 (7.1) | 22 (5.0) | <0.001 |
| NYHA class III/IV, n (%)b | 51 (21.3) | 33 (12.3) | 67 (15.3) | 102 (23.0) | 0.001 |
| Previous HF, n (%) | 81 (33.9) | 82 (30.5) | 106 (24.3) | 155 (34.9) | 0.004 |
| Previous MI, n (%) | 70 (29.3) | 62 (23) | 98 (22.4) | 142 (32.0) | 0.005 |
| COPD, n (%) | 56 (23.4) | 54 (20.1) | 83 (19) | 103 (23.2) | 0.358 |
| PAD, n (%) | 30 (12.6) | 16 (5.9) | 28 (6.4) | 37 (8.3) | 0.020 |
| Stroke, n (%) | 25 (10.5) | 18 (6.7) | 39 (8.9) | 56 (12.6) | 0.063 |
| Renal failure, n (%) | 72 (30.1) | 11 (4.1) | 18 (4.1) | 126 (28.4) | <0.001 |
| Radiological pleural effusion, n (%) | 57 (23.8) | 64 (23.8) | 243 (55.6) | 253 (57) | <0.001 |
| Peripheral oedema, n (%) | 105 (43.9) | 123 (45.7) | 270 (61.8) | 297 (66.9) | <0.001 |
| Previous use of diuretics, n (%) | 150 (62.8) | 149 (55.4) | 209 (47.8) | 321 (72.3) | <0.001 |
| Previous use of beta-blockers, n (%) | 63 (26.4) | 67 (24.9) | 98 (22.4) | 125 (28.2) | 0.265 |
| Previous use of ACEIs, n (%) | 126 (52.7) | 130 (48.3) | 153 (35) | 211 (47.5) | <0.001 |
| Previous use of ARBs, n (%) | 10 (4.2) | 10 (3.7) | 17 (3.9) | 11 (2.5) | 0.582 |
| Previous use of statins, n (%) | 84 (35.1) | 92 (34.2) | 115 (26.3) | 151 (34) | 0.029 |
| Vital signs | |||||
| Heart rate, b.p.m. | 98 ± 28 | 105 ± 31 | 105 ± 29 | 99 ± 30 | 0.001 |
| SBP, mmHg | 155 ± 39 | 159 ± 38 | 150 ± 34 | 145 ± 34 | <0.001 |
| DBP, mmHg | 83 ± 21 | 87 ± 22 | 85 ± 19 | 79 ± 19 | <0.001 |
| Electrocardiogram | |||||
| Atrial fibrillation, n (%) | 85 (35.6) | 106 (39.4) | 203 (46.5) | 209 (47.1) | 0.008 |
| QRS >120 ms, n (%) | 77 (32.2) | 87 (32.3) | 113 (25.9) | 145 (32.7) | 0.103 |
| Laboratory | |||||
| Haemoglobin, g/dL | 12.5 ± 1.8 | 13.2 ± 1.9 | 12.7 ± 1.8 | 12.2 ± 1.9 | <0.001 |
| Serum creatinine, mg/dL | 1.55 ± 0.73 | 1.02 ± 0.26 | 1.02 ± 0.31 | 1.55 ± 0.65 | <0.001 |
| BUN, mg/dL | 38.5 ± 15 | 19.2 ± 3.6 | 18.8 ± 3.9 | 37.7 ± 13.1 | <0.001 |
| Uric acid, mg/dL | 8.5 ± 2.3 | 7.0 ± 2 | 7.2 ± 2 | 8.7 ± 2.5 | <0.001 |
| Sodium, meq/L | 139.3 ± 3.6 | 139.4 ± 4.7 | 138.7 ± 4.6 | 138.9 ± 4.8 | 0.145 |
| Troponin I, ng/mLa | 0 (0–0) | 0 (0–0) | 0 (0–0) | 0 (0–0) | 0.001 |
| Troponin I >0.2 ng/mL, n (%) | 68 (28.5) | 60 (22.3) | 67 (15.3) | 107 (24.1) | <0.001 |
| BNP, pg/mLa | 128 (70–226) | 110 (72–176) | 156 (94–270) | 208 (101–378) | <0.001 |
| Relative lymphocyte count, % | 19 ± 11 | 20 ± 11 | 18 ± 10 | 17 ± 11 | 0.008 |
| CA125, U/mLa | 19 (13–27) | 19 (14–25) | 102 (62–183) | 97 (56–169) | <0.001 |
| Total cholesterol, mg/dL | 176.3 ± 43.2 | 182.8 ± 42.3 | 167 ± 44.5 | 160.4 ± 43.3 | <0.001 |
| Triglycerides, mg/dL | 137.1 ± 71.7 | 127 ± 59.2 | 107.8 ± 50.3 | 118.3 ± 53 | <0.001 |
| Echocardiography | |||||
| LVEF, % | 53.2 ± 14.3 | 52.2 ± 15.5 | 48.4 ± 15.8 | 49.3 ± 15.5 | <0.001 |
| LVEF ≤50%, n (%) | 88 (36.8) | 109 (40.5) | 220 (50.3) | 230 (51.8) | <0.001 |
| LAD, mm | 41.3 ± 7.6 | 42.2 ± 7.2 | 44.2 ± 8 | 43.8 ± 7.7 | <0.001 |
| LVDD, mm | 54.9 ± 9 | 55.2 ± 9.9 | 56.2 ± 9.9 | 55.8 ± 9.7 | 0.307 |
| Medical treatment | |||||
| Beta-blockers, n (%) | 136 (56.9) | 152 (56.5) | 270 (61.8) | 238 (53.6) | 0.104 |
| Diuretics, n (%) | 239 (100) | 269 (100) | 437 (100) | 444 (100) | 1.00 |
| HDLDs, n (%) | 58 (24.3) | 45 (16.7) | 90 (20.6) | 146 (32.9) | <0.001 |
| Furosemide equivalent dose, mg/daya | 80 (40–80) | 80 (40–80) | 80 (40–80) | 80 (40–120) | <0.001 |
| Mineralcorticoid receptor inhibitors, n (%) | 39 (16.3) | 56 (20.8) | 144 (33) | 83 (18.7) | <0.001 |
| ACEIs, n (%) | 86 (36) | 119 (44.2) | 190 (43.5) | 152 (34.2) | 0.008 |
| ARBs, n (%) | 76 (31.8) | 92 (34.2) | 123 (28.1) | 143 (32.2) | 0.351 |
| Oral anticoagulants, n (%) | 86 (36) | 106 (39.4) | 188 (43) | 183 (41.2) | 0.334 |
| Nitrates, n (%) | 56 (23.4) | 40 (14.9) | 60 (13.7) | 107 (24.1) | <0.001 |
| Digoxin, n (%) | 46 (19.2) | 64 (23.8) | 129 (29.5) | 105 (23.6) | 0.022 |
| Outcomes | |||||
| All-cause mortality, n (%) | 102 (42.7) | 79 (29.4) | 141 (32.3) | 239 (53.8) | <0.001 |
| Cardiovascular mortality, n (%) | 75 (31.4) | 50 (18.6) | 97 (22.2) | 182 (41.0) | <0.001 |
| Follow-up time, yearsa | 1.5 (0.7–3.4) | 2.1 (0.9–4.2) | 2.0 (0.6–3.9) | 1.4 (0.5–3.0) | <0.001 |
Values are expressed as mean ± SD, unless otherwise specified; categorical variables are presented as percentages.
CA125 and BUN categories: C1, CA125 ≤35 U/mlL and BUN above the median; C2, CA125 ≤35 U/mL and BUN below the median; C3, CA125 >35 U/mL and BUN below the median; C4, CA125 >35 U/mL and BUN above the median.
ACEI, angiotensin-converting enzyme inhibitor; ADHF, acute decompensated heart failure; AHF, acute heart failure; ARB, angiotensin II receptor blocker; BNP, brain natriuretic peptide; BUN, blood urea nitrogen; CA125, carbohydrate antigen 125; COPD, chronic obstructive pulmonary disease; DBP, diastolic blood pressure; HDLD, high-dose loop diuretic; HF, heart failure; LAD, left atrial diameter; LOS, length of stay; LVDD, left ventricular diastolic diameter; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NYHA, New York Heart Association; PAD, peripheral artery disease; SBP, systolic blood pressure.
aValue presented as the median (interquartile range).
bLast NYHA functional class measured under clinically stable conditions.
Table 2.
Baseline characteristics of the population according to furosemide equivalent dose at discharge
| Variables | Furosemide equivalent dose <120 mg/day (n = 1050) | Furosemide equivalent dose ≥120 mg/day (n = 339) | P-value |
|---|---|---|---|
| Demographic and medical history | |||
| Age, years | 72 ± 12 | 74 ± 11 | 0.029 |
| Male, n (%) | 510 (48.6) | 170 (50.1) | 0.618 |
| Previous admission for AHF, n (%) | 746 (71) | 165 (48.7) | <0.001 |
| Hypertension, n (%) | 812 (77.3) | 277 (81.7) | 0.095 |
| Diabetes mellitus, n (%) | 408 (38.9) | 171 (50.4) | <0.001 |
| Dyslipidaemia, n (%) | 469 (44.7) | 163 (48.1) | 0.286 |
| Current smoker, n (%) | 129 (12.3) | 35 (10.3) | 0.384 |
| Previous smoker, n (%) | 216 (20.6) | 70 (20.6) | 1.000 |
| Ischaemic heart disease, n (%) | 378 (36) | 149 (44) | 0.010 |
| Valvular heart disease, n (%) | 238 (22.7) | 97 (28.6) | 0.029 |
| ADHF, n (%) | 714 (68) | 247 (72.9) | 0.104 |
| Acute pulmonary oedema, n (%) | 225 (21.4) | 66 (19.5) | 0.490 |
| Hypertensive AHF, n (%) | 98 (9.3) | 21 (6.2) | 0.075 |
| Previous HF, n (%) | 265 (25.2) | 159 (46.9) | <0.001 |
| Previous MI, n (%) | 256 (24.4) | 116 (34.2) | 0.001 |
| COPD, n (%) | 222 (21.1) | 74 (21.8) | 0.819 |
| PAD, n (%) | 76 (7.2) | 35 (10.3) | 0.083 |
| Stroke, n (%) | 105 (10) | 33 (9.7) | 1.000 |
| Renal failure, n (%) | 145 (13.8) | 82 (24.2) | <0.001 |
| Radiological pleural effusion, n (%) | 423 (40.3) | 194 (57.2) | <0.001 |
| Peripheral oedema, n (%) | 568 (54.1) | 227 (67) | <0.001 |
| Previous use of diuretics, n (%) | 559 (53.2) | 270 (79.6) | <0.001 |
| Previous use of beta-blockers, n (%) | 270 (25.7) | 83 (24.5) | 0.668 |
| Previous use of ACEI, n (%) | 451 (43) | 169 (49.9) | 0.028 |
| Previous use of ARB, n (%) | 37 (3.5) | 11 (3.2) | 1.000 |
| Previous use of statins, n (%) | 334 (31.8) | 108 (31.9) | 1.000 |
| Vital signs | |||
| Heart rate, b.p.m. | 104 ± 31 | 97 ± 25 | <0.001 |
| SBP, mmHg | 153 ± 36 | 145 ± 36 | 0.001 |
| DBP, mmHg | 85 ± 21 | 78 ± 18 | <0.001 |
| Electrocardiogram | |||
| Atrial fibrillation, n (%) | 452 (43) | 151 (44.5) | 0.659 |
| QRS >120 ms, n (%) | 311 (29.6) | 111 (32.7) | 0.278 |
| Laboratory | |||
| Haemoglobin, g/dL | 12.7 ± 1.9 | 12.4 ± 1.8 | 0.005 |
| Serum creatinine, mg/dL | 1.23 ± 0.53 | 1.45 ± 0.68 | <0.001 |
| BUN, mg/dL | 27.1 ± 12.8 | 32 ± 16.1 | <0.001 |
| Uric acid, mg/dL | 7.7 ± 2.3 | 8.3 ± 2.6 | <0.001 |
| Sodium, meq/L | 139.1 ± 4.5 | 138.7 ± 4.8 | 0.133 |
| Troponin I, ng/mLa | 0 (0–0.1) | 0 (0–0.06) | 0.083 |
| Troponin I >0.2 ng/mL, n (%) | 230 (21.9) | 72 (21.2) | 0.821 |
| BNP, pg/mLa | 140 (84–252) | 195 (98–356) | <0.001 |
| Relative lymphocyte count, % | 19 ± 11 | 17 ± 9 | 0.001 |
| CA125, U/mLa | 51 (24–113) | 70 (29–165) | <0.001 |
| Total cholesterol, mg/dL | 172.2 ± 44 | 161.5 ± 44 | <0.001 |
| Triglycerides, mg/dL | 121.6 ± 58.4 | 114.6 ± 56.3 | 0.051 |
| Echocardiography | |||
| LVEF, % | 50.9 ± 15.3 | 48.3 ± 15.9 | 0.008 |
| LVEF ≤50%, n (%) | 465 (44.3) | 182 (53.7) | 0.003 |
| LAD, mm | 42.5 ± 7.2 | 45.1 ± 9 | <0.001 |
| LVDD, mm | 55.1 ± 9.5 | 57.5 ± 10.1 | <0.001 |
| Medical treatment | |||
| Beta-blockers, n (%) | 621 (59.1) | 175 (51.6) | 0.016 |
| Diuretics, n (%) | 1050 (100.0) | 339 (100.0) | 1.00 |
| Furosemide equivalent dose, mg/daya | 60 (40–80) | 120 (120–120) | <0.001 |
| Mineralcorticoid receptor inhibitors, n (%) | 223 (21.2) | 99 (29.2) | 0.003 |
| ACEI, n (%) | 426 (40.6) | 121 (35.7) | 0.125 |
| ARB, n (%) | 334 (31.8) | 100 (29.5) | 0.459 |
| Oral anticoagulants, n (%) | 430 (41) | 133 (39.2) | 0.611 |
| Nitrates, n (%) | 179 (17) | 84 (24.8) | 0.002 |
| Digoxin, n (%) | 254 (24.2) | 90 (26.5) | 0.386 |
| Outcomes | |||
| All-cause mortality, n (%) | 352 (33.5) | 209 (61.7) | <0.001 |
| Cardiovascular mortality, n (%) | 259 (24.7) | 145 (42.8) | <0.001 |
| Follow-up time, yearsa | 1.8 (0.7–3.6) | 1.5 (0.5–3.5) | 0.048 |
Values are expressed as mean ± standard deviation, unless otherwise specified; categorical variables are presented as percentages.
ACEI, angiotensin-converting enzyme inhibitor; ADHF, acute decompensated heart failure; AHF, acute heart failure; ARB, angiotensin II receptor blocker; BNP, brain natriuretic peptide; BUN, blood urea nitrogen; CA125, carbohydrate antigen 125; COPD, chronic obstructive pulmonary disease; DBP, diastolic blood pressure; HF, heart failure; LAD, left atrial diameter; LVDD, left ventricular diastolic diameter; LVEF, left ventricular ejection fraction; MI, myocardial infarction; PAD, peripheral artery disease; SBP, systolic blood pressure.
aValue presented as the median (interquartile range).
Loop diuretics and mortality
At a median follow-up of 1.72 years (IQR = 0.61–3.55), 561 (40.4%) patients died. Of these 561 deaths, 404 (72%) were documented as being CV-related deaths. As regards the entire population, HDLD was independently associated with all-cause mortality (HR 1.23, 95% CI 1.01–1.50; P = 0.04). As a main effect, continuous FED (transformed as FEDsqrt) was positively associated with mortality, although such an association did not achieve statistical significance (HR 1.03, 95% CI 0.99–1.06, P = 0.16).
Furosemide equivalent dose and all-cause mortality across carbohydrate antigen 125–blood urea nitrogen categories
Continuous FED was tested against mortality with a df(4) RCS, and interacting with CA125–BUN categories (P for interaction = 0.0034). The P-value for linearity supports the lack of linearity in the risk function for FED (P = 0.001). Figure 1 shows a differential adjusted risk between the continuum of FED and mortality across CA125–BUN categories, in terms of HR, with the value of FED 40 mg/day used as reference. As FED increased above 40 mg/day, the HR for mortality increased in C1 (low CA125/high BUN) and C3 (high CA125/low BUN) categories (Figure 1A and C, respectively). In C2 (low CA125/low BUN), increases in FED translated into a neutral effect on mortality (Figure 1B). However, increases in FED for patients in C4 (high CA125/high BUN) were associated with marginal survival benefit (Figure 1D).
Figure 1.
Adjusted hazard ratios (and their 95% pointwise confidence intervals) for the effect of FED on mortality at each BUN–CA125 category. FED is modelled with df(4) RCS. Hazard ratios are calculated against the value of 40 mg/day as reference point. BUN, blood urea nitrogen; CA125, carbohydrate antigen 125; CI, confidence interval; FED, furosemide equivalent dose; HR, hazard ratio; RCS, restricted cubic splines.
The analysis dichotomizing FED into HDLD ( ≥120 mg/day) also revealed a significant interaction with CA125–BUN categories (P for interaction <0.001). The corresponding adjusted estimates are presented in Table 3. HDLDs were associated with an increased risk of mortality in patients with low CA125 and BUN above the median (C1), but not in those below the median (C2). Conversely, in patients with high CA125, the administration of HDLDs showed a survival benefit only if BUN was above the median (C4); for those with BUN below the median, it became a significant risk factor for mortality.
Table 3.
Multivariable regression estimates indicating the effect of high-dose loop diuretics on all-cause and cardiovascular mortality, according to carbohydrate antigen 125–blood urea nitrogen, carbohydrate antigen 125–creatinine, and carbohydrate antigen 125–estimated glomerular filtration rate categories
| Dual-marker variables | Hazard ratio (95% CIs) | P-value | Harrell C-statistic | Omnibus P-value |
|---|---|---|---|---|
| All-cause mortality | ||||
| CA125 categories–BUN categoriesa | HDLDs | |||
| C1 | 2.29 (1.51–3.46) | 0.000 | 0.770 | <0.001 |
| C2 | 1.22 (0.73–2.04) | 0.448 | ||
| C3 | 1.94 (1.36–2.76) | 0.000 | ||
| C4 | 0.73 (0.55–0.98) | 0.034 | ||
| CA125–eGFRb | HDLDs | |||
| C1 | 2.41 (1.40–4.14) | 0.002 | 0.741 | <0.001 |
| C2 | 1.67 (1.12–2.48) | 0.011 | ||
| C3 | 1.46 (1.11–1.92) | 0.007 | ||
| C4 | 0.76 (0.54–1.09) | 0.135 | ||
| CV mortality | ||||
| CA125 categories–BUN categoriesa | HDLDs | |||
| C1 | 2.42 (1.54–3.78) | 0.000 | 0.7535 | 0.003 |
| C2 | 0.95 (0.51–1.75) | 0.861 | ||
| C3 | 1.21 (0.75–1.95) | 0.432 | ||
| C4 | 0.86 (0.61–1.20) | 0.366 | ||
BUN, blood urea nitrogen; CA125, serum carbohydrate antigen 125; eGFR, estimated glomerular filtration rate; HDLD, high-dose loop diuretics (furosemide equivalent doses ≥120 mg/day).
aCA125 and BUN categories: C1, CA125 ≤35 U/mL and BUN above the median; C2, CA125 ≤35 U/mL and BUN below the median; C3, CA125 >35 U/mL and BUN below the median; C4, CA125 >35 U/mL and BUN above the median.
bCA125 and eGFR: C1, CA125 ≤35 U/mL and eGFR <45 mL/min/m2; C2, CA125 ≤35 U/mL and eGFR ≥ 45 mL/min/m2; C3, CA125 >35 U/mL and eGFR ≥45 mL/min/m2; C4, CA125 >35 U/mL and eGFR <45 mL/min/m2.
In order to understand the disease course and the potential basis for the above HRs, we estimated the baseline hazard function among those patients with and without HDLD and plotted against follow-up time. Figure 2 shows the adjusted mortality rates (expressed as per 10 PYs) at each BUN–CA125 category. The highest death rates correspond to patients on HDLDs which belong to groups C1 and C3 (10 deaths per 10 PYs, approximately) (Figure 2A and C). For patients in C2, death rates among those on HDLDs were similar (Figure 2B). For C4 (Figure 2D), death rates were higher for those on HDLDs = 0 (8.5 deaths per 10 PYs, approximately) compared with those taking HDLDs (6 deaths per 10 PYs). These figures are also telling us that overall, the death rate seems to be highest ∼1 year after discharge, and it decreases after that time to plateau at ∼4 years.
Figure 2.
Mortality rates (expressed as 10 PYs) estimated for patients with and without high-dose loop diuretics at each BUN–CA125 category, and plotted against follow-up time. BUN, blood urea nitrogen; CA125, carbohydrate antigen 125; FED, furosemide equivalent dose; PYs, person-years.
Furosemide equivalent dose and cardiovascular mortality across carbohydrate antigen 125–blood urea nitrogen categories
High-dose loop diuretics also proved to be independently associated with CV mortality, with an effect that varied according to BUN–CA125 categories. The direction and strength of the association are similar to those of all-cause mortality (Table 3).
Furosemide equivalent dose and all-cause mortality across serum carbohydrate antigen 125–estimated glomerular filtration rate categories
In a sensitivity analysis, a composite variable (CA125–eGFR) with four categories was created, with eGFR dichotomized at 45 mL/min/m2, and CA125 at 35 U/mL. The multivariable model for all-cause mortality that included CA125–eGFR had the same set of covariates as the CA125–BUN model. The P-value for the interaction was significant (P = 0.001). This sensitivity analysis confirmed the presence of a differential effect of HDLD on mortality, with estimates pointing to the same direction as in the CA125–BUN model (Table 3). It is worth mentioning that the effect of HDLDs in C4 was not significant (P = 0.135); moreover, the discriminative accuracy of using eGFR instead of BUN decreased (Harrell's C-statistics = 0.741 vs. 0.770) (Table 3).
Discussion
The principal finding of this hypothesis-generating study is that the mortality risk associated with the prescription of HDLDs at discharge is differentially dependent on CA125 and BUN serum values. Loop diuretics are viewed as a double-edged sword; on the one hand, they are very effective in relieving symptoms of systemic and pulmonary congestion in patients with AHF; on the other hand, their use, particularly in high doses, has been associated with increased mortality.3–6 In the absence of well-designed randomized studies, it has been very difficult to determine if the associated increased risk in mortality merely represents a spurious association due to confounding by indication, as has been suggested by recent findings.24,25 or a real effect. In reference to this topic, direct roles, by promoting renal dysfunction and stimulating multiple neurohormonal systems [including the renin–angiotensin–aldosterone system (RAAS) activity], have been proposed as crucial factors explaining the diuretic-associated detrimental effects.25–28 A recent study showed in a selected cohort of 2456 compensated HF patients with LVEF ≤35% that high BUN (>21 mg/dL) identified those with an increased risk of mortality when HDLDs ( ≥160 mg/day) were prescribed.9 Given that urea tubular reabsorption is largely dependent on neurohormonal activation,29,30 these authors proposed that an elevated BUN level, in addition to being a marker related to reduction of glomerular filtration, may act as surrogate for RAAS activity.7–9,30
In other respects, it is also known that diuretics sometimes improve renal function,26,27 and recent studies have highlighted the role of venous congestion rather than reduced arterial renal perfusion in the pathophysiology of renal function impairment observed in HF.10–13 For instance, recent works have shown that elevated intra-abdominal and central venous pressure, but not cardiac index, were related to the degree of impairment in glomerular filtration.10–12 Along this line, Damman et al. showed in 30 patients with stable chronic HF that following a diuretic withdrawal, a subtle urinary volume increase occurred, paralleling an increase in markers of tubular renal injury.13 CA125, a glycoprotein released by mesothelial cells in response to a mechanical or inflammatory stimulus, has been shown to be a reliable marker of systemic congestion.15−18 In fact, serum levels of this glycoprotein were significantly related to the presence of mesothelial effusion and peripheral oedema, independent of age, gender, and renal function.17 Interestingly, CA125 provided additional prognostic information beyond natriuretic peptides.17 In addition, other factors such as wide availability, low cost, standardized measurement, and long half-life support the use of this biomarker in routine clinical practice.16–18
In this study, conciliating both previous pathophysiological postulates, we found that the high mortality risk associated with the use of HDLDs is strongly dependent on levels of BUN and serum CA125. We found that in patients with CA125 <35 U/mL (no important fluid overload), HDLDs were associated with high mortality in patients with BUN above the median (C1), but not in those with BUN below the median (C2), reproducing the results recently published by Testani et al.9 in patients with stable chronic HF, a scenario where the majority of patients exhibit normal CA125 values.18 Thus, in the absence of important fluid overload, higher BUN levels may help to identify those patients in which the potentially beneficial effect of HDLD does not produce the potentially deleterious effect on renal function and/or neurohormonal activation.
Conversely, in patients with CA125 ≥35 U/mL, the direction of the association mediated by BUN levels was divergent. Indeed, the use of HDLDs in patients with high BUN (C4) was associated with improved survival, while in those with low BUN it was associated with higher mortality (C3). Based on previous experimental studies where selective congestion of the renal veins induced an increase in neurohormonal parameters,31,32 we speculate that the survival benefit associated with the use of HDLD suggests that renal dysfunction/neurohormonal activation largely depends on venous congestion in C4 patients, a situation where an aggressive decongestion would result in a net positive prognostic effect.
Finally, CA125 >35 U/mL and BUN below the median (C3) may help to identify those patients with fluid overload or tissue redistribution where renal venous congestion is not important, a situation where an adequate diuretic response would be expected following the first weeks after hospitalization. Based on a previous result showing that CA125 undergoes important modifications following the first weeks after discharge (especially for those with high values during hospitalization),18 we speculate that the excess risk associated with the use of HDLDs in this category (high CA125 in the absence of renal dysfunction/neurohormonal activation) stems from the fact that most of these patients would control fluid overload within the first weeks following discharge (normalizing CA125 values) and, therefore, move either to C2 (normal CA125 and BUN) or to C1 (normal CA125 but elevated BUN).
Our findings underscore the importance of including a surrogate for systemic/pulmonary congestion as part of the equation relating HDLDs to mortality. Indeed, we believe that this hypothesis-generating study constitutes a first step to delineate further clinical research lines in order to: (i) select those patients who benefit from the use of HDLDs; and (ii) carefully titrate the intensity of diuretic therapy for those patients deemed at risk for their deleterious effects. A clinical instrument able to perform these two tasks represents an unmet need in the management of patients discharged after an episode of AHF, a situation where a residual fluid overload may still be present.14
Limitations
Given the observational nature of this study, the contamination of our results due to confounding by indication cannot be excluded. Furthermore, the fact that the treating physicians were not blinded to the value of these markers makes this study prone to channelling bias, even in the absence of specific recommendations about the use of CA125 and BUN for guiding patient therapy. In order to minimize such unintentional influences, and within the available resources, we developed a well-adjusted multivariable model by including the most important predictors of mortality in AHF using a state of the art survival methodology.21 Is also worth mentioning the possibility that the sample size may have been insufficient to test the interaction effects with appropriate statistical power. We have assumed throughout the study that CA125 is a reliable surrogate for fluid overload, and also renal venous congestion, a presumption that needs to be corroborated with carefully designed experimental studies. BUN values are influenced by prior administration of HDLDs and other factors such as protein catabolism and diet, factors that were not accounted for in this study and might act as important confounders. Furthermore, the lack of serial measurements and the temporal dissociation between the variables precludes evaluating the updated effect of FED, BUN, and CA125 on mortality. Finally, we cannot unravel the complex mechanisms underlying these results.
Conclusions
Following a hospital discharge for AHF, the higher mortality risk associated with the use of HDLDs appears to be dependent on levels of CA125 and BUN. In patients with normal CA125, HDLD use was associated with higher mortality if BUN was above the median but not in those where it was below the median. Conversely, in patients with high CA125, the direction of the association mediated by BUN levels was the opposite (the HDLD group showed an association with increased survival if BUN was above the median, but an association with increased mortality in those with BUN below the median). Further studies are needed to corroborate our results and to provide robust experimental evidence about the complex association between HDLD dose, renal function, systemic congestion, neurohormonal activation, and subsequent mortality.
Supplementary material
Supplementary material is available at European Journal of Heart Failure online.
Funding
The Ministerio de Ciencia e Innovación, Instituto de Salud Carlos III (FEDER), Red HERACLES RD06/0009/1001.
Conflict of interest: G.C.F. has received research support from the National Institutes of Health, and has received honoraria from Novartis (consultant), Medtronic (consultant), and CHF solutions (consultant). The other authors have no conflicts of interest to declare.
References
- 1.Dickstein K, Cohen-Solal A, Filippatos G, McMurray JJ, Ponikowski P, Poole-Wilson PA, Strömberg A, van Veldhuisen DJ, Atar D, Hoes AW, Keren A, Mebazaa A, Nieminen M, Priori SG, Swedberg K. ESC Committee for Practice Guidelines (CPG). ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2008: the Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2008 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association of the ESC (HFA) and endorsed by the European Society of Intensive Care Medicine (ESICM) Eur Heart J. 2008;29:2388–2442. doi: 10.1093/eurheartj/ehn309. [DOI] [PubMed] [Google Scholar]
- 2.Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS, Ganiats TG, Jessup M, Konstam MA, Mancini DM, Michl K, Oates JA, Rahko PS, Silver MA, Stevenson LW, Yancy CW. American College of Cardiology Foundation; American Heart Association. 2009 Focused update incorporated into the ACC/AHA 2005 Guidelines for the Diagnosis and Management of Heart Failure in Adults. A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines Developed in Collaboration With the International Society for Heart and Lung Transplantation. J Am Coll Cardiol. 2009;53:e1–e90. doi: 10.1016/j.jacc.2008.11.013. [DOI] [PubMed] [Google Scholar]
- 3.Ahmed A, Husain A, Love TE, Gambassi G, Dell'Italia LJ, Francis GS, Gheorghiade M, Allman RM, Meleth S, Bourge RC. Heart failure, chronic diuretic use, and increase in mortality and hospitalization: an observational study using propensity score methods. Eur Heart J. 2006;27:1431–1439. doi: 10.1093/eurheartj/ehi890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Peacock WF, Costanzo MR, De Marco T, Lopatin M, Wynne J, Mills RM, Emerman CL. ADHERE Scientific Advisory Committee and Investigators Impact of intravenous loop diuretics on outcomes of patients hospitalized with acute decompensated heart failure: insights from the ADHERE registry. Cardiology. 2009;113:12–19. doi: 10.1159/000164149. [DOI] [PubMed] [Google Scholar]
- 5.Domanski M, Norman J, Pitt B, Haigney M, Hanlon S, Peyster E. Studies of Left Ventricular Dysfunction. Diuretic use, progressive heart failure, and death in patients in the Studies Of Left Ventricular Dysfunction (SOLVD) J Am Coll Cardiol. 2003;42:705–708. doi: 10.1016/s0735-1097(03)00765-4. [DOI] [PubMed] [Google Scholar]
- 6.Eshaghian S, Horwich TB, Fonarow GC. Relation of loop diuretic dose to mortality in advanced heart failure. Am J Cardiol. 2006;97:1759–1764. doi: 10.1016/j.amjcard.2005.12.072. [DOI] [PubMed] [Google Scholar]
- 7.Lindenfeld J, Schrier RW. Blood urea nitrogen a marker for adverse effects of loop diuretics? J Am Coll Cardiol. 2011;58:383–385. doi: 10.1016/j.jacc.2011.01.054. [DOI] [PubMed] [Google Scholar]
- 8.Testani JM, Coca SG, Shannon RP, Kimmel SE, Cappola TP. Influence of renal dysfunction phenotype on mortality in the setting of cardiac dysfunction: analysis of three randomized controlled trials. Eur J Heart Fail. 2011;13:1224–1230. doi: 10.1093/eurjhf/hfr123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Testani JM, Cappola TP, Brensinger CM, Shannon RP, Kimmel SE. Interaction between loop diuretic-associated mortality and blood urea nitrogen concentration in chronic heart failure. J Am Coll Cardiol. 2011;58:375–382. doi: 10.1016/j.jacc.2011.01.052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Mullens W, Abrahams Z, Francis GS, Sokos G, Taylor DO, Starling RC, Young JB, Tang WH. Importance of venous congestion for worsening of renal function in advanced decompensated heart failure. J Am Coll Cardiol. 2009;53:589–596. doi: 10.1016/j.jacc.2008.05.068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Guglin M, Rivero A, Matar F, Garcia M. Renal dysfunction in heart failure is due to congestion but not low output. Clin Cardiol. 2011;34:113–116. doi: 10.1002/clc.20831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Mullens W, Abrahams Z, Skouri HN, Francis GS, Taylor DO, Starling RC, Paganini E, Tang WH. Elevated intra-abdominal pressure in acute decompensated heart failure. J Am Coll Cardiol. 2008;51:300–306. doi: 10.1016/j.jacc.2007.09.043. [DOI] [PubMed] [Google Scholar]
- 13.Damman K, Ng Kam Chuen MJ, MacFadyen RJ, Lip GY, Gaze D, Collinson PO, Hillege HL, van Oeveren W, Voors AA, van Veldhuisen DJ. Volume status and diuretic therapy in systolic heart failure and the detection of early abnormalities in renal and tubular function. J Am Coll Cardiol. 2011;57:2233–2241. doi: 10.1016/j.jacc.2010.10.065. [DOI] [PubMed] [Google Scholar]
- 14.Gheorghiade M, Follath F, Ponikowski P, Barsuk JH, Blair JE, Cleland JG, Dickstein K, Drazner MH, Fonarow GC, Jaarsma T, Jondeau G, Sendon JL, Mebazaa A, Metra M, Nieminen M, Pang PS, Seferovic P, Stevenson LW, van Veldhuisen DJ, Zannad F, Anker SD, Rhodes A, McMurray JJ, Filippatos G. European Society of Cardiology; European Society of Intensive Care Medicine. Assessing and grading congestion in acute heart failure: a scientific statement from the acute heart failure committee of the heart failure association of the European Society of Cardiology and endorsed by the European Society of Intensive Care Medicine. Eur J Heart Fail. 2010;12:423–433. doi: 10.1093/eurjhf/hfq045. [DOI] [PubMed] [Google Scholar]
- 15.D'Aloia A, Faggiano P, Aurigemma G, Bontempi L, Ruggeri G, Metra M, Nodari S, Dei Cas L. Serum levels of carbohydrate antigen 125 in patients with chronic heart failure: relation to clinical severity, hemodynamic and doppler echocardiographic abnormalities, and short-term prognosis. J Am Coll Cardiol. 2003;41:1805–1811. doi: 10.1016/s0735-1097(03)00311-5. [DOI] [PubMed] [Google Scholar]
- 16.Núñez J, Núñez E, Consuegra L, Sanchis J, Bodí V, Martínez-Brotons A, Bertomeu-González V, Robles R, Bosch MJ, Fácila L, Darmofal H, Llàcer A. Carbohydrate antigen 125: an emerging prognostic risk factor in acute heart failure? Heart. 2007;93:716–721. doi: 10.1136/hrt.2006.096016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Núñez J, Sanchis J, Bodí V, Fonarow GC, Núñez E, Bertomeu-González V, Miñana G, Consuegra L, Bosch MJ, Carratalá A, Chorro FJ, Llàcer A. Improvement in risk stratification with the combination of the tumour marker antigen carbohydrate 125 and brain natriuretic peptide in patients with acute heart failure. Eur Heart J. 2010;31:1752–1763. doi: 10.1093/eurheartj/ehq142. [DOI] [PubMed] [Google Scholar]
- 18.Núñez J, Núñez E, Sanchis J, Bodí V, Fonarow GC, Miñana G, Palau P, Bertomeu-González V, Carratalá A, Mainar L, Chorro FJ, Llàcer A. Antigen carbohydrate 125 and brain natriuretic peptide serial measurements for risk stratification following an episode of acute heart failure. Int J Cardiol. 2011 doi: 10.1016/j.ijcard.2011.02.001. in press. [DOI] [PubMed] [Google Scholar]
- 19.Nieminen MS, Böhm M, Cowie MR, Drexler H, Filippatos GS, Jondeau G, Hasin Y, Lopez-Sendon J, Mebazaa A, Metra M, Rhodes A, Swedberg K, Priori SG, Garcia MA, Blanc JJ, Budaj A, Cowie MR, Dean V, Deckers J, Burgos EF, Lekakis J, Lindahl B, Mazzotta G, Morais J, Oto A, Smiseth OA, Garcia MA, Dickstein K, Albuquerque A, Conthe P, Crespo-Leiro M, Ferrari R, Follath F, Gavazzi A, Janssens U, Komajda M, Morais J, Moreno R, Singer M, Singh S, Tendera M, Thygesen K. ESC Committe for Practice Guideline (CPG). Executive summary of the guidelines on the diagnosis and treatment of acute heart failure: the Task Force on Acute Heart Failure of the European Society of Cardiology. Eur Heart J. 2005;26:384–416. doi: 10.1093/eurheartj/ehi044. [DOI] [PubMed] [Google Scholar]
- 20.Levy WC, Mozaffarian D, Linker DT, Sutradhar SC, Anker SD, Cropp AB, Anand I, Maggioni A, Burton P, Sullivan MD, Pitt B, Poole-Wilson PA, Mann DL, Packer M. The Seattle Heart Failure Model: prediction of survival in heart failure. Circulation. 2006;113:1424–1433. doi: 10.1161/CIRCULATIONAHA.105.584102. [DOI] [PubMed] [Google Scholar]
- 21.Royston P, Lambert PC. Flexible Parametric Survival Analysis Ssing Stata: Beyond the Cox Model. College Station, TX: Stata Press; 2011. [Google Scholar]
- 22.Royston P, Sauerbrei W. Multivariable Model-building: A Pragmatic Approach to Regression Analysis Based on Fractional Polynomials for Modelling Continuous Variables. Chichester, UK: Wiley; 2008. [Google Scholar]
- 23.Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509. [Google Scholar]
- 24.Yilmaz MB, Gayat E, Salem R, Lassus J, Nikolaou M, Laribi S, Parissis J, Follath F, Peacock WF, Mebazaa A. Impact of diuretic dosing on mortality in acute heart failure using a propensity-matched analysis. Eur J Heart Fail. 2011;13:1244–1252. doi: 10.1093/eurjhf/hfr121. [DOI] [PubMed] [Google Scholar]
- 25.Damman K, O'Connor CM. Dangerous diuretics or death defying drugs? Eur J Heart Fail. 2011;13:1157–1158. doi: 10.1093/eurjhf/hfr120. [DOI] [PubMed] [Google Scholar]
- 26.Felker GM, O'Connor CM, Braunwald E. Heart Failure Clinical Research Network Investigators. Loop diuretics in acute decompensated heart failure: necessary? Evil? A necessary evil? Circ Heart Fail. 2009;2:56–62. doi: 10.1161/CIRCHEARTFAILURE.108.821785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gottlieb SS. Diuretics: are our ideas based on knowledge? J Am Coll Cardiol. 2011;57:2242–2243. doi: 10.1016/j.jacc.2010.11.062. [DOI] [PubMed] [Google Scholar]
- 28.Francis GS, Siegel RM, Goldsmith SR, Olivari MT, Levine TB, Cohn JN. Acute vasoconstrictor response to intravenous furosemide in patients with chronic congestive heart failure Activation of the neurohumoral axis. Ann Intern Med. 1985;103:1–6. doi: 10.7326/0003-4819-103-1-1. [DOI] [PubMed] [Google Scholar]
- 29.Sands JM. Mammalian urea transporters. Annu Rev Physiol. 2003;65:543–566. doi: 10.1146/annurev.physiol.65.092101.142638. [DOI] [PubMed] [Google Scholar]
- 30.Schrier RW. Blood urea nitrogen and serum creatinine: not married in heart failure. Circ Heart Fail. 2008;1:2–5. doi: 10.1161/CIRCHEARTFAILURE.108.770834. [DOI] [PubMed] [Google Scholar]
- 31.Doty JM, Saggy BH, Sugerman HJ. Effect of increased renal venous pressure on renal function. J Trauma. 1999;47:1000–1003. doi: 10.1097/00005373-199912000-00002. [DOI] [PubMed] [Google Scholar]
- 32.Kishimoto T, Maekawa M, Miyazaki M, Yamamoto K, Ueda J. Effects of renal venous pressure elevation on renal hemodynamics, urine formation, and renin release. Jap Circ J. 1972;36:439–448. doi: 10.1253/jcj.36.439. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.