Does pre-albumin predict in-hospital mortality in heart failure?

Letters to the Editor 758 [8] Ormiston JA. Stent deformation following simulated side-branch dilatation: a comparison of five stent designs. Catheter Cardiovasc Interv 1999;47:258. [9] Uren NG, Schwarzacher SP, Metz JA, et al. Predictors and outcomes of stent thrombosis. An intravascular ultrasound registry. Eur Heart J 2002;23:124–32. [10] Romagnoli E, Sangiorgi GM, Cosgrave J, Guillet E, Colombo A. Drug-eluting stenting: the case for post-dilation. J Am Coll Cardiol Intv 2008;1:22–31. 0167-5273/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2012.09.170 Does pre-albumin predict in-hospital mortality in heart failure?☆ Patrícia Lourenço a,⁎, Sérgio Silva a, Fernando Friões a, Margarida Alvelos a, Marta Amorim a, Paulo Torres-Ramalho a, Maria José Teles b, João Tiago Guimarães b,c, Paulo Bettencourt a a b c Serviço de Medicina Interna, Centro Hospitalar São João, Faculdade de Medicina da Universidade do Porto, Unidade I&D Cardiovascular do Porto, Portugal Serviço de Patologia Clínica, Centro Hospitalar São João, Faculdade de Medicina da Universidade do Porto, Portugal Departamento de Bioquímica, Faculdade de Medicina da Universidade do Porto, Portugal a r t i c l e i n f o Article history: Received 19 August 2012 Accepted 25 September 2012 Available online 16 October 2012 Keywords: Heart failure In-hospital mortality Pre-albumin Malnutrition Inflammation Malnutrition, inflammation and the so called wasting syndrome predict poor outcome in heart failure (HF) [1]. Albumin and total cholesterol have been used as markers of malnutrition and wasting syndrome in HF [2]. Lower serum cholesterol, but mainly hypoalbuminemia were reported to associate with hospital death in acute HF [3,4]. Although albumin is widely used to assess nutritional status, it is now recognized that it is relatively insensitive to nutritional variations [2,5]. Pre-albumin, with a 2 day half-life, is the best marker for protein malnutrition [5,6]. Low pre-albumin levels predict higher mortality in different clinical settings namely hemodialysis patients and in intensive care units [7,8]. The prognostic value of pre-albumin in HF is largely unknown and was never studied in the acute HF setting. We aimed to study if low pre-albumin predicted in-hospital mortality in patients admitted with acute HF. During a 20-month period, all patients admitted in the Internal Medicine ward with the primary diagnosis of HF were eligible. Patients with acute coronary syndromes; those whose complaints were considered by the attending physician as attributable to causes other than HF; and those with no echocardiographic structural or functional cardiac abnormalities were excluded from the study. The European Society of Cardiology guidelines were used for HF diagnosis. Both patients with systolic dysfunction and those with HF with preserved ejection fraction were studied. Fasting venous blood samples were collected from all patients within 48 h of admission. A comprehensive echocardiographic assessment was performed within 72 h of admission; a multi-frequency matrix probe (Vivid6, GE Healthcare, Chalfont St Giles) was used. ☆ This study was supported by a grant from “Fundação para a Ciência e a Tecnologia”, project PIC/IC/82773/2007. ⁎ Corresponding author at: Serviço de Medicina Interna, Hospital S. João, Alameda Professor Hernâni Monteiro, 4202-451 Porto, Portugal. Tel.: + 351 225512200; fax: + 351 225512332. E-mail address: [email protected] (P. Lourenço). The outcome under study was in-hospital death. Pre-albumin was assayed using particle-enhanced immunonephelometric assays on a BN®II laser nephelometer (Siemens, Lisboa, Portugal). All patients provided written informed consent to participate in the study. The study protocol conforms to the ethical guidelines of the declaration of Helsinki and was approved by the local ethic committee. The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology. Patients dying in-hospital and those discharged alive were compared. A χ2 test was used for categorical variables, an independent sample t-test for normally distributed continuous variables, and the Mann–Whitney U test when the distribution was skewed. Correlations between admission albumin, pre-albumin, total cholesterol and CRP were made by the use of the Pearson's or Spearman's correlation coefficient. A logistic regression analysis was used to evaluate predictors of inhospital mortality. Adjustments were made, one at a time, for each of the variables prognostically associated in a univariate approach. A multivariate model was built taking into consideration variables with confounding effect or interaction. Pre-albumin, SBP b 115 mm Hg, admission albumin, total cholesterol, and CRP entered the model. A total of 659 patients were studied. Median age was 79 years (23 to 100 years). HF aetiology was ischemic in 39.2%. Table 1 shows patient characteristics: demographic, comorbidities, clinical and laboratory parameters. Median and interquartile range (IQR) of length of hospital stay was 8 (6 to 12) days. Thirty-two patients (4.8%) died in the hospital. Patients dying in-hospital and those discharged alive are compared in Table 1. Patients that died in-hospital were older, had lower admission systolic blood pressure (SBP), worse admission renal function and higher neuro-humoral activation. Patients with intra-hospital death also had lower total cholesterol, lower albumin, and lower pre-albumin. Pre-albumin and albumin showed a positive correlation (Rho =+0.55, p b 0.001). Albumin and pre-albumin had a positive correlation with total cholesterol (Rho= 0.28 and Rho= 0.36, respectively). Both albumin and pre-albumin showed a negative correlation with C-reactive protein (CRP) (Rho of −0.27 and −0.37, respectively). Older age, lower admission systolic blood pressure, higher neuro-humoral activation, lower albumin, pre-albumin and total cholesterol, all were found to be associated with in-hospital death (Table 2). CRP showed no prognostic value, but interaction and a confounding effect was seen (data not shown). Pre-albumin 759 Letters to the Editor Table 1 Characteristics of study patients and comparison between patients dying in-hospital and those discharged alive. Characteristics All patients (659) In-hospital death (32) Discharged alive (627) p-Value Demographic data Male sex, n (%) Age, median (IQR) 290 (44.0) 79 (72–84) 12 (37.5) 83 (78–88) 278 (44.3) 78 (72–84) 0.56 0.005 Comorbidities Ischemic aetiology, n (%) Chronic HF, n (%) Atrial fibrillation, n (%) Hypertension history, n (%) 258 (39.2) 583 (88.5) 302 (46.5) 476 (76.4) 9 28 17 14 (28.1) (87.5) (56.7) (58.3) 249 (39.7) 555 (88.5) 285 (46.0) 462 (76.1) 0.19 1.00 0.34 0.08 Previous medication ACEi or ARB, n (%) Beta blocker, n (%) Spironolactone, n (%) 403 (61.5) 315 (48.2) 92 (14.1) 15 (48.4) 11 (35.5) 7 (22.6) 388 (62.2) 304 (48.9) 85 (13.7) 0.18 0.20 0.26 Clinical characteristics at admission Systolic blood pressure (mm Hg), median (IQR) BMI (Kg/m2), median (IQR) NYHA IV (vs II/III) Left ventricular systolic function, n (%) Preserved LVSF Mild LVSD Moderate LVSD Severe LVSD Laboratory parameters Haemoglobin (g/dL), median (IQR) Sodium (mEq/L), median (IQR) Creatinine (mg/dL), median (IQR) Urea (mg/dL), median (IQR) CRP (mg/L) median (IQR) Albumin (g/L), mean (SD) Pre-albumin (mg/dL), mean (SD) BNP (pg/mL), median (IQR) Total cholesterol (mg/dL), median (IQR) 130 (113–150) 25.1 (22.6–27.7) 397 (60.7) 284 (44.9) 35 (5.5) 93 (14.7) 221 (34.9) 11.7 (10.4–13.3) 139 (136–141) 1.38 (1.10–1.82) 68 (48–98) 20.1 (9.5–55.4) 35.2 (4.9) 18.0 (7.1) 1664.1 (942.2–2832.8) 149 (124–184) 112 (92–120) 23.4 (22.5–27.7) 21 (67.7) 10 0 2 14 (38.5) (0) (7.7) (53.8) 11.7 (10.6–13.3) 137 (133–141) 1.46 (1.22–2.41) 82 (60–151) 29.4 (14.3–53.6) 32.0 (4.0) 13.2 (5.2) 2981.5 (1949.8–4816.4) 112 (91–145) 131 (115–152) 25.2 (22.6–27.8) 376 (60.4) 274 (45.1) 35 (5.8) 91 (15.9) 207 (34.1) 11.7 (10.4–13.3) 139 (136–141) 1.37 (1.10–1.81) 67 (47–95) 19.5 (9.2–55.4) 35.4 (4.9) 18.2 (7.1) 1601.0 (918.6–2761.1) 150 (126–186) b 0.001 0.31 0.53 0.14 1.00 0.14 0.08 0.002 0.16 b 0.001 b 0.001 b 0.001 b 0.001 ACEi: angiotensin converting enzyme inhibitors; ARB: angiotensin receptor blockers; BMI: body mass index; BNP: B-type natriuretic peptide; CRP: C-reactive protein; HF: heart failure; IQR: interquartile range; LVSD: left ventricular systolic dysfunction; LVEF: left ventricular systolic function; NYHA: New York Heart Association; SD: standard deviation. remained independently associated with in-hospital death when one at a time adjustments were made with other prognostic associated variables. Patients with admission pre-albumin b 15 mg/ dL had a 3.22 (95% CI: 1.28–8.14), p value = 0.01 SBP-, albumin-, cholesterol- and CRP-independent higher risk of in-hospital death (Table 2). Nutritional status, as assessed by pre-albumin, was an independent predictor of in-hospital mortality. Patients with admission prealbumin b 15 mg/dL had a SBP-, albumin-, total cholesterol- and CRP-independent more than three-fold higher risk of in-hospital death. Normal pre-albumin has been variably considered from above 13 mg/dL [2], 15 mg/dL [5] and 17 mg/dL [8]. Pre-albumin associates with outcome in critically ill [8] and several groups of non-HF, noncritically-ill patients. The association of hypoalbuminemia with in-hospital HF death was previously suggested [3,4]; however, the pathophysiological significance of this association has been questioned. Two main confounders are to be considered: malnutrition and inflammation. Albumin is a late and nonsensitive nutritional marker, nevertheless a malnourished state is unquestionably associated with hypoalbuminemia development. Inflammation also associates with hypoalbuminemia development in that the inflammatory stimulus reprioritizes hepatic protein synthesis towards acute phase proteins. Moreover, albumin itself has important physiological antioxidant properties [9]. Pre-albumin is also mainly hepatically synthesized and experiences the same synthesis inhibition in inflammatory settings [5]. However, contrarily to albumin, no intrinsic anti-inflammatory properties have been described. Our study confirmed a direct correlation of albumin and prealbumin with total cholesterol, and showed that the association was stronger for pre-albumin. It also confirmed an inverse correlation with inflammation, again stronger for pre-albumin. To the best of our knowledge, and as far as an extensive literature review could show, this is the first study showing simultaneously the correlation between pre-albumin and inflammation/malnutrition together with its strong and independent association with inhospital mortality. The study design does not allow the establishment of causality relations and our results need to be reproduced in further studies due to its single centre design and sample size. Lower admission pre-albumin was an independent predictor of in-hospital mortality in HF. Whether this biomarker has an intrinsic prognostic role, beyond that of associated malnutrition, needs to be further studied. References [1] Anker SD, Ponikowski P, Varney S, et al. Wasting as independent risk factor of survival in chronic heart failure. Lancet 1997;349:1050–3. [2] Blas PA, Gutiérrez PC, Frutos VA, Peris PG, López MG, Torres Segovia FJ. Insuficiencia cardíaca, malnutrición e inflamación. Prevalencia y aspectos relevantes en su valoración. Rev Clin Esp 2006;206:122–8. [3] Arques S, Roux E, Stolidi P, Gelisse R, Ambrosi P. Usefulness of serum albumin and serum total cholesterol in the prediction of in-hospital death in older patients with severe, acute heart failure. Arch Cardiovasc Dis 2011;104:502–8. [4] Kinugasa Y, Kato M, Sugihara S, et al. A simple risk score to predict in-hospital death of elderly patients with acute decompensated heart failure — hypoalbuminemia as an additional prognostic factor. Circ J 2009;73:2276–81. [5] Beck FK, Rosenthal TC. Prealbumin: a marker for nutrition evaluation. Am Fam Physician 2002;65:1575–8. 760 Letters to the Editor Table 2 Predictors of in-hospital death: univariate and multivariate approaches. Characteristics Odds ratio (95% CI) univariate Demographic data Male sex Age (per year) p-Value 0.75 (0.36–1.57) 1.05 (1.01–1.10) 0.45 0.01 Comorbidities Ischemic aetiology Chronic HF Atrial fibrillation Hypertension history 0.59 (0.27–1.31) 0.91 (0.31–2.66) 1.54 (0.73–3.22) 0.44 (0.19–1.01) 0.20 0.86 0.25 0.05 Previous medication ACEi or ARB Beta blocker Spironolactone 0.57 (0.28–1.18) 0.58 (0.27–1.22) 1.84 (0.77–4.41) 0.13 0.15 0.17 Clinical characteristics at admission Admission systolic blood pressure (b 115 mm Hg) BMI (per kg/m2) NYHA IV (vs II/III) LVSD (vs HFPEF) 0.96 (0.94–0.98) 0.98 (0.90–1.07) 1.38 (0.64–2.98) 1.32 (0.59–2.95) b 0.001 0.70 0.41 0.50 Laboratory parameters Anaemia at admission Hyponatremia upon admission (b135 mEq/L) Creatinine N 1.5 mg/dL Urea N50 mg/dL CRP (per mg/L) Albumin b 35 g/L Pre-albumin b 15 mg/dL BNP (per 100 pg/mL) Total cholesterol b 150 mg/dL 1.03 (0.50–2.15) 2.41 (1.12–5.18) 1.55 (0.76–3.16) 2.84 (0.98–8.22) 1.00 (1.00–1.01) 3.24 (1.48–7.13) 4.80 (1.98–11.80) 1.02 (1.01–1.03) 4.23 (1.70–10.51) 0.93 0.02 0.22 0.05 0.94 0.003 0.001 b 0.001 0.002 Odds ratio (95% CI) multivariate p-Value 2.64 (1.14–6.10) 0.02 1.00 (0.99–1.00) 2.51 (0.92–6.89) 3.22 (1.28–8.14) 0.23 0.07 0.01 3.59 (1.17–11.05) 0.03 ACEi: angiotensin converting enzyme inhibitors; ARB: angiotensin receptor blockers; BMI: body mass index; BNP: B-type natriuretic peptide; CRP: C-reactive protein; HF: heart failure; HFPEF: heart failure with preserved ejection fraction; IQR: interquartile range; LVSD: left ventricular systolic dysfunction; NYHA: New York Heart Association; SD: standard deviation. [6] Devoto G, Gallo F, Marchello C, et al. Prealbumin serum concentration as a useful tool in the assessment of malnutrition in hospitalized patients. Clin Chem 2006;52:2281–5. [7] Sreedhara R, Avram MM, Blanco M, Batish R, Avram MM, Mittman N. Prealbumin is the best nutritional predictor of survival in hemodialysis and peritoneal dialysis. Am J Kidney Dis 1996;28:937–42. 0167-5273/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2012.09.169 [8] Devakonda A, George L, Raoof S, Esan A, Saleh A, Bernstein LH. Transthyretin as a marker to predict outcome in critically ill patients. Clin Biochem 2008;41:1126–30. [9] Roche M, Rondeau P, Singh NR, Tarnus E, Bourdon E. The antioxidant properties of serum albumin. FEBS Lett 2008;582:1783–7.