Relation of Unrecognized Hypervolemia in Chronic Heart Failure to Clinical Status, Hemodynamics, and Patient Outcomes Ana Silvia Androne, MD, Katarzyna Hryniewicz, MD, Alhakam Hudaihed, MD, Donna Mancini, MD, John Lamanca, PhD, and Stuart D. Katz, MD, MS Clinically unrecognized intravascular volume overload 0.01) and greatly increased risk of death or urgent may contribute to worsening symptoms and disease cardiac transplantation during a median follow-up of progression in patients with chronic heart failure (CHF). 719 days (1-year event rate 39% vs 0%, p <0.01 by The present study was undertaken to prospectively com- log-rank test). Systolic blood pressure was significantly pare measured blood volume status (determined by ra- lower in hypervolemic patients than in those with nor- diolabeled albumin technique) with clinical and hemo- movolemia or hypovolemia (107 ؎ 2vs119؎ 2mm and hypotension was independently ,(0.008 ؍ dynamic characteristics and patient outcomes in 43 Hg, p nonedematous ambulatory patients with CHF. Blood vol- associated with increased risk of hypervolemia in mul- ume analysis demonstrated that 2 subjects (5%) were tivariate analysis (odds ratio 2.64 for a 10-mm Hg hypovolemic (mean deviation from normal values ؊20 decrease in systolic blood pressure, 95% confidence -These findings dem .(0.025 ؍ subjects (30%) were normovolemic (mean interval 1.13 to 6.19, p 13 ,(6% ؎ deviation from normal values ؊1 ؎ 1%), and 28 sub- onstrate that clinically unrecognized hypervolemia is jects (65%) were hypervolemic (mean deviation from frequently present in nonedematous patients with CHF normal values ؉30 ؎ 3%). Physical findings of conges- and is associated with increased cardiac filling pressures tion were infrequent and not associated with blood vol- and worse patient outcomes. ᮊ2004 by Excerpta ume status. Increased blood volume was associated with Medica, Inc. (Am J Cardiol 2004;93:1254–1259) ؍ increased pulmonary capillary wedge pressure (p ntravascular volume can be directly measured non- jects between 21 and 80 years of age with CHF for Ͼ3 Iinvasively with isotopic tracer and dye dilution months’ duration, with stable New York Heart Asso- techniques.1 Previous reports of blood volume mea- ciation class II to IV symptoms for Ͼ2 months, and surements with chronic heart failure (CHF) were con- left ventricular ejection fraction Յ35% were eligible ducted in small study populations without a detailed for the study. Criteria for exclusion were acute decom- description of physical examination findings and pensated heart failure, severe renal dysfunction (se- background medical therapy.2–6 Based on the findings rum creatine Ͼ2.5 mg/dl or history of nephrotic syn- of previous hemodynamic studies,7–10 we hypothe- drome), severe hepatic dysfunction (serum liver sized that clinically unrecognized increases in blood enzymes Ͼ3 times the upper limits of normal or volume may be common, even in nonedematous pa- history of cirrhosis), pregnancy, history of thyroid tients with CHF. The present study was undertaken to disease, and patients with known history of allergy to noninvasively measure blood volume after iodine-125 iodine or iodinated products. Background cardiac albumin injection in ambulatory nonedematous pa- medications included stable doses of diuretics, tients with CHF. Our goal was to quantify blood digoxin, renin-angiotensin system inhibitors (angio- volume derangements and to compare measured blood tensin-converting enzyme inhibitors and/or angioten- volume status with clinical and hemodynamic charac- sin receptor antagonists), and/or ␤-adrenergic receptor teristics of heart failure. antagonists for Ͼ2 months before the study. The pro- METHODS tocol was approved by the institutional review board Study group: Forty-three consecutive nonedema- at Columbia Presbyterian Medical Center. All subjects tous ambulatory patients with CHF were studied. Sub- gave written informed consent before participation. Blood volume analysis: Blood volume was deter- mined after intravenous administration of iodine-131– 1,11,12 From the Department of Internal Medicine, Yale University School of labeled albumin as previously described. After Medicine, New Haven, Connecticut; and the Department of Medi- obtaining a baseline sample of 5 ml of venous blood as cine, Columbia University College of Physicians & Surgeons, New a control for background radiation, 10 to 25 ␮Ci of York, New York. Manuscript received November 25, 2003; revised iodine-131–labeled albumin (Volumex, Daxor Corp., manuscript received and accepted January 23, 2004. Address for reprints: Stuart D. Katz, MD, Yale University School of New York City, New York) was injected into a pe- Medicine, 135 College Street, Suite 301, New Haven, Connecticut ripheral vein from a specialized prefilled flow cham- 06510. E-mail: [email protected]. ber designed to ensure Ͼ99.8% delivery of the radio- 1254 ©2004 by Excerpta Medica, Inc. All rights reserved. 0002-9149/04/$–see front matter The American Journal of Cardiology Vol. 93 May 15, 2004 doi:10.1016/j.amjcard.2004.01.070 isotope dose. Five milliliters of venous blood were collected from an indwelling catheter placed in a forearm vein of the contralateral arm at 12, 18, 24, 30, and 36 minutes after isotope injection. Spun hematocrit was determined from each sample and plasma radioactivity of each sample was measured in a semiautomated counter (BVA-100 Blood Volume Analyzer, Daxor Corp.). Plasma volume was determined as the zero-time volume of distribution of the radiolabeled albumin ob- tained by semilogarithmic extrapolation of values mea- sured from the 5 samples.1,11 Blood volume and red blood cell volumes were calculated from the plasma volume measurement, and the measured hematocrit cor- rected for trapped plasma and mean body hematocrit and then compared with normal values for age, gender, height, and weight based on the ideal weight sys- tem.11,13,14 The coefficient of variance for this analytic technique is Ͻ3.5%.11,15 Normovolemia was prospec- tively defined as a measured blood volume within Ϯ8% of the predicted normal value.11 Clinical assessment of volume status: The clinical volume status was assessed immediately before blood volume analysis. Blood pressure and heart rate were measured with an automated cuff device with patients in the supine position (Dinamapp, Critikon, Tampa, Florida). Jugular venous pressure was assessed in a seated position and was considered elevated if venous FIGURE 1. Frequency distribution (number of subjects) of red pressure waves were visible above the clavicle. Pres- blood cell (RBC) volume (top panel), plasma volume (middle ence or absence of elevated jugular venous pressure, panel), and blood volume (bottom panel) in 43 patients with inspiratory rales, S3 gallops, and hepatomegaly were CHF. recorded as dichotomous variables. Clinical hypervol- emia was defined empirically as the presence of Ն1 physical finding of congestion (elevated jugular ve- analysis, or Fisher’s exact test as appropriate (Stata nous pressure, rales, or hepatomegaly) and the pres- software version 8.0, College Station, Texas). Diuretic ence of Ն1 congestive symptom (orthopnea, paroxys- dose and brain natriuretic peptide levels were not mal nocturnal dyspnea, abdominal bloating, nocturia). normally distributed, so the natural log transformation Patient outcomes determination: Patient outcomes or nonparametric testing was used in statistical anal- as of April 1, 2003 (death or urgent cardiac transplan- yses. Multivariate logistic regression models were ex- tation, nonurgent cardiac transplantation, or no event) plored to identify clinical predictors of hypervolemia. were determined for all patients by direct contact with Age, gender, and New York Heart Association class the patient or their family or physician. were included in all models. Other variables with Hemodynamic measurements: Seventeen subjects univariate p values Յ0.20 were added in stepwise who were undergoing outpatient cardiac transplanta- fashion and retained in the model if the adjusted p tion evaluations had elective outpatient right-sided value was Յ0.20. The Hosmer-Lemeshow goodness- cardiac catheterization performed for assessment of of-fit statistic was used to test model assumptions. pulmonary vascular resistance on the same day as Survival data were analyzed by the Kaplan-Meier blood volume analysis with standard clinical tech- method and stratified log-rank tests. Observations for niques. The catheter tip position was confirmed by patients with elective cardiac transplantation were fluoroscopy and hemodynamic monitoring. A mean censored at the time of surgery. For all analyses, a p pulmonary capillary wedge pressure of Ն15 mm Hg value Ͻ0.05 was considered statistically significant. was used as the clinically derived criterion to infer hypervolemia. RESULTS Plasma brain natriuretic peptide measurements: In Blood volume analysis: Blood volume analysis 18 subjects, plasma brain natriuretic peptide was mea- demonstrated that 2 subjects (5%) were hypovolemic sured with a calibrated automated quantitative fluo- (mean deviation from normal blood volume values rescent sandwich immunoassay device (Biosite Diag- Ϫ20 Ϯ 6%), 13 subjects (30%) were normovolemic nostic, San Diego California).16 (mean deviation from normal blood volume values Ϫ1 Statistical analysis: All values are presented as Ϯ 1%), and 28 subjects (65%) were hypervolemic means Ϯ SEM. Clinical characteristics of subjects (mean deviation from normal blood volume values grouped by blood volume status (hypervolemia