Methods

We retrospectively reviewed 272 consecutive patients with advanced heart failure undergoing elective implantation of a HeartMate I (n = 147) or a HeartMate II (n = 125) LVAD (both Thoratec Corp., Pleasanton, California) from November 2000 to August 2010 at the Columbia University Medical Center. Patients who underwent emergent LVAD implantation because of cardiogenic shock were excluded from the study. Preoperative clinical characteristics and laboratory examinations including serum albumin concentration were obtained from a clinical database. For patients with multiple laboratory measurements before the implantation, the results obtained at the date closest to the implantation were used for the present study.

We divided patients into 2 groups: those with preoperative hypoalbuminemia defined as albumin concentration <3.5 g/dl and those with normal albumin level (≥3.5 g/dl). Survival after LVAD implantation was compared between the groups. To determine the impact of hypoalbuminemia as a risk factor for mortality, we performed a multivariate analysis of the various preoperative factors including albumin levels associated with mortality.

Furthermore, we classified patients with and without preoperative hypoalbuminemia according to the dynamics in postoperative albumin levels to investigate the effect of changes in albumin levels during LVAD support on post-LVAD survival. Postoperative albumin levels were collected within 7 days from the end of observation or the device removal due to transplant, recovery, or death. Thus, subgroups of patients consisted of normal-to-normal (both pre- and postoperative normal albumin levels), normal-to-hypo (preoperative normal albumin level but postoperative hypoalbuminemia), hypo-to-normal (preoperative hypoalbuminemia but normalization of albumin postoperatively), and hypo-to-hypo (both pre- and postoperative hypoalbuminemia) groups. The study was approved by the Institutional Review Board of Columbia University Medical Center and all ethical guidelines outlined by the 1975 Declaration of Helsinki.

Data are presented as mean ± SD and frequency (percentage). Normality was evaluated for each variable on the basis of normal distribution plots and histograms and by the Kolmogorov-Smirnov test. Student unpaired t test was used to compare the variables between the groups. Categorical variables were compared using chi-square test. A Cox proportional hazard model was used to assess the impact of preoperative factors on postoperative mortality. Variables that achieved statistical significance in the univariate analysis were subsequently included in a multivariate analysis. Survival after LVAD implantation was compared using Kaplan-Meier methods with log-rank test. All statistical analyses were performed using SPSS 19.0 software (IBM, Armonk, New York).

Results

Among the entire cohort, 147 patients (54.0%) showed normal albumin concentration and 125 patients (46.0%) showed hypoalbuminemia (<3.5 g/dl). Clinical characteristics of patients with and without hypoalbuminemia are listed in Table 1 .

Of the 147 patients with normal albumin levels before LVAD implantation, 107 patients (39.3% of total cohort) remained to have normal albumin levels postoperatively (normal-to-normal group), whereas 40 patients (14.7% of total cohort) developed hypoalbuminemia during LVAD support (normal-to-hypo group). Of patients with preoperative hypoalbuminemia, 81 patients (29.8% of total cohort) showed a normalization of albumin levels during LVAD support (hypo-to-normal group), whereas 44 patients (16.2% in total cohort) remained to have hypoalbuminemia postoperatively (hypo-to-hypo group).

Kaplan-Meier survival analysis revealed that patients with normal preoperative albumin levels showed better survival compared with that of patients with hypoalbuminemia during a mean observation period of 8.4 ± 10.2 months (survival at 3 and 12 months: 93.2% vs 82.4% and 88.4% vs 75.2%, respectively, p <0.001; Figure 1 ). After subdividing patients according to the device types, survival benefits of albumin levels were still seen in both patients who underwent HeartMate I and HeartMate II implantations (12-month survival, normal albumin level vs hypoalbuminemia; 96.1% vs 53.2% and 92.9% vs 78.9%, respectively; both p <0.001).

Kaplan-Meier survival curves of patients with and without preoperative hypoalbuminemia. Blue line indicates patients with normal albumin concentration defined as ≥3.5 g/dl and red line indicates patients with hypoalbuminemia (<3.5 g/dl). Dotted lines indicate 3- and 12-month mortalities of both groups.

Subanalysis of survival comparing patients classified according to dynamics in albumin levels during LVAD support is shown in Figure 2 . Patients in the normal-to-normal group (3-month survival, 98.1%) showed better survival than any other subgroup (all p <0.001). Of note, patients with improvement of albumin levels from preoperative hypoalbuminemia showed better survival than that of normal-to-hypo and hypo-to-hypo groups (at 3 months: 92.6% vs 65.0% and 63.6%, respectively; both p <0.001). The post-LVAD survival of the normal-to-hypo group was not significantly different from the hypo-to-hypo group, indicating that levels of albumin had prognostic impact also in patients on LVAD support.

Kaplan-Meier survival curve of subgroup of patients classified by pre- and postoperative albumin levels. Blue line indicates hypo-to-hypo group, red line indicates normal-to-normal group, green line indicates normal-to-hypo group, and orange line indicates hypo-to-normal group. Survival at 3 months in normal-to-normal, normal-to-hypo, hypo-to-normal, and hypo-to-hypo groups was 98.1, 65.0, 92.6, and 63.6%, respectively.

Table 2 lists the preoperative factors associated with mortality after LVAD implantation. Variables found to be associated with mortality during LVAD support by univariate analyses were selected for inclusion into a subsequent multivariate analysis. Blood urea nitrogen and aspartate aminotransferase were selected for inclusion into the multivariate analysis but not creatinine or alanine aminotransferase because these variables were selected in a previous study for LVAD risk stratification. Multivariate analysis revealed that preoperative albumin levels were independently associated with postoperative mortality. The analysis revealed that with every 0.1 g/dl increase in albumin concentration, the postoperative mortality rate decreased by 4.8% adjusted for platelet count and aspartate aminotransferase.

Table 2

Preoperative factors associated with postoperative mortality

Variable	Univariate		Multivariate
	HR (95% CI)	p	HR (95% CI)	p
Age (yrs)	1.034 (1.010–1.059)	0.006	1.026 (0.999–1.053)	0.058
Men	1.023 (0.488–2.142)	0.952
Body mass index (kg/m 2 )	1.006 (0.951–1.065)	0.828
Coronary artery disease	1.151 (0.121–2.543)	0.729
Diabetes mellitus	1.077 (0.573–2.024)	0.819
Hyperlipidemia ∗	1.167 (0.625–2.179)	0.628
Hypertension	0.385 (0.131–1.130)	0.082
Peripheral vascular disease	1.364 (0.738–2.519)	0.322
Renal failure (chronic kidney disease stage ≥3)	0.856 (0.465–1.577)	0.618
Albumin (g/dl)	0.462 (0.266–0.803)	0.006	0.521 (0.290–0.934)	0.029
White blood cell (×10 3 /μl)	1.014 (0.947–1.085)	0.695
Hematocrit (%)	0.983 (0.935–1.033)	0.503
Platelets (×10 3 /μl)	0.994 (0.990–0.998)	0.004	0.996 (0.992–1.000)	0.030
Total bilirubin (mg/dl)	1.126 (0.899–1.410)	0.301
Sodium (mEq/L)	0.971 (0.956–1.026)	0.598
Potassium (mEq/L)	1.017 (0.618–1.674)	0.947
Blood urea nitrogen (mg/dl)	1.019 (1.004–1.034)	0.014	1.014 (0.997–1.032)	0.100
Creatinine (mg/dl)	1.659 (1.082–2.546)	0.020
Aspartate aminotransferase (IU/L)	1.001 (1.000–1.002)	0.009	1.001 (1.000–1.002)	0.025
Alanine aminotransferase (IU/L)	1.001 (1.000–1.002)	0.006

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Race/Ethnic Disparities in Risk Factor Control and Survival in the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) Trial
2. Relation Between Preoperative Renal Dysfunction and Cardiovascular Events (Stroke, Myocardial Infarction, or Heart Failure or Death) Within Three Months of Isolated Coronary Artery Bypass Grafting
3. Long-Term Effect of Continuing Sports Activity in Competitive Athletes With Frequent Ventricular Premature Complexes and Apparently Normal Heart
4. Trauma in Patients With Continuous-Flow Left Ventricular Assist Devices

Stay updated, free articles. Join our Telegram channel

Join