Abstract
Background:
Significant weight loss due to cardiac cachexia is an independent predictor of mortality in many heart failure (HF) clinical trials. The impact of significant weight loss while on the waitlist for heart transplant (HT) has yet to be studied with respect to post-transplant survival.
Methods:
Adult HT recipients from 2010–2021 were identified in the UNOS registry. Patients who experienced an absolute weight change from the time of listing to transplant were included and classified into 2 groups by percent weight loss from time of listing to time of transplant using a cut off of 10%. The primary endpoint was 1-year survival following HT.
Results:
5,951 patients were included in the analysis, of which 763 (13%) experienced ≥10% weight loss from the time of listing to transplant. Weight loss ≥10% was associated with reduced 1-year post-transplant survival (86.9% vs. 91.0%, long-rank p=0.003). Additionally, weight loss ≥10% was an independent predictor of 1-year mortality in a multivariable model adjusting for significant risk factors (adjusted HR 1.23, 95% CI 1.04–1.46). In secondary analyses, weight loss ≥10% was associated with reduced 1-year survival independent of hospitalized status at time of transplant as well as obesity status at listing (i.e., body mass index (BMI) <30 kg/m2 and BMI ≥30 kg/m2).
Conclusions:
Preoperative weight loss ≥10% is associated with reduced survival in patients listed for HT. Nutrition interventions prior to transplant may prove beneficial in this population.
Keywords: Cardiac transplant, nutrition, cachexia
Introduction
The International Society for Heart and Lung Transplantation (ISHLT) guidelines support intentional reduction in body weight among those patients who meet the World Health Organization (WHO) body mass index (BMI) criteria for obesity in order to optimize outcomes after heart transplant (HT) (1). In practice, BMI ≥35 kg/m2 is considered a relative contraindication to HT in most centers given the association of class II-III obesity with post-operative complications as well as poor long-term survival (2). It is notable however that patients with advanced heart failure who meet criteria for overweight or class I obesity (BMI 30–34.9 kg/m2) demonstrate improved survival compared to their normal and underweight counterparts in a phenomenon known as the obesity paradox (3). Given the association of advanced heart failure with heightened inflammation, anabolic-catabolic imbalance, immobility, gut edema and malabsorption, overweight and/or class I obesity may confer a protective effect against the consequences of wasting in advanced heart failure (4). Among all patients with advanced heart failure regardless of BMI, significant unintentional weight loss known as cardiac cachexia, defined as edema-free weight loss at least ≥5% of body weight over a year or less, is associated with reduced survival (4). Whether the development of cardiac cachexia prior to HT is associated with survival after HT is unknown.
Methods
All adult, single organ patients who underwent HT from January 1st, 2010-March 4th, 2021, were identified in the United Network for Organ Sharing (UNOS) registry, with follow up data through June 4th, 2021. The exposure of interest was weight loss from the time of listing to the time of transplant. To assess this, we divided the cohort into 2 groups by percent weight loss from the time of listing to the time of transplant using a cut off of 10%. While cardiac cachexia is defined as an unintentional weight loss of at least 5%, we chose a cut off of 10% in this population to account for potential fluid-related weight loss that may occur with pharmacologic and mechanical HF therapies prior to HT. Notably, for the primary analysis, we excluded patients with missing data for weight at the time of listing and at transplant (i.e., weight change of zero). The primary outcome was 1-year mortality following HT. Secondary outcomes included length of stay for the post-transplant index hospitalization, need for permanent renal replacement therapy, and the incidence of post-transplant infection. The Institutional Review Board of Columbia University Irving Medical Center previously determined analyses using the UNOS registry (de-identified public data set) as exempt.
Statistical Analysis
We compared baseline characteristics by weight loss category (≥10% vs. <10% weight loss). Categorical variables were summarized as counts (percentages) and were compared by weight loss group using the chi-square test. Continuous variables were summarized as medians (interquartile range) and were compared by weight loss group using the Wilcoxon rank sum test as the data were non-parametric. For time to event analysis, we utilized the Kaplan-Meier method to estimate time to death after HT and compared these estimates by weight loss group using the log-rank test. We generated Cox proportional hazards models to examine 1-year mortality following HT. We tested the impact of weight loss (≥10%) along with potential confounding variables in a univariate analysis. Those variables determined to be univariate predictors (p <0.2) were included in the multivariable analysis and included weight loss ≥10%, age, sex, obesity (defined using the WHO obesity criteria), former smoking, diabetes, renal dysfunction (CKD stage 3 or greater), heart failure etiology (non-ischemic vs. other), pulmonary capillary wedge pressure (PCWP), presence of temporary mechanical support (MCS) at time of transplant, intensive care unit (ICU) disposition at time of transplant, and waitlist time. Secondarily, we examined the impact of weight loss by the patient’s hospitalized status at the time of transplant (i.e., hospitalized vs. non-hospitalized) and by BMI category (i.e., BMI <30 versus BMI ≥30 kg/m2) using Kaplan Meier analysis and Cox proportional hazards modeling as above. We additionally used Cox proportional hazards modeling to determine clinical predictors of weight loss ≥10% while on the waitlist. To assess for nonlinearity between percent weight change on the waitlist as a continuous variable and post-transplant mortality we fitted restricted cubic spline models with three knots. There were missing data for cardiac hemodynamics. Missing data were assumed to be missing completely at random and handled with multiple imputation using a Markov chain Monte Carlo method to generate ten imputations to allow for multivariable analysis of the entire cohort. Sensitivity analyses were performed assessing the impact of various weight loss cut-offs, time on the waitlist, analyzing every transplant recipient (i.e., those with and without weight loss, and adjusting for significant changes in wedge pressure. A p-value of <0.05 was considered statistically significant in all analyses. Statistical analyses were performed using SAS Version 9.4 (Cary, NC) software.
Results
Baseline Characteristics
From 2010–2021, there were 24,804 adults who underwent HT. After excluding those with missing weight data (n=18,853), 5,951 patients were included in this analysis (Table 1). The median age in this cohort was 55 years with 22.7% reporting female sex. Notably, the majority of the cohort (71.1%) was classified overweight or obese at listing (BMI ≥25 kg/m2). The median waitlist time for the overall cohort was 253 days. The majority of the cohort had left ventricular assist device (LVAD) at the time of listing (56.4%) and were outpatient at the time of transplant (56.4%).
Table 1.
Baseline characteristics by weight loss status (≥10% weight loss versus <10% weight loss)
| Variable | Total (N=5951) | ≥10% Weight Loss (N=763) | <10% Weight Loss (N=5188) | p-value |
|---|---|---|---|---|
| Demographics | ||||
| Age | 55 (45–62) | 55 (43–61) | 55 (45–62) | 0.10 |
| Female sex | 1349 (22.7%) | 192 (25.2%) | 1157 (22.3%) | 0.08 |
| Race | 0.12 | |||
| White | 3856 (64.8%) | 484 (63.4%) | 3372 (65.0%) | |
| Black | 1374 (23.1%) | 199 (26.1%) | 1175 (22.7%) | |
| Hispanic | 499 (8.4%) | 53 (7.0%) | 446 (8.6%) | |
| Other | 222 (3.7%) | 27 (3.5%) | 195 (3.8%) | |
| Blood type | 0.10 | |||
| A | 2243 (37.7%) | 260 (34.1%) | 1983 (38.2%) | |
| B | 826 (13.9%) | 117 (15.3%) | 709 (13.7%) | |
| AB | 240 (4.0%) | 27 (3.5%) | 213 (4.1%) | |
| O | 2642 (44.4%) | 359 (47.1%) | 2283 (44.0%) | |
| Body mass index (kg/m2) | <0.0001 | |||
| Underweight (<18.5) | 146 (2.5%) | 18 (2.4%) | 128 (2.5%) | |
| Normal (18.5–24.9) | 1571 (26.4%) | 121 (15.9%) | 1450 (28.0%) | |
| Overweight (25–29.9) | 2120 (35.6%) | 240 (31.5%) | 1880 (36.2%) | |
| Class I obesity (30–34.9) | 1600 (26.9%) | 241 (31.6%) | 1359 (26.2%) | |
| Class II obesity (35–39.9) | 514 (8.6%) | 143 (18.7%) | 371 (7.2%) | |
| Heart failure etiology | 0.005 | |||
| Ischemic | 1917 (32.2%) | 218 (28.6%) | 1699 (32.8%) | |
| Non-ischemic | 3563 (59.9%) | 462 (60.7%) | 3100 (59.8%) | |
| Restrictive/infiltrative | 184 (3.1%) | 37 (4.9%) | 147 (2.8%) | |
| Congenital | 169 (2.8%) | 25 (3.3%) | 144 (2.8%) | |
| Re-transplant | 118 (2.0%) | 20 (2.6%) | 98 (1.9%) | |
| Medical comorbidities | ||||
| ICD prior to transplant | 4683 (78.7%) | 588 (77.1%) | 4095 (77.1) | 0.24 |
| Prior smoking | 2818 (47.4%) | 8354 (46.4%) | 2464 (47.5%) | 0.57 |
| Diabetes | 1711 (28.8%) | 212 (27.8%) | 1499 (28.9%) | 0.52 |
| Prior stroke | 349 (5.9%) | 34 (4.5%) | 315 (6.1%) | 0.08 |
| Renal dysfunction (CKD 3 or greater) | 2060 (34.6%) | 278 (36.4%) | 1782 (34.4%) | 0.26 |
| Hemodynamic (at listing) | ||||
| PA systolic (mmHg) | 42 (32–53) | 45 (35–55) | 42 (32–53) | <0.0001 |
| PA diastolic (mmHg) | 20 (15–27) | 22 (16–29) | 20 (15–27) | <0.0001 |
| PA mean (mmHg) | 29 (22–37) | 30 (24–38) | 29 (21–36) | <0.0001 |
| PCWP (mmHg) | 20 (13–26) | 21 (15–27) | 19 (13–25) | <0.0001 |
| Cardiac index (L/min/m2) | 2.07 (1.71–2.48) | 2.03 (1.59–2.47) | 2.08 (1.72–2.49) | 0.006 |
| Clinical characteristics at transplant | ||||
| Waitlist time (days) | 253 (102–527) | 199 (79–469) | 262 (106–536) | <0.0001 |
| UNOS status | <0.0001 | |||
| 1A (old allocation system) | 3149 (52.9%) | 472 (61.9%) | 2677 (51.6%) | |
| 1B (old allocation system) | 1294 (21.7%) | 118 (15.5%) | 1176 (22.7%) | |
| 2 (old allocation system) | 75 (1.3%) | 7 (0.9%) | 68 (1.3%) | |
| 1 | 101 (1.7%) | 20 (2.6%) | 81 (1.6%) | |
| 2 | 584 (9.8%) | 79 (10.4%) | 505 (9.7%) | |
| 3 | 347 (5.8%) | 43 (5.6%) | 304 (5.9%) | |
| 4 | 359 (6.0%) | 21 (2.8%) | 338 (6.5%) | |
| 6 | 42 (0.7%) | 3 (0.4%) | 39 (0.8%) | |
| Mechanical support | ||||
| IABP | 476 (8.0%) | 82 (10.8%) | 394 (7.6%) | 0.003 |
| ECMO | 53 (0.9%) | 18 (2.4%) | 35 (0.7%) | <0.0001 |
| LVAD | 3356 (56.4%) | 352 (46.1%) | 3004 (57.9%) | <0.0001 |
| Hospitalization status at transplant | <0.0001 | |||
| Intensive care unit | 1506 (25.5%) | 277 (36.7%) | 1229 (23.4%) | |
| Hospitalized | 1070 (18.1%) | 184 (24.4%) | 886 (17.2%) | |
| Non-hospitalized | 3328 (56.4%) | 293 (38.8%) | 3035 (58.9%) | |
CKD = chronic kidney disease; ECMO = extracorporeal membrane oxygenation; IABP = intraaortic balloon pump; ICD = implantable cardioverter defibrillator; LVAD = left ventricular assist device; PA = pulmonary artery; PCWP = pulmonary capillary wedge pressure; UNOS = United Network for Organ Sharing
Of the total cohort, 13% (n=763) experienced weight loss ≥10% from the time of listing to the time of transplant. The average weight loss among those who lost ≥10% body weight was 13.6 kg and the average weight change among those who lost <10% body weight was a weight gain of 2.0 kg.
Weight Loss and Post-Transplant Survival
Among the total cohort of patients who underwent HT, there were 539 deaths at 1 year of follow up, corresponding to 90.5% overall 1-year survival. Those patients with >10% weight loss had a survival estimate at 1-year of 86.9% compared to 91.0% among those with weight loss 10% was a significant univariate predictor of 1-year post-transplant mortality in this cohort (HR 1.33, 95% CI 1.14–1.56) (Table 2). Other univariate predictors included age, female sex, obesity, former smoking, diabetes, renal dysfunction, non-ischemic cardiomyopathy, PCWP at listing, temporary MCS at transplant, ICU status at transplant, and wait time. In a multivariable model adjusting for the above predictors, weight loss >10% remained a significant predictor of 1-year mortality in the overall cohort, with 23% increased risk of death at 1-year post-HT (HR 1.23, 95% CI 1.04–1.46, p=0.02). Female sex, obesity, former smoking, diabetes, renal dysfunction, non-ischemic cardiomyopathy, PCWP at listing, and wait time remained significant predictors in the multivariable model (Table 2). 8 Further analysis demonstrated a gradient of effect, with an increasing risk of post-transplant mortality with increasing degree of weight loss of 5%, 7.5% and 10%, respectively (HR 1.13, 95% CI 1.001–1.28, HR 1.19, 95% CI 1.04–1.37, and HR 1.33, 95% CI 1.14–1.56, respectively). Notably, when we repeated the analysis to include all patients in the cohort i.e., those with and without an absolute weight change, the effect of weight loss > 10% was similar (HR 1.26, 95% CI 1.01–1.45, p=0.001, 95% CI 1.001–1.28, HR 1.19, 95% CI 1.04–1.37, and HR 1.33, 95% CI 1.14–1.56, respectively). When examined as a continuous variable, percent weight loss demonstrated a non-linear association with risk of post-transplant mortality. The risk of posttransplant mortality increased for any weight loss where the extremes of weight loss (especially greater than 10%) and weight gain were both associated with increased hazards for death at 1- year post-HT (Figure 2).
Table 2.
Univariable and multivariable Cox proportional hazards models for 1-year mortality
| HR | 95% CI | p-value | HR | 95% CI | p-value | |
|---|---|---|---|---|---|---|
|
| ||||||
| Weight loss ≥10% | 1.33 | 1.14–1.56 | 0.0003 | 1.23 | 1.04–1.47 | 0.02 |
| Female gender | 1.11 | 0.96–1.27 | 0.15 | 1.16 | 1.01–1.36 | 0.04 |
| Non-white race | 1.07 | 0.95–1.21 | 0.25 | |||
| Age (per 10-year increase) | 1.01 | 1.00–1.01 | 0.02 | 1.00 | 0.99–1.01 | 0.72 |
| Underweight (BMI<18.5 kg/m2) vs. normal | 1.24 | 0.87–1.76 | 0.24 | |||
| Obesity (BMI≥30) vs. normal | 1.30 | 1.16–1.47 | <0.0001 | 1.21 | 1.07–1.38 | 0.004 |
| Non-ischemic (vs. other) | 0.74 | 0.66–0.83 | <0.0001 | 0.79 | 0.69–0.90 | 0.0003 |
| Prior LVAD | 0.96 | 0.85–1.08 | 0.47 | |||
| PCWP at listing (per 1 unit increase) | 0.88 | 0.985–0.999 | 0.03 | 0.99 | 0.985–999 | 0.02 |
| Temporary MCS (IABP or ECMO) | 1.35 | 1.08–1.68 | 0.008 | 1.26 | 0.97–1.63 | 0.09 |
| ICD | 1.08 | 0.93–1.26 | 0.29 | |||
| Smoking | 1.21 | 1.08–1.37 | 0.001 | 1.20 | 1.06–1.37 | 0.005 |
| Prior stroke | 1.15 | 0.91–1.46 | 0.25 | |||
| Diabetes | 1.27 | 1.12–1.44 | 0.0002 | 1.19 | 1.04–1.37 | 0.01 |
| CKD Stage 3 or greater | 1.35 | 1.20–1.52 | <0.0001 | 1.31 | 1.15–1.49 | <0.0001 |
| ICU at transplant | 1.18 | 1.03–1.35 | 0.02 | 1.14 | 0.97–1.35 | 0.11 |
| Wait time (per 100-day increase) | 1.02 | 1.002–1.03 | 0.03 | 1.03 | 1.01–1.04 | 0.002 |
BMI = body mass index; CKD = chronic kidney disease; ECMO = extracorporeal membrane oxygenation; IABP = intraaortic balloon pump; ICD = implantable cardioverter defibrillator; ICU = intensive care unit; LVAD = left ventricular assist device; PA = pulmonary artery; MCS = mechanical circulatory support; PCWP = pulmonary capillary wedge pressure
Figure 2.
Restricted cubic spline analysis, nonlinear relationship between the hazard ratio for post-transplant mortality and percent weight change while on the wait list. The risk of mortality increases with increased weight loss while on the waitlist and there is an increase at the extreme of weight gain as well. The red lines indicate the 95% confidence intervals with a reference of zero weight change.
Weight Loss and Clinical Outcomes of Interest
Compared to those with <10% weight loss, individuals with weight loss ≥10% had a longer post-transplant length of stay as well as a higher incidence of post-transplant dialysis at 1 year (16.0% vs. 12.3%, p=0.004) (Table 3). There were no differences between groups in the incidence of allograft rejection, and/or post-transplant infection at 1 year.
Table 3.
Clinical outcomes by weight loss category (<10% vs. ≥10%)
| Outcome | Total (N=5951) | Weight loss ≥5% (N=1732) | Weight loss <5% (N=4219) | p-value |
|---|---|---|---|---|
| Length of stay (days) | 16 (11–24) | 17 (12–27) | 16 (11–24) | 0.008 |
| Post-transplant dialysis | 758 (12.7%) | 122 (16.0%) | 636 (12.3%) | 0.004 |
| Rejection at 1-year | 730 (13.4%) | 97 (14.2%) | 633 13.3%) | 0.53 |
| Post-transplant infection | 1591 (29.2%) | 196 (28.6%) | 1395 (29.3%) | 0.72 |
Weight Loss and Post-Transplant Survival by Hospitalized Status
We analyzed the impact of hospitalized status on the relationship between weight loss and 1-year survival. Of those who underwent HT, 44% were hospitalized at the time of HT and 66% were ambulatory. Regardless of inpatient versus outpatient disposition at the time of HT, weight loss ≥10% was associated with worse 1-year survival in time-to-event analysis (log-rank p=0.0004, Figure 3). In the predictive model for 1-year mortality, weight loss ≥10% was significantly associated with increased risk for 1-year post-transplant mortality whether the patient was transplanted from the hospitalized or outpatient setting (HR 1.30, 95% CI 1.06–1.61; HR 1.64, 95% CI 1.30–2.06, respectively). Notably, compared to those who were outpatient with weight loss <10% prior to transplant, being hospitalized with weight loss <10% was not associated with increased risk for 1-year mortality in this cohort.
Figure 3.
Kaplan Meier analysis demonstrating one-year mortality post-heart transplant by weight loss group (≥10% versus <10%) stratified by hospitalized status (inpatient versus outpatient)
Weight Loss and Post-Transplant Survival by Obesity Status
We also assessed the impact of obesity status (i.e., BMI <30 kg/m2 vs. BMI ≥30 kg/m2) on the association between weight loss and 1-year mortality after heart transplant. Comparing 1-year survival by BMI category and weight loss status, survival was worse among those who experienced weight loss ≥10% regardless of obesity status (log rank p<0.0001) (Figure 4). In the predictive model, weight loss ≥10% was associated with increased risk for 1-year mortality in both obese and non-obese patients (HR 1.66, 95% CI 1.34–2.05 and HR 1.26, 95% CI 1.01–1.58, respectively). Notably, regardless of weight loss, obesity status at the time of listing (BMI >30 kg/m2) was associated with worse post-transplant survival, including those patients who lost <10% of their listing weight (HR 1.27, 95% CI 1.11–1.45). Further analysis including categorizing patients as underweight, normal/overweight and/or obese patients demonstrated that weight loss ≥10% was associated with worse survival within any BMI category (Supplemental File). In this model, obesity with both categories of weight loss (≥10% and <10%) was associated with higher risk for mortality at 1 year. The risk, however, was greater among obese patients those who experienced weight loss ≥10% (HR 1.68, 95% CI 1.36–2.07) compared to those obese patients with <10% weight loss (HR 1.28, 95% CI 1.12–1.47) (Supplemental Table 1).
Figure 4.
1 Kaplan Meier analysis demonstrating one-year mortality post-heart transplant by weight loss group (≥10% versus <10%) stratified by obesity status (obese versus non-obese)
Clinical predictors of weight loss
In this exploratory analysis, we sought to determine clinical predictors of weight loss ≥10% from the time of listing to HT (Table 3, Figure 5). We found that overweight, class I, and class II obesity were significant predictors of weight loss ≥10% in the multivariable model. Higher PCWP, Stage III CKD, and ICU status at the time of transplant were also significant predictors of greater weight loss. Notably, those with an LVAD were less likely to experience weight loss ≥10%.
Figure 5.
Logistic regression analysis demonstrating clinical predictors of weight loss ≥10%. Displayed are odds ratios with 95% confidence
Sensitivity Analyses
Three sensitivity analyses were performed. The initial sensitivity analysis adjusted weight loss for duration of waitlist time, where the cohort was separated based on the 25th quartile or 1.85% weight loss/100 waitlist days. The outcomes were similar to the primary analysis in that there was a signal that individuals with more than 1.85% weight loss/100 waitlist days had an increased risk of mortality (HR 1.13, 95% CI 0.98–1.29, p=0.10). The next analysis limited the cohort to patients who were on the waitlist for at least 60 days, to allow an adequate amount of time for non-fluid weight loss. In this cohort of 5,041 patients, the findings were similar with a 54% increase in risk of post-transplant mortality among individuals with ≥10% weight loss (95% CI 1.31–1.82, p<0.0001). For the final sensitivity analysis, we sought to determine if there was a difference in the amount of fluid weight loss between weight loss groups (i.e., weight loss ≥10% versus weight loss <10%) using a decrease PCWP of at least 5 mm Hg as a surrogate for fluid weight loss. In those with available data (n = 5,321), we found that the proportion of patients with a decrease in PCWP of at least 5 mmHg was similar between groups, with 27.9% in the weight loss ≥10% group and 29.5% in the weight loss <10% group experiencing a decrease in PCWP of at least 5 mm Hg (p=0.42).
Discussion
This analysis of a contemporary cohort of adult HT recipients in the United States found that weight loss ≥10% from the time of listing to transplant was an independent predictor of 1-year post-transplant mortality. Notably, weight loss ≥10% prior to transplant was associated with worse survival regardless of patient location at the time of transplant (i.e., inpatient versus outpatient) and obesity status at the time of listing (i.e., BMI <30 kg/m2 versus BMI ≥ 30 kg/m2). In addition, weight loss ≥10% was associated with longer length of stay post-transplant and a higher incidence of dialysis at 1-year post-transplant.
While our study is the first to study the impact of significant weight loss on post-operative survival after HT, numerous studies have reported similar associations between weight loss and survival in populations of patients with chronic advanced heart failure (5). Cardiac cachexia, or the unintentional weight loss of at least 5% of non-edematous body weight is a well-established independent predictor of poor prognosis in advanced HF populations (6–8). The majority of the data the association between the development of cardiac cachexia comes from cohorts of patients with HFrEF (6–8). However, a recent analysis restricted to patients with preserved ejection fraction (HFpEF) reported that weight loss ≥5% within the 6 months following a HF hospitalization was associated with more than 5-fold risk for mortality compared to those who lost less weight (9). In addition, in an analysis of patients enrolled in the TOPCAT trial, significant weight loss ≥5% over a median follow-up period of 3 years was an independent predictor of mortality (10). The pathophysiology of cardiac cachexia includes a complex disarray of metabolic, immunologic, and neuroendocrine processes that culminate in a state of heightened inflammation and catabolism leading to malnutrition and involuntary weight loss associated with poor quality of life, functional capacity, and increased mortality (4,11,12). Given their advanced disease, patients listed for HT are likely at the highest risk for HF wasting syndromes, which is evident by fact that 13% of our cohort experienced at least 10% weight loss from the time of listing to HT. Accordingly, patients with at least 10% weight loss had more severe disease with worse hemodynamics, higher incidence of temporary MCS, and consequently a higher UNOS status at listing. Our analysis is the first to show that in the pre-HT population, the development of cardiac cachexia while on the waitlist, as indicated by weight loss ≥10%, is an independent predictor of 1-year post-transplant mortality.
As an exploratory analysis, we examined clinical predictors of weight loss ≥10% and found that patients with overweight and obesity (any class) were most likely to lose weight while on the waitlist. Importantly underweight BMI at listing was also a significant risk factor for significant weight loss. Given the association between underweight with poor outcomes post-transplant, including mortality, infection, and bleeding complications, this is an especially important group to monitor for the development of cachexia (2).
It is important to note that weight loss associated with cardiac cachexia specifically refers non-fluid weight loss (11). Prior studies of patients with cardiac cachexia have shown that weight loss ≥5% was independent predictor of survival even after controlling for plasma volume, peripheral edema, and diuretic dose, indicating that loss of lean mass is prognostic in this population (8). Notably, in the UNOS database, we are unable to precisely categorize weight loss that was due to fluid versus weight loss due to lean and/or fat mass however we attempted to address this using several methods. First, we set the cutoff for significant weight loss at 10% instead of the traditional cut-off of 5% due to the inability to adjust for certain parameters that may indicate loss of fluid weight (i.e., right atrial pressure). In addition, we examined PCWP as a surrogate for volume status, demonstrating equal proportions of patients with a decrease in PCWP of at least 5 mm Hg in both weight loss groups such that fluid-related weight loss likely did not differ significantly between weight loss groups. In addition, our sensitivity analyses, which indexed weight loss to wait list time and restricted the analysis to only those on the wait list for at least 60 days to allow adequate time for non-fluid weight loss, showed similar findings to our primary analysis in that a greater degree of weight loss was associated with poorer prognosis. Despite this, future studies that incorporate body composition analysis, using gold standard techniques such as computed tomography (CT) and/or magnetic resonance imaging MRI), are imperative to precisely quantify the loss of lean and/or fat mass versus fluid mass and determine the associations of these measures with outcomes in patients listed for HT (11).
Given the higher prevalence of malnutrition in the hospital setting, we analyzed the impact of hospitalization status on the association between weight loss and survival after HT (13). In this secondary analysis, we found that, compared to those who were outpatient with <10% weight loss, those with weight loss ≥10% demonstrated increased risk for mortality at 1 year post-transplant regardless of inpatient versus outpatient disposition at the time of transplant (Figure 4). The association between reduced survival with weight loss even in the outpatient setting emphasizes that significant weight loss is a risk factor for poor outcomes regardless of the immediate acuity of disease. These data demonstrate that patients with this HF phenotype in any care setting may benefit from closer monitoring and nutrition interventions to prevent ongoing weight loss.
It is also important to note that patients with advanced HF are at risk for cardiac cachexia regardless of initial BMI. As such, we examined the impact of weight loss on survival by obesity status. This secondary analysis found that weight loss ≥10% was associated with worse survival in both obese and non-obese patients. Our findings are consistent with those of both the GISSI-HF and ValHeFT trials in which weight loss ≥5% over 1 year was associated with reduced survival even after controlling for baseline BMI (8). While obesity itself was a risk factor for increased mortality regardless of degree of weight loss, obesity at listing with ≥10% weight loss prior to transplant was associated with the highest risk for 1-year mortality post-transplant of any subgroup. Therefore, while it still is important to treat morbid obesity prior to HT given its associations with poor post-operative outcomes, it is likely just as important to recognize significant unintentional weight loss (i.e., cardiac cachexia) as a marker of increased risk even in patients who meet criteria for obesity at listing. These patients may be missed clinically.
Lastly, in addition to increased risk for post-transplant mortality, we found weight loss ≥10% prior to transplant to be associated with a longer length of stay at the index hospitalization post-HT and a greater incidence of post-transplant dialysis. Given the greater severity of illness including a higher incidence of chronic kidney disease in those patients with ≥10% weight loss, these findings are not surprising. Notably, there are studies that demonstrate that the cachexia associated with advanced HF and/or chronic kidney disease may exacerbate the other end organ dysfunction, which may explain this association (14).
Limitations
There are several important limitations of this analysis that must be considered in interpreting the results. First are the inherent limitations related to the UNOS database, including coding errors and missing data, and we only assessed weight loss among those who underwent HT. Notably, we excluded a significant number of patients for our primary analysis due to the same weight being present at listing and at the time of transplantation, which limited our sample size. Reassuringly, the principal findings were consistent when the whole cohort was analyzed. One important limitation specific to this analysis is the inability to distinguish between loss of fluid weight and loss of non-fluid weight. Given the prognostic implications of reduction in lean mass as opposed to fluid mass, and unintentional versus intentional weight loss, these are important considerations (11). To address this, we attempted to account for fluid weight by including PCWP as a covariate in our predictive models and in a sensitivity analysis which demonstrated that the change in PCWP from the time of listing to the time of transplant was similar between both weight loss groups, though this data was not available for all patients and may not be fully representative. While right atrial pressures would have been a more accurate surrogate for fluid status, this datapoint was not available in this database. In addition, we are unable to distinguish between intentional and unintentional weight loss. We attempted to account for this by using a more stringent definition for cardiac cachexia using a cutoff of 10% rather than 5%. Finally, our exploratory analyses stratified by BMI and hospitalization status are underpowered to predict mortality in this analysis and should be considered hypothesis-generating.
Conclusion
Despite the limitations, the data from this analysis indicate that the development of cardiac cachexia (i.e., weight loss ≥10%) while on the waitlist for heart transplant is associated with a high-risk clinical phenotype and is an independent predictor of 1-year survival after HT. Patients had reduced survival with weight loss ≥10% regardless of inpatient versus outpatient disposition and obesity status at the time of transplant. We believe that this analysis supports the need for screening and potential treatment of cardiac cachexia in patients listed for HT to optimize post-transplant outcomes.
Supplementary Material
Figure 1.
Kaplan Meier analysis demonstrating one-year mortality post-heart transplant by weight loss group (≥10% versus <10%).
Table 4.
Univariable and multivariable predictors of weight loss greater than 10% while on the waitlist
| OR | 95% CI | p-value | OR | 95% CI | p-value | |
|---|---|---|---|---|---|---|
| Female gender | 1.09 | 0.95–1.24 | 0.21 | |||
| Non-white race | 0.95 | 0.84–1.07 | 0.38 | |||
| Age (per 10-year increase) | 0.99 | 0.99–1.01 | 0.16 | 1.00 | 0.95–1.006 | 0.96 |
| Underweight (BMI<18.5 kg/m2) vs. normal | 0.71 | 0.44–1.13 | 0.15 | 0.71 | 0.44–1.17 | 0.14 |
| Overweight (30>BMI≥25 kg/m2) vs. normal | 1.44 | 1.23–1.68 | <0.001 | 1.59 | 1.34–1.88 | <0.001 |
| Class I Obese (35>BMI≥30 kg/m2) vs. normal | 2.11 | 1.80–2.48 | <0.001 | 2.41 | 2.02–2.88 | <0.001 |
| Class II Obesity (BMI≥35) vs. normal | 4.61 | 3.72–5.70 | <0.001 | 6.16 | 4.86–7.80 | <0.001 |
| Non-ischemic (vs. other) | 1.09 | 0.97–1.26 | 0.13 | 1.16 | 1.01–1.32 | 0.03 |
| Prior LVAD | 0.50 | 0.45–0.56 | <0.001 | 0.52 | 0.45–0.60 | <0.001 |
| PCWP at listing (per 1 unit increase) | 1.00 | 1.01–1.03 | <0.001 | 1.02 | 1.01–1.03 | <0.001 |
| Temporary MCS (IABP or ECMO) | 1.79 | 1.49–2.16 | <0.001 | 0.92 | 0.73–1.15 | 0.45 |
| ICD | 0.91 | 0.80–1.05 | 0.19 | 1.94 | 0.80–1.10 | 0.43 |
| Smoking | 0.96 | 0.86–1.08 | 0.49 | |||
| Prior stroke | 0.79 | 0.61–1.01 | 0.06 | 0.77 | 0.58–1.10 | 0.06 |
| Diabetes | 1.12 | 0.99–1.26 | 0.08 | 1.06 | 0.92–1.21 | 0.45 |
| CKD stage 3 or greater | 1.27 | 1.13–1.42 | <0.001 | 1.16 | 1.01–1.32 | 0.03 |
| ICU at transplant | 2.16 | 1.91–2.44 | <0.001 | 1.63 | 1.39–1.92 | <0.001 |
| Wait time (per 100-day increase) | 0.95 | 0.94–0.96 | <0.001 | 0.97 | 0.96–0.99 | <0.001 |
BMI = body mass index; CKD = chronic kidney disease; ECMO = extracorporeal membrane oxygenation; EGFR = estimated glomerular filtration rate; IABP = intraaortic balloon pump; ICD = implantable cardioverter defibrillator; ICU = intensive care unit; LVAD = left ventricular assist device; PA = pulmonary artery; MCS = mechanical circulatory support; PCWP = pulmonary capillary wedge pressure
Footnotes
Disclosures:
None
Supplemental Material:
References
- 1.Mehra MR, Canter CE, Hannan MM et al. The 2016 International Society for Heart Lung Transplantation listing criteria for heart transplantation: A 10-year update. J Heart Lung Transplant 2016;35:1–23. [DOI] [PubMed] [Google Scholar]
- 2.Doumouras BS, Fan CS, Mueller B et al. The effect of pre-heart transplant body mass index on posttransplant outcomes: An analysis of the ISHLT Registry Data. Clin Transplant 2019;33:e13621. [DOI] [PubMed] [Google Scholar]
- 3.Lavie CJ, Alpert MA, Arena R, Mehra MR, Milani RV, Ventura HO. Impact of obesity and the obesity paradox on prevalence and prognosis in heart failure. JACC Heart Fail 2013;1:93–102. [DOI] [PubMed] [Google Scholar]
- 4.von Haehling S, Ebner N, Dos Santos MR, Springer J, Anker SD. Muscle wasting and cachexia in heart failure: mechanisms and therapies. Nat Rev Cardiol 2017;14:323–341. [DOI] [PubMed] [Google Scholar]
- 5.Bielecka-Dabrowa A, Ebner N, Dos Santos MR, Ishida J, Hasenfuss G, von Haehling S. Cachexia, muscle wasting, and frailty in cardiovascular disease. Eur J Heart Fail 2020;22:2314–2326. [DOI] [PubMed] [Google Scholar]
- 6.Anker SD, Negassa A, Coats AJ et al. Prognostic importance of weight loss in chronic heart failure and the effect of treatment with angiotensin-converting-enzyme inhibitors: an observational study. Lancet 2003;361:1077–83. [DOI] [PubMed] [Google Scholar]
- 7.Pocock SJ, McMurray JJ, Dobson J et al. Weight loss and mortality risk in patients with chronic heart failure in the candesartan in heart failure: assessment of reduction in mortality and morbidity (CHARM) programme. Eur Heart J 2008;29:2641–50. [DOI] [PubMed] [Google Scholar]
- 8.Rossignol P, Masson S, Barlera S et al. Loss in body weight is an independent prognostic factor for mortality in chronic heart failure: insights from the GISSI-HF and Val-HeFT trials. Eur J Heart Fail 2015;17:424–33. [DOI] [PubMed] [Google Scholar]
- 9.Kamisaka K, Kamiya K, Iwatsu K et al. Impact of weight loss in patients with heart failure with preserved ejection fraction: results from the FLAGSHIP study. ESC Heart Fail 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ferreira JP, Rossignol P, Claggett BL et al. Weight changes in heart failure with preserved ejection fraction: findings from TOPCAT. Clin Res Cardiol 2021. [DOI] [PubMed] [Google Scholar]
- 11.Carbone S, Billingsley HE, Rodriguez-Miguelez P et al. Lean Mass Abnormalities in Heart Failure: The Role of Sarcopenia, Sarcopenic Obesity, and Cachexia. Curr Probl Cardiol 2020;45:100417. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Pureza V, Florea VG. Mechanisms for cachexia in heart failure. Curr Heart Fail Rep 2013;10:307–14. [DOI] [PubMed] [Google Scholar]
- 13.Kaegi-Braun N, Mueller M, Schuetz P, Mueller B, Kutz A. Evaluation of Nutritional Support and In-Hospital Mortality in Patients With Malnutrition. JAMA Netw Open 2021;4:e2033433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cicoira M, Anker SD, Ronco C. Cardio-renal cachexia syndromes (CRCS): pathophysiological foundations of a vicious pathological circle. J Cachexia Sarcopenia Muscle 2011;2:135–142. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.