|
|
||||||||
a Division of Cardiac Surgery, Department of Surgery, The Johns Hopkins Medical Institutions Baltimore, Maryland
b Section of Cardiac Surgery, Department of Surgery, Yale University School of Medicine, New Haven, Connecticut
Accepted for publication November 18, 2011.
* Address correspondence to Dr Yuh, Section of Cardiac Surgery, Yale University School of Medicine, 330 Cedar St, Boardman 204, New Haven, CT 06510 (Email: david.yuh{at}yale.edu).
| Abstract |
|---|
|
|
|---|
Methods: Adult patients undergoing OHT between 1998 and 2008 were identified in the United Network for Organ Sharing registry. The impact of metabolic risk factors (hypertension, diabetes mellitus, and obesity) on mortality post-OHT was evaluated in a Cox proportional hazards regression analysis adjusted for other variables associated with survival in univariate analysis (exploratory p value < 0.2). Kaplan-Meier survival estimates were compared with the log-rank test.
Results: A total of 15,960 eligible patients underwent OHT during the study period. There were 6,368 (39.9%) patients with none of these risk factors, 6,138 (38.5%) with 1 risk factor, 2,811 (17.6%) with 2 risk factors, and 643 (4.0%) who had all 3 risk factors. After adjusting for other significant variables influencing survival, each individual risk factor independently increased the likelihood of mortality post-OHT (hypertension: HR 1.10 [1.03 to 1.17]; diabetes: HR 1.22[1.13 to 1.31]; obesity: HR 1.17 [1.10 to 1.26], each p < 0.01). There was an exponential trend of increasing mortality with the addition of each risk factor (r2 = 0.99, p < 0.001) such that patients with all 3 risk factors had a 63% increased mortality compared with those with no risk factors (HR 1.63 [1.42 to 1.88], p < 0.001). There was also a significant trend in declining 5-year survival rates with an increasing number of risk factors: 0 (74.7%), 1 (71.3%), 2 (68.2%), and 3 (63.1%) (p < 0.001).
Conclusions: This large-cohort study demonstrates that an increasing number of metabolic risk factors in OHT recipients is associated with exponential increases in postoperative mortality rates.
| Introduction |
|---|
|
|
|---|
| Patients and Methods |
|---|
|
|
|---|
Study Design
All patients undergoing OHT between 1998 and 2008 were identified in the UNOS registry. The study was limited to adult (
18 years) recipients. Patients undergoing retransplantation or multivisceral transplantation were excluded, as were patients in whom the presence or absence of these metabolic risk factors was not recorded. The primary endpoint was all-cause mortality after OHT.
Data and Statistical Analyses
Baseline characteristics were initially compared between patients with all 3 risk factors and patients with no risk factors. All recipient, transplant, and donor covariates were then evaluated in univariate Cox proportional hazards regression analysis to determine their impact on the primary endpoint of all-cause mortality after OHT. These variables were assessed at the time of transplant, not at the time of listing. These covariates included UNOS status, age, serum creatinine, serum bilirubin, body mass index, sex, race, mechanical ventilation, hypertension, bridging with ventricular assist devices, intraaortic balloon pump, intensive care unit stay prior to OHT, recipient serum cytomegalovirus status, biatrial versus bicaval anastomotic technique, year of transplant, days on waitlist, ischemic time, donor age, donor body mass index, donor sex, donor race, donor serum cytomegalovirus status, donor cigarette use, donor diabetes, donor hypertension, donor inotropes, human leukocyte antigen matching, cytomegalovirus status matching, race matching, gender matching, donor to recipient body mass index ratio, and blood type matching. Those factors found to be associated with mortality (exploratory p value < 0.2) were entered into a multivariable Cox regression model. Variables with more than 20% missing data were excluded from this multivariate model. The impact of metabolic risk factors on mortality was then evaluated in the multivariate model, adjusting for significant variables identified in the univariate analysis. All hazard ratios were presented with 95% confidence intervals.
In addition, the Kaplan-Meier method was used to model survival, and survival curves were compared using the log-rank test. The
2 test was performed for categoric data and the Student t test for continuous data. All analyses were performed using STATA software, version 11 (StataCorp LP, College Station, TX).
| Results |
|---|
|
|
|---|
|
|
Predictors of Mortality After OHT
Given these significant yet expected baseline differences, a comprehensive univariate Cox regression analysis was conducted on all plausible recipient, donor, and transplant covariates to test their respective impact on mortality. The only variable that had more than 20% missing data was donor inotrope use, a factor that was not significantly associated with mortality post-OHT (p = 0.54). A total of 23 covariates were found to be associated (p < 0.2) with mortality in univariate analysis, including the number of metabolic risk factors (hazard ratio [HR] 1.17 [1.13 to 1.20], p < 0.001) (Table 2). These variables were entered into a multivariate analysis to evaluate the independent effect of these risk factors on mortality. In multivariable analysis adjusting for all significant covariates, the number of risk factors as a continuous variable was found to independently impact mortality, such that each additional risk factor was associated with a 16% higher risk of mortality after OHT (HR 1.16 [1.12 to 1.20], p < 0.001). This effect was also seen when the number of risk factors was examined as a categoric variable. Moreover, the largest gap in increased mortality was seen in patients with 2 risk factors versus those with all 3 risk factors. Indeed, patients with 1 risk factor had 15% increased mortality versus those with 0 risk factors, those with 2 risk factors had 16% higher mortality than those with 1 risk factor, and those with all 3 risk factors had a 32% higher risk as compared with those with 2 risk factors. This exponential trend in increasing mortality with increasing risk factors was highly significant (r2 = 0.99, p < 0.001) (Fig 2).
|
|
|
|
| Comment |
|---|
|
|
|---|
Another important finding was that, as expected, the patients with all 3 risk factors were at baseline much different from those with no risk factors. As such, we decided to conduct our multivariate analysis adjusting for all other variables that were associated with survival as these populations significantly differed. After adjusting, we found that each of these individual risk factors increased the risk of mortality post-OHT. The addition of each risk factor was found to significantly impact mortality both as a continuous variable as well as a categoric variable. Perhaps most notably, there was a highly significant exponential trend in increasing mortality with an increasing number of risk factors.
Prevalence of Metabolic Disorders in the OHT Population
The observed rates of diabetes, hypertension, and obesity in our study were slightly lower than what would be predicted in this patient population based on epidemiologic data [3]. Also, only 4% of patients had all 3 risk factors. While this may be due to an inability to capture all of these patients using this particular dataset, it may also be a reflection of hesitancy to offer OHT to these patients. For instance, in a review of over 27,000 OHT recipients, obese patients were found to have longer wait times and have lower odds of undergoing OHT as compared with non-obese patients [7]. This finding was echoed in our study as well where patients with all 3 risk factors had longer wait times for OHT than patients with no risk factors. This may indicate particular biases in transplanting certain populations.
Impact of Individual Metabolic Risk Factors on Mortality
The data regarding the effects of obesity on OHT outcomes are mixed. Several prior investigations have found worse graft and patient survival in obese patients [8–10]. A recent UNOS analysis found that a BMI of 30 to 34.9 did not represent an increased risk of mortality after OHT although patients with a BMI of 35 or higher did have worse survival [11]. Our study indicates a higher mortality risk in obese individuals with BMI 30 kg/m2 or greater, although the magnitude of this risk was lower than that observed with diabetes.
With regard to pre-OHT diabetes, single-institution studies have also shown conflicting results [12–14]. A large multiinstitutional study utilizing the UNOS registry found that complicated diabetics specifically had worse survival as compared with nondiabetics and uncomplicated diabetics [15]. Although we did not stratify our diabetic cohort by severity, diabetes as a whole did impact mortality significantly in our analysis, as it was independently associated with a 22% increased risk. This was the largest adverse effect of the 3 metabolic risk factors.
The impact of pre-OHT hypertension on outcomes of OHT is much less studied than the other 2 risk factors. In our study pre-OHT hypertension conferred a 10% increased risk of mortality independent of the effects of diabetes or obesity. Although pre-OHT hypertension is not well studied, hypertension after OHT is a well-described complication that arises frequently and is commonly due to immunosuppressive agents [16]. Post-OHT hypertension is associated with cardiac allograft vasculopathy, the development of which represents an increased mortality risk to the recipient [17].
Impact of Metabolic Risk Factors in Combination
An important question is whether the increased post-OHT mortality when these metabolic risk factors are present in combination simply represents the sum of the increased mortality seen with each individual risk factor. In our study we noted that the hazard ratio for mortality in patients with all 3 risk factors was significantly higher than the sum of the ratios for its components. Additionally, although diabetes had the largest independent adverse effect (22% increased mortality risk), this was only 5% higher than the risk conferred by obesity and 12% higher than the risk from hypertension. In other words, no 1 risk factor had a predominant influence on the observed increased risk seen in patients with all 3 risk factors. Moreover, the largest jump in mortality was seen in going from 2 to 3 risk factors. The exponential trend seen in increasing mortality with an increasing number of risk factors was also highly significant. These data collectively suggest that while the presence of each of these risk factors is associated with an independent increase in mortality, their adverse impact is exponentially greater when all are present in combination.
This observed phenomenon of an exponential, and not linear, rise in mortality with an increasing number of risk factors is known as multiplicative risk. Potential reasons for this observation in our study were that several unmeasured risk factors tend to cluster as more of the measured risk factors are present. These unmeasured mortality risk factors include a prothrombotic and proinflammatory state [18].
Implications in OHT
There may be several potential clinical implications of this data. Foremost, the importance of lifestyle changes and medical therapy for these independent risk factors is underscored by this data. In addition, endothelial dysfunction is known to arise when multiple metabolic risk factors are present in combination. This may warrant closer surveillance protocols in recipients with several of these metabolic disorders as coronary endothelial dysfunction has been shown to predict allograft vasculopathy and cardiovascular mortality after OHT [19].
Limitations
A limitation of this study is that the results are not generalizable to every OHT recipient. Moreover, the UNOS database is limited to OHT recipients within the United States and the impact of multiple metabolic risk factors in the international population where these disorders likely follow different patterns is unknown. In addition, the impact of these risk factors should they develop post-OHT is unknown as this analysis was limited to pretransplant risk factors. Other limitations include those associated with the UNOS database, including errors in data entry and reporting biases.
Conclusions
This study analyzed over 15,000 patients to determine the impact of an increasing number of metabolic risk factors on mortality after OHT. Although the presence of each risk factor studied was associated with a significant increase in mortality, the largest adverse effect was seen when all 3 risk factors were present in combination. This exponential effect was greater in magnitude than would be expected in summing the individual effects of the components. Therefore, although different institutions define different levels of patient risk that they are willing to consider for OHT, these data should provide added caution as to transplanting those with all 3 risk factors, namely hypertension, diabetes mellitus, and obesity, particularly if other significant risk factors are present. Nonetheless, these data merit further prospective investigations to better delineate the impact of metabolic risk factors, especially when present in combination, on the OHT population, and to perhaps gain insight into the pathophysiological mechanisms as to how this clustering of metabolic disorders impacts cardiac allograft function and patient survival in this setting.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
S. Westaby and M. Deng Continuous flow blood pumps: the new gold standard for advanced heart failure? Eur J Cardiothorac Surg, July 1, 2013; 44(1): 4 - 8. [Full Text] [PDF] |
||||
![]() |
S. Westaby Cardiac transplant or rotary blood pump: Contemporary evidence J. Thorac. Cardiovasc. Surg., January 1, 2013; 145(1): 24 - 31. [Full Text] [PDF] |
||||
![]() |
S. Westaby Editorial Comment: Multimodal intervention in the management of left ventricular assist device outpatients Eur J Cardiothorac Surg, December 1, 2012; 42(6): 1033 - 1034. [Full Text] [PDF] |
||||
![]() |
N. C. Cavarocchi and P. Mather Invited Commentary Ann. Thorac. Surg., March 1, 2012; 93(3): 724 - 725. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |