CLINICAL
AND
TRANSLATIONAL RESEARCH
Graft Survival After Cardiac Transplantation for Alcohol Cardiomyopathy D. Marshall Brinkley, MD,1,4 Eric Novak, MS,2 Veli K. Topkara, MD,3 and Edward M. Geltman, MD 2 Background. Alcohol cardiomyopathy (ACM) constitutes up to 40% of patients with non-ischemic dilated cardiomyopathy. Transplant-free survival is worse for patients with ACM versus idiopathic dilated cardiomyopathy (IDCM) with continued exposure. The prognosis for patients with ACM after cardiac transplantation is unknown. Methods. We evaluated adults who underwent single-organ, cardiac transplantation from 1994 to 2009 with a diagnosis of ACM (n=134) or IDCM (n=10,243) in the Organ Procurement Transplantation Network registry. KaplanMeier curves were generated by cohort for time until graft failure, cardiac allograft vasculopathy, and hospitalization for rejection. A Cox proportional hazards model was created to determine factors associated with each outcome. Results. Patients with ACM were more likely to be males (P G0.0001), minorities (P G0.0001), and smokers (P = 0.0310) compared with IDCM. Overall graft survival was lower for the ACM cohort (P = 0.0001). After multivariate analysis, ACM was not independently associated with graft survival (HR 1.341, 95% CI 0.944Y1.906, P = 0.1017). Creatinine, total bilirubin, minority ethnicity, graft under-sizing, life support, diabetes, and donor age were independent predictors of graft failure. There were no significant differences between primary cause of death, vasculopathy, or rejection. Conclusions. There was no association between ACM and graft survival in this large registry study, but poorer overall survival in the ACM cohort was associated with other recipient characteristics. Keywords: Alcoholic, Dilated cardiomyopathy, Heart transplant. (Transplantation 2014;98: 465Y469)
lcoholism is a major cause of dilated cardiomyopathy phenotype, constituting up to 40% of disease at some referral centers (1). An estimated 30% of alcoholics have echocardiographic evidence of systolic dysfunction (2), but increased left ventricular volume or impaired filling may
A
The authors declare no funding or conflicts of interest. 1 Division of Cardiovascular Disease, Department of Medicine, Beth Israel Deaconess Medical Center, Boston, MA. 2 Division of Cardiology, Department of Medicine, Washington University School of Medicine, St. Louis, MO. 3 Division of Cardiology, Department of Medicine, Columbia University Medical Center, New York, NY. 4 Address correspondence to: D. Marshall Brinkley, M.D., Division of Cardiovascular Disease, Department of Medicine, Beth Israel Deaconess Medical Center, 185 Pilgrim Road, West Campus, Baker 4, Boston, MA 02215. E-mail:
[email protected]. D.M.B participated in designing the research, analyzing the data, interpreting the results, and writing the article. E.N. participated in analyzing the data. V.K.T. participated in analyzing the data. E.M.G. participated in designing the research, interpreting the results, and writing the article. Supplemental digital content (SDC) is available for this article. Direct URL citations appear in the printed text, and links to the digital files are provided in the HTML text of this article on the journal’s Web site (www.transplantjournal.com). Received 7 January 2014. Accepted 21 January 2014. Copyright * 2014 by Lippincott Williams & Wilkins ISSN: 0041-1337/14/9804-465 DOI: 10.1097/TP.0000000000000083
Transplantation
& Volume 98, Number 4, August 27, 2014
precede any detectable change in ejection fraction (3, 4). The relationship between alcohol intake and heart failure is complex, and likely depends on interaction between exposure and susceptibility. Although light to moderate intake (e30 g/day) is associated with a decreased incidence of heart failure(5, 6), intake of greater than 80 g/day for more than 5 years is linked to alcoholic cardiomyopathy (ACM) (7Y9). There are conflicting data regarding transplant-free survival. Authors have reported both a better and worse prognosis when compared to idiopathic dilated cardiomyopathy (IDCM), but continued exposure may be a key confounder. Stratifying ACM cases according to relapse indicates that continued drinking portends a worse prognosis (7, 8). Indeed, even decreasing alcohol intake without complete abstinence can improve ejection fraction (10). Patients with progressive disease despite rehabilitation may require advanced therapies. According to International Society of Heart and Lung Transplantation (ISHLT) guidelines, patients with a history of alcohol abuse can be listed for transplant after abstinence, but no minimum period is prescribed (11). Up to 50% of liver transplant recipients for alcoholic cirrhosis relapse (12). Recidivism and comorbidities like cirrhosis or malignancy could potentially lead to poorer outcomes for patients with ACM who undergo cardiac transplant. On the other hand, unfounded misgivings could inappropriately bias physicians not to refer these patients for transplantation. To www.transplantjournal.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
465
466
www.transplantjournal.com
TABLE 1.
Transplantation
& Volume 98, Number 4, August 27, 2014
Baseline characteristics
Characteristic
Idiopathic (n=10,243)
Recipient Age Creatinine, mg/dL Total bilirubin, mg/dL BMI Male Ethnicity White Black Hispanic Other Graft under-sizing TPG 912, mm Hg Smoking history Inotrope Life support Diabetes Donor Age Ischemic time, hr Male
Alcoholic (n=134)
Mean (SD), %
Mean (SD), %
P
49.5 (12.1) 1.29 (0.64) 1.17 (1.12) 26.4 (5.00) 71.6
49.6 (10.3) 1.38 (0.59) 1.30 (1.19) 25.6 (4.89) 94.0
0.9224 0.1229 0.2076 0.0659 G0.0001a G0.0001a
66.1 23.4 7.3 3.2 4.4 28.1 13.0 50.4 23.4 17.0
44.0 35.8 17.2 3.0 3.3 29.6 59.1 59.0 23.9 18.0
0.8220 0.7353 0.0310a 0.0554 0.9182 0.8045
30.9 (12.4) 3.01 (1.01) 70.2
32.2 (13.3) 3.08 (0.98) 70.9
0.1979 0.4012 0.9244
a PG0.05. BMI, body mass index; SD, standard deviation; TPG, transpulmonary gradient.
our knowledge, the prognosis for ACM after cardiac transplantation has not been described. We utilized United Network of Organ Sharing (UNOS) data to examine graft survival for ACM versus IDCM to determine the prognosis of ACM after cardiac transplantation.
RESULTS We identified 134 patients with ACM and 10,243 with IDCM. Table 1 lists patient and donor characteristics at transplant registration. Patients with ACM were more likely be male (94% vs. 72%, PG0.0001), a minority (36% vs. 23% black and 17% vs. 7% Hispanic, PG0.0001), and former smoker (59% vs. 13%, P=0.0310) than those with IDCM. Log-rank analysis of Kaplan-Meier survival curves for each cohort (Fig. 1) revealed that overall graft survival was significantly worse for patients with ACM (P=0.0001). Estimated 1-, 3-, 5-, and 10-year survival for the ACM cohort was 83.2% (95% CI 75.6Y88.6), 73.3% (64.4Y80.3), 61.1% (51.1Y69.7), and 40.1% (29.4Y50.5) versus 88.2% (87.5Y88.8), 80.7% (79.9Y81.5), 73.6% (72.6Y74.5), and 55.2% (53.9Y56.4) for the IDCM cohort. To account for unbalanced cohort characteristics at baseline, a Cox proportional hazards model was created (Table 2). Diagnosis of ACM was not independently associated with graft survival (HR 1.341, 95% CI 0.944Y1.906, P=0.1017). Creatinine (PG0.0001), total bilirubin (PG0.0001), minority ethnicity (PG0.0001), graft under-sizing (P=0.0064), life support (PG0.0001), diabetes (PG0.0001), and donor age (PG0.0001) were all strong, independent predictors of graft survival.
To further explore the lower overall graft survival for ACM, we evaluated the primary cause of death by system (SDC, Table S1, http://links.lww.com/TP/A957). There were no significant differences in causes of death between cohorts, including cardiovascular, pulmonary, neurologic, infectious, hepatic failure, renal failure, rejection, cancer, other, and unknown. There was a trend toward more death from rejection in the ACM cohort (P=0.0786). We then examined the quantitative outcomes of time until CAV
FIGURE 1.
Graft survival with number at risk.
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Brinkley et al.
* 2014 Lippincott Williams & Wilkins
TABLE 2.
467
Multivariate predictors of graft failure
Variable Recipient Age Creatinine, mg/dL Total bilirubin, mg/dL BMI Female (vs. male) Ethnicity (vs. white) Black Hispanic Other Alcoholic (vs. idiopathic) Graft under-sizing TPG 912, mm Hg Inotrope support Life support Diabetes Donor Age Ischemic time Female (vs. male)
HR
95% CI
P-value
0.997 1.116 1.074 0.998 1.082
0.993Y1.001 1.064Y1.171 1.037Y1.113 0.990Y1.007 0.985Y1.187
0.0915 G0.0001a G0.0001a 0.6792 0.1001 G0.0001a
1.475 0.968 1.075 1.341 1.276 1.065 1.049 1.244 1.260
1.341Y1.622 0.814Y1.150 0.834Y1.386 0.944Y1.906 1.071Y1.520 0.973Y1.166 0.965Y1.140 1.124Y1.377 1.124Y1.407
0.1017 0.0064a 0.1695 0.2608 G0.0001a G0.0001a
1.013 1.012 0.964
1.009Y1.016 0.971Y1.054 0.876Y1.059
G0.0001a 0.5856 0.4434
a
PG0.05. BMI, body mass index; CI, confidence interval; HR, hazard ratio; TPG, transpulmonary gradient.
(Fig. 2) and time until hospitalization for rejection (Fig. 3). Kaplan-Meier survival curves were not significantly different between cohorts for either CAV (P=0.5234) or hospitalization for rejection (P=0.1423). Accordingly, ACM was not an independent predictor of time until CAV (HR 1.205, 95% CI 0.765Y1.899, P=0.4208) or time until hospitalization for rejection (HR 1.495, 95% CI 0.935Y2.390, P=0.0927). Additional analyses were performed to evaluate potential limitations. To evaluate for confounding from a potential relationship between ACM and ethnicity, we performed a second
set of Cox analyses including an interaction term for both. There was no significant interaction between diagnosis and ethnicity on graft survival (P=0.4469), time to CAV (P=0.6527), or time to hospitalization for rejection (P=0.1078). The hazard ratio of graft survival for ACM versus IDCM among whites was 1.153 (95% CI 0.637Y2.089) and minorities 1.534 (95% CI 0.993Y2.369). Because smoking was excluded in the primary analysis as a result of missing data, we performed another Cox analysis of graft survival in patients with smoking data. Neither smoking history (HR 0.906, 95% CI 0.731Y1.124, P=0.3691)
FIGURE 2. at risk.
FIGURE 3. at risk.
Cardiac allograft vasculopathy with number
First hospitalization for rejection with number
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
468
www.transplantjournal.com
nor ACM (HR 1.180, 95% CI 0.519 Y2.683, P =0.6933) were associated with graft survival in this subset.
DISCUSSION We examined graft survival using the OPTN registry to determine the prognosis of alcoholic cardiomyopathy after cardiac transplantation. Although the ACM cohort had decreased graft survival compared to IDCM, there was no independent association with diagnosis. Rather, other recipient factors accounted for the observed overall survival difference. Creatinine, total bilirubin, minority ethnicity, graft under-sizing, life support, diabetes, and donor age all independently predicted graft survival, consistent with previous studies (13Y15). Based on OPTN data as of September 7, 2012, the 1-, 3-, and 5-year graft survival for ACM is comparable to that for congenital heart disease (81%, 72%, and 66%), a well-accepted indication for transplantation. Whereas recipient survival would be an alternative outcome, graft survival informs organ allocation decisions and may be less subject to retransplantation bias against patients with a history of alcoholism. Our findings were internally consistent. Recipient causes of death were similar between cohorts. We found no significant difference in time until CAV or association with ACM. This is important, for the ACM cohort were more likely to be males, smokers, and minorities, all previously shown risk factors for CAV (16, 17). Medication noncompliance is a common concern for patients with a history of substance abuse. We found no significant difference in hospitalization for rejection or association with ACM, despite a trend toward more deaths from rejection. We can only speculate that compliance with immunosuppression was thus not a major issue. Missing data for smoking did not appear to be a limitation, for neither smoking nor ACM were significant predictors when analyses were repeated in the subset with data. In addition, there was similar death from malignancy and cardiovascular disease and a similar rate of CAV in both cohorts. However, we cannot exclude a potential effect. Minority ethnicity also did not appear to be a confounder. There was no statistical interaction between ethnicity and diagnosis for any outcome examined. This is an observational study and subject to the limitations thereof. Although we utilized the largest U.S. dataset available, the ACM cohort size was relatively small. Risks that were more common may have masked an effect of diagnosis on survival, but the hazards associated with established predictors were not large. Secondly, ACM remains a clinical diagnosis, so diagnostic uncertainty may have led to crossover, biasing toward the null hypothesis. There was no quantitative alcohol consumption history available for correlation. The survival curves for hospitalization for rejection diverge initially and then converge again when there are fewer events. Whether hospitalization for rejection would become statistically different with a larger sample is uncertain. However, there was no statistically significant difference in death from rejection. We cannot exclude the effect of unmeasured variables that may be associated with ACM and impact outcomes, such as psychosocial factors. Finally, transplant candidates are selected carefully, so these results cannot be extrapolated to all patients with ACM. There was a low prevalence of ACM in this
Transplantation
& Volume 98, Number 4, August 27, 2014
population compared the literature, consistent with this notion. Prevalence may also vary significantly by region. European centers have reported an ACM prevalence of 23% to 40% of all dilated cardiomyopathies (1, 7, 8, 18), yet U.S. centers have reported only 3% to 21% (19 Y21). Data from alcoholic liver disease (ALD) parallel our findings and provide insight for candidate selection in ACM. Graft survival after liver transplantation for ALD is similar to non-ALD, despite a high relapse rate (22, 23). Although the duration of abstinence before transplant predicts recidivism later, no specific time exhibits good receiver-operator characteristics (24, 25). Other factors like poor social support, family history of alcoholism, other substance abuse, and depression also predict recidivism (24, 26, 27). Current ISHLT guidelines do not require a specific length of abstinence before listing. Our finding that ACM does not predict graft survival suggests that these guidelines are sufficiently stringent. In conclusion, there was association between ACM and graft survival after cardiac transplantation in this large, U.S. registry study. The lower overall survival observed in this cohort appears attributable to other recipient characteristics. This implies that patients with ACM are reasonable candidates for transplant using current screening guidelines. For patients deemed suboptimal candidates, ventricular assist devices may offer an alternative, as destination therapy, bridge-to-recovery, or bridge-to-decision. More work is needed to elucidate the pathophysiology of alcohol cardiomyopathy and predict recovery.
MATERIALS AND METHODS Design The data reported here were supplied by UNOS as the contractor for the Organ Procurement and Transplantation Network (OPTN) and were used in compliance with the Washington University School of Medicine IRB regulations. The interpretation and reporting of these data are the responsibility of the authors, not OPTN or the U.S. Government. We selected adults who underwent only cardiac transplantation from 1994 to 2009 with a diagnosis code of ACM (n=134) or IDCM (n=10,243). Baseline characteristics (Table 1) were determined at the time of registration. Graft undersizing was defined as (recipient weightjdonor weight)/(recipient weight) greater than 30%, as previously described (14). Life support was defined as ventilator, extracorporeal membrane oxygenation, ventricular assist device, or intra-aortic balloon pump support. For continuous variables, values outside the prespecified ranges of creatinine 0.1 to 15, bilirubin 0.1 to 15, BMI 10 to 50, and transpulmonary gradient 0 to 30 were excluded as clinically improbable outliers. Graft survival was measured from date of transplant to date of recipient death or retransplantation. Cardiovascular death included infarction, coronary artery disease, arrhythmia, cardiogenic shock, aortic aneurysm, and sudden death. Stroke was classified as neurologic. Hospitalization for rejection and cardiac allograft vasculopathy (CAV) were reported by each center as yes, no, or unknown during follow-up. Dates for CAV and first hospitalization for rejection were imputed from the first follow-up date with an event.
Statistical Analysis All statistical analyses were performed using SAS version 9.3. Mean values and frequency counts were tabulated by diagnosis. Continuous and categorical variables were compared with t tests and Fisher exact tests, respectively. Kaplan-Meier survival curves were generated for each cohort, and time to graft failure, CAV, and hospitalization for rejection were evaluated by the logrank test. Multivariate analyses were then performed to determine if cohort differences remained after adjusting for covariates. Cox proportional hazards models were created using the listed variables (Table 2). Smoking history was
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Brinkley et al.
* 2014 Lippincott Williams & Wilkins
excluded in the main analysis because of missing data for 70% of cases. Hazard ratios were reported (expressed per unit for continuous variables) with 95% confidence intervals. Assumptions of proportionality were checked with Schoenfeld residual plots, and multicollinearity was assessed through tolerance estimates. Cox models were repeated with an interaction term between race and diagnosis. Differences in primary cause of death were evaluated with Fisher exact test. Significance for all tests was evaluated at type I error rate of 5% (>=0.05).
ACKNOWLEDGMENTS We thank the Division of Cardiology at Washington University for statistical support and Jeffrey Crippin, M.D., for editorial comments.
12. 13.
14.
15.
16. 17.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
McKenna CJ, Codd MB, McCann HA, et al. Alcohol consumption and idiopathic dilated cardiomyopathy: a case control study. Am Heart J 1998; 135: 833. Urbano-Marquez A, Estruch R, Navarro-Lopez F, et al. The effects of alcoholism on skeletal and cardiac muscle. N Engl J Med 1989; 320: 409. Kupari M, Koskinen P, Suokas A, et al. Left ventricular filling impairment in asymptomatic chronic alcoholics. Am J Cardiol 1990; 66: 1473. Urbano-Ma´rquez A, Estruch R, Ferna´ndez-Sola´ J, et al. The greater risk of alcoholic cardiomyopathy and myopathy in women compared with men. JAMA 1995; 274: 149. Bryson CL, Mukamal KJ, Mittleman MA, et al. The association of alcohol consumption and incident heart failure: the Cardiovascular Health Study. J Am Coll Cardiol 2006; 48: 305. Djousse´ L, Gaziano JM. Alcohol consumption and risk of heart failure in the Physicians’ Health Study I. Circulation 2007; 115: 34. Fauchier L, Babuty D, Poret P, et al. Comparison of long-term outcome of alcoholic and idiopathic dilated cardiomyopathy. Eur Heart J 2000; 21: 306. Gavazzi A, De Maria R, Parolini M, et al. Alcohol abuse and dilated cardiomyopathy in men. Am J Cardiol 2000; 85: 1114. Lazarevi( AM, Nakatani S, Nesˇkovi( AN, et al. Early changes in left ventricular function in chronic asymptomatic alcoholics: relation to the duration of heavy drinking. J Am Coll Cardiol 2000; 35: 1599. Nicola´s JM, Ferna´ndez-Sola` J, Estruch R, et al. The effect of controlled drinking in alcoholic cardiomyopathy. Ann Intern Med 2002; 136: 192. Mehra MR, Kobashigawa J, Starling R, et al. Listing criteria for heart transplantation: International Society for Heart and Lung Transplantation guidelines for the care of cardiac transplant candidatesV2006. J Heart Lung Transplant 2006; 25: 1024.
18.
19. 20.
21. 22.
23.
24.
25.
26.
27.
469
Tome S, Lucey MR. Timing of liver transplantation in alcoholic cirrhosis. J Hepatol 2003; 39: 302. Russo MJ, Iribarne A, Easterwood R, et al. Post-heart transplant survival is inferior at low-volume centers across all risk strata. Circulation 2010; 122: S85. Stehlik J, Feldman DS, Brown RN, et al. Interactions among donor characteristics influence post-transplant survival: a multi-institutional analysis. J Heart Lung Transplant 2010; 29: 291. Singh TP, Givertz MM, Semigran M, et al. Socioeconomic position, ethnicity, and outcomes in heart transplant recipients. Am J Cardiol 2010; 105: 1024. Botha P, Peaston R, White K, et al. Smoking after cardiac transplantation. Am J Transplant 2008; 8: 866. Costanzo MR, Naftel DC, Pritzker MR, et al. Heart transplant coronary artery disease detected by coronary angiography: a multiinstitutional study of preoperative donor and recipient risk factors. J Heart Lung Transplant 1998; 17: 744. Prazak P, Pfisterer M, Osswald S, et al. Differences of disease progression in congestive heart failure due to alcoholic as compared to idiopathic dilated cardiomyopathy. Eur Heart J 1996; 17: 251. Fuster V, Gersh BJ, Giuliani ER, et al. The natural history of idiopathic dilated cardiomyopathy. Am J Cardiol 1981; 47: 525. Kasper EK, Agema WRP, Hutchins GM, et al. The causes of dilated cardiomyopathy: a clinicopathologic review of 673 consecutive patients. J Am Coll Cardiol 1994; 23: 586. Piano MR. Alcoholic cardiomyopathy: incidence, clinical characteristics, and pathophysiology. Chest 2002; 121: 1638. Belle SH, Beringer KC, Detre KM. Liver transplantation for alcoholic liver disease in the United States: 1988 to 1995. Liver Transpl Surg 1997; 3: 212. Burra P, Senzolo M, Adam R, et al. Liver transplantation for alcoholic liver disease in Europe: a study from the ELTR (European Liver Transplant Registry). Am J Transplant 2010; 10: 138. DiMartini A, Day N, Dew MA, et al. Alcohol consumption patterns and predictors of use following liver transplantation for alcoholic liver disease. Liver Transpl 2006; 12: 813. Yates WR, Martin M, LaBrecque D, et al. A model to examine the validity of the 6-month abstinence criterion for liver transplantation. Alcohol Clin Exp Res 1998; 22: 513. Dew MA, DiMartini AF, Steel J, et al. Meta-analysis of risk for relapse to substance use after transplantation of the liver or other solid organs. Liver Transpl 2008; 14: 159. Pfitzmann R, Schwenzer J, Rayes N, et al. Long-term survival and predictors of relapse after orthotopic liver transplantation for alcoholic liver disease. Liver Transpl 2007; 13: 197.
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.