Back to Annual Meeting Abstracts
Impact of ABO Compatibility on Heart Transplantation Outcomes in a Nationwide Cohort Study Over the Last Decade.
Jawitz, Oliver K.1, Gong, Nicole2, Bellumkonda, Lavanya3, Jacoby, Daniel3, Dries, Daniel3, Lee, Forrester3, Yuh, David D.1, Bonde, Pramod4
1Surgery, Yale School of Medicine, New Haven, CT, USA, 2MD/PhD Program, Penn State College of Medicine, Hershey, PA, USA, 3Yale Center for Advanced Heart Failure, Mechanical Circulatory Support and Heart Transplantation, Yale School of Medicine, New Haven, CT, USA, 4Surgery, Yale Center for Advanced Heart Failure, Mechanical Circulatory Support and Heart Transplant, Yale School of Medicine, New Haven, CT, USA
Objective: Immunological incompatibility has implications for primary graft failure, rate of rejection and survival in heart transplantation. The impact of ABO compatible but non-identical versus ABO identical match on survival following heart transplantation has not been systematically investigated in a large cohort study.
Methods: We used a nationwide, decade long sample (2000-2010) from the United States contained within the United Network for Organ Sharing (UNOS) database. Stratification was between ABO identical and ABO compatible heart transplantations for univariate and multivariate analyses. The primary end-point was graft failure from all causes. Cumulative and 30-day post-transplant survival was compared between groups using a Cox proportional hazard model and a logistic regression model, respectively.
Results: Of the 17,951 patients that met inclusion criteria, 2,684 (~15%) underwent ABO compatible heart transplantation. ABO compatible recipients were generally sicker than ABO identical recipients before transplant as a larger proportion were UNOS Status 1A (p < 0.05): 50.3% (1,350/2,684) vs. 37.8% (5,771/15,267). A larger proportion of ABO compatible recipients were also in the ICU before transplant and on mechanical ventilatory support compared to ABO identical recipients (p < 0.05), further suggesting that ABO compatible grafts were preferentially reserved for sicker recipients. In univariate analysis, ABO compatible transplants were associated with decreased 30-day, 1-year, 3-year, and 5-year survival post-transplant (p < 0.05). Compatible transplants were also associated with higher incidences of primary graft failure as cause of death (p < 0.05). There was no statistically significant difference, however, in rates of acute graft rejection between the two cohorts (p = 0.53). In addition, multivariate analysis (Table 1) did not demonstrate adverse outcomes in terms of decreased graft survival (hazard ratio 0.99, p = 0.87). Variables coding for pre-operative hemodynamic instability remained the predictors of mortality.
Conclusions: In the past decade, ABO compatible donor hearts were preferentially given to sicker transplant recipients. Transplantation using ABO compatible adult hearts does not result in adverse outcomes with respect to graft survival. Therefore, ABO compatible and ABO identical heart transplant matches should be viewed equally in clinical decision-making and to maximize efficiency within the available donor pool.
Multivariate Cox Proportional Hazards Regression Model
|Variable*||Hazard Ratio (95% Confidence Limits)||p-Value**|
|ABO compatible (vs. identical)||0.99 (0.89-1.10)||0.87|
|Recipient ethnicity - Black (vs. White ethnicity)||1.42 (1.30-1.56)||<0.05|
|Ventilatory support at transplant (vs. no life support)||1.88 (1.50-2.37)||<0.05|
|ECMO at transplant (vs. no life support)||2.60 (1.72-3.83)||<0.05|
|Ischemic time||1.09 (1.06-1.13)||<0.05|
|Recipient in ICU before transplant (vs. not in hospital)||1.24 (1.10-1.39)||<0.05|
|Total bilirubin||1.03 (1.02-1.04)||<0.05|
|ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit. *Non-significant (p>0.05) variables not included in table: recipient gender, donor ethnicity, IABP and IV inotropes at transplant, waitlist status at transplant. **p-Value based on multivariate Cox proportional hazards regression model, using factors significant on univariate analysis (p < 0.05 considered statistically significant).|
Back to Annual Meeting Abstracts