Secondary Logo

Primary Graft Dysfunction

The Devil Is in the Details

Foroutan, Farid, HBSc, PhD candidate1; Ross, Heather J., MD1

doi: 10.1097/TP.0000000000002221

Primary graft dysfunction (PGD) remains a critical challenge in heart transplantation, impacting graft and recipient survival and there is a critical need for future studies to evaluate the predictive power of the PGD consensus instrument using robust regression models.

1 Ted Rogers Centre for Heart Research, Multi-Organ Transplant Program, Toronto General Hospital, University Health Network, Toronto, Ontario, Canada.

Received 13 February 2018.

Accepted 22 February 2018.

The authors declare no conflicts of interest.

F.F. and H.J.R. participated in writing of the paper.

Correspondence: Farid Foroutan, HBSc, PhD candidate, Ted Rogers Centre for Heart Research, Multi-Organ Transplant Program, Toronto General Hospital, University Health Network, 11 PMB 137, 585 University Ave, ON M5G 2C4. (

Heart transplant clinicians and researchers have long been interested in understanding the infrequent but potentially devastating and deadly challenge of primary graft dysfunction (PGD). Before 2014, there was a lack of consensus on the definition of PGD resulting in variable reported incidence rates (2.8%-23%), management strategies, and probabilities for early mortality. The mortality risk associated with PGD was as high as 61.5% when defined as the need for mechanical circulatory support (MCS) within the first 30 days.1 The lack of common nomenclature created difficulty in comparing incidence and mortality over time and across centers and limited the ability to develop potential interventions that could impact PGD. The International Society for Heart and Lung Transplantation (ISHLT) developed a consensus-based definition and classification tool for PGD in 2014.2 Since that time, the ISHLT PGD definition and classification system has been cited in more than 70 publications either as an outcome or predictor of mortality.3-7 Unfortunately, some of these studies either modified the definition and classification criteria (deviating away from diagnosing PGD within the first 24 hours), focused only on severe cases (eg, requirement for extracorporeal membrane oxygenation posttransplant), or created spurious associations between pretransplant factors and PGD because of poor statistical analyses.

Singh et al8 are to be congratulated on the study “ISHLT primary graft dysfunction incidence, risk factors and outcome: a UK national study.” It is the largest study to date to utilize robust methodology and statistical analyses to assess all PGD severity levels while remaining true to the ISHLT criteria. The authors, thereby, identify risk factors that guide clinicians in identifying transplant candidates at increased risk for PGD. Singh et al8 followed a cohort of 450 adult heart transplant recipients observing 4 cases of mild, 72 moderate, and 81 severe cases of PGD. The large sample of patients allowed for more precise estimates of 30-day mortality in patients with severe PGD (Figure 1), providing the statistical power necessary to tease out associations between specific pretransplant factors and each severity level of PGD. Using multivariable logistic regression analysis, the authors identified donor age, implant time, female donor–to–male recipient sex mismatch, and preoperative MCS as risk factors for PGD. Only implant time, female donor–to–male recipient sex mismatch, and previous sternotomy were independent predictors for severe PGD. For example, the risk of any PGD was 21.3% (3.8% for severe PGD) for a 20-year-old sex-matched donor in a nondiabetic recipient without preoperative MCS undergoing transplantation with a minimal implant time (defined as the time between removal from cold storage or organ care system until the heart is reperfused in the recipient) of 54 minutes. This risk increased to 45.7% (11% for severe PGD) for a male recipient bridged with MCS (ventricular assist device or extracorporeal membrane oxygenation) transplanted with a female donor heart. Importantly, most of the identified predictors of PGD are modifiable. This allows transplant clinicians to consider these risks and make decisions around donor/recipient matching at transplant. For example, sex matching for recipients on MCS support pretransplant would result in a 24% risk reduction for all severity levels of PGD (7% risk reduction for severe PGD). Of course, this is a balance, specifically in critical patients where the risk of dying may not allow the luxury of time necessary to find a sex-matched donor. In situations where the predicted risk of PDG is high, clinicians should be on alert and ready to intervene with supportive care as needed.



Despite this study's robust methodology and powerful statistical analyses, there are some areas that require further investigation. For example, the authors stated that recipient diabetes is an independent predictor of PGD. However, careful examination of the results shows that the 95% confidence interval (0.9993-4.1720) crosses the boundary of no effect for diabetes. No other study has identified recipient diabetes as a predictor of PGD when using robust statistical analysis (overfitting of regression models1,9). Based on the findings from Singh et al8 a baseline risk of 21.3% for all severity levels of PGD may decrease by 0.3% or increase by 32% for diabetic recipients (stronger than the independent impact of sex-mismatch or MCS support pretransplant). However, recipient diabetes seems to be a predictor of graft loss within and beyond the first-year posttransplant.10,11

Singh et al8 identified their final 4 predictors of PGD from an initial list of 13 plausible predictors. Notably, their initial list of predictors did not include cold ischemic time or the percentage difference in predicted left ventricle mass between recipients and donors (a known marker of mismatching and mortality risk12). Among implant time, explant time, and cold ischemic time, Singh et al8 only included the former 2 as potential predictors of PGD. It could be argued that the most important time variable, however, is cold ischemic time.2 In comparison with implant and explant time, cold ischemic time contributes to the bulk of total ischemic time (previously shown to be a strong predictor of mortality13). Its omission from the final model may have impacted the results by reducing generalizability for the independent impact of the other PGD predictors.

PGD remains a critical challenge in heart transplantation, impacting graft and recipient survival. Singh et al8 have advanced our understanding of the factors leading to PGD. Such investigations are needed as we continue to extend donor criteria, including donation after circulatory death and ex vivo donors, and to assess the impact of opiate overdose, as the cause of donor death, on graft function. By utilizing the ISHLT PGD consensus definition, investigations will be easier to compare, contrast, and interpret. There is a critical need for future studies to evaluate the predictive power of the PGD consensus instrument using robust regression models with patient important outcomes. Only then can we evaluate the need to give or withhold (un)necessary therapies for patients classified under the current PGD instrument.

Back to Top | Article Outline


1. Segovia J, Cosio MD, Barcelo JM, et al. RADIAL: a novel primary graft failure risk score in heart transplantation. J Heart Lung Transplant. 2011;30:644–651.
2. Kobashigawa J, Zuckermann A, Macdonald P, et al. Report from a consensus conference on primary graft dysfunction after cardiac transplantation. J Heart Lung Transplant. 2014;33:327–340.
3. Braulio R, Sanches MD, Teixeira Junior AL, et al. Associated clinical and laboratory markers of donor on allograft function after heart transplant. Braz J Cardiovasc Surg. 2016;31:89–97.
4. Lima EB, Cunha CR, Barzilai VS, et al. Experience of ECMO in primary graft dysfunction after orthotopic heart transplantation. Arq Bras Cardiol. 2015;105:285–291.
5. Sabatino M, Vitale G, Manfredini V, et al. Clinical relevance of the International Society for Heart and Lung Transplantation consensus classification of primary graft dysfunction after heart transplantation: epidemiology, risk factors, and outcomes. J Heart Lung Transplant. 2017;36:1217–1225.
6. Squiers JJ, Saracino G, Chamogeorgakis T, et al. Application of the International Society for Heart and Lung Transplantation (ISHLT) criteria for primary graft dysfunction after cardiac transplantation: outcomes from a high-volume centre†. Eur J Cardiothorac Surg. 2017;51:263–270.
7. Dronavalli VB, Rogers CA, Banner NR. Primary cardiac allograft dysfunction-validation of a clinical definition. Transplantation. 2015;99:1919–1925.
8. Singh SSA, Banner NR, Rushton S, et al. ISHLT primary graft dysfunction incidence, risk factors and outcome: a UK national study. Transplantation. 2019;103:342–349.
9. Peduzzi P, Concato J, Kemper E, et al. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49:1373–1379.
10. Foroutan F, Alba AC, Guyatt G, et al. Predictors of 1-year mortality in heart transplant recipients: a systematic review and meta-analysis. Heart. 2018;104:151–160.
11. Russo MJ, Chen JM, Hong KN, et al. Survival after heart transplantation is not diminished among recipients with uncomplicated diabetes mellitus: an analysis of the United Network of Organ Sharing database. Circulation. 2006;114:2280–2287.
12. Reed RM, Netzer G, Hunsicker L, et al. Cardiac size and sex-matching in heart transplantation : size matters in matters of sex and the heart. JACC Heart Fail. 2014;2:73–83.
13. Lund LH, Khush KK, Cherikh WS, et al. The registry of the International Society for Heart and Lung Transplantation: thirty-fourth adult heart transplantation report—2017; focus theme: allograft ischemic time. J Heart Lung Transplant. 2017;36:1037–1046.
Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.