Harrison, Ewen

Oniscu, Gabriel

Wigmore, Stephen

Roebuck, Amanda

2024-11-22T15:10:19Z

2024-11-22T15:10:19Z

2023-11-25

INTRODUCTION: Liver Transplantation across the globe has a supply and demand issue. Ever- advancing improvements in road safety and medical care mean a reduced pool of “ideal donors” and the population age is increasing rapidly over time. There is a growing need for donor livers as the number of critically unwell recipients awaiting liver transplantation increases. The extended criteria donor liver has been increasingly utilised in an attempt to balance the equation. There are many studies looking at combinations of variables associated with complications, graft survival and recipient survival but no universal consensus on how to define the limits for each of these variables when choosing a donor. Matching these marginal liver grafts with an appropriate recipient adds another challenge in the quest to deliver the best outcomes for patients. Increasing donor age and donation after cardiac death (DCD) liver transplantations are variables that are widely studied and reported in the literature to negatively affect graft and recipient survival outcomes and are the focus of this analysis. The aim of this study was to determine the impact of these variables on outcomes and to establish whether other variables as confounders enhanced or limited these effects. METHODS: A retrospective analysis of NHSBT (National Health Service Blood and Transplant service) whole liver transplantation data was performed using R studio version 4.1.1. Donor and recipient demographics were scrutinised. Recipients <18 years old and living donors were actively removed from the dataset. Univariable analysis with logistic regression for graft and recipient survival outcomes and Kaplan Meier survival curves for time to event analysis was used to identify variables of interest to pass to multivariable analysis. Consideration was then given to variables that should be included or excluded on the basis of literature and clinical grounds. Missing data analysis was performed, and data filtered to include transplantation between 2000- 2015. Multivariable analysis was performed using a Cox proportional hazards model. Primary explanatory variables of interest were donor age and DCD transplantation and other variables resulting in significant difference to graft and recipient survival were noted. Variables with significance but not widely reported in the literature were excluded and those with low or no significance were still included if they were felt of importance in the literature. RESULTS: Both increasing donor age >60 years old (HR 1.29, CI 1.09-1.53, p=0.004) and DCD liver transplantation (when compared to donation after brainstem death transplantation, HR 1.69, CI 1.40-2.04, p<0.001) negatively impacted graft survival at both univariable and multivariate analysis. Other variables after adjusting for confounders to negatively impact on graft survival include Hepatitis C positive recipient status (HR 1.47, CI 1.27-1.71, p<0.001), recipient inpatient status pre- transplant (HR 1.17, CI 1.01-1.36, p=0.032), “suboptimal” donor organ appearance (HR 1.32, CI 1.16-1.51, p<0.001) and increasing CIT >8 hours (HR 1.23, CI 1.05- 1.45, p=0.009). For recipient survival increasing donor age caused poorer recipient survival on both univariable (HR 1.23, CI 1.08-1.40, p=0.002) and multivariate analysis (HR 1.25, CI 1.06-1.46, p=0.007). For DCD transplantations there was no statistically significant difference in recipient survival on univariable or multivariable analysis (HR 1.09, CI 0.89-1.34, p=0.400). Other variables correlating with poorer recipient survival after adjustment for confounding factors in the cox proportional hazards model include HCV positive recipients (HR 1.37, CI 1.20-1.57, p<0.001) and recipient inpatient status pre-op (HR 1.32, CI 1.14-1.52, p<0.001). DISCUSSIONS: Donor livers utilised from >60 year olds were more likely to be transplanted to a recipient in better clinical condition when considering variables such as ventilation status, renal support, blood group compatibility and lifestyle scores to indicate fitness. This correlates with many studies which suggest careful recipient characteristic selection to avoid further host stress to the graft. This was also evident in DCD liver transplantations. Older donor livers were also more likely to be CMV positive, to have died from intracranial haemorrhage (rather than trauma causes in younger donors) and were more likely DCD grafts all compounding the risks of poorer outcomes. The arbitrary cut-off chosen for the older group of donors was >60 years and this corresponded with the median in the literature and as a result this study demonstrated similar outcomes. It is possible that choosing an older cohort or cut-off would have resulted in a more significant negative impact on survival. In DCD liver transplantation it was evident that these grafts less likely to be transplanted to Hepatitis C virus positive recipients or to recipients with poorer clinical state pre-transplantation as with older donors, again all in an effort it seems to reduce increased host stress on a graft already exposed to some ischaemic reperfusion injury. Care was taken when designing statistical analysis to choose variables which were either significant on univariable analysis or important in the literature to allow a condensed model to be chosen and to optimise the fit and predictive power for survival outcomes. It was worth nothing the relationships that naturally occurred between variables, such as CMV status as this was more likely positive in older donors. This means these were difficult to separate in cause and effect analysis. Future work to enhance this study could include development of a more complete or robust dataset and data recording strategies to eliminate the large volume of missing data. A warm ischaemic time variable would be an essential addition to fully explore the effects of DCD transplantation, but as there are now machine and normothermic regional perfusion techniques that are well established and have developed a good evidence base this has become less relevant. CONCLUSIONS: In this study, increasing donor age was found to adversely affect both graft and recipient survival outcomes. DCD transplantation resulted in poorer outcomes for graft survival but had no significant impact on recipient survival. This correlated with much of the wider literature in this field, although the range of analytic methods and strategies across multiple studies makes it difficult to standardise a universal consensus on how to limit these negative outcomes. These extended donor characteristics are the focus of many donor-recipient matching protocols to limit these effects on survival. Machine perfusion and normothermic perfusion procedures help to limit the impact of these donors and are now the focus of the next stage of development of techniques to bridge the waiting list gaps for liver transplantation in the United Kingdom.

https://hdl.handle.net/1842/42688

http://dx.doi.org/10.7488/era/5382

en

The University of Edinburgh

Liver Transplantation

ideal donors

donor livers

extended criteria donor liver

marginal liver grafts

donation after cardiac death (DCD) liver transplantations

National Health Service Blood and Transplant service

whole liver transplantation

Kaplan Meier survival curves

CMV status

DCD transplantation

Marginal donor characteristics and outcomes after liver transplantation

Thesis or Dissertation

Doctoral

MSc(R) Master of Science by Research