Incorporation of Donor Risk Into Liver Allocation Algorithms


Corresponding author: Andries E. Braat

To the Editor:

We would like to thank Avolio et al. [1] for the comments on our article on the Eurotransplant Donor Risk Index (ET-DRI) [2]. The Authors address an important issue: the practical incorporation of donor risk into liver allocation algorithms. It is suggested that the ET-DRI should first be compared to existing donor-to-recipient-matching (D2Rm)-models, such as D-MELD [3], BAR [4] or SOFT [5], for applicability and accuracy before using it in liver allocation.

We fully agree that transplantation outcome depends on donor, transplant and recipient risk factors. Of course, combining all these factors would lead to the best assessment of pretransplant risk and therefore best estimation of posttransplant results (note that all risk models are from observational studies, and have never been tested prospectively to confirm predictive value). The first step is to look into donor and transplant factors, as did Feng et al. [6] for the UNOS region. In our current study we describe all significant donor- and transplant risk factors within the Eurotransplant-region, and combine these into one number/score—the ET-DRI. The scores mentioned by Avolio et al. only include some of these factors or factors that were not significant in our study; D-MELD only includes donor age, BAR includes donor age and cold ischemia time (CIT), and SOFT includes donor age, cause of death, creatinine, allocation and CIT. The interesting thing about the ET-DRI (and DRI) is that it is the strongest risk indicator when looking at donor and transplant factors, and it is a continuous and validated score. It provides one number that represents the hazard ratio (HR) of that specific liver allograft, including the transplant factors allocation and CIT. Step 2 will be to investigate significant risk factor in recipients. Ultimately, research will be needed to see how to best integrate both (donor/transplant and recipient) scores to develop a D2R model, as stated by Avolio et al.

Scores like D-MELD, BAR and SOFT do not take the strongest donor factors into account: split-liver (HR1.67) and donation after cardiac death (DCD) (HR1.71). Although such allografts are scarce for the total Eurotransplant region (6.5%), they account for almost 25% in, for example, the Netherlands, and in our opinion should therefore be incorporated in donor risk stratification.

Clearly, further research is needed, especially for recipient risk factors. The ET-DRI (DRI) gives a validated specific HR for every single allograft, which is necessary to correct for in such research. Currently the only factor used in allocation is the extended criteria donor (ECD). We showed the superiority of ET-DRI over ECD for use within the Eurotransplant region. Further steps to be taken are development of a (validated) recipient risk index and the incorporation of donor and recipient risk scores into a risk stratifying model.

Finally we agree that scores may be helpful for allocation, but cannot replace individual assessment and decision making of the transplantation team. In addition, the socioethical choice of weighting utility versus equity remains a major factor of impact of how allocation will be steered in different countries. ET-DRI is a tool, rather than a means, which can be used to do this.


The authors of this Letter to the Editor have no conflict of interest to disclose as described by theAmerican Journal of Transplantation. This manuscript was not prepared or funded by a commercial organization in any way.