Reply: Additional viewpoints from transplant surgeons on Banff consensus recommendations for the assessment of steatotic donor livers

We thank Drs Cao and Gu for their thoughtful response to our publication. The 3-step algorithm for assessing donor steatosis was based on the approach of 3 experienced liver transplant pathologists with the goal of developing a standardized approach that was lacking. Diagrams, digital images with a range of steatosis, and extensive discussion were used. Cases were then selected to include a severity spectrum from minimal to severe large droplet fat (LDF) on the local pathologist’s assessment. As noted by Drs Cao and Gu, this turned out not to be the case. The cases represent the spectrum of donor livers biopsied for the assessment of steatosis in the United Kingdom and are mirrored in a yet unpublished follow-up study using biopsies performed in the United States. Thus, it is possible that many cases with a gross appearance of severe steatosis are not biopsied based on clinical or visual assessment. This raises the question: are livers suitable for transplantation being overlooked based on clinical and visual inspection of the degree of steatosis? The timing of the specimen is outside the control of pathologists. Procurement organizations tend to biopsy early, while implanting surgeons perform on the back table shortly before transplantation. LDF% does not change during storage; however, smaller droplet fat (SDF) may.1 SDF is not assessed as part of the Banff 3-step algorithm, but in view of defatting strategies under investigation, it will be important to understand the processes involved, as this organelle (lipid droplet) is required for liver regeneration.2,3 Therefore, biopsy timing is important. A rapid paraffin section is generally of better quality but takes longer (≈4 h) than a frozen section (<1 h). A good frozen section is satisfactory, a poor frozen section is not. A rapid paraffin service would necessitate an early biopsy or utilization of machine perfusion to safely increase preservation time. Another option is to standardize the quality of frozen sections. Frozen section quality is important for the pathologist and digital algorithms. We agree that further validation of the 3-step approach and digital algorithms based on this approach should be undertaken based on outcome data personalized to recipient factors (being able to tolerate postreperfusion syndrome and short periods of graft dysfunction) to determine cutoffs for safe transplantation. This “validation metric” will also need to be further refined based on the recipient clinical team’s ability to manage complications that arise when transplanting fatty livers. Looking to the future the reason given for biopsy may change to “should this liver be optimized/defatted by extracorporeal perfusion?”