Real-World Use of Hypothermic Machine Perfusion: Trends and Impact on Short-Term Kidney Allograft Outcomes

Introduction: Hypothermic machine perfusion (HMP) is widely used in kidney transplantation to reduce ischemia-reperfusion injury and reduce delayed graft function (DGF). However, existing evidence largely reflects continuous perfusion, whereas real-world practice often includes interrupted or partial perfusion due to logistical constraints. The impact of these alternative HMP strategies on graft outcomes has not been well characterized. We aimed to characterize the use of these HMP strategies and evaluate their impact on early kidney transplant outcomes.
Methods: Data were obtained from the Scientific Registry of Transplant Recipients (SRTR) and included all adult single kidney transplants performed between December 5, 2014 and March 2, 2024. Cases with missing pump data were excluded, yielding a final cohort of 108,236 transplants. Kidney preservation was categorized as static cold storage (SCS), continuous HMP, or partial HMP. Partial HMP was defined as any noncontinuous perfusion strategy, including upfront-only, interrupted, and end-ischemic perfusion sequences. Temporal trends in perfusion strategy were examined. Primary outcomes included delayed graft function (DGF) and primary nonfunction (PNF). Multivariable logistic regression models were used to assess the association between HMP strategy and outcomes, adjusting for donor and recipient characteristics.
Results: Over time the total number of kidneys procured for transplant increased by 6.33 per month (95% CI 5.9-6.76, p<0.001), as did HMP utilization (5.06/month, 95% CI 4.75-5.37, p<0.001). All types of HMP strategies increased over time, with up front perfusion representing the largest increase (1.78/month, 95% CI 1.64-1.92) followed by continuous (1.35/month, 95% CI 1.17-1.54).

End ischemic (0.31/month, 95% CI 0.25-0.36) and interrupted perfusion (0.11/month, 95% CI 0.06-0.16) increased only slightly over time. SCS also increased by 1.27 kidneys/month (95% CI 1.05-1.49). On adjusted analyses, continuous HMP was associated with lower odds of DGF compared with SCS (OR 0.75, 95% CI 0.72–0.78), as was partial HMP (OR 0.89, 95% CI 0.86–0.93). Further stratified, all HMP strategies were associated with lower odds of DGF compared to SCS except up front HMP (OR 0.99, 95% CI 0.95-1.03).

On univariate analysis, partial HMP was associated with increased risk of PNF compared to SCS (OR 1.42, 95% CI 1.20-1.68) but this association was no longer statistically significant after adjusting for donor and recipient factors.  
Conclusion: The majority of HMP data in kidney transplantation is derived from studies using continuous, whereas real-world logistical constraints often require partial perfusion. In this study, we found that continuous, interrupted, and end-ischemic perfusion were associated with lower odds of DGF, whereas the odds of DGF with up front HMP did not differ significantly from SCS. These findings suggest that the timing and continuity of perfusion may influence early graft function. Continuous perfusion may be associated with more favorable early outcomes when feasible, underscoring the importance of optimizing real-world perfusion practices as utilization continues to expand.