An updated review of 20 observational studies links normothermic regional perfusion to fewer ischemic biliary complications after controlled DCD liver transplantation. Ischemic cholangiopathy risk was lower than with non-NRP recovery (RR 0.16), alongside favorable graft and recipient outcomes. No significant differences were found versus brain-death donors, but equivalence was not established. Evidence certainty ranged from moderate to very low, with residual confounding limiting causal conclusions.
Why regional perfusion matters for liver transplantation
Controlled donation after circulatory death (cDCD) expands the liver donor pool, but the interval of warm ischemia around circulatory arrest can damage the graft before recovery. Bile ducts are particularly vulnerable. Ischemic cholangiopathy, characterized by non-anastomotic biliary strictures despite a patent hepatic artery, can lead to repeated interventions, graft failure, and retransplantation. A 2026 systematic review by Ramírez-Esteban and colleagues examines whether normothermic regional perfusion (NRP) is associated with better liver transplant outcomes than non-NRP cDCD recovery and how those outcomes compare with donation after brain death (DBD).
What NRP does—and what this review evaluated
NRP restores oxygenated blood flow to organs in situ at near-physiological temperature before procurement. The rationale is to support metabolic recovery after warm ischemia and permit assessment of organ function before acceptance. Abdominal NRP and thoracoabdominal NRP were grouped together in this review. These techniques differ from ex situ machine perfusion, which supports an organ after removal from the donor. The distinction matters because some studies used additional normothermic or hypothermic oxygenated machine perfusion, potentially influencing outcomes independently of NRP.
A larger evidence base, with important limits
The authors searched five databases and screened 715 records after duplicate removal. Twenty observational studies met the inclusion criteria, representing 1,776 NRP-cDCD liver recipients across European and North American cohorts. Sixteen studies contributed to the quantitative synthesis, encompassing 1,512 NRP recipients; four were retained for descriptive discussion because of overlap, unavailable incidence data, inability to isolate NRP, or a different technique comparison. Consequently, the headline total should not be interpreted as the sample size contributing to every pooled outcome. Individual comparisons drew on smaller subsets of studies and patients.
The strongest finding: fewer ischemic biliary complications
Compared with non-NRP cDCD recovery, NRP was associated with a substantially lower risk of ischemic cholangiopathy: risk ratio 0.16, with a 95% confidence interval of 0.09–0.28. Six studies contributed to this comparison, with no detected statistical heterogeneity. The authors reported pooled incidences of 1.8% with NRP and 12.7% without NRP, alongside an estimated number needed to treat of approximately nine. These findings identify biliary protection as the clearest signal in the review. The number needed to treat is a contextual estimate derived from observational evidence, rather than a guaranteed benefit for an individual donor or recipient.
Graft loss, mortality, and other complications
NRP was also associated with less graft loss, with a risk ratio of 0.41 (95% CI 0.32–0.54), and lower recipient mortality, with a risk ratio of 0.46 (0.34–0.62). Hepatic artery thrombosis was reduced in the pooled comparison, with a risk ratio of 0.51 (0.30–0.85). Other biliary complications, including anastomotic strictures and leaks, had a risk ratio of 0.49 (0.32–0.73). Primary non-function likewise favored NRP, with a risk ratio of 0.51 (0.28–0.94). These associations are clinically relevant, but they should not be presented as randomized evidence that NRP alone caused the improvements.
Why pooled percentages do not always match risk ratios
The primary non-function results illustrate an important statistical distinction. The review reported marginal pooled incidences of approximately 3.0% in both NRP and non-NRP cDCD groups, despite a comparative risk ratio favoring NRP. The apparent discrepancy reflects different study sets: incidence estimates included all studies reporting that outcome, whereas relative effects used the studies contributing directly to the comparison. Within the comparative pool, crude rates were approximately 2.0% versus 3.1%. Comparing the separate incidence estimates as though they came from one matched population would therefore misrepresent the analysis.
Early dysfunction depends on the comparator
Early allograft dysfunction was lower overall with NRP, with a risk ratio of 0.65 (95% CI 0.45–0.94), but heterogeneity was substantial at I² = 74%. Exploratory subgroup analysis found a benefit against cold-storage recovery pathways, while comparisons against ex situ normothermic machine perfusion did not show a significant advantage. Only two studies provided isolated normothermic machine-perfusion comparators. The authors also noted that perfusion can wash out biomarkers used in early dysfunction definitions, making biochemical differences difficult to separate from true functional recovery. ICU and hospital length of stay did not differ significantly.
Comparison with brain-death donors
None of the pooled outcomes differed significantly between NRP-cDCD and DBD recipients. For ischemic cholangiopathy, the preferred rare-event analysis gave a risk ratio of 0.70, with a wide 95% confidence interval of 0.16–3.01. Pooled incidences were 1.8% with NRP and 1.2% with DBD. These results are encouraging, but absence of a statistically significant difference does not establish equivalence or noninferiority. The confidence intervals allow clinically meaningful differences in either direction, and the authors rated all NRP-versus-DBD outcomes as very-low-certainty evidence.
How the authors tested the findings
The investigators used random-effects analyses and examined alternative statistical estimators, rare-event methods, influential studies, and higher-risk cohorts. They also addressed overlapping populations to reduce double counting. A sensitivity analysis excluded grafts exposed to ex situ machine perfusion in either arm, helping separate the NRP association from that additional preservation strategy. The favorable direction and statistical significance persisted for the principal comparisons against non-NRP cDCD recovery. For ischemic cholangiopathy, the restricted analysis produced a risk ratio of 0.14 (95% CI 0.08–0.26). These checks support robustness, although they cannot eliminate unmeasured confounding.
How strong is the evidence?
No included study was judged to have low overall risk of bias: seven were rated moderate and thirteen serious using ROBINS-I. Donor selection, recipient characteristics, center experience, and procurement practices may have influenced both the use of NRP and transplant outcomes. Fewer than ten studies contributed to each pooled comparison, preventing formal publication-bias testing. Mortality and graft loss were summarized using event-count risk ratios rather than pooled time-to-event hazard ratios. Outcome definitions and follow-up varied, and remaining cohort overlap could not be definitively resolved without individual-patient data.
What this means for perfusion teams
The review rated certainty as moderate for the ischemic cholangiopathy reduction, low for most other favorable comparisons against non-NRP cDCD recovery, and very low for early allograft dysfunction and comparisons with DBD. For perfusion professionals, the findings strengthen the rationale for NRP within established multidisciplinary donation and transplantation programs, particularly its potential to protect the biliary system. They do not identify an optimal circuit configuration, prove superiority of abdominal versus thoracoabdominal NRP, or define procedural targets. Those questions require dedicated technical and clinical investigations.
The clinical takeaway
This updated synthesis provides a consistent and substantial association between NRP and fewer ischemic biliary complications after cDCD liver transplantation, accompanied by encouraging graft and recipient outcomes. The strongest interpretation is that NRP appears to improve the performance of cDCD liver recovery pathways in the populations studied. Claims that it definitively matches DBD outcomes or guarantees survival benefits go beyond the evidence. Randomized or pragmatic registry-based studies, standardized biliary outcome definitions, individual-patient analyses, and longer follow-up would help establish the durability and causal contribution of these benefits.
Keywords: Normothermic Regional Perfusion, Controlled Donation After Circulatory Death, Liver Transplantation, Ischemic Cholangiopathy, Graft Survival, Biliary Complications, Thoracoabdominal NRP, Abdominal NRP, Organ Preservation, Machine Perfusion.
Source: Ramírez-Esteban J, Sarrió E, Ruiz-Pacheco A, Cots A, Romero-García N, Badenes R. Transplant International. 2026;39:17154. doi:10.3389/ti.2026.17154.
Original educational summary of an open-access article published under CC BY. Cover artwork is an AI-generated conceptual illustration, not a procedural diagram.
