Multicentre parallel-group open-label non-inferiority randomized clinical trial of recooling versus no recooling after normothermic machine perfusion in liver transplantation: ARANMP-LT protocol

Normothermic machine perfusion (NMP) has emerged as a preservation strategy for liver grafts, sustaining aerobic metabolism and allowing real-time viability assessment of marginal grafts that would otherwise be discarded¹. Despite these advantages, a critical procedural step remains inherited from the cold storage paradigm: after NMP, the liver is commonly recooled with cold preservation solutions before implantation. However, the physiological necessity of this step has never been validated in the context of NMP.

The transition from normothermia to hypothermia and back to normothermia during reperfusion imposes a thermal shock on cellular and subcellular structures. This temperature cycling triggers oxidative stress, may compromise mitochondrial integrity and ATP stores, and degrades the endothelial glycocalyx^2,3. Avoiding recooling eliminates this thermal cycle, reduces the handling of cold preservation fluids and shortens the interval between graft removal from the perfusion device and reperfusion. A retrospective cohort study has described a standardized no-recooling technique associated with comparable or improved outcomes⁴, but no randomized comparison exists.

ARANMP-LT (ClinicalTrials.gov ID NCT07692490, registered 2 July 2026) is an investigator-initiated parallel-group open-label randomized clinical trial across multiple Italian liver transplant centres comparing implantation with and without graft recooling in adult recipients of deceased donor livers preserved with normothermic or sequential hypothermic–normothermic machine perfusion. Paediatric, retransplant, acute liver failure and multiorgan recipients are excluded. Participants are randomized 1 : 1 after the transplant team has definitively decided to proceed with transplantation, using a centrally generated sequence stratified by site with permuted blocks. In the experimental group, the liver is transferred to the surgical field immediately after removal from the perfusion device, without exposure to cold fluids, and a standardized portal flush with 1 l of 5% albumin at room temperature (20–25°C) precedes reperfusion, clearing residual perfusate and stabilizing vascular resistance. Control grafts are flushed with cold preservation solution according to local practice. The two strategies are compared in Table 1.

Table 1.

Key differences between standard recooling and no-recooling implantation after NMP

	Standard recooling strategy	No-recooling strategy
Thermal and implantation pathway	Graft re-cooled after NMP, followed by implantation and subsequent rewarming at reperfusion	Graft transferred directly to the surgical field without cold exposure; implantation proceeds after a portal flush at 20–25°C
Portal flush	Cold preservation solution	1 l of 5% albumin at 20–25°C
Expected implications	Additional cooling–rewarming cycle and more complex implantation sequence	Avoidance of thermal cycling, simplified implantation workflow, and potential attenuation of ischaemia–reperfusion injury

NMP, normothermic machine perfusion.

A non-inferiority framework was adopted. The primary outcome is the proportion of recipients free from early allograft dysfunction within 7 days after surgery, defined according to Olthoff criteria⁵. Non-inferiority will be declared if the lower limit of the one-sided 95% confidence interval for the between-group difference, computed by the Farrington–Manning score method, lies above −7.5 percentage points, a margin reflecting clinical judgement and consistent with two previous machine perfusion trials.

Secondary outcomes comprise the histopathological severity of ischaemia–reperfusion injury, evaluated on paired pre- and post-reperfusion wedge biopsies; hepatic artery and portal vein thrombosis; primary non-function; biliary complications within 90 days, including strictures and ischaemic cholangiopathy; post-reperfusion syndrome and intraoperative haemodynamic instability; and patient and graft survival at 90 days.

Assuming a control arm success rate of 65% and a planning alternative of a 10-percentage-point advantage for the experimental strategy, with one-sided α = 0.05 and 80% power, 85 evaluable patients per group are required (170 patients in total). The trial is currently recruiting, with competitive enrolment across centres over 24 months and 3 months of follow-up. Blinding of the operating team is not feasible. The primary outcome is nevertheless derived from objective laboratory values; allocation is concealed until the decision to transplant has been taken, and histopathological scoring is performed blind to allocation.

The full protocol and the completed SPIRIT 2025 checklist are provided in the Supplementary material.

ARANMP-LT addresses a decision taken in every machine-perfused liver transplantation, never before subjected to randomized evaluation. Should non-inferiority be established, the implantation sequence could be simplified without a demonstrated cost in early graft function.

Funding

This research received no external funding.

Author contributions

Salvatore Gruttadauria (Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Validation, Writing—original draft, Writing—review & editing), Roberta Vella (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review & editing), Fabrizio Di Francesco (Conceptualization, Investigation, Validation, Writing—original draft, Writing—review & editing), Duilio Pagano (Conceptualization, Investigation, Methodology, Validation), Marco Barbara (Data curation, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review & editing), Monica Rizzo (Data curation, Investigation, Methodology, Writing—review & editing), Ivan Vella (Conceptualization, Validation, Writing—review & editing), Sergio Calamia (Conceptualization, Investigation, Validation, Writing—review & editing), Alessandro Tropea (Conceptualization, Investigation, Validation, Writing—review & editing), Pasquale Bonsignore (Conceptualization, Investigation, Validation, Writing—review & editing), Caterina Accardo (Conceptualization, Validation, Writing—review & editing), Gaetano Burgio (Conceptualization, Validation, Writing—review & editing), Enzo Andorno (Project administration, Validation, Writing—review & editing), Marco Miggino (Resources, Writing—review & editing), Stefania Camagni (Project administration, Validation, Writing—review & editing), Domenico Pinelli (Project administration, Validation, Writing—review & editing), Amedeo Carraro (Project administration, Validation, Writing—review & editing), Alex Borin (Resources, Writing—review & editing), Matteo Cescon (Project administration, Validation, Writing—review & editing), Matteo Serenari (Resources, Writing—review & editing), Umberto Cillo (Project administration, Validation, Writing—review & editing), Jacopo Lanari (Resources, Writing—review & editing), Fabrizio Di Benedetto (Project administration, Validation, Writing—review & editing), Paolo Magistri (Resources, Writing—review & editing), Giuseppe Maria Ettorre (Project administration, Validation, Writing—review & editing), Sara Ingallinella (Supervision, Validation, Writing—review & editing), Francesco Frongillo (Project administration, Validation, Writing—review & editing), Marco Maria Pascale (Project administration, Validation, Writing—review & editing), Davide Ghinolfi (Project administration, Resources, Validation, Writing—review & editing), Giulia Cirillo (Project administration, Validation, Writing—review & editing), Fabio Melandro (Project administration, Visualization, Writing—review & editing), Gianluca Mennini (Project administration, Validation, Writing—review & editing), Francesco Tandoi (Project administration, Validation, Writing—review & editing), Irene Scalera (Project administration, Visualization, Writing—review & editing), Giuseppe Tisone (Project administration, Validation, Writing—review & editing), Roberta Angelico (Project administration, Validation, Writing—review & editing), Marco Vivarelli (Project administration, Visualization, Writing—review & editing), Federico Mocchegiani (Project administration, Validation, Writing—review & editing), and Sergio Li Petri (Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Resources, Validation, Writing—review & editing])

Disclosure

The authors declare no conflict of interest.

Supplementary material

Supplementary material is available at BJS Open online.

Data availability

This article describes the protocol of an ongoing randomized clinical trial; no participant data had been generated or analysed at the time of submission.