Liver resection is the golden standard in the treatment of hepatic malignancies. The size and function of the remnant liver is a major concern. If the future liver remnant (FLR) is below 30 % of the initial liver volume, the risk of post hepatectomy liver insufficiency rises. Several techniques have been developed to increase the size of FLR before liver resection. In this study a new technique ARAPS (portal vein embolization with radio frequency ablation) is compared to portal vein embolization alone for accelerated liver growth in the FLR. This is done in a randomized controlled trial.

Curative intent liver resection is the gold standard for treatment of hepatic malignancies, including metastases from colorectal cancer (CRLM), but liver resection can only be performed if future liver remnant (FLR) volume is at least 25 %. Unfortunately, many cancer patients have multiple liver lesions at the time of diagnosis, rendering radical resection impossible due to a marginal FLR, thereby increasing the risk of posthepatectomy liver failure and subsequent death.

Several techniques have been designed to increase the size of the FLR, allowing more patients to undergo surgery. These techniques include portal vein embolization (PVE), portal vein ligation (PVL), two-stage partial hepatectomy, and the associating liver partition with PVL for staged hepatectomy (ALPPS). The ALPPS procedure seems to be superior to the other techniques for achieving increased size and accelerated growth of the FLR. Especially the kinetic growth rate (KGR) is superior to what has been previously described. The major drawback of ALPPS is, however, that the patient is subjected to two open procedures that are associated with high morbidity and mortality.

In a previous experimental study on rats, The investigators have shown that PVE combined with radiofrequency ablation (RFA), hereafter called Associated Radiofrequency Ablation and Portal Vein Ligation for Staged Hepatectomy (ARAPS), can induce accelerated growth of the FLR as potent as that seen after ALPPS. In humans, the ARAPS procedure can (the first step), in contrast to the two-stage ALPPS procedure, be performed percutaneously, and the investigators expect this to reduce the high morbidity and mortality associated with the ALPPS. In the present study, the investigators want to examine the effect of ARAPS compared with PVE alone on liver regeneration and function of the FLR.

Study rationale In order to allow more patients to undergo radical surgery for CRLM, implementation of new methods to enhance liver regeneration is vital.

Aim In the present study, the investigators test the hypothesis that the ARAPS approach is superior to PVE alone in achieving growth of the FLR in patients with CRLM.

Hypothesis the investigators expect to see higher growth of the FLR in the ARAPS arm compared with the PVE arm one week after intervention.

Study design The study is an investigator-initiated prospective, randomized study. Any patient eligible for this study will be offered enrollment. Patients enrolled in the study will be randomized (1:1) to either the intervention (ARAPS) or control (PVE) as

Recruitment of study subjects Potential study subjects will be given the option to participate after the study is presented during their first visit to the outpatient clinic. The visit will take place in a consultation room, where the surgeon will provide verbal and written information regarding the study prior to enrollment. Patients are allowed a companion for this interview. At least 24 hours are given before an informed consent is obtained. Patient consent can be withdrawn at any time, and all data collected for the project, will be destroyed without any consequence for their subsequent treatment.

Data analysis plan Power calculations Total liver volume is ~1500 mL. Assuming that FLR is 20 % of the total liver volume, FLR comprises 300 mL. Based on previous literature, the investigators expect a growth in the FLR of 10 % (30 mL, KGR ≈ 4.3 ml/day) in the PVE arm and 50 % (150 mL, KGR ≈ 21.4 ml/day) in the ARAPS arm. At alpha=0.05, power 80 %, and a standard deviation of 200 mL, 12 patients in each arm is needed to show a statistically significant difference between the two interventions. In order to be able to deal with potential dropout of up to 20%, the investigators aim to include 15 patients in each arm; 30 patients in total.

Statistical methods Liver regeneration will be measured as the increase in anatomical (CT) and functional (PET) liver volume seven and thirteen days following intervention (quantified in milliliters). Further, the KGR will be calculated as volume increase per day (ml/day). Differences in the two arms will be expressed as means or medians with associated standard deviations or interquartile ranges (depending on the distribution of data). Differences between the two arms will be analyzed using a t-test or non-parametric test where appropriate. P<0.05 will be considered statistically significant.

Data on patients Data on age, diagnosis, comorbidity, and tumor imaging material will be passed on from the medical records in order to assess if the patient is eligible for inclusion in the study. Information on sex, tumor size, tumor histology, number and size of metastases, medications, tobacco/alcohol exposure, and blood sample results will be retrieved on from the patient's medical records. Radiological, biochemical, and pathological reports will also be retrieved for use in the statistical analyses.

Study procedures Intervention Portal vein embolization The standard-of-care procedure, which would be offered to the patients regardless of this study, is PVE. PVE will be used to redirect portal flow to segments of the FLR: The portal venous system is assessed through the tumor-bearing part of the liver (i.e. the part that is later going to be removed). The portal vein is assessed under guidance by ultrasonography, and a catheter is then advanced under fluoroscopic guidance for subsequent embolization. In the case of a planned extended right hepatectomy, the portal branches to segment 4 should be selectively embolized. As embolization material plugs are preferred.

Radio-frequency ablation Under guidance of ultrasonography a single-electrode RFA-needle is placed in the part of the liver that has just been deportalized. The needle is placed in normal parenchyma, and within a period of ten minutes a necrosis of approx. 3 cm is created. In literature, complication rates following RFA treatment are low, making it a feasible procedure.

Both treatments will be performed in general anesthesia. Patients will be equipped with two bandages, which they have to wear until after the second PET/CT scan in order to blind them to the treatment received.

PET/CT scanning Accurate liver volume estimation relies on precise segmentation of the liver by multi-phase CT images according to the Couinaud classification. A thicknesses of ≥ 1 mm will be used to ensure accurate results. Total liver volume, volume of the deportalized part of the liver, and volume of the FLR will be estimated before and after intervention. Positron emission tomography (PET) with the radioactively labelled galactose analogue 2-[18F]fluoro-2-deoxy-D-galactose (FDGal) can be used to quantify whole-liver and regional hepatic metabolic function as well as to measure functional liver volume. The method is validated, simple and with insignificant day-to-day variation. The PET scan is performed after a 4h fast, but the patient is allowed to drink water. In Aarhus, the scanner is a 64-slice Siemens Biograph TruePoint PET/CT camera (Siemens AG, Erlangen, Germany). FDGal (100 MBq) is injected intravenously and a low-dose CT scan (50 effective mAs with CAREDose4D, 120 kV, pitch 0.8, slice thickness 5 mm) is performed for attenuation correction of the PET data and anatomical co-registration. A static PET scan of the mean tissue radioactivity concentration is performed from 10-20 minutes after tracer injection (12). The PET data are corrected for radioactive decay back to injection time and reconstructed with resolution modeling (4 iterations, 21 subsets and 2 mm (18F-FDGal) or 3 mm (18F-FDG) Gaussian filter) and 336 x 336 matrix. The standardized uptake value (SUV; unit-less) is calculated dividing tissue radioactivity concentration (kBq/mL tissue) with the ratio between injected dose (kBq) and body weight (mL tissue, assuming an average tissue density of 1 g/mL tissue). Whole-liver and regional hepatic FDGal-SUVs as well as the functional liver volume are estimated.

Potential benefits, risks, and side-effects Potential benefits As the potential impact of this study is extensive, including the possibility for more patients with CRLM to undergo curative-intent resection, it is our opinion that the benefits by far outweigh the few potential side effects. Only two PET and two CT scans with minimal radiation will be conducted. The blood samples drawn is part of the standard work-up procedure. No additional blood samples will be drawn on the basis of participation in this study.

Radiation dose from PET/CT scans Patients participating in the present study are not sustained to more radiation from imaging modalities than patients not participating. From the low-dose CT-scans, the patient receives an irradiation dose of up to 1 mSv. For FDGal the effective radiation dose is 0.032 mSv/MBq, the administration of 100 MBq FDGal thus gives a dose of 3.2 mSv. The total irradiation dose that each patient receives from two PET and two CT scans is approximately 8.4 mSv which equals three times the average yearly background dose in Denmark. This presents an estimated risk of 1:10.000 for cancer development due to participation in the study. These doses should be seen in connection with the potential benefit for the patient in participating in the study, as mentioned above.

Interventions Potential risks related to the RFA treatment are extremely sparse, and include bleeding, hepatic abscesses, hepatic infarcts, and bile duct injuries. PVE is a safe procedure that causes few adverse effects and can be performed in an outpatient setting. Most large series report 0 % procedure related mortality. Risks include bleeding, infection and displacement of the embolization material. Especially the latest is extremely rare.

Further patient treatment Following calculation of KGR and increase in FLR, patients are re-evaluated with regard to the timing of the second-stage partial hepatectomy. Some patients, especially in the PVE arm, are expected to need further liver hypertrophy before advancing to the partial hepatectomy. Further hypertrophy in a period of 1-3 weeks and a new CT evaluation might be necessary. In some patients, the hypertrophy might be impaired or absent. These patients will be evaluated with regard to other treatment options including salvage ALPPS, which have previously been used in patients with impaired hypertrophy following PVE. The resection rate in each arm will be reported as a secondary endpoint.

Time schedule This study will commence on 1. November 2019. Total study duration is expected to be 24 months for subject recruitment, and an additional six months for data analysis. The investigators expect to finalize the data analysis and present our findings during 2021. The project will begin at Aarhus University Hospital, while the other participating centers will be enrolled as appropriate.

Publications Both positive, negative, and inconclusive results will be published, and presented at scientific meetings. Results will be presented at international scientific congresses, and the investigators will aim for publications in high-impact journals with focus on oncology, radiology, and liver surgery. The study will be registered at www.clinicaltrials.gov.

Financial aspects This study is investigator-initiated, and supported by several Danish foundations (Vissing Fonden [322.000 DKK (Danish Kroner)], Knud & Edith Eriksens Mindefond [50.000 DKK], Læge Sofus Carl Emil Friis og Hustru Olga Doris Friis' Legat [148.000 DKK], Familien Erichsens Mindefond [98.000 DKK], Tømmerhandler Vilhelm Bangs Fond [98.000 DKK], Riisfort Fonden [198.000 DKK], and Neye Fonden [98.000 DKK]). Grants from these foundations are given to the principal investigator Prof. Frank Viborg Mortensen, and are administered by Aarhus University Hospital. The grants support running costs related to the project (CT/PET scans, blood samples, RFA needles, etc). The researchers participating in this study are independent from any financial interests. No fee will be paid to the study subjects. Salaries for the scientific staff are already financed by their respective employers.

Ethical considerations The study will be conducted according to the Helsinki Declaration, and following approval by the local ethics committee. The project is registered at the Danish Data Protection Agency in accordance with the Central Denmark Region common registration system. No patient will be included before these permissions are granted. All information involving patients will be handled according to national law on personal data information. The study is covered by the Danish Patient Compensation Authority.

Inclusion Criteria:

1. Patients with Colorectal liver metastasis; initially considered unresectable or borderline resectable due to an expected postoperative FLR volume of <30 %
2. Extrahepatic metastatic disease is not an exclusion criteria if it can be treated with curative intent

Exclusion Criteria:

1. Liver cirrhosis
2. Significant comorbidity rendering subjects unsuitable for major surgery