Patients with colorectal livermetasteses and heavy tumour burden and progression on 1st line chemotherapy have no other available treatment in Norway today other than 2nd line chemotherapy. The Investigators will randomize patients to HAI-floxuridine (FUDR), or liver-Tx, in addition to 2nd line chemotherapy versus 2nd line chemotherapy alone (Excalibur 1) or systemic chemotherapy with HAI/FUDR versus systemic chemotherapy alone (Excalibur 2). Primary endpoint is overall survival at 2yrs.

1.1 Background - Disease Colorectal cancer (CRC) is the second most frequent malignant disease in Norway (Cancer in Norway 2017). About 50% of the patients will have metastatic disease at time of diagnosis or develop metastatic disease later on. Liver metastases are the most frequent site of metastatic disease. Liver resection is considered the only curative treatment option in CRC patients with liver metastases, however only about 20% of the patients are candidates for liver resection. The treatment option for the majority of the patients is palliative chemotherapy with median overall survival from start of chemotherapy of about 2 years and 10-12 months from starting second line chemotherapy.

1.2 Liver resections for Colorectal Liver metastases (CRLM) While high-quality data (randomized trials) are lacking, it is generally accepted that the only curative treatment for colorectal liver metastases (CRLM) is surgery. Liver resections are generally well tolerated and safe 1, but some patients recur early and probably have limited or no benefit from surgery. These are hard to identify upfront. Even following three decades of systematic liver surgery for CRLM, there is a lack of robust prognostic scoring systems that have sufficient discriminatory power to serve as selectors for surgery or non-operative treatment 2, 3. Even among patients with very poor prognostic scores, there are some who will survive five years following surgery 4, and even without surgery 5. Over the decades, the definition of resectability/un-resectability has been steadily modified. Today, any configuration of metastases can be deemed resectable as long as a resection will leave behind a working liver volume of at least 20-30 % of the estimated total liver volume with a functioning arterial inflow, portal venous inflow, draining bile duct and draining hepatic vein.

1.3 The grey zone As resections are generally well tolerated and adequate prognostication is wanting, there is a tendency to offer resections to patients who have borderline resectable CRLM or who exhibit other non-favourable traits like large or multiple metastases. For patients who have early recurrence of disease, such resections represent a net loss of quality-of-life and an unwanted expenditure for society. Exploring the optimal treatment modality for patients in this grey zone, i.e. with uncertain benefit from surgery, is important to avoid unnecessary resections and providing the optimal treatment for patients in a critical situation.

1.4 Systemic chemotherapy for CRLM Palliative chemotherapy is in general the only treatment option for the vast majority of non-resectable patients. The expected median overall survival (OS) from start of first line chemotherapy is about 2 years and the 5 years OS is about 10%, although, longer median OS has been obtained in selected patients with good performance status (ECOG 0-1), no (K)RAS or BRAF mutations and left-sided tumors 6-10. OS from start of second line chemotherapy is 10-12 months 11.

1.5 Liver transplant for CRLM Liver transplant (LTX) has emerged as a possible solution for some patients with unresectable CRLM who otherwise have good prognosis based on available scorings systems 12, 13. In patients with non-resectable liver only metastases the investigators have previously shown 5 year OS of 56% compared to 9% in a similar cohort of patients starting first line chemotherapy 6, but due to lack of donor organs this will never become the backbone of any treatment modality for a disease as prevalent as CRLM. However, LTX is probably the best treatment option in highly selected patients with non-resectable CRLM liver only disease.

1.6 Hepatic artery infusion (HAI) chemotherapy for CRLM The biological rationale for intra-arterial chemotherapy is that the hepatic artery rather than the portal vein is responsible for most of the blood supply to liver tumors. Hepatic Artery Infusion (HAI) of a cytotoxic drug floxuridine (FUDR) that has a very high first-pass extraction (ca 95 %) in the liver has shown promising results in selected series for several decades 14-16. It was developed at Memorial Sloan Kettering Cancer Center (MSKCC, New York, USA) but is currently unavailable in the European Union, because floxuridine is not registered. HAI has however not gained foothold as a standard treatment option for CRLM, and most publications stem from a very few centres. The reasons for this lack of dissemination are unknown but could well be related to the complexity of the treatment algorithm and the lack of modern randomized trials. In Europe several centers in The Nederlands have recently started the HAI treatment procedure as adjuvant treatment in CRC patients who have received liver resection. (Buisman FE et al. Ann. Surg. Oncol. 2019 26: 4599-4607. Of the 20 patients included in the study in The Nederlands two patients had Clavien-Dindo complication grade III with reoperation due to replacement of a pump with slow flow-rate and a flipped pump. The treatment administered both at MSKCC and the two centers in The Nederlands consist of 0.12 mg FUDR/kg/day + 35.000 IE heparin + 25 mg dexamethasnone in a total volume of 35ml NaCL administered as a continue infusion for 14 days with dose reduction if liver function is affected (Table 2). At day 15 the pump is emtied and refiled with a low dose heparin solution for continuous infusion to avoid coagulation of the catheter. A new cycle is started at day 29. The HAI treatment has been combined with both oxaliplatin and irinotecan regimens combined with 5-FU as systemic chemotherapy 14,16,17 .HAI has also been combined with systemic gemcitabine-oxaliplatin regimen in patients with non-resectable intrahepatic cholangiocarcinoma18 In the study by D'Angelica in non-resectable CRC patients the response rate was 76%, median overall survival was 38 months and 23 of 49 patients became resectable and received a liver resection. Patients having a liver resection had a 3 year overall survival of 80%. In the study by Pak 33 of 64 non-resectable CRC patients received a liver resection with 5 year overall survival of 36%. These results are better compared to what has been reported by systemic chemotherapy only with median overall survival of about 24 months in most studies. Optimal treatment for patients with CRLM in the grey zone is therefore not yet defined and there is a definte need for further studies.

To optimize treatment for patients with a large tumour burden and borderline resectability, the investigators will compare three treatment modalities in a randomized controlled trial in patients that have progressive disease on 1st line of chemotherapy treatment.

1.7 Rationale for the Study and Purpose The target population for this study will be patients who based on traditional preoperative criteria have a very dismal prognosis. They will - according to the inclusion criteria - have a large tumour burden and have shown progression on 1st line systemic chemotherapy treatment. Based on previous trials, only 30 % of this patient group will be alive after two years. These patients have today only one treatment modality available: 2nd line systemic chemotherapy. Response can, however, only be expected in a small minority of these patients. As of today, they are not acceptable for inclusion into any of the liver transplant protocols, and hepatic aretery infusion (HAI) chemotherapy treatment is not offered in Norway (or any other European country, save the Netherlands, as far as the investigators know).

With such a dismal outcome for these patients, an alternative modality that has the potential to improve survival is highly warranted. While transplantation has such a potential, the access to donor organs will allways limit the real-life use of such a treatment and the inclusion of a transplant group in this trial is primarily for proof-of-principle reasons: to benchmark what the investigators have reason to believe is the optimal treatment.

The use of HAI chemotherapy with FUDR has some inherent risks. A laparotomy is necessary to apply the catheter and intra hepatic infusion of FUDR has been reported to cause biliary inflammation and necrosis in a small fraction of the patients. The risk for the latter is however significantly reduced by concommittant steroid infusion. The experiences published from the MSKCC (se further) does however suggest that the drug has few systemic side effects as the first-pass effect in the liver is close to complete, i.e. there is minimal release of active drug into the systemic circulation.

The IMP for use in this protocol does not have a marketing authorization in Europa. The institution that has pioneered the HAI treatment in CRC is Memorial Sloan Kettering Hospital (MSKCC) in New York and the dose is identical to the dose that used in several studies from MSKCC 14,16 and The Nederlands17.

1.8 Research hypothesis In patients with large tumour burden and/or borderline resectability of colorectal liver metastases and progression on 1st line systemic chemotherapy, overall survival following systemic therapy combined with hepatic artery infusion chemotherapy (HAI), or liver transplantation, is better than following conventional systemic chemotherapy alone.

NOTE: Following approval from the necessary regulation bodies in Norway, protocol adjustments were made in May 2023, specifically widening the inclusion criteria (stratum 2c) and allowing continued inclusion following the primary planned 45 patients.

The difference between 1 and 2 is the presence of a transplant alternative in Excalibur 1. For the purpose of prognosis stratification, patients in Excalibur 2 will be randomized in three strata (a, b and c). The intervention is identical, and they will be analysed as a merged group together with the corresponding groups of Excalibur 1.

Included patients must fulfil the following criteria

1. Primary histology: verified adenocarcinoma in colon or rectum

2. Liver metastases

   1. Not possible or feasible to resect at time of inclusion.

   2. Resection will require 10 % or more response in index lesions.

   3. And one of the following:

      • Insufficient response on current line chemotherapy and in need of next line systemic chemotherapy or major change of active agents as judged by treating oncologist (I, IIa, IIb).
      • Treatment stopped due to toxicity, and hence in need of next line systemic chemotherapy (I, IIa, IIb).
      • Stable disease or partial response (RECIST) is achieved following first cycle of 1st. line conventional chemotherapy (4 doses), but minimal probability of reaching liver surgery (IIc) due to any of the following:

   i. > 6 lesions with bi-lobar distribution and CEA > 1.000, or ii. > 10 lesions with bi-lobar distribution and at least one lesion with a diameter > 5 cm, or iii. > 15 lesions with bi-lobar distribution

3. Chemotherapy

   1. Patients must have received at least one line of systemic chemotherapy at time of inclusion in the study. Planned for next line chemotherapy (I, IIa, IIb).
   2. If patients have commenced next line chemotherapy, randomization can only be allowed prior to first evaluation on next line chemotherapy regimen (I, IIa, IIb).
   3. For IIc, patients must have undergone one cycle of systemic conventional chemotherapy and only have stable disease or partial response at first evaluation, but with a response insufficient for resection.

4. The patient

   1. Good performance status, ECOG 0 or 1.
   2. Satisfactory blood tests: Hb >9 g/dl, neutrophiles >1.0 (after any G-CSF), TRC >75, Bilirubin<1.5 x upper normal level, ASAT, ALAT<5 x upper normal level, Creatinine <1.25 x upper normal level. Albumin above lower normal level.
   3. Women of childbearing potential (WOCBP) must have a confirmed menstrual cycle and a negative highly sensitive pregnancy test prior to inclusion, or two negative pregnancy tests two weeks apart
   4. WOCBP must agree to use a highly effective method of contraception (see section 6.1.2) for the entire period of exposure to the IMP in the trial, plus for one menstrual cycle/30 days after the last exposure due to the genotoxic potential of the IMP
   5. Men that may have sexual relations with a WOCBP during the trial must agree to use a condom during intercourse for the entire period of exposure plus for one sperm cycle / 90 days after the last exposure due to the genotoxic potential of the IMP

5. Signed informed consent and expected cooperation of the patients for treatment and follow up must be obtained and documented according to GCP, and national/local regulations.

Exclusion Criteria

Any of the following criteria will exclude participation in the trial:

1. Arterial anatomy not suited for HAI pump-line insertion.

2. A primary tumour in situ that is either a

   1. Rectal tumour scheduled for radiation therapy with fractionation 2 Gy x 25, or
   2. A right-sided or transverse colonic tumour

3. Previous or current bone or CNS metastatic disease.

4. Patients with known intolerance or allergy to any ingredient of the IMP to be used as standard therapy for that patient

5. Breastfeeding women

6. Patients with a psychiatric condition that makes participation in the trial impossible or unethical

7. Patients in a poor nutritional state, those with depressed bone marrow function or those with potentially serious infections must be excluded.

8. Any other reason why, in the opinion of the investigators, the patient should not participate.

Exclusion Excalibur I Any of the following will preclude inclusion into Excalibur I (but not into Excalibur IIa/b/c)

1. BRAF positivity
2. Any sign of extra-hepatic metastatic disease or local recurrence on PET/CT scan, and on CT or MRI thorax/abdomen/pelvis dated within 6 weeks prior to the trial hospital MDT meeting (exception allowed for <3 resectable lung lesions all < 15mm).
3. Liver lesion >10cm
4. Patient BMI > 30
5. Any previous non-colorectal malignancy within latest five years with the exception of basal cell carcinoma of the skin.
6. Age > 70 years
7. Liver metastatic ingrowth to the diaphragm determined by CT-scan and/or MRI/or ultrasound
8. Any primary tumour in situ

Stratification of Excalibur II:

For patients eligible for Excalibur II, randomization will be in stratum:

IIa for patients following a switch to next-line chemotherapy with none of the following features, and IIb for patients following a switch to next-line chemotherapy with any of the following features IIc for patients still on 1st line chemotherapy with none or any of the following features

• Non-curable pulmonary disease
• Non-hilar, non-regional lymph node metastases
• Limited and resectable peritoneal disease