Patients with intrahepatic cholangiocarcinoma (IHC) have relatively aggressive tumors, and the prognosis for most of these patients is dismal. Surgery is the only option that can offer potential cure, but only an estimated 20-25 % are amenable to resection. Down-staging conventional chemotherapy has a relatively low response rate (< 50 %). Patients will be included into the respective treatment arms based on their tumour characteristics and disease stage, but also based on their ability/preferences, as HAI-FUDR/DEX requires going to Oslo every fortnight for the duration of the treatment and SIRT has some limitations regarding tumour distribution.

Data from the MSKCC has suggested a clinically relevant benefit from adding intrahepatic chemotherapy to systemic therapy. HAI-FUDR/DEX is not approved in Norway and can only be evaluated in a protocolized trial. Given the risk of distant disease progression with IHC, the addition of conventional systemic chemotherapy makes good clinical sense, and data from MSKCC supports this approach. SIRT is another modality also applied trans-arterially and directly into the tumour. This treatment is approved in Norway and available in Bergen and in Oslo. It is far less cumbersome to deliver and maintain than HAI-FUDR/DEX. The efficacy and safety of the two treatment groups, HAI-FUDR/DEX and SIRT, will be compared in a parallel cohort (non-randomized) design

This is a parallel (twin) cohort study aiming to assess the efficacy and safety of selective internal radiation therapy (SIRT) and hepatic artery infusion (HAI) with floxuridine (FUDR) and dexamethasone (DEX) in combination with modern systemic chemotherapy. The study groups will consist of patients with intrahepatic cholangiocarcinoma (IHC) not amenable to surgical resection.

Thus, for many patients with IHC, effective treatment options are limited. Liver-directed hepatic arterial chemotherapy delivered through a surgically implanted HAI pump has been evaluated in several small series and appears to have greater efficacy than systemic therapy alone. Although used predominantly in patients with hepatic colorectal metastases, HAI chemotherapy has shown efficacy in patients with primary liver cancer. HAI-FUDR/DEX is not approved in Norway (nor the EU), but is used in an ongoing RCT at Oslo University Hospital for the treatment of colorectal cancer liver metastases (CRLM) (Excalibur).

Based on the tumor characteristics and tumor stage, as well as the patient's ability to undergo HAI-FUDR/DEX treatment, we will divide patients who are not amenable to upfront resection but who are potentially curable, into two down-staging treatment groups: systemic chemotherapy in combination with either SIRT or HAI with FUDR/DEX. If allocated to SIRT, a standard work-up is performed in Oslo, with SIRT being performed at the same site 1-3 weeks later.

Trans-arterial treatment modalities If allocated to HAI-FUDR/DEX, a laparotomy is scheduled within three weeks from inclusion. Upon laparotomy, hilar lymph nodes are cleared completely and cholecystectomy is performed. Placement of The Tricumed IP2000V pump and catheter will be as per current routine. The patient is expected to stay at the hospital ward for 3-6 days after the pump is implanted. During this time they will undergo a standard control of catheter patency and absence of any collateral which is performed by direct angiography and isotope scanning 3-5 days post surgery.

The Tricumed IP2000V pump is used outside its approved indication, and experience is limited to other clinical trials using the HAI pump for administration of FUDR. The Tricumed IP2000V pump is a constant flow implantable infusion pump intended to store and slowly deliver FUDR/DEX intrahepatically. The infusion pump is contoured with smooth radius to increase the comfort for patients. It contains a fluid reservoir in the form of a titanium bellow that is pressurized by n-butane at constant pressure. The n-butane is in phase change between a liquid and a vapor and at body temperature of 37 °C provides a constant pressure of 2.5 bar. The body temperature of the patient provides the system with energy. This drive system provides the implantable infusion pump with an inexhaustible source of energy. The drug therapy is therefore delivered at a substantially constant rate over the whole refill interval. Through use of a gaseous drive medium, arduous warming up of the implantable infusion pump is no longer necessary before implantation.

Surgical complications related to HAI pump placement are uncommon (<10%), but include hepatic artery thrombosis, pump pocket infection, and arterial hemorrhage at the site of arterial catheter insertion. Prior to the first administration of HAI-FUDR/DEX an additional technetium-99-labeled macro aggregated albumin nuclear medicine scan and a CT angiography are performed to confirm bilobar hepatic perfusion and rule out extrahepatic perfusion. A total of 5-10 ml contrast agent will be infused, with a negligible effect on renal function, which is additional to the standard of care. HAI-chemotherapy toxicities include ulcer disease and biliary sclerosis, which can both be largely avoided by imaging prior to treatment and monitoring of liver tests and dosages adjustments, if needed. Systemic side effects with HAI chemotherapy of FUDR are rare (<1%).

Hepatic Artery Infusion with Floxuridine and Dexamethasone (HAI-FUDR/DEX) Treatment will be provided in 28-day cycles. The Hepatic Artery infusion is dosed with 0.12 mg/kg floxuridine together with 35,000 U of heparin, 25 mg of dexamethasone and enough normal saline for a total volume of 35 ml. This is to be delivered as a 14-day continuous infusion through the pump. For patients who are more than 25% above ideal body weight, the actual dose of floxuridine is calculated by using a weight that averages the patient's actual weight and their ideal weight. HAI-FUDR/DEX is provided at Day 1 of each cycle, and the remaining FUDR in the pump is emptied at Day 15 and then re-filled with 35,000 units of heparin in normal saline (q.s. 35 ml), and infused over the next 14 days.

The patient's "reference value" is defined as the value obtained on the first day of the most recent FUDR dose. Variations within hospital reference value range shall not prompt any dose adjustments. Bilirubin, alkaline phosphatase or AST increases outside of hospital reference value range may necessitate FUDR dose modification. To determine if a FUDR dose modification is necessary in these cases, the patient reference value is compared to either the value obtained on the day the pump was emptied (e.g. day 14) or the value obtained on the day of planned pump filling (e.g. day 28), whichever is higher. Percentages listed under FUDR Dose refer to percentage of last dose of FUDR administered.

If patient develops a total bilirubin > 50 μmol/L, the pump should be emptied and dexamethasone 25 mg plus heparin 35,000 units and saline 35 ml placed in the pump for 14 days. Once there is no longer evidence of toxicity, the dexamethasone dose should be tapered in increments of 5 mg every 14 days. Tapering will continue unless enzymes increase. FUDR should be permanently discontinued unless the patient responded previously and there is evidence of disease progression (increasing CEA, worsening CT scan, worsening clinical status) while off therapy AND bilirubin has returned to < 26 μmol/L. In this case, FUDR can be restarted as follows: use 25% of the last FUDR dose given with dexamethasone, heparin, and saline in the pump for 7 days. Pump should be emptied after 7 days, and patients given a 3-week rest period. This treatment and treatment schedule should continue as long as bilirubin remains < 26 μmol/L and liver enzyme values do not increase.

Epigastric pain unresponsive to oral H2 blocker use is suggestive of gastroduodenal irritation or ulcer. Severe pain should prompt workup with an upper gastrointestinal endoscopy. Serum amylase should be checked along with the routine blood (screening profile, creatinine, and CBC) in patients with abdominal pain. If an ulcer or gastroduodenitis is documented, therapy should be held for one month to allow healing. If abdominal pain is severe, the pump should be emptied of FUDR until results of workup are available.

If the physician feels the patient cannot tolerate systemic therapy, it can be held for 1-2 weeks. If the patient has elevated liver function tests, systemic therapy can be held and dexamethason can be placed in the pump with heparin saline at the discretion of the Principal Investigator.

Systemic chemotherapy is administered at the patients local hospital some days following pump re-fill (approximately at Day 2-3 and 16-17 of each cycle, with the exception of Day 3, Cycle 1, if patient will receive HAI-FUDR/DEX). The Oncologist at the patients local hospital will decide the type and dosage of the patients systemic therapy.

At Day 1 and Day 15 of each cycle, a brief physical examination is conducted, ECOG score obtained, vital signs are measured, adverse events / adverse device effects and toxicity assessments are made and blood sampling for safety laboratory variables are taken. In addition, weight in kg is measured at Day 1.

Selective Internal Radiation Therapy (SIRT) Selective Internal Radiation Therapy (SIRT) (also named trans-arterial radio-embolization, TARE) is approved in Norway, however, used in a limited degree. Recent research on hepatocellular carcinoma has shown the importance of personalized dosimetry to obtain high tumour radiation dose, while limiting the dose to surrounding liver tissue, yielding improved response and survival (Dosisphere study), it is likely that these findings can be applied also to cholangiocarcinoma.

Radioactive Yttrium 90 (Y-90) labeled particles are injected selectively into the hepatic artery branches that are feeding arteries of the tumour(s). Y-90 is a β-emitter with a half-life of 64 hours and tissue penetration of 2-10 mm. This allows for a high radiation dose in the liver lesion while limiting systemic side effects.

Actual treatment will be preceded by a work-up procedure mapping the tumor(s), and the arterial anatomy and flow. Based on the distribution of 99Tch-MAA injected into the designated feeders on Spect-CT, any lung shunt will be quantified and a treatment plan established in close collaboration with specialists in nuclear medicine and radiation physicists. The aim will be a tumor dose of at least 200 Gy, liver tissue to be ablated should receive at least 150 Gy. The dose to the lung may not exceed 30 Gy, while the dose to non-tumorous liver should be kept below 40 Gy. In the setting of radiation lobectomy, treating tumors while also stimulating liver parenchyma growth, this threshold will be higher. Tumor characteristics such as number, size and hypervascularity all influence dose delivery.

Also underlying liver disease will need to be taken into account during dose planning.

SIRT will be given at Day 1 in Cycle 1. Initiation of systemic chemotherapy will not take place until two weeks post-SIRT, so the first dose of systemic chemotherapy will be given on Cycle 1, Day 15, and then every 2 weeks thereafter. Dexamethasone will be provided preoperatively. Patients who experience significant nausea following SIRT may be treated with p.o. methylprednisolone for up to one week following SIRT.

Patients who have response (RECIST 1.1) on SIRT at 6 months following treatment initiation and then recur or progress in a manner that would have made them eligible for SIRT per inclusion criteria (see above), can be assessed for re-treatment with SIRT.

CT chest, abdomen and pelvis, blood samples for efficacy evaluation (CA 19-9 and CEA), and Quality of Life Questionnaires: EORTC QLQ-C30 and EQ-5D-5L, will be taken every 8 weeks (Day 1 of odd numbered cycles).

Re-assessments for possible resection or treatment modification will be performed every 8 weeks at regional MDT-meetings.

Treatment will continue until death or tumour progression. Resected patients will be followed until death or recurrence or at least for three years if no recurrence.

Data for patients who are amenable to upfront resection and patients with non-liver/non-liver hilum metastatic disease will be retrieved from the UNICORN cohort (a national cohort study in establishment, which will include all cholangiocarcinoma patients independent of treatment) as comparison groups.

Overall risk-benefit conclusion For the applicable group of patients few, if any, treatment options are remaining, and the benefits are thus considered to outweigh the risks

Inclusion Criteria:

1. Intra-hepatic cholangiocarcinoma. Diagnosis confirmed by biopsy, cytology or previous resection.

2. Not amenable for upfront resection. Defined as:

   1. A tumour that is technically not resectable with R0 margins (i.e. where resection will not yield an FLR of sufficient size and function) without reconstruction of portal or liver vein, or artery.
   2. Any multifocality (more than one tumour) irrespective of distance between assumed primary and other lesions
   3. Recurrent tumour following resection
   4. Radiologically or cytology-proven malignant regional lymph nodes

3. Disease confined to the liver or associated with limited, resectable porta hepatis lymph node metastases

4. Radiologically measurable disease with at least one lesion > 2 cm in greatest diameter

5. Physical performance score WHO/ECOG stage 0/1

6. Age > 18 years

7. Assumed ability to tolerate at least one full cycle of GemOx

8. For eligibility to HAI-FUDR/DEX treatment, patients must be willing and able to go to Oslo every fortnight

9. Women of childbearing age and potential must be willing to use highly effective contraception during the study and for a period after the study, as defined in this protocol. Male patients or male patients who have female partners of childbearing age and potential must be willing to use highly effective contraception during the study and for a period after the study, as defined in this protocol. Highly effective methods of birth control are defined as those which result in a low failure rate (i.e. less than 1% per year) when used consistently and correctly.

Exclusion Criteria:

1. Any non-liver malignant deposit (except for resectable, hilar lymph nodes)
2. Serum bilirubin, creatinine or INR outside of normal range
3. Haemoglobin < 7 g/dL and thrombocytes < 75 × 109/L
4. Liver failure (if cirrhosis, Child-Pugh B or C)
5. Clinical evidence of portal hypertension (non-surgically related ascites, gastro-oesophageal varices, portal vein thrombosis)
6. History of peripheral neuropathy
7. More than 70 % of liver consisting of tumour
8. History of other malignancy past three years except localized/early stage cancer that has been adequately resected.
9. Pregnant or lactating women
10. Expected life expectancy less than three months.
11. Inability to comply with study routines or follow-up procedures
12. Inability to read and comprehend Norwegian
13. Arterial anatomy unsuited for SIRT or HAI, respectively
14. Any reason why, in the view of the investigators, the patient should not be included