The most frequently used products in CHE are doxorubicin (36%), cisplatin (31%), and epirubicin (12%). But until recently, there were no obvious reasons to use one product over another. In fact, systemic chemotherapy is considered ineffective in HCC [hepatocellular carcinoma], which does not allow any argument in favour of the product. Moreover, 2 randomised trials comparing the molecules (doxorubicin vs. epirubicin) proved to be negative in terms of survival.

Cytotoxicity of different anticancer agents on HCC cell lines have been compared in order to select the best candidate for CHE. Eleven chemotherapy molecules have been tested, including those more frequently used in CHE. Among them, idarubicin (an anthracycline) proved to be the most effective in vitro by far. The superiority of idarubicin (as opposed to doxorubicin) was noted especially on the SNU-449 line, which is known for its resistance to several chemotherapy agents. The best cytotoxicity of idarubicin can be explained by 2 mechanisms: 1) idarubicin has a better intracellular penetration than the other anthracyclines. This is probably due to its more considerable lipophily, facilitating thus its passage through the membrane made up of a double lipid layer, 2) idarubicin is resistant to the multidrug resistance system (MDR). The MDR mechanism, which is often noted in HCC, consists of membrane pumps transporting the molecule outside the cell. These two particularities could explain a more significant accumulation of idarubicin in the HCC cells, and thus better efficacy. It is interesting to note that orally administered idarubicin (5 mg/day for 21 days) has proved to be less toxic and is effective in HCC. Currently, idarubicin is used to treat leukaemia. Its toxicity profile (especially, haematological and cardiac) is known.

On these grounds, A pilot study has been conducted in order to assess the tolerance and efficacy of lipiodol-based CHE using a 10 mg dose of idarubicin in 21 patients with unresectable HCC. These preliminary data reveal that CHE with idarubicin is effective and less toxic.

Idarubicin can be loaded in microbeads. A phase I study (IDASPHERE) has been conducted on DC Beads® microbeads (300-500µm) loaded with idarubicin (dose increased from 5 to 25 mg). The DLT [dose-limiting toxicity] and MTD [maximum tolerated dose] have been determined in 21 patients using a CRM. The MTD of idarubicin was assessed at 10 mg. In our study, the idarubicin-loaded beads did not give rise to any specific toxicity-related problem. The 10 mg dose is compatible with the known toxicity profile of idarubicin: cumulative cardiotoxicity of doxorubicin is noted from 550 mg/m², whereas that of idarubicin is noted from 93 mg/m². There is thus a 5.9:1 ratio between their cumulative toxicities. The most frequently used dose (and also the weakest one) for the doxorubicin-based CHE is 50 mg. The equivalent of the idarubicin dose would thus be: 50 mg (doxorubicin) / 5.9 (doxorubicin/idarubicin ratio) = approx. 10 mg of idarubicin.

It has been already demonstrated that hepatic extraction of idarubicin is better than those of doxorubicin and daunorubicin in an animal sarcoma model. In this study, AUC 0-48h and AUC 0-72h were 1.35 times higher with idarubicin, proving that its intra-hepatic penetration was 35% higher.

The randomised phase II PRECISION V study compared conventional CHE (cCHE) with CHE by doxorubicin beads (DC Bead®) in patients with HCC. It is currently the largest randomised trial on CHE published. The PRECISION V data can be thus used to compare the other studies in terms of efficacy and tolerance.

To continue our preliminary study and the phase I IDASPHERE study, investigators wish to assess thus the efficacy and confirm the tolerance of idarubicin-loaded beads for the CHE of HCC according to a protocol similar to PRECISION V, as part of a single-arm phase II study.

By using a 2-step Fleming plan (Fleming, 1982) with a unilateral alpha risk of 5% and 90% potency, it is necessary to include 86 assessable patients.

On the 1st step: 43 patients will be included (+/- 2 patients, if non-assessable patient(s)

• If 10 patients or less present an objective response, the trial will be discontinued on grounds of futility (H1 rejected)
• If 18 patients or less present an objective response, the trial will be discontinued on grounds of efficacy (H0 rejected)

If not, we proceed with the 2nd step including 43 additional patients. If 29 patients or more present an objective response, the treatment will be considered as effective (H0 rejected)

Considering a 5% ratio of visual loss or non-assessable patients, 91 patients will be included.

Inclusion Criteria:

• - Histologically diagnosed HCC or HCC diagnosed according to the EASL criteria
• Measurable targets according to the mRECIST v1.1 criterion
• Preserved liver function (in case of Child-Pugh A or B7 cirrhosis)
• Tumour not subject to interventive care (liver transplant, surgical resection or percutaneous destruction)
• BCLC A/B without portal or extra-hepatic invasion
• No prior treatment by chemotherapy, radiotherapy or transarterial embolisation (with or without chemotherapy)
• Age ≥ 18 years
• WHO 0 or 1
• Laboratory test: platelets ≥ 50,000 mm3, N ≥ 1,000/mm3, creatininaemia ≤ 150 µmol/L, PT ≥ 50%
• No heart failure (isotope or ultrasound VEF > 50%)

Exclusion Criteria:

• - Advanced tumour (vascular or extra-hepatic invasion including brain metastasis or diffuse HCC with liver invasion > 50%)
• History of other type of cancer except cancer known to be in remission for more than 5 years (in this case, HCC histological proof is required), or basal-cell carcinoma or in situ cervix uteri cancer properly treated with curative treatment
• Advanced liver disease (Child B8, B9 and C, bilirubinaemia > 3 mg/dL, SGOT and SGPT > 5 x ULN or 250 U/L)
• Previous treatment by idarubicin and/or doxorubicin
• Idarubicin contraindications (cardiopathy with myocardial failure, serious kidney or liver failure, yellow fever vaccine)
• Concurrent disease or uncontrolled severe clinical condition
• Uncontrolled severe infection
• Patient requiring long-term anticoagulant treatment
• Thrombosis of the portal vein or a 3-segment region or more
• Hepatofugal portal venous flow
• Presence of serious atheromatosis
• Presence of collateral vascular ways potentially affecting the normal regions during embolisation
• Presence of arthritis of the hepatic artery branches to be treated
• Presence of arterioportal or arterial subhepatic fistula that cannot be embolised by coils
• Pregnancy or breastfeeding
• Absence of effective contraception (for men and women of childbearing age)
• Patient who cannot be regularly monitored on account of psychological, social, family- or geography-related reasons
• Concomitant participation of a patient in another study