Randomized Phase II Trial of Chemoembolization and Sorafenib

Transcatheter arterial chemoembolization (TACE) is the most widely performed procedure for
patients with unresectable HCC. Although TACE can induce tumor necrosis, tumor recurrence and
metastasis is not uncommon and likely due to stimulation of angiogenesis immediately after
TACE. Plasma VEGF levels are significantly elevated following TACE procedures, usually
peaking 24 to 48 hours after treatment.

Sorafenib, a multikinase inhibitor, has been shown to increase survival in patients with
advanced HCC, presumably due to its predominant and strong antiangiogenic activities thereby
preventing tumor growth. However, despite its targeted approach, sorafenib is not without
toxicities; consequently most patients are unable to remain on full dose throughout the
course of their treatment.

Because of sorafenib's antiangiogenic properties, it has been suggested that sorafenib could
be used in combination with TACE to counteract the post-TACE angiogenic release and therefore
prevent or minimize the risk of tumor recurrence. The possible synergy between TACE and
sorafenib has been tested in numerous clinical studies. Although the safety profile has been
clearly established, the efficacy of this combination therapy has yet to be demonstrated.
Several combination methods have been tested, i.e., continuously where sorafenib is
administered before the first TACE and then continuously throughout the planned TACE
treatments, or sequentially where sorafenib is administered after the completion of TACE
(usually 4 days after). Here, the investigators postulate that the sequence of the
combination could have a significant impact on patient outcomes. Specifically, the
investigators hypothesize that the continuous method is superior to that of the sequential
protocol because the presence of sorafenib before the first TACE will preempt the peak of
angiogenesis from TACE. To that end, the investigators propose to measure VEGF levels
serially in order to detect differences between the two methods.

Inclusion Criteria:

- Unresectable hepatocellular carcinoma with liver-predominant disease, or patients with
hepatocellular carcinoma who refuse surgery. No more than 30% of the cohort should
have macrovascular invasion and/or asymptomatic extrahepatic disease. Multifocal HCC
is acceptable, no diffuse HCC with greater than 70% tumor burden.

- Patient is at least 18 years of age.

- ECOG performance status of 0-1.

- Child-Pugh class of A or B (up to 7).

- Adequate end-organ function as manifested by:

- Absolute neutrophil count of ≥ 1500/mm3 and platelets ≥ 50,000/mm3

- Creatinine ≤ 2.0 mg/dL

- AST, ALT < 8X ULN

- Total bilirubin of ≤ 3 mg/dL

- Albumin > 2.0 g/dL

- INR < 2.0

- Leukocyte count > 3000 cells/mm3

- Patients who have been treated with previous hepatic surgery, radiofrequency ablation
(RFA), percutaneous ethanol injection (PEI), or cryoablation are eligible if target
lesion(s) have not been treated and local therapy is completed at least 6 weeks prior
to entry.

- Patients with HBV are eligible if controlled with medication. Patients with HCV must
have completed treatment with sustained viral response before being eligible for
study.

- Patients with asymptomatic HIV infection are not eligible.

- Willingness of male and female subjects, who are not surgically sterile or
post-menopausal, to use reliable methods of birth control for the duration of the
study and for 30 days after the last dose of study medication.

- Patient must have signed informed consent prior to registration on study.

- Resolution of all acute toxic effects of any prior local treatment to CTC adverse
event Grade 1 or 0.

- At least one tumor lesion can be accurately measured in at least one dimension
according to RECIST (Appendix D). The lesion has not been previously treated with
local therapy (such as surgery, radiation therapy, RFA, PEI, or cryoablation) unless
it has shown progression in the interim.

Exclusion Criteria:

- Patients unable to swallow oral medications.

- Prior embolization, systemic, or radiation therapy for HCC (liver), or patients who
have taken sorafenib for greater than 2 weeks. Patients randomized to the sequential
arm who started taking sorafenib within 2 weeks of the planned TACE procedure will
have to stop sorafenib for 2 weeks before undergoing TACE. This will prevent a
possible confounding effect of sorafenib on patients randomized to the sequential arm.

- Tumor burden in the liver exceeding 70%.

- Complete occlusion of the main portal venous system. Partial or branch portal vein
occlusion allowed if without reversal of flow.

- Ascites refractory to diuretic therapy (minimal or trace ascites on imaging is
acceptable).

- Previous or concurrent cancer that is distinct in primary site or histology from HCC,
except cervical carcinoma in situ, treated basal cell carcinoma, superficial bladder
tumors (Ta, Tis, and T1). Any cancer curatively treated more than 3 years prior to
entry is permitted.

- Patients with clinically significant gastrointestinal bleeding within 30 days prior to
study entry.

- History of bleeding within the past 4 weeks (unless deemed by PI as clinically
insignificant, as for example, a brief episode of epistaxis).

- Any contraindication to doxorubicin or mitomycin-C administration.

- Evidence of severe or uncontrolled systemic diseases.

- Congestive cardiac failure greater than NYHA class 2 (Appendix E), myocardial
infarction within 6 months, active coronary artery disease, cardiac arrhythmias
requiring anti-arrhythmic therapy other than beta blockers or digoxin, unstable
angina, or laboratory or clinical finding that in the view of the principal
investigator makes it undesirable for the patient to participate in the study.

- Any prior history of hypertensive crisis or hypertensive encephalopathy.

- History of stroke or transient ischemic attack within 6 months prior to study
enrollment.

- Inadequately controlled hypertension (defined as systolic blood pressure of > 150/100
mmHg on antihypertensive medications). Patients with treated hypertension are
eligible. Patients noted to be hypertensive in our clinic, but with documented normal
blood pressure in the office of their referring physician, close follow-up with their
referring providers, and a plan for coordination between the referring physician and
our team for management of blood pressure while on therapy, are eligible.

- Significant vascular disease (e.g. aortic aneurysm, aortic dissection, peripheral
vascular disease).

- History of organ allograft

- Presence of grade 2 or higher hepatic encephalopathy.

- Current, recent (within 4 weeks of the first infusion of this study), or planned
participation in an additional experimental drug.

- Evidence of bleeding diathesis or coagulopathy or on warfarin. Note: if a patient has
been on Coumadin for a period of 1 month and has been stable, they may be accepted
into the protocol.

- Presence of clinically evident central nervous system or brain metastases.

- Major surgical procedure, open biopsy, or significant traumatic injury within 28 days
prior to Day 1, or anticipation of need for major surgical procedure during the course
of the study.

- Pregnant (positive pregnancy test) or lactating.

- Inability to comply with the study and/or follow-up procedures.

- Life expectancy of less than 12 weeks.

- Patients with concomitant HIV infection or AIDS-related serious acute or chronic
illness.

- Substance abuse, medical, psychological, or social conditions that may interfere with
the patient's participation in the study or evaluation of study results.

- Active clinically serious infections (greater than Grade 2)

- Patients with obvious and/or symptomatic extrahepatic disease. Findings of uncertain
significance, such as lung lesions less than 10 mm in diameter or enlarged periportal
lymph nodes will not exclude patients.

- Any contraindication to an arterial procedure such as impaired clotting tests
(platelet count < 50,000/mm3 or prothrombin activity < 50%).

- Any contraindication to sorafenib administration