Trial Information

A Phase I-II Study to Evaluate, in Patients Operated for a Hepatic Resection of Colorectal Liver Metastases, a High Intensity Focused Ultrasound (HIFU) Procedure: Feasibility, Safety and Accuracy.

Justification and interest of the study:

Colorectal liver metastases (CLM) are a major public health issue (16 000 deaths each year
in France). To date, complete resection of CLM offers the only potential curative approach.
However, only a minority of patients (10 to 20%) is candidates. 5-year survival rates after
liver resection are ranged from 20 to 35%. Consequently, alternative locally targeted
therapies (radiofrequency, cryotherapy, laser, microwave) have been developed and evaluated.
Their use presents many limitations: invasive procedures requiring the intraparenchymal
introduction of a probe, impossibility of an accurate monitoring in real-time, small size
destruction, time-consuming, high recurrence rate. The development of a non-invasive and
more accurate technique combined to surgery is required.

High Intensity Focused Ultrasound (HIFU) is a new approach, which enables to generate
irreversible cell death through coagulative necrosis in a few seconds. There's no cooling
effect of the perfusion because of the shortness of the phenomenon.

The surgical team of the Centre Leon Berard, in collaboration with the INSERM U556 one, has
undertaken a research program on CLM treatment by HIFU. A new and very powerful device,
without the previously named limitations, has been developed. Preclinical studies have
revealed the interest, the feasibility and the safety of this process. These results enable
considering preceding the program with an early clinical study.

Experimental design:

Prospective, monocentric phase I/II study evaluating a surgical medical device (SMD).

HIFU SMD will be evaluated in clinical conditions, in patients undergoing hepatectomy for
colorectal liver metastases. Several HIFU "shootings" (the number will be function of the
advancement of the study) will be performed on the liver witch will be resected

Objectives and main assessment criteria:

1. st part: Phase I - on the healthy liver to remove. Feasibility 1, Safety 2, Tolerance
3. Success = 1 Ability to perform shootings, supplementary duration of intervention <
30 minutes; 2 Asepsis, absence of lesion of nearly tissues; 3 Preservation of
hemodynamic and respiratory constants.

2. nd part: Phase IIa - on the healthy liver to remove, distant then close to the vascular
structures of the liver.

Ability of targeting shootings. Success = distance from the epicentre of the HIFU
lesion to the mark previously positioned in the liver ≤ 5mm.

3. rd part: Phase IIb - on metastases to remove. Ability, by summation of HIFU lesions, to
generate " macro-lesions " including a metastasis and insuring a sufficient safety
margin in the healthy tissue.

Success = macro-lesion generated in negative margins.

Number of patients:

1. st part: The inclusion from 2 to 6 patients is required. The next step (phase II) will
depend on:

- 1st analysis after 2 patients:

- If < 2 failures: continuation,

- If 2 failures: cessation of the study.

- 2nd analysis after 6 patients:

- If < 3 failures: continuation (phase II),

- If ≥ 3 failures: cessation.

2. nd part: Three patients will be successively included in each of the 2 landings
(appendix 1):

- If no failure: continuation (2nd landing / Phase IIb),

- If ≤ 2 failures: + 3 patients,

- If > 2 failures on 6 patients: cessation,

- If ≤ 2 failures on 6 patients: continuation (2nd landing / phase IIb),

- If 3 failures: cessation.

3. rd part:

- p0 = 0.70 = upper limit of the success rate resulting in the inefficacy of HIFU
multi-shootings.

- p1 = 0.90 = lower limit of the success rate resulting in the efficacy of HIFU
shootings.

With a 5% risk of type I error and a 80% power, 28 metastases have to be targeted by HIFU in
order to conclude in one-sided situation. Taking into account that an average of 1.5
metastases will be targeted in each patient, 20 patients will be enrolled. 24 observed
successes will allow concluding in the reject of H0 to accept H1.