The Interventional Centre

and

Surgical Department,

Rikshospitalet

Cryoablation of liver tissue

Monitoring, techniques and tumour effects

Tom Mala

Thesis submitted for the degree of Doctor of Medicine

July 7th, 2003

FACULTY OF MEDICINE

UNIVERSITY OF OSLO

Acknowledgements…………………………………………..… IV

List of papers…………………………………………………… V

Abbrevations…………………………………………………… VI

Introduction…………………………………………………….. 1

Cryoablation……………………………………………………. 2

Mechanisms of tissue injury.……………………………….. 3

“Thermal history” of ablated tissue………………………… 4

The cryolesion………………………………………………. 5

Monitoring of the freezing process…………………………. 6

Clinical studies……………………………………………… 9

Interventional access….…………………………………….. 10

Summary of introduction…………………………………… 11

Aims of the study………………………………………………. 13

Summary of papers…………………………………………….. 14

Discussion……………………………………………………… 19

General comments..………………………………………… 19

Comments to experimental studies…………………………. 19

Comments to clinical studies……………………………….. 22

Summary……………………………………………………. 25

Conclusions…………………………………………………….. 26

Tables ………………………………………………………….. 27

Errata/comments…………………………………………………… 32

Reference list…………………………………………………… 33

Acknowledgements

This study was carried out at the Interventional Centre and the Surgical Department, Rikshospitalet, Oslo, and financially supported by the Research Council of Norway.

The thesis is the result of the skills and contributions of many persons. First, I will express my sincere gratitude to my supervisors Prof.dr.med Odd Søreide, dr.med Ivar Gladhaug and Bjørn Edwin. Prof.dr.med. Odd Søreide initiated and provided financing of the study. His help and encouragement has been invaluable. Specially thanks for the prompt and thorough response and guidance. I deeply appreciate the help, encouragement and thorough mentorship provided by dr.med Ivar Gladhaug. His door has always been open. My sincere thank to Bjørn Edwin for his invaluable role particularly in the clinical part of the thesis. I also thank him for teaching and including me in other fields of surgery.

Thanks to Eigil Samset who played a major role in the early parts of the thesis and dr.scient Laurs Aurdal for his help and enlightening comments. The assistance and encouragement from Lars Frich has been of great help and is highly appreciated.

I am thankful to Prof.dr.med Erik Fosse and the Interventional Centre for financing and working facilities. The enthusiasm of the staff at the Centre has been of great help. I particularly want to thank Terje Tillung, Per Kristian Hol, and dr.med Sumit Roy for assistance in MRI. The help from Helga Teigland and Ann Mari Halstensen in animal treatment, and the surgical assistance from Carina Olofsson, Isabelita Fiksdal, and Linda Nes are also greatly appreciated.

I am in debt to the Surgical Department, dr.med Øystein Mathisen, and Prof.dr.med Anstein Bergan for their help and support in the clinical part of the thesis. I am also thankful to Prof.dr.med Ole Petter Clausen for assistance in histopathologic analyses and to Laura Killingbergtrø for assistance in patient follow-up.
Finally, I will express my sincere appreciation to friends and family. Specially thanks to Terje Mørland for support and helpful advices.

List of papers

I. Samset E, Mala T, Edwin B, Gladhaug I, Søreide O, Fosse E.
Validation of estimated 3D temperature maps during hepatic cryosurgery. Magn Reson Imaging 2001; 19: 715-721.

II. Mala T, Samset E, Aurdal L, Edwin B, Gladhaug I, Søreide O.
Magnetic resonance imaging estimated three-dimensional temperature distribution in liver cryolesions. A study of cryolesion characteristics assumed necessary for tumour ablation. Cryobiology 2001; 43: 268-275.

III. Mala T, Frich L, Aurdal L, Edwin B, Clausen OP, Søreide O, Gladhaug I. Intraoperative contrast-enhanced MR-imaging as predictor of tissue damage during cryoablation of porcine liver. In press, Magn Reson Imaging.

IV. Mala T, Frich L, Aurdal L, Edwin B, Clausen OP, Søreide O, Gladhaug I. Hepatic vascular occlusion increases tissue destruction during cryoablation of porcine liver. Submitted.

V. Mala T, Edwin B, Samset E, Gladhaug I, Hol PK, Fosse E, Mathisen Ø, Bergan A, Søreide O. Magnetic- Resonance- Guided percutaneous cryoablation of hepatic tumours. Eur J Surg 2001; 167: 610-617.

VI. Mala T, Edwin B, Tillung T, Hol PK, Søreide O, Gladhaug I.
Percutaneous cryoablation of colorectal liver metastases: potentiated by two consecutive freeze-thaw cycles. Cryobiology 2003; 46: 99-102.

VII. Mala T, Edwin B, Mathisen Ø, Tillung T, Fosse E, Bergan A, Søreide O, Gladhaug I. Cryoablation of colorectal liver metastases. Minimally invasive tumour control. Submitted.

Abbrevations

MRI Magnetic resonance imaging

MR Magnetic resonance

CT Computer tomography

CEA Carcinoembryonal antigen

RFA Radiofrequency ablation

2D/3D Two/three dimensional

CVP Central venous pressure

HIFU High intensity focused ultrasound

Introduction

The most frequent liver neoplasms considered for surgical treatment are metastases (secondary tumours) from colorectal cancer and hepatocellular carcinoma (primary tumour). Benign solid tumours rarely require surgical intervention. In Norway, the most frequent liver neoplasms evaluated for surgical treatment are colorectal metastases (66). Hepatocellular carcinoma is uncommon with an annual incidence less than 5/100 000 (67,75).

Resection is the only established treatment that may cure patients with malignant liver tumours. The 5 - year survival following resection of colorectal metastases confined to the liver is 25 - 38 % (Table 1). However, only 10 - 15 % of the patients with such metastases are eligible for resection (98). Causes of non-resectability include multiple lesions, distribution and location of tumours, previous liver resections, extrahepatic tumour growth, and general medical status of the patients. Without resection the survival is poor (Table 2). For patients with non-resectable colorectal liver metastases the only established treatment available is palliative chemotherapy. In general, current chemotherapeutic regimes do not cure patients. Survival of patients receiving chemotherapy is median 12 months (73).

Criteria of tumour resectability may vary between centres depending on local traditions and experience in hepatic surgery. New and less documented treatment modalities may affect judgements made in regard to resectability of liver tumours - particularly due to the often less invasive character and apparent simplicity of such procedures. If not evaluated by surgeons experienced in hepatic surgery patients may thus be offered less effective treatment. Long-term outcome of any treatment intended to improve patient survival should be compared to the natural history of the disease in question, ideally in similar groups of untreated and treated patients (123).

Tumour ablation is a treatment modality based on destruction of tumour tissue. Following lethal injury the non-viable tissue is left in situ to be replaced by inflammatory processes and subsequent fibrosis (69). Several techniques have been used for destruction of liver tumours, including cryoablation, radiofrequency ablation (RFA), microwave- and laser ablation, high-intensity focused ultrasound (HIFU), embolisation, and ethanol injection (18,21).

Cryoablation, i.e. the use of low temperatures to induce cellular necrosis, was among the first of the thermal ablative techniques widely used as treatment for liver tumours. Cryotherapy and cryosurgery are other terms used to describe this technique of in situ tumour destruction. In this thesis, cryoablation is explored experimentally in liver tissue and clinically in patients with colorectal liver metastases.

Cryoablation1

Cryoablation, although explored as a treatment modality for liver tumours for several years, was not extensively used until intraoperative ultrasonography became available in the late 1980s (29). Ultrasonography enables real-time imaging of the freezing process within parenchymatous organs.

The most frequent indications for cryoablation of liver tumours include non-resectable metastases from colorectal and neuroendocrine primary tumours, and hepatocellular carcinoma (38,104,113). The technique has also been used to freeze resection margins of uncertain radicality (23,41,111,124).

Theoretically, there may be two indications for performing cryoablation: curation or palliation of disease. Published reports indicate that most centres perform ablation alone or in combination with resection of all established liver tumours in curative intent. Effects of palliative procedures, i.e. incomplete ablation of tumours or ablation of selected tumours only, are less extensively documented. The aim of such procedures could be to improve patient survival, delay or prevent complications caused by progressive tumour growth, or to reduce tumour related symptoms (8).

Knowledge about the mechanisms of tissue destruction induced by low temperatures is necessary to develop adequate techniques for tumour cryoablation. Intraprocedural monitoring is mandatory to allow exact placement of probes used for freezing, to ensure that all target tissue is ablated, and that no damage is made to other structures.

1 Cryo: icy cold (Greek), ablatus: taken/carried away (Latin), (Webster`s encyclopedic unabridged dictionary of the English language, 1989)

Mechanisms of tissue injury

The mechanisms of tissue injury and destruction induced by low temperatures are complex. Several theories have been proposed to describe these mechanisms. Direct (immediate) and vascular (delayed) cellular injury have been described in detail.

Direct cellular injury.

The mechanisms of direct cellular injury depend on the rate of freezing. Fast cooling rates occur centrally within the ice-ball and induce intracellular ice-crystals that are considered lethal to the cells (30). Water molecules are trapped within cells and ice-crystals nucleate. Although the precise mechanisms by which intracellular ice-crystal formation destroy cells are unknown, intracellular ice-crystals are thought to be deleterious to cellular membranes and organelles (30,46,72).

In the periphery of the frozen tissue ice-crystals initially form in the extracellular spaces. Water is withdrawn from the cytoplasm and cells dehydrate due to the hyperosmotic extracellular environment induced by ice-crystal formation (solution effect injury). The cells shrink, membrane damage occurs and the intracellular machinery is affected. This type of injury is associated with slow freezing rates and may not always be lethal to cells (30,46). A study of tissue morphology after freezing of resected normal and malignant livers found that tumour cells dehydrate less for the same rate of freezing than normal cells, indicating different susceptibility for different tissues to the solution effect (10).

Vascular injury

Microvascular damage is an important mechanism of the tissue injury induced by low temperatures (30,46). During freezing ice-crystals form and propagate along the vascular system. At low cooling rates the cells surrounding small vessels dehydrate. Water leaves the cell and freezes in adjacent blood vessels. The diameter of the blood vessels may expand by a factor of more than 2 (92). This expansion is believed to cause damage to the endothelium and consequently thrombosis resulting in ischemia of tissue nutritioned by the damaged vessels (92). A study of rat liver showed that temperatures lower than 0 0C caused complete shut-down of microvascular perfusion and thus irreversible perfusion failure (101). Reperfusion injury is also a suggested mechanism of vascular induced tissue damage (46).

It takes several hours until the vascular effect is established (30). An experiment made to separate the direct effect of freezing and the delayed effect of blood flow deprivation, showed that tumour cells excised immediately after exposure to low temperatures were successfully transplanted, while cells excised two days later did not survive transplantation (30).

Large vessels, on the other hand, seem to tolerate freezing well (25,33). Tumour location close to large vessels can be an indication for cryoablation as such locations may preclude safe or adequate tumour resection.

Other mechanisms of tissue damage induced by low temperatures

Cells exposed to sublethal freezing temperatures can undergo programmed cell death (apoptosis) (43). This may be an important cause of cellular death in the periphery of the cryolesion and, theoretically, promotion of apoptosis in these regions could increase tissue injury (46,132). Apoptosis following freezing of malignant tissue is subject of ongoing research and needs to be further elucidated.

Sensibilisation of the immune system to destroyed tissue is another suggested mechanism of tissue injury following cryoablation (46,47,113). It is supposed that the remaining sublethally damaged tumour load may be destroyed by direct or indirect immunologic mechanisms. The experimental evidence for such mechanisms are sparse, and clinical effects are not established.

“Thermal history” of ablated tissue

Tissue necrosis induced by low temperatures depends on the final tissue temperature, the rate of freezing, the rate of thawing, duration of freezing, and the use of repeated freeze-thaw cycles (34). The term "thermal history" has been used to point out that the damage induced by low temperatures is subject to complex interactions, and caused by the entire time-temperature history experienced by the tissue during freezing (46).

The lethal effects of freezing increase as temperature drops. Attempts have been made to establish a “lethal temperature” at which all tissue is completely destroyed. Freezing to such temperatures in all target tissue would ensure adequate tumour ablation (113). To define such a temperature in vivo and to separate the effects of temperature from other mechanisms of tissue destruction, such as from microvascular damage is difficult.

Temperatures lower than – 40 0C to – 50 0C are consider lethal to cells. Temperatures in this range may ensure intracellular ice-crystal formation (30,32,72). Such temperatures are therefore recommended in all tumour tissue and ideally in a 1 cm zone of normal liver tissue surrounding the tumour. Theoretically, temperature monitoring could be used as indicator of adequacy of ablation. However, cellular susceptibility to low temperatures varies between different types of neoplastic tissue, and tumour cell survival has been reported following freezing to temperatures lower than – 50 0C (30,46,113). The exact value of the “lethal temperature” of malignant liver tissue is thus not agreed upon.

Recrystalization and increased cellular damage seems to be induced by prolonged duration of freezing particularly in regions with temperatures higher than – 40 0C (30). Similarly, experimental studies have shown increased cellular damage by prolonged duration of thawing as compared to hastened thawing. This is probably caused by increased ice-crystal growth (recrystallisation) induced by the slow thaw (30,46). The effect of rate of freezing has been described previously (34).

Repeated freeze-thaw cycles increase tissue necrosis in experimental studies (19,30,37). Few clinical reports have documented such an effect of twin freeze cycles in the treatment of liver tumours. One clinical study, however, found increased levels of serum transaminases following two freeze-thaw cycles as compared to single freeze ablations indicating increased hepatocellular damage (116).

All parts of the freeze-thaw cycle may be of importance during cryosurgery (30,34). Protocols used for cryoablation of liver tumours should therefore be carefully designed to maximise the deleterious effects of freezing in the target region (Table 3). Hepatic vascular inflow occlusion during freezing is claimed to potentiate cryoablation (see later). Few studies exist to document such an effect (19,54,78,79).

The cryolesion

We use the term cryolesion for description of the tissue damage induced by low temperatures and the term ice-ball for the intraoperative frozen region1.