Microwave Ablation Versus Resection for Resectable Colorectal Liver Metastases

Overview

This study aims to prove that a strategy of first line local ablation of colorectal liver metastases with microwaves is not inferior to liver resections in terms of survival rates at three years with secondary endpoints being survival at five and ten years, interventional complication rates, length of stay, ablation precision measurements, need for further interventions and health-economic analysis. A cohort of 100 patients treated with CT guided microwave ablation of 1-5 metastases \<31mm in size will be followed and compared with propensity scored matched controls from the Swedish liver surgery registry - Sweliv. The study is a multi-institutional effort by the Hepato Pancreatico Biliary (HPB) units in Stockholm Sweden, Bern Switzerland and Groningen in the Netherlands.

BACKGROUND Surgery Surgical resection of operable colorectal liver metastases is the current gold standard. With advancement in surgical techniques and technology the 3-year survival rate for this group of patients is now in the range of 60% \[1\] where the survival without resection is virtually non-existent. Palliative chemotherapy can prolong survival for 12-24 months, but very rarely much more than this \[2\]. Operability is decided by the general health condition of the patient, the global liver function and tumour associated factors like the number, size and distribution of the liver metastases. It is uncommon to disqualify surgery because of the general health of the patient, but is sometimes a reality with advanced age and severe comorbidities. The combination of size, number and distribution of liver metastases determines the extent of the resection and thereby the size of the liver that will remain after surgery. This is a critical factor as postoperative liver failure is often fatal. With an otherwise healthy liver, approximately 75% of the parenchyma can be resected and the remaining 25% will within a year has regenerated so that the total liver volume is around 80-90% of the original liver \[3\]. This regenerative capacity is used to make complex resections in stages, there by manipulating an initially small liver remnant to a sufficient size for survival. Liver surgery, however, comes with a price. For small local resections and laparoscopic minor hepatectomies the morbidity of surgery is 20-30% with a quarter being serious. For major and complex procedures the morbidity is 40-50% with the same proportion of serious complications requiring further surgical intervention or death \[4\]. With advanced patient support and state of the art surgical technique, the mortality following surgery is small, typically in the range of 1-2%. The patients spend 5-14 days in hospital and are incapacitated for a month or two after open surgery, significantly less after laparoscopic surgery where the patient usually can return home after 2-7 days in hospital and with an early return to normal physical activities within two weeks. Another factor to take into account is the frequency of liver recurrences. After resection, 50-80% of the patients will get new liver metastases \[5\]. With previous major resections the choice of following surgical strategies gets more limited. Ablations Local ablation techniques have evolved from local injection of cytotoxic agents like ethanol and acetic acid to thermal ablations with freezing of the tissue or heating with radiofrequency energy, laser beams or microwave energy. Ablations can also be done by destroying cell membranes with electricity in short intense bursts, the irreversible electroporation technique, causing localised cell death with very little heat thereby preserving collagen structures and other surrounding proteins, which will result in patent vessels, bile-ducts and nerve sheaths \[6\]. The ablative procedures have two critical points; the correct localization of the energy and the correct amount of energy. Ablation in the wrong place will decrease the efficacy as will an incorrect size of the ablation. Targeting of lesions has improved with computer assisted targeting and fusion of radiological modalities so that as much information as possible can be used in the placement of the energy source. Percutaneous ultrasound targeting has until recently been the most common means of placing ablation devices in liver lesions. Local recurrence rates of 20-30% are typically reported with this approach. With more recent fused and computer assisted techniques the local recurrence rate is often stated to be around 5-10% \[7\]. The main limitation with ablative techniques is the size of the lesions. Many studies have shown a good result with lesion less than 30mm in diameter and increasing recurrence rates with larger lesions. A single ablation can usually only be up to 40-50mm in diameter. Larger lesions can be treated with multiple applications but the geometries of ovoid ablations in irregular tumours makes complete ablations difficult. Destroying lesions with ablative techniques spares liver parenchyma and allows for treating patients that are not resectable in situations with multiple lesions engaging all or most of the liver segments. This approach has greatly increased the survival in this group of patients. Sparing unaffected liver parenchyma also increases the flexibility when there are liver recurrences of metastatic disease. This happens in up to 50 percent of patients after local resections and ablations and in a third of patients after major hepatectomies and with a focused ablative approach more functioning liver parenchyma can be saved and the options for treatment of recurrences are wider. Lesions can be targeted with a percutaneous approach using ultrasound fused with magnetic resonance imaging or computer tomography, or in the CT with guidance techniques, or laparoscopically if needed. Very rarely an open approach is needed. This means that patients usually goes home the same day or the day after treatment and has returned to normal physical activities within a week. Morbidity is in the area of 10% with a fourth being serious complications requiring some kind of intervention. Mortality is typically very low and under 1%. Local thermal ablations have been shown to be equal to resections in the treatment in hepatocellular carcinoma when lesions are less than 30mm in size. These data come from retrospective analyses of single centre experiences and are as such of low scientific grade of evidence \[8,9\]. Study design considerations There has not been any well-powered randomized study comparing an ablative strategy with a resective. Doing a non-inferiority randomized study with sufficient power would need up to a thousand patients in each group, thus making this study virtually impossible to conduct. A way to come around this is to use propensity score techniques to compare cases with controls. The investigators have well documented controls in the swedish liver registry allowing for propensity score matching of the data. With this design only 100 patients will have to be treated, as has been computed with a power calculation with survival curves after resective treatment with colorectal liver metastases. Data was extracted from "Livermet survey", an on-line prospective database with more than 25000 hepatectomies for colorectal cancer in 69 countries. Data for a first resection of a single metastasis smaller than 30mm (best case scenario) was extracted and survival curves computed. With a non-inferiority assumption, with a power of 90% and the non-inferiority limit being not worse than 5% reduction in 3-year survival, a sample size of 100 was computed, see figure below. CLINICAL AND ECONOMICAL CONSEQUENCES If the study thesis holds there would be a great impact on the treatment of colorectal liver metastases where approximately half could be treated with a minimal invasive and low-complication ablation strategy, greatly reducing patient morbidity and increasing possibilities for further treatment modalities in case of tumour recurrences in the liver. With much shorter hospital stay and sick-leave there is a great potential for big savings in health-related costs for health service providers. From Engstrand et al. In manuscript. This shows the distribution of 272 liver metastases from colorectal cancer among a population based cohort of 1032 patients that were diagnosed with that primary in the greater Stockholm area in 2008. Approximately half of all patients eligible for resection could potentially be treated with ablation. STUDY DESIGN European multicentre prospective cohort study with propensity score matching for number or tumours, age, gender and response to chemotherapy (no chemo - response/stable disease - progression). Patients with colorectal liver metastases are evaluated at a weekly liver multidisciplinary conference and a treatment strategy is decided. In this decision process patients that are resectable and have tumours of 30mm or less and not more than 5 in number, and deemed as both ablatable and resectable, will be offered treatment with an ablative strategy using state of the art targeting and microwave ablation devices. During the study inclusion phase, which can be sequential or with specific weeks during the year (one in two/three/four/five depending on local considerations), all patients that fulfill the inclusion criteria must be entered into the database. All ablations will be done with computer assisted CT targeting (Cascination AG, Bern, Ch) allowing for precise image documentation and evaluation of tumour, ablation device placement and resulting ablation zones. Repeat ablation is allowed and a new multidisciplinary team conference (MDT) discussion is mandated with loss of disease control. Patient characteristics and procedure related parameters will be recorded and outcome measured as overall survival at three years using survival analysis. Secondary outcomes will be tumour state (disease free interval), complications, local recurrence rates, accuracy in placement of ablative device, analyses of ablation volumes, length of stay and further need of treatment with ablations or resections. Treatment with chemotherapy will be as decided at the MDT and with the same protocol that is used with neoadjuvant or adjuvant treatment with surgical resection. Patients will give their informed consent before entering the study. Patient data will be stored in a secure password-protected server with Red-cap software belonging to Karolinska Institutet at Danderyd Hospital. Procedural data will be stored together with image series downloaded directly from the ablation aiming device to Artorg Centre at the University of Bern, responsible for volumetric and precisional data analyses in the study. Patient recruitment will primarily be from the greater Stockholm area. All treatment of swedish patients will be at Danderyd Hospital, Stockholm Sweden. A second centre will be at Inselspital in Bern, Switzerland, and a third in Groningen in the Netherlands. With an optimal inclusion rate the study will have recruited the necessary patients within one year. TIME FRAME September 1: finalization of protocol and submission to ethics board November 2015: Start inclusion November 2017?: Inclusion complete 2018: Publication of intervention specific data like workflow, precision and ablation volumes. 2020: Three year follow-up finished and files closed for main study 2021: Publication of results 2023: Evaluation and publication of 5-year survival data 2028: Evaluation and publication of 10-year survival data. Database finally closed.