Stereostactic microwave ablation as an alternative to surgical resection for colorectal cancer liver metastates

Abstract

Colorectal cancer (CRC) implies a substantial global burden of disease with a relevant impact on the general population and on healthcare systems in terms of morbidity, mortality, quality of life, and costs. A raising CRC incidence emphasises the need to refine screening and prevention strategies, and design optimal algorithms for treatment indication and outcome prediction. Around 25% to 30% of patients with CRC develop colorectal cancer liver metastases (CRLM) at any time point during their disease, with high variation in the disease presentation, severity, chronology and response to treatment. This and the increasing quantity and quality of available therapeutic options, enhance the complexity of defining treatment algorithms and designing feasible studies leading to meaningful results. Considering the high rate of tumour recurrence after initial CRLM treatment with curative intent, applying low-morbidity local treatments enhancing the possibilities of repeat treatments, are gaining importance. As such, thermal ablation (TA) promises high rates of local tumour control and favourable oncological outcomes comparable to the gold-standard surgical resection. Nevertheless, results from highquality prospective comparative studies are missing, hampering the integration of TA as a valid treatment alternative into current guidelines. The aims of the studies included in this thesis were to investigate i) non-inferiority in overall survival (OS), and compare healthcare consumption, costs and treatment-associated morbidity, when treating patients with potentially resectable CRLM with TA versus resection, while applying highlevel navigation technology for stereotactic microwave ablation (SMWA), and ii) the potential of a novel algorithm for computation of 3D quantitative ablation margins (QAM) to enhance treatment success and predict local tumour control after SMWA.

Study I was a population-based analysis comparing 3-year OS after microwave ablation (MWA) versus resection using data from a nationwide Swedish patient registry. After adjusting for factors known to affect the treatment type and OS (confounding by indication) using propensity score (PS) analysis, 3-year OS probabilities were similar in patients treated with MWA (n = 70) (76%, CI 59% to 86%) versus resection (n = 201) (3- year OS 76%, CI 68% to 83%), with a change in the hazard of death of 1.43 (CI 0.77 to 2.65) induced by the treatment type in a multivariable model.

Studies II, III and IV were analyses or sub-analyses of a prospective, multi-centre cohort study (MAVERRIC study), comparing patients with ≤ 5 CRLM ≤ 3cm in size, qualifying for both SMWA and resection and deliberately treated with SMWA (study group), to a contemporary cohort of patients treated with resection, extracted from a Swedish nationwide patient registry (control group). The primary outcome of 3-year OS after a prospective follow-up of 3 years was analysed in Study IV. PS analyses yielded comparable groups with a balanced distribution of baseline characteristics across the study (n = 98) and control (n = 158) cohorts. Three-year OS was non-inferior after SMWA (78%, CI 68% to 85%) versus resection (76% (CI 69% to 82%), with a hazard ratio (HR) of 1.09 (CI 0.69 to 1.51) for the treatment type (SMWA over resection). In the Swedish subgroup of patients included into the MAVERRIC study, a particular inclusion pattern (patients amenable to both ablation and resection treated with SMWA every even week and with resection every odd week) created a quasi-randomised situation, where healthcare related costs and OS were analysed (Study III). Overall costs (all inpatient hospital admissions, outpatient visits, oncological treatments and radiological imaging) from the time of index treatment indication and two years onwards, were significantly reduced in the SMWA versus resection cohorts. Two-year OS and disease-free survival were similar, while hepatic recurrence-free survival was shorter and hepatic re-treatments more frequent after SMWA. Morbidity and length of hospital stay were significantly reduced, and re-treatment significantly more frequent, after MWA / SMWA versus resection, in Studies I, III and IV.

Study II was a secondary outcome-analysis applying a novel QAM metric on a subgroup of patients treated with SMWA within the MAVERRIC study. 3D-QAM was retrospectively computed to 65 CRLM treated with SMWA, and varying definitions investigated in a multivariable model. 3D-QAM was the most relevant factor affecting the occurrence of local recurrence within one year of treatment.

In conclusion, OS at 3 years may be considered similar after SMWA versus resection in patients with potentially resectable small CRLM, with significantly reduced morbidity, time spent in medical facilities and healthcare related costs. In an ageing and more comorbid population, this supports the role of TA as a valid low-morbidity, tissue-sparing treatment alternative, enhancing options for re-treatments in case of hepatic recurrences. This and the potential of innovative technology to enhance safety, efficacy and reproducibility of results, might aid decision-making when designing individualised treatment algorithms for patients with CRLM.