Advanced image analytics predicting clinical outcomes in patients with colorectal liver metastases: A systematic review of the literature.

BACKGROUND: To better select patients with colorectal liver metastases (CRLM) for an optimal selection of treatment strategy (i.e. local, systemic or combined treatment) new prognostic models are warranted. In the last decade, radiomics has emerged as a field to create predictive models based on imaging features. This systematic review aims to investigate the current state and potential of radiomics to predict clinical outcomes in patients with CRLM. METHODS: A comprehensive literature search was conducted in the electronic databases of PubMed, Embase, and Cochrane Library, according to PRISMA guidelines. Original studies reporting on radiomics predicting clinical outcome in patients diagnosed with CRLM were included. Clinical outcomes were defined as response to systemic treatment, recurrence of disease, and survival (overall, progression-free, disease-free). Primary outcome was the predictive performance of radiomics. A narrative synthesis of the results was made. Methodological quality was assessed using the radiomics quality score. RESULTS: In 11 out of 14 included studies, radiomics was predictive for response to treatment, recurrence of disease, survival, or a combination of outcomes. Combining clinical parameters and radiomic features in multivariate modelling often improved the predictive performance. Different types of individual features were found prognostic. Noticeable were the contrary levels of heterogeneous and homogeneous features in patients with good response. The methodological quality as assessed by the radiomics quality score varied considerably between studies. CONCLUSION: Radiomics appears a promising non-invasive method to predict clinical outcome and improve personalized decision-making in patients with CRLM. However, results were contradictory and difficult to compare. Standardized prospective studies are warranted to establish the added value of radiomics in patients with CRLM.

Hepatic vascular inflow occlusion is associated with reduced disease free survival following resection of colorectal liver metastases.

BACKGROUND: Hepatic vascular inflow occlusion (VIO) can be applied during resection of colorectal liver metastases (CRLM) to control intra-operative blood loss, but has been linked to accelerated growth of micrometastases in experimental models. This study aimed to investigate the effects of hepatic VIO on disease-free and overall survival (DFS and OS) in patients following resection for CRLM. METHODS: All patients who underwent liver resection for CRLM between January 2006 and September 2015 at our center were analyzed. Hepatic VIO was performed if deemed indicated by the operating surgeon and severe ischemia was defined as ≥20 min continuous or ≥45 min cumulative intermittent VIO. Cox regression analysis was performed to identify predictive factors for DFS and OS. RESULTS: A total of 208 patients underwent liver resection for CRLM. VIO was performed in 64 procedures (31%), and fulfilled the definition of severe ischemia in 40 patients. Patients with severe ischemia had inferior DFS (5-year DFS 32% vs. 11%, P < 0.01), and inferior OS (5-year OS 37% vs. 64%, P < 0.01). At multivariate analysis, a high clinical risk score (Hazard ratio (HR) 1.60 (1.08-2.36)) and severe ischemia (HR 1.89 (1.21-2.97)) were independent predictors of worse DFS. Severe ischemia was not an independent predictor of OS. CONCLUSION: The present cohort study suggests that prolonged hepatic VIO during liver resection for CRLM was associated with reduced DFS. A patient-tailored approach seems advisable although larger studies should confirm these findings.

Regional and inter-hospital differences in the utilisation of liver surgery for patients with synchronous colorectal liver metastases in the Netherlands.

BACKGROUND: The objective of this study was to map referral patterns in patients with synchronous colorectal liver metastases (SCLM) and to investigate if type, volume and location of the hospital of diagnosis are associated with whether or not patients underwent liver resection. METHODS: This population-based study includes all patients diagnosed with SCLM between 2008 and 2012, based on the Netherlands Cancer Registry. To study inter-hospital variation, the proportion of patients undergoing liver surgery was calculated per hospital of diagnosis. Multivariable multilevel logistic regression analysis was used to investigate the association between hospital characteristics and liver resection. RESULTS: Of 10,520 patients with SCLM, 12% (n = 1259) underwent liver surgery. Of these patients, 58% (n = 733) were referred to another hospital to undergo liver surgery. In 53% of the patients (n = 647), liver resection was performed in a university hospital, in 39% (n = 482) in a dedicated liver centre and in 8% (n = 102) in a general hospital. There was a large inter-hospital variation in the proportion of patients undergoing liver resection (2-26%). In a multilevel logistic regression model, the odds of undergoing liver surgery were higher when patients were diagnosed in hospitals where liver surgery was performed compared with the general hospitals (dedicated liver centre: odds ratio 1.36 [95% confidence intervals 1.08-1.70], university hospital: odds ratio 1.69 [95% confidence intervals 1.22-2.34]). CONCLUSION: There is a large inter-hospital and inter-regional variation in the utilisation of liver resection. Patients diagnosed with SCLM in expert centres had a higher chance of undergoing liver resection.

[Local approval procedures act as a brake on RCTs]. / Lokale toestemmingsprocedures zetten een rem op RCT's.

OBJECTIVE: Large multicentre randomised controlled trials (RCTs) in the Netherlands are increasingly being impeded by major differences between local approval procedures. However, no national agenda exists as yet to improve this situation. The existence of major local differences in processing time and documentation required has been reported previously but little is known about the costs incurred and whether or not specific certifications and research contracts are mandatory. The current study evaluated these aspects of local procedures for obtaining approval of two oncological multicentre RCTs. DESIGN: Retrospective, descriptive. METHOD: All local procedures for obtaining approval of two randomised clinical trials were evaluated: the CAIRO5 and CHARISMA trials initiated by the Dutch Colorectal Cancer Group (DCCG). We objectified time between approval by the Medical Ethics Review Committee (METC) and final approval by the Board of Directors (RvB), the type and number of documents needed, and costs charged. RESULTS: The median time interval between the approval by the Medical Ethics Review Committee and the approval by the Board of Directors was 90 days (range 4-312). The number of documents required per centre ranged from 6-20. The costs charged ranged from € 0-€ 1750, and amounted to € 8575 for all procedures combined. No costs were charged by the majority of the centres. CONCLUSION: The approval procedures for multicentre clinical trials in the Netherlands demonstrate major differences. Processing times, documentation required and costs are unpredictable; greater uniformity is highly desirable in this context.