The measurement and clinical impact of hepatic steatosis in patients undergoing resection for colorectal liver metastases

Abstract

Introduction Approximately one third of patients undergoing resection for colorectal liver metastases (CLM) have hepatic steatosis. Steatosis may be associated with worse peri-operative outcome, but its effect on long-term survival is unknown. In the general population, steatosis is associated with excess adiposity and insulin resistance, but predictors for its presence in CLM patients are poorly defined. Magnetic Resonance (MR) imaging modalities appear to be the most reliable non-invasive method of quantifying intra-hepatic fat. Methods Three streams of research: (i) analysis of the international LiverMetSurvey database (n = 16,779) to investigate the effect of clinician-reported steatosis on survival following resection for CLM in two clinical settings - with and without pre-operative chemotherapy; (ii) a prospective clinical study (n = 37) to develop a novel method of digital quantification of histological fat, propose this as the new referent standard, and relate with clinical- and CT-derived anthropometric measures, and circulating insulin-related biomarkers, including Homeostasis Model of Insulin Resistance (HOMA-IR) and adipokine hormones; and (iii) two pilot studies to develop MR imaging protocols (chemical shift and proton MR spectroscopy, 1HMRS) for the pre-operative quantification of intra-hepatic fat.Results (i) Comprehensive analysis of LiverMetSurvey data, using statistical methods to reduce biases and confounding, showed that compared to those with normal hepatic histology, patients with steatosis not receiving pre-operative chemotherapy have equivalent peri-operative mortality, and paradoxically, improved cancer-specific survival (HR 0.81, 95% CI: 0.72-0.91). (ii) Digital quantification of histological fat was established as reliable and reproducible. Multivariate linear regression models showed this new standard was best predicted by the following (non-invasive) biomarkers: HOMA-IR (per doubling = 2.41, P = 0.030) and serum adiponectin (per doubling = 0.197, P = 0.011). (iii) The In-Out ratio on routine pre-operative MR scans correlated with liver fat percentage on digital histology (rho = 0.626), but methods to address multi-level biases are in development. Protocols were developed for 1HMRS, with early readouts.Discussion There is a need to develop MR imaging protocols to accurately quantify intra-hepatic fat in patients with CLM non-invasively, and facilitate reliable and reproducible studies examining relationships between steatosis, chemotherapy and post-resection outcome. Such evidence is required before considering changes in practice, which would have the potential to affect many patients.

Date of Award	1 Aug 2014
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Andrew Renehan (Supervisor)