Repeat liver resection for colorectal metastases.

BACKGROUND: Following resection of colorectal liver metastases (CLMs) up to 75 per cent of patients develop recurrent liver metastases. Although repeat resection remains the only curative therapy, data evaluating the outcome are deficient. This study analysed postoperative morbidity, mortality and independent predictors of survival following repeat resection of CLMs. METHODS: Data on surgical treatment of primary and recurrent CLMs between 1994 and 2010 were collected retrospectively, and compared with those for single hepatic resections carried out during the same period. Independent predictors of survival were evaluated by means of univariable and multivariable Cox regression models. RESULTS: In this interval 1026 primary resections of CLMs were performed and 94 patients underwent repeat CLM excision. Overall postoperative morbidity and mortality rates were low (15·8 and 1·3 per cent respectively), with no statistical difference in patients undergoing repeat surgery (P = 0·072). Compared with single liver resections, overall survival was improved in repeat resections (P = 0·003). Multivariable analysis revealed that size of primary CLM over 50 mm was an independent predictor of survival (hazard ratio (HR) 2·61; P = 0·008). Only major hepatic resection was associated with poorer outcome following repeat surgery (HR 2·62; P = 0·009). International Union Against Cancer stage, number of CLMs, age at surgery and need for intraoperative transfusion had no impact on survival after repeat resection. CONCLUSION: Recurrent CLM surgery is feasible with similar morbidity and mortality rates to those of initial or single CLM resections.

Liver resection of colorectal liver metastases in elderly patients.

BACKGROUND: The percentage of elderly patients with colorectal liver metastases (CLM) has increased. Liver resection remains the only curative therapy; data evaluating the outcome in this age group is limited. Aim of the present study was to determine if postoperative morbidity, mortality, and other independent predictors influence survival in patients ≥ 70 years undergoing liver resection for CLM. METHODS: Clinical data on primary tumor and metastases of 939 patients after liver resection for CLM between 1994 and 2008 were retrospectively collected and subdivided in three age-groups (≥ 70, 40-69, <40). Independent predictors of survival were evaluated with overall and age-specific univariate and multivariate Cox regression models. RESULTS: A total of 939 patients underwent liver resection for CLM, 20.3% aged ≥ 70 years. Overall postoperative mortality and morbidity were 1.08 and 14.82%, revealing no age-related differences. With 5-year survival of 31.8% in the elderly and 37.5% in the mid-age population, age ≥ 70 years was linked with decreased survival (Hazard Ratio [HR] = 1.305; P = 0.0186). Multivariate overall analyses showed size of CLM > 50 mm (HR = 1.376; P = 0.0060), a high amount of transfusion during surgery (HR = 1.676; P = 0.0110), duration of surgery >210 min (HR = 1.241; P = 0.0322), primary UICC (International Union Against Cancer) stage IV (HR = 2.297; P < 0.0001), and performance of repeat resections (HR = 0.652; P = 0.0107) as independent predictors of survival. In the elderly group, effects of UICC IV (HR = 3.260; P = 0.0148) and high numbers of transfusions (HR = 3.647; P = 0.0129) were confirmed; the others did not show statistical significance. CONCLUSIONS: Resection of CLM at older age is feasible with morbidity and mortality rates similar to those in younger patients. Although age ≥ 70 was shown to be associated with poorer overall outcome, reasonable 5-year survival was observed.

The hepatocyte nuclear factor 6 (HNF6) and FOXA2 are key regulators in colorectal liver metastases.

The molecular causes leading to secondary liver malignancies are unknown. Here we report regulation of major hepatic nuclear factors in human colorectal liver metastases and primary colonic cancer. Notably, the genes coding for HNF6, HNF1beta, and C/EBPgamma were selectively regulated in liver metastases. We therefore studied protein expression of regulated transcription factors and found unacetylated HNF6 to be a hallmark of colorectal liver metastases. For its known interaction with HNF6, we investigated expression of FOXA2, which we found to be specifically induced in colorectal liver metastases. By electromobility shift assay, we examined DNA binding of disease regulated transcription factors. Essentially, no HNF6 DNA binding was observed. We also searched for sequence variations in the DNA binding domains of HNF6, but did not identify any mutation. Furthermore, we probed for expression of 28 genes targeted by HNF6. Mostly transcript expression was repressed except for tumor growth. In conclusion, we show HNF6 protein expression to be driven by the hepatic environment. Its expression is not observed in healthy colon or primary colonic cancer. HNF6 DNA binding is selectively abrogated through lack of post-translational modification and interaction with FOXA2. Targeting of FOXA2 and HNF6 may therefore enable mechanism-based therapy for colorectal liver metastases.