Liver resection and ablation for squamous cell carcinoma liver metastases.

BACKGROUND: Limited evidence exists to guide the management of patients with liver metastases from squamous cell carcinoma (SCC). The aim of this retrospective multicentre cohort study was to describe patterns of disease recurrence after liver resection/ablation for SCC liver metastases and factors associated with recurrence-free survival (RFS) and overall survival (OS). METHOD: Members of the European-African Hepato-Pancreato-Biliary Association were invited to include all consecutive patients undergoing liver resection/ablation for SCC liver metastases between 2002 and 2019. Patient, tumour and perioperative characteristics were analysed with regard to RFS and OS. RESULTS: Among the 102 patients included from 24 European centres, 56 patients had anal cancer, and 46 patients had SCC from other origin. RFS in patients with anal cancer and non-anal cancer was 16 and 9 months, respectively (P = 0.134). A positive resection margin significantly influenced RFS for both anal cancer and non-anal cancer liver metastases (hazard ratio 6.82, 95 per cent c.i. 2.40 to 19.35, for the entire cohort). Median survival duration and 5-year OS rate among patients with anal cancer and non-anal cancer were 50 months and 45 per cent and 21 months and 25 per cent, respectively. For the entire cohort, only non-radical resection was associated with worse overall survival (hazard ratio 3.21, 95 per cent c.i. 1.24 to 8.30). CONCLUSION: Liver resection/ablation of liver metastases from SCC can result in long-term survival. Survival was superior in treated patients with liver metastases from anal versus non-anal cancer. A negative resection margin is paramount for acceptable outcome.

Exploring the Presence of Core Skills for Surgical Practice Through Simulation.

OBJECTIVE: The ability to simulate procedures in silico has transformed surgical training and practice. Today's simulators, designed for the training of a highly specialized set of procedures, also present a powerful scientific tool for understanding the neural control processes that underpin the learning and application of surgical skills. Here, we examined whether 2 simulators designed for training in 2 different surgical domains could be used to examine the extent to which fundamental sensorimotor skills transcend surgical specialty. DESIGN, SETTING & PARTICIPANTS: We used a high-fidelity virtual reality dental simulator and a laparoscopic box simulator to record the performance of 3 different groups. The groups comprised dentists, laparoscopic surgeons, and psychologists (each group n = 19). RESULTS: The results revealed a specialization of performance, with laparoscopic surgeons showing the highest performance on the laparoscopic box simulator, while dentists demonstrated the highest skill levels on the virtual reality dental simulator. Importantly, we also found that a transfer learning effect, with laparoscopic surgeons and dentists showing superior performance to the psychologists on both tasks. CONCLUSIONS: There are core sensorimotor skills that cut across surgical specialty. We propose that the identification of such fundamental skills could lead to improved training provision prior to specialization.

Ablation or Resection for Colorectal Liver Metastases? A Systematic Review of the Literature.

Background: Successful use of ablation for small hepatocellular carcinomas (HCC) has led to interest in the role of ablation for colorectal liver metastases (CRLM). However, there remains a lack of clarity about the use of ablation for colorectal liver metastases (CRLM), specifically its efficacy compared with hepatic resection. Methods: A systematic review of the literature on ablation or resection of colorectal liver metastases was performed using MEDLINE, Cochrane Library, and Embase until December 2018. The aim of this study was to summarize the evidence for ablation vs. resection in the treatment of CRLM. Results: This review identified 1,773 studies of which 18 were eligible for inclusion. In the majority of the studies, overall survival (OS) and disease-free survival (DFS) were significantly higher and local recurrence (LR) rates were significantly lower in the resection groups. On subgroup analysis of solitary CRLM, resection was associated with improved OS, DFS, and reduced LR. Three series assessed the outcome of resection vs. ablation for technically resectable CRLM, and showed improved outcome in the resection group. In fact, there were no studies showing a survival advantage of ablation compared to resection in the treatment of CRLM. Conclusions: Resection remains the "gold standard" in the treatment of CRLM and should not be replaced by ablation at present. This review supports the use of ablation only as an adjunct to resection and as a single treatment option when resection is not safely possible.

Expectations of outcomes in patients with colorectal cancer.

BACKGROUND: Understanding patients' expectations of their treatment is critical to ensure appropriate treatment decisions, and to explore how expectations influence coping, quality of life and well-being. This study aimed to examine these issues related to treatment in patients with colorectal cancer. METHODS: A literature search from January 1946 to September 2016 was performed to identify available data regarding patients' expectations of outcomes following colorectal cancer treatment. A narrative synthesis of the evidence was planned. RESULTS: Of 4337 items initially identified, 20 articles were included in the review. In studies presenting data on overall and short-term survival, patients considerably overestimated prognosis. Patients also had unrealistic expectations of the negative aspects of chemotherapy and stomas. There was marked discordance between patients' and clinicians' expectations regarding chemotherapy, end-of-life care, bowel function and psychosocial outcomes. Level of education was the most consistent factor influencing the accuracy of patients' expectations. CONCLUSION: Patients with colorectal cancer frequently have unrealistic expectations of treatment. Marked disparities exist between patients' and clinicians' expectations of outcomes.

Operating list composition and surgical performance.

BACKGROUND: Recent reviews suggest that the way in which surgeons prepare for a procedure (warm up) can affect performance. Operating lists present a natural experiment to explore this phenomenon. The aim was to use a routinely collected large data set on surgical procedures to understand the relationship between case list order and operative performance. METHOD: Theatre lists involving the 35 procedures performed most frequently by senior surgeons across 38 private hospitals in the UK over 26 months were examined. A linear mixed-effects model and matched analysis were used to estimate the impact of list order and the cost of switching between procedures on a list while controlling for key prognosticators. The influence of procedure method (open versus minimally invasive) and complexity was also explored. RESULTS: The linear mixed-effects model included 255 757 procedures, and the matched analysis 48 632 pairs of procedures. Repeating the same procedure in a list resulted in an overall time saving of 0·98 per cent for each increase in list position. Switching between procedures increased the duration by an average of 6·48 per cent. The overall reduction in operating time from completing the second procedure straight after the first was 6·18 per cent. This pattern of results was consistent across procedure method and complexity. CONCLUSION: There is a robust relationship between operating list composition and surgical performance (indexed by duration of operation). An evidence-based approach to structuring a theatre list could reduce the total operating time.

Impact of a National Controlled Donation After Circulatory Death (DCD) Program on Organ Donation in the United Kingdom: A 10-Year Study.

Organ transplantation is the most successful treatment for some forms of organ failure, yet a lack of organs means many die on the waiting list. In the United Kingdom, the Organ Donation Taskforce was set up to identify barriers to organ donation and in 2008 released its first report (Organ Donation Taskforce Report; ODTR). This study assesses the success since the ODTR and examines the impact of the United Kingdom's controlled donation after circulatory death (DCD) program and the controversies surrounding it. There were 12 864 intended donation after brain death (DBD) or DCD donors from April 2004 to March 2014. When the 5 years preceding the ODTR was compared to the 5 years following, intended DCD donors increased 292% (1187 to 4652), and intended DBD donors increased 11% (3327 to 3698). Organs retrieved per intended DBD donor remained static (3.30 to 3.26), whereas there was a decrease in DCD (1.54 to 0.99) due to a large rise in donors who did not proceed to donation (325 to 2464). The majority of DCD donors who proceeded did so within 30 min from time of withdrawal. Our study suggests further work on converting eligible referrals to organ donation and exploring methods of converting DCD to DBD donors.

A systematic examination of preoperative surgery warm-up routines.

BACKGROUND: Recent evidence indicates that a preoperative warm-up is a potentially useful tool in facilitating performance. But what factors drive such improvements and how should a warm-up be implemented? METHODS: In order to address these issues, we adopted a two-pronged approach: (1) we conducted a systematic review of the literature to identify existing studies utilising preoperative simulation techniques; (2) we performed task analysis to identify the constituent parts of effective warm-ups. We identified five randomised control trials, four randomised cross-over trials and four case series. The majority of these studies reviewed surgical performance following preoperative simulation relative to performance without simulation. RESULTS: Four studies reported outcome measures in real patients and the remainder reported simulated outcome measures. All but one of the studies found that preoperative simulation improves operative outcomes-but this improvement was not found across all measured parameters. While the reviewed studies had a number of methodological issues, the global data indicate that preoperative simulation has substantial potential to improve surgical performance. Analysis of the task characteristics of successful interventions indicated that the majority of these studies employed warm-ups that focused on the visual motor elements of surgery. However, there was no theoretical or empirical basis to inform the design of the intervention in any of these studies. CONCLUSIONS: There is an urgent need for a more rigorous approach to the development of "warm-up" routines if the potential value of preoperative simulation is to be understood and realised. We propose that such interventions need to be grounded in theory and empirical evidence on human motor performance.

Clinical outcomes of left hepatic trisectionectomy for hepatobiliary malignancy.

BACKGROUND: Left hepatic trisectionectomy (LHT) is a challenging major anatomical hepatectomy with a high complication rate and a worldwide experience that remains limited. The aim of this study was to describe changes in surgical practice over time, to analyse the outcomes of patients undergoing LHT for hepatobiliary malignancy, and to identify factors associated with morbidity and mortality. METHODS: A cohort study was undertaken of patients who underwent LHT at a single tertiary hepatobiliary referral centre between January 1993 and March 2013. Univariable and multivariable analysis was used to identify factors associated with short- and long-term outcomes following LHT. RESULT: Some 113 patients underwent LHT for colorectal liver metastasis (57), hilar cholangiocarcinoma (22), intrahepatic cholangiocarcinoma (12) and hepatocellular carcinoma (11); 11 patients had various other indications. Overall morbidity and 90-day mortality rates were 46.0 and 9.7 per cent respectively. Overall 1- and 3-year survival rates were 71.3 and 44.4 per cent respectively. Total hepatic vascular exclusion and intraoperative blood transfusion were independent predictors of postoperative morbidity, whereas blood transfusion was the only factor predictive of in-hospital mortality. Time period analysis revealed a decreasing trend in blood transfusion, duration of hospital stay, and postoperative morbidity and mortality in the last 5 years. CONCLUSION: Morbidity, mortality and long-term survival after LHT support its use in selected patients with a significant tumour burden.

Inconsistent reporting of minimally invasive surgery errors.

INTRODUCTION: Minimally invasive surgery (MIS) is a complex task requiring dexterity and high level cognitive function. Unlike surgical 'never events', potentially important (and frequent) manual or cognitive slips ('technical errors') are underresearched. Little is known about the occurrence of routine errors in MIS, their relationship to patient outcome, and whether they are reported accurately and/or consistently. METHODS: An electronic survey was sent to all members of the Association of Surgeons of Great Britain and Ireland, gathering demographic information, experience and reporting of MIS errors, and a rating of factors affecting error prevalence. RESULTS: Of 249 responses, 203 completed more than 80% of the questions regarding the surgery they had performed in the preceding 12 months. Of these, 47% reported a significant error in their own performance and 75% were aware of a colleague experiencing error. Technical skill, knowledge, situational awareness and decision making were all identified as particularly important for avoiding errors in MIS. Reporting of errors was variable: 15% did not necessarily report an intraoperative error to a patient while 50% did not consistently report at an institutional level. Critically, 12% of surgeons were unaware of the procedure for reporting a technical error and 59% felt guidance is needed. Overall, 40% believed a confidential reporting system would increase their likelihood of reporting an error. CONCLUSION: These data indicate inconsistent reporting of operative errors, and highlight the need to better understand how and why technical errors occur in MIS. A confidential 'no blame' reporting system might help improve patient outcomes and avoid a closed culture that can undermine public confidence.

Cost-utility analysis of operative versus non-operative treatment for colorectal liver metastases.

BACKGROUND: Surgical resection of colorectal liver metastases (CRLMs) is the standard of care when possible, although this strategy has not been compared with non-operative interventions in controlled trials. Although survival outcomes are clear, the cost-effectiveness of surgery is not. This study aimed to estimate the cost-effectiveness of resection for CRLMs compared with non-operative treatment (palliative care including chemotherapy). METHODS: Operative and non-operative cohorts were identified from a prospectively maintained database. Patients in the operative cohort had a minimum of 10 years of follow-up. A model-based cost-utility analysis was conducted to quantify the mean cost and quality-adjusted life-years (QALYs) over a lifetime time horizon. The analysis was conducted from a healthcare provider perspective (UK National Health Service) in a secondary care (hospital) setting. RESULTS: Median survival was 41 and 21 months in the operative and non-operative cohorts respectively (P < 0·001). The operative strategy dominated non-operative treatments, being less costly (€22,200 versus €32,800) and more effective (4·017 versus 1·111 QALYs gained). The results of extensive sensitivity analysis showed that the operative strategy dominated non-operative treatment in every scenario. CONCLUSION: Operative treatment of CRLMs yields greater survival than non-operative treatment, and is both more effective and less costly.

Management of gallbladder dyskinesia: patient outcomes following positive 99mtechnetium (Tc)-labelled hepatic iminodiacetic acid (HIDA) scintigraphy with cholecystokinin (CCK) provocation and laparoscopic cholecystectomy.

AIMS: To evaluate clinical outcomes in patients with typical biliary pain, normal ultrasonic findings, and a positive (99m)technetium (Tc)-labelled hepatic iminodiacetic acid analogue (HIDA) scintigraphy with cholecystokinin (CCK) provocation indicating gallbladder dyskinesia, as per Rome III criteria, undergoing laparoscopic cholecystectomy (LC). METHODS AND MATERIALS: Consecutive patients undergoing LC for gallbladder dyskinesia were identified retrospectively. They were followed up by telephone interview and review of the electronic case records to assess symptom resolution. RESULTS: One hundred consecutive patients (median age 44; 80% female) with abnormal gallbladder ejection fraction (GB-EF <35%) were followed up for a median of 12 months (range 2-80 months). Following LC, 84% reported symptomatic improvement and 52% had no residual pain. Twelve percent had persisting preoperative-type pain of either unchanged or worsening severity. Neither pathological features of chronic cholecystitis (87% of 92 incidences when histology available) nor reproduction of pain on CCK injection were significantly predictive of symptom outcome or pain relief post-LC. CONCLUSION: In one of the largest outcome series of gallbladder dyskinesia patients in the UK with a positive provocation HIDA scintigraphy examination and LC, the present study shows that the test is a useful functional diagnostic tool in the management of patients with typical biliary pain and normal ultrasound, with favourable outcomes following surgery.

Impact of margin status and neoadjuvant chemotherapy on survival, recurrence after liver resection for colorectal liver metastasis.

BACKGROUND: This study was designed to determine the impact of positive margin and neoadjuvant chemotherapy (NAC) on recurrence and survival after resection of colorectal liver metastasis (CRLM). METHODS: Prospective analysis of 1,255 patients undergoing resection of CLRM was undertaken. The impact of NAC, site of recurrence, and survival between R0 and R1 groups was analysed. RESULTS: The R0 and R1 resection rates were 68.9 % (n = 865) and 31.1 % (390). The median OS for R0 group was 2.7 years (95 % CI 2.56-2.85) and R1 group 2.28 years (CI 2.06-2.52; P < 0.001). The median DFS for R0 group was 1.52 years (CI 1.38-1.66) and R1 group 1.04 years (CI 0.94-1.19; P < 0.001). The intrahepatic recurrence was higher in R1 group 132 (33.8 %) versus 142 (16.4 %) [P = 0.0001]. A total of 103 (11.9 %) patients in R0 group underwent redo liver resection for recurrence compared with 66 (16.9 %) patients in R1 group (P = 0.016). NAC did not impact recurrence rate (57.8 % vs. 61.5 %, P = 0.187) and redo liver surgery between R0 and R1 groups (13 % vs. 17 %, P = 0.092). Within the R1 group, the intrahepatic recurrence rates were similar with and without NAC (33.9 % vs. 33.7 %, P = 0.669). However, DFS was longer in the no chemotherapy group than the chemotherapy group. CONCLUSIONS: R1 resections increase the likelihood of recurrence in the liver and redo liver surgery. NAC does not seem to improve survival in margin positive patients or have an impact on recurrence or reduce need for redo liver surgery for recurrence. In patients with R1 resection, neoadjuvant chemotherapy may have adverse outcome on disease free survival.

Outcome after liver resection in patients presenting with simultaneous hepatopulmonary colorectal metastases.

BACKGROUND: The most common sites of metastasis from colorectal cancer (CRC) are hepatic and pulmonary; they can present simultaneously (hepatic and pulmonary metastases) or sequentially (hepatic then pulmonary metastases, or vice versa). Simultaneous disease may be aggressive, and thus may be approached with caution by the clinician. The aim of this study was to determine the outcomes following hepatic and pulmonary resection for simultaneously presenting metastatic CRC. METHODS: A retrospective review was undertaken of a prospectively maintained database to identify patients presenting with simultaneous hepatopulmonary disease who underwent hepatic resection. Patients' electronic records were used to identify clinicopathological variables. The log rank test was used to determine survival, and χ(2) analysis to determine predictors of failure of intended treatment. RESULTS: Fifty-nine patients were identified and underwent hepatic resection; median survival was 45·4 months and the 5-year survival rate 38 per cent. Twenty-two patients (37 per cent) did not have the intended pulmonary intervention owing to progression or recurrence of disease. Thirty-seven patients who progressed to hepatopulmonary resection had a median survival of 54·2 months (5-year survival rate 43 per cent). Those who had hepatic resection alone had a median survival of 24·0 months (5-year survival rate 30 per cent). Failure to progress to pulmonary resection was predicted by heavy nodal burden of primary colorectal disease and bilobar hepatic metastases. Redo pulmonary surgery following pulmonary recurrence did not confer a survival benefit. CONCLUSION: Selected patients with simultaneous hepatopulmonary CRC metastases should be considered for attempted curative resection, but some patients may not receive the intended treatment owing to progression of pulmonary disease after hepatic resection.

Performance of prognostic scores in predicting long-term outcome following resection of colorectal liver metastases.

BACKGROUND: Ten-year survival appears to define cure following resection of colorectal liver metastases (CRLMs). Various scores exist to predict outcome at 5 years. This study applied several scores to a patient cohort with 10 years of actual follow-up to assess their performance beyond 5 years. METHODS: The study included consecutive patients who underwent liver resection at a single institution between 1992 and 2001. The ability of eight prognostic scoring systems to predict disease-free (DFS) and disease-specific (DSS) survival was analysed using the C-statistic. RESULTS: Among 286 patients, the 1-, 3-, 5- and 10-year actual DSS rates were 86.6, 58.3, 39.5 and 24.5 per cent respectively. Seventy patients underwent 105 further resections for recurrent disease, of which 84.8 per cent were within 5 years of follow-up. Analysis of C-statistics showed only one score--the Rees postoperative index--to be a significant predictor of DFS and DSS at all time points. The remaining scores performed less well, and regularly showed no significant improvement in predictive accuracy over what would be expected by chance alone. No score yielded a C-statistic in excess of 0.8 at any time point. CONCLUSION: Although available risk scores can predict DFS and DSS, none does so with sufficient discriminatory accuracy to identify all episodes of recurrent disease. A non-negligible proportion of patients develop recurrent disease beyond 5 years of follow-up and so surveillance beyond this point may be advantageous.

Nodular regenerative hyperplasia (NRH) complicating oxaliplatin chemotherapy in patients undergoing resection of colorectal liver metastases.

INTRODUCTION: Sinusoidal obstructive syndrome (SOS) is well associated with the use oxaliplatin-based chemotherapy, and represents a spectrum of hepatotoxicity, with nodular regenerative hyperplasia (NRH) representing the most significant degree of injury. The aim of this study was to determine the prevalence of NRH in patients undergoing resection of colorectal liver metastases (CRLM) and to determine its impact on outcome. METHODS: From January 2000 to December 2010, some 978 first primary liver resections were performed for CRLM. A prospectively maintained database was analysed to identify all patients with evidence of NRH in the non-tumour portion of their histopathology specimens. Clinical data of these patients was reviewed and outcomes assessed. RESULTS: Five patients exhibited NRH (four males, one female) with a median age of 69 years (range: 35-74). Three patients presented with synchronous hepatic metastases, and two with metachronous lesions. All received at least 6 cycles of oxaliplatin as either adjuvant or neo-adjuvant chemotherapy. Only one patient developed a post-operative complication namely transient hepatic failure that required a 4-day stay in the intensive care unit. The median hospital stay was 6 days (range: 6-14 days). There were no 90-day mortalities. One patient is alive and disease free at 55 months, the remaining 4 died of recurrent disease between 37 and 70 months following diagnosis of their primary tumours. CONCLUSIONS: NRH is not an uncommon finding amongst patients with SOS with all patients having received oxaliplatin-based chemotherapy. Data on outcome would suggest no increased morbidity and mortality associated with the presence of NRH.

Variation in referral practice for patients with colorectal cancer liver metastases.

BACKGROUND: Half of patients with colorectal cancer develop liver metastases. There remains great variability between hospitals in rates of liver resection for colorectal cancer liver metastases (CLM). This study aimed to determine how many patients with potentially resectable CLM are not seen by specialist liver surgeons. METHODS: Patients presenting with new CLM in a cancer network consisting of a tertiary centre and seven attached hospitals were studied prospectively over 12 months. Data were collected retrospectively for patients who did not have a complete data set. Outcomes for patients referred to the liver tertiary centre were collated. The radiology of tumours deemed inoperable by the local colorectal specialist teams was reviewed by specialist liver surgeons and radiologists. RESULTS: In total, 631 patients with CLM were assessed. Prospective data were complete for 241 patients, and 64 (26.6 per cent) of these were referred to the specialist liver team for consideration of resection. No decision was documented for 16 patients (6.6 per cent). Of those not referred, 30 (18.6 per cent) were deemed unfit or refused and 131 (81.4 per cent) were thought inoperable. Referral rates varied between hospitals (13-43.6 per cent). Of 131 patients deemed fit but inoperable by the colorectal specialist teams, 38 (29.0 per cent) were deemed operable and 20 (15.3 per cent) had equivocal imaging when assessed retrospectively by liver specialists. In total, 142 of the 631 patients were referred to liver specialists for consideration of treatments, and 107 (75.4 per cent) treated with curative intent. CONCLUSION: A considerable number of patients with potentially resectable CLM are not assessed by specialist liver teams. Improved referral rates could greatly improve resection rates for CLM, which may improve outcomes for patients with colorectal cancer.

Fatty liver disease as a predictor of local recurrence following resection of colorectal liver metastases.

BACKGROUND: Obesity and tissue adiposity constitute a risk factor for several cancers. Whether tissue adiposity increases the risk of cancer recurrence after curative resection is not clear. The present study analysed the influence of hepatic steatosis on recurrence following resection of colorectal liver metastases. METHODS: A prospective cohort of patients who had primary resection of colorectal liver metastases in two major hepatobiliary units between 1987 and 2010 was studied. Hepatic steatosis was assessed in non-cancerous resected liver tissue. Patients were divided into two groups based on the presence of hepatic steatosis. The association between hepatic steatosis and local recurrence was analysed, adjusting for relevant patient, pathological and surgical factors using Cox regression and propensity score case-match analysis. RESULTS: A total of 2715 patients were included. The cumulative local (liver) disease-free survival rate was significantly better in the group without steatosis (hazard ratio (HR) 1·32, 95 per cent confidence interval 1·16 to 1·51; P < 0·001). On multivariable analysis, hepatic steatosis was an independent risk factor for local liver recurrence (HR 1·28, 1·11 to 1·47; P = 0·005). After one-to-one matching of cases (steatotic, 902) with controls (non-steatotic, 902), local (liver) disease-free survival remained significantly better in the group without steatosis (HR 1·27, 1·09 to 1·48; P = 0·002). Patients with steatosis had a greater risk of developing postoperative liver failure (P = 0·001). CONCLUSION: Hepatic steatosis was an independent predictor of local hepatic recurrence following resection with curative intent of colorectal liver metastases.

Repeat hepatic resection for colorectal liver metastases.

BACKGROUND: Some 75-80 per cent of patients undergoing liver resection for colorectal liver metastases develop intrahepatic recurrence. A significant number of these can be considered for repeat liver surgery. This study examined the outcomes of repeat liver resection for the treatment of recurrent colorectal metastases confined to the liver. METHODS: Patients who underwent repeat liver resection in a single tertiary referral hepatobiliary centre were identified from a database. Clinicopathological variables were analysed to assess factors predictive of survival. RESULTS: A total of 195 patients underwent repeat resection between 1993 and 2010. Median age was 63 years, and the median interval between first and repeat resection was 13·8 months. Thirty-three patients (16·9 per cent) underwent completion hemihepatectomy or extended hemihepatectomy and the remainder had non-anatomical or segmental resection. The 30-day mortality rate was 1·5 per cent, and the overall 30-day morbidity rate was 20·0 per cent. Overall 1-, 3- and 5-year survival rates were 91·2, 44·3 and 29·4 per cent respectively. Tumour size 5 cm or greater was the only independent predictor of overall survival (relative risk 1·71, 95 per cent confidence interval 1·08 to 2·70; P = 0·021). Neoadjuvant chemotherapy before resection, perioperative blood transfusion, bilobar disease, R1 resection margin and multiple metastases were among factors that did not significantly influence survival. CONCLUSION: Repeat hepatic resection remains the only curative option for patients presenting with recurrent colorectal liver metastases.