Abstract Aim Prior studies have concluded that advanced age alone should not be an absolute contraindication to pancreaticoduodenectomy (PD). This study aimed to compare PD outcomes by patient age and quantify these risks. Method Data were extracted from the Recurrence After Whipple’s (RAW) study, a multicentre retrospective cohort study of outcomes of PD performed for pancreatic head malignancy (29 centres in 8 countries, n = 1484). Patients were grouped by their age and the following were compared: median length of stay (LoS), 30-day readmission rate, 90-day mortality rate, unplanned return to theatre rate, provision of nutritional support, five-year recurrence, and five-year survival. Comparisons were made using the Mann-Whitney U test and Fisher’s exact test. Results Concerning patient age, 7.1% were <50 years, 15.7% were 50-59, 36.7% were 60-69, 34.1% were 70-79 and 6.3% were ≥80. 90-day mortality rates were 3.8%, 3.0%, 1.7%, 5.7% and 7.4%, respectively, and five-year survival was 31.1%, 41.2%, 32.5%, 28.5% and 21.9%, respectively. When those aged <70 were compared to those ≥70, median LoS (13 vs 12.5 days, p = 0.8), 30-day readmission (9.6% vs 8.2%, p = 0.4), unplanned return to theatre rate (4.2% vs 6.0%, p = 0.15) and five year recurrence (62.4% vs 58.2%, p = 0.1) were similar but 90-day mortality (2.3% vs 5.8%, p = 0.0006) and five-year survival (34.6% vs 27.5%, p = 0.005) rates were significantly lower in the former. Conclusions Older patients should be considered for curative-intent PD if they have an appropriate performance status but one must consider the slightly higher rates of peri-operative mortality and reduced five-year survival.

Abstract Background Pancreatoduodenectomy (PD) is recommended in fit patients with a resectable pancreatic ductal adenocarcinoma (PDAC) of the pancreatic head. Early rates of recurrence are high and long-term survival is uncommon. This study aimed to investigate the association between histological lymph node ratio (LNR) and long-term survival. Method Data were extracted from the Recurrence After Whipple’s (RAW) study, a multicentre retrospective cohort study of outcomes of PD performed for pancreatic head malignancy (29 centres in 8 countries, n = 1484). Patients with histologically confirmed PDAC were identified, and five-year and overall survival rates were compared by LNR (number of positive resected nodes/total number of resected nodes). Analyses were performed using Fisher’s exact test, the Kaplan-Meier method, and the log-rank test. Results In total, 885 patients (59.6%) were found to have PDAC. Concerning LNR, 412 patients (48.3%) had a LNR <0.2 and 437 (51.5%) had a LNR ≥0.2 (unknown: 36 cases). Five-year survival was significantly higher in the former (34.0% vs 10.5%, p<0.0001). LNR was 0, >0.0-0.2 and >0.2 in 116 (17.5%), 261 (39.5%) and 284 (43.0%) patients, respectively (patients with missing data were excluded). Median survival was 646 (95% confidence interval: 526-766), 532 (483-581) and 438 (395-481) days, respectively (p<0.00001). Conclusions In our multicentre study of PD patients with confirmed PDAC, a histological LNR ≥0.2 was associated with reduced five-year survival. Patients with a LNR of >0.0-0.2 and >0.2 lived a median of 114 and 208 days less than those with a LNR of 0.0.

The intent of treatment in the majority of patients with metastatic colorectal cancer (mCRC) remains palliative, with most unsuitable for potentially radical resection. Approximately 20% of patients with mCRC are potentially suitable for primary resection or metastasis directed therapy including surgical resection, ablation or stereotactic radiotherapy (RT), which may improve their outcomes. We carried out a retrospective analysis of patients with mCRC treated at University College London Hospital (UCLH), evaluating the use of multimodal, potentially curative therapy. All cases undergoing metastectomies were discussed within a regional specialised Hepato-Biliary MDT at the Royal Free Hospital or Thoracic MDT at UCLH. Data on clinicopathological characteristics, multimodal treatments and outcomes of all patients with mCRC treated consecutively from January 2013 to April 2017 were retrospectively collected. Primary procedure was defined as colorectal resection or radical rectal chemoradiotherapy. Overall survival (OS) was defined as time in months from diagnosis to either death or last follow-up date, and was calculated using the Kaplan Meier method. The prognostic value of baseline factors was assessed using the Cox regression model. A total of 130 patients with mCRC (median age: 62 years (range 19-82) were treated during the study period. The majority of primary tumours were left-sided (65%; 85/130), with the liver the most frequent metastatic site (77%; 100/130) and liver-only metastases in 55/130 (42%). Of the study population with available mutational analysis results (81%; 105/130), 52% (55/105) were KRAS WT, 11% (11/99) BRAF mutant, and 5% of those with availability of mismatch repair status (4/75) were deficient. 50% (65/130) had ≥2 lines of systemic chemotherapy, with 35% (46/130) receiving targeted treatment as part of first-line therapy (either VEGF or EGFR directed). 57% (74/130) underwent a primary tumour or metastasis directed procedure. Median OS for all patients was 24.9 months. Median OS for those who underwent any procedure (surgery/ ablation/ RT) was 33.6 months (95% CI 29.9-44.9), compared with 12.5 months for those who did not (95% CI 7.79-18.7). Univariate analysis showed that patients who had any procedure were significantly less likely to die (HR 0.18, 95% CI 0.12-0.29) compared with patients who did not. With increasing numbers of procedures, survival incrementally increased, with median OS 29.2m, 33.2m, and 43.4m respectively for 1x, 2x or ≥3 procedures (log-rank p< 0.0001). Multivariate analysis found that left sidedness, undergoing a primary procedure and metastatic surgery had a positive impact on survival. A large proportion of patients with mCRC undergoing multimodal therapies within a good multi-disciplinary setting demonstrated significant improvements in their survival. Where appropriate, upfront primary resection/ radical RT or secondary metastasis directed therapy, particularly metastasectomy, should be considered to improve survival outcomes.

Background: Pancreaticoduodenectomy is increasingly being offered as treatment for periampullary tumours in patients 80 years and older. Our aim was to evaluate clinical outcomes of pancreaticoduodenectomy in octogenarians from two high-volume tertiary HPB centres. Materials and Methods: A retrospective case-control analysis of a prospectively-maintained database, between January 2008 and December 2017, was performed. Octogenarians who underwent pancreaticoduodenectomy were matched with consecutively-operated younger patients with 1:1 ratio, based on extent of surgery (venous, arterial or additional resection). Pre-operative comorbidities, intra-operative variables, post-operative complications and mortality were compared, with statistical significance at p<0.05. Results: 88 octogenarians (median age 81 years,range:80-90) who underwent pancreaticoduodenectomy (either classical or pylorus-preserving) were compared to similar number of controls (median age 66 years,range:16-79). Gender, ASA (octogenarians median grade 2 [range:1-4] versus median grade 2 [range:1-3] for controls) and performance status (median score 0 [range:0-2] for both groups) were comparable (p=1.00,p=0.18 and p=0.11 respectively). Charlson Comorbidity Index scores were higher overall for octogenarians (median 7 [range:6-10] versus 5 [range2-9],p=0.001). Median LOS was comparable, 15 days (range:5-69) for octogenarians and 12 days (range:5-78) for controls (p=0.12), as was median ICU stay, 2 days (range:0-58) for octogenarians versus 2 days (range:0-25) (p=0.40). Negative resection margin (R0) rates were 64% for octogenarians and 59% for controls (p=0.53), with no significant difference in lymph node status (p=0.53). Overall post-operative complication rate was higher in octogenarians (60% versus 49%); however this was not statistically significant (p=0.17). Similarly, complication rates across the Clavien-Dindo classification categories were not statistically significant. 30-day and 90-day mortality were 2.3% and 8% for octogenarians versus 1.1% and 3.4% for controls (p=1.0 and p=0.33 respectively). Conclusion: Despite higher 90-day mortality, pancreaticoduodenectomy outcomes in octogenarians are comparable to their younger counterparts. Pancreaticoduodenectomy should therefore be offered as a curative surgical option in this cohort, in specialised centres after meticulous pre-operative assessment.

Background: Primary graft dysfunction (PGD) remains a major obstacle after lung transplantation. Large animal models of isolated left lung transplantation are essential to study PGD. However, most studies do not clamp the contralateral native lung in their models to avoid right heart failure. In this study, we developed an innovate approach to selectively assess reperfusion injury in the transplanted lung only and also investigated the impact of pulmonary flow. Materials and Methods: We performed 12 left orthotopic lung transplants (n = 6 9 donor + recipient in low and high flow group). Donor lungs were harvested, stored for 24 h on ice followed by a left lung transplantation in a recipient animal and were observed for 6 h after reperfusion. Invasive catheters for pulmonary artery (PAP) and left atrial pressure (LAP) monitoring were inserted. Cardiac output (CO) and flow to the transplanted left lung were measured with flowmeters (Transonic ) around left and main pulmonary artery (PA). In group 1 (high flow), the right PA was partially clamped to allow 50% of the CO towards the transplanted lung. In group 2, no vessels were clamped. Results: Survival after 6 h was 100% in both groups. In table 1 the data are depicted. The pO2/FiO2 ratio of the transplanted lung was significantly lower and the mPAP was higher (p < 0.05) in the high flow versus low flow group. The wet/dry ratios of the right native lung and the left transplanted lung were not significantly different after 6 h of reperfusion. Conclusions: Single-lung transplantation in a large animal setting is demanding, but feasible. High flow through the pulmonary artery leads to a lower P/F ratio, indicative of more severe PGD. The lack of difference in lung water accumulation may be related to too short reperfusion times. Our innovative approach to control the flow over the transplanted lung and to monitor the function of the right native lung allows to further study the physiology and treatment of PGD. FG002 DEFERIPRONE MARKEDLY ATTENUATES BOTH KIDNEY GRAFT PRESERVATION AND ISCHEMIA/ REPERFUSION INJURY IN A NOVEL AUTOTRANSPLANTATION MODEL IN PIGS

BACKGROUND:The oncological impact of surgical complications has been studied in visceral and pancreatic cancer. AIM:To investigate the impact of complications on tumour recurrence after resections for pancreatic neuroendocrine tumours. METHODS:We have retrospectively analysed 105 consecutive resections performed at the Royal Free London Hospital from 1998 to 2014, and studied the long-term outcome of nil-minor (<3) versus major (≥3) Clavien-Dindo complications (CD) on disease-free (DFS) and overall survival (OS). RESULTS:The series accounted for 41 (39%) pancreaticoduodenectomies, two (1.9%) central, 48 (45.7%) distal pancreatectomies, eight (7.6%) enucleations, four (3.8%) total pancreatectomies. Sixteen (15.2%) were extended to adjacent organs, 13 (12.3%) to minor liver resections. Postoperative complications presented in 43 (40.1%) patients; CD grade 1 or 2 in 23 (21.9%), grades ≥3 in 20 (19%). Among 25 (23.8%) pancreatic fistulas, 14 (13.3%) were grades B or C. Thirty-four (32.4%) patients developed exocrine, and 31 (29.5%) endocrine insufficiency. Seven patients died during a median 27 (0-175) months follow up. Thirty-day mortality was 0.9%. OS was 94.1% at 5 years. Thirty tumours recurred within 11.7 (0.8-141.5) months. DFS was 44% at 5 years. At univariate analysis, high-grade complications were not associated with shorter DFS (p = 0.744). At multivariate analysis, no parameter was independent predictor for DFS or OS. The comparison of nil-minor versus major complications showed no DFS difference (p = 0.253). CONCLUSION:From our series, major complications after P-NETs resection are not associated to different disease recurrence; hence do not require different follow up or adjuvant regimens.

Aims: Radiofrequency ablation(RFA) is an effective therapeutic option for a variety of malignant liver lesions. Early studies comparing outcome from operative and percutaneous techniques have demonstrated no significant difference, but have tended to incorporate small patient cohorts with multiple types of liver tumours, including primary and metastatic. This study aimed to compare open/surgical and radiological/percutaneous RFA techniques in patients with only liver metastases from colorectal cancer(CLRM). Methods: A retrospective review of CLRM case notes, computerised records and prospective departmental databases was performed between June 2008 and October 2014. Results: Thirty-one patients underwent intra-operative RFA at the same time as liver resection during the study period(21M/10F; median age 64 years; median length of stay(LOS) 9 days; median follow-up 30 months. There were 10 post-operative complications(nine Clavien-Dindo grade II/one grade IIIa). Seventy-one patients underwent percutaneous RFA(44M/27F; median age 71 years; median LOS 2 days; median follow-up 28 months). Two patients developed post-procedure complications(both Clavien-Dindo grade I). Upon comparison of open and percutaneous cohorts, there was a significant difference in the rate of local recurrence at the ablation site(3.2%vs.33.8% respectively; p = 0.000); however, rates of complete ablation(87%vs.85%; p = 0.468), alternate site hepatic recurrence(41.9%vs.52.1%; p = 0.349), extrahepatic recurrence(25.8%vs.29.6%; p = 0.701) and overall survival(87%vs.95%; p = 0.171) were similar. Lesion number was also similar(1(1–6) vs. 1(1–3); p = 0.373) but there was a significantly larger mean lesion size in the percutaneous group(14.5mm vs.20.2mm; p = 0.018). Conclusions: These results suggest that open RFA may be a more effective technique as it is associated with a lower local recurrence rate, albeit with no effect upon overall survival. However, the percutaneous group had a significantly larger mean lesion size and this together with the fact that patients undergoing percutaneous procedures are often less fit, or have more advanced disease, may partly explain these results. Further studies with large, well-matched cohorts are required to further investigate these findings.

Introduction Patients with cirrhosis are frequently malnourished and this has a detrimental effect on outcome in terms of complications, survival after liver transplantation and overall liver-related mortality. The detection of malnutrition and its active management is, therefore, pivotal to these patients' well-being and survival. The aim of this study was to design and develop a simple, nutritional screening tool for use in patients with cirrhosis -The Royal Free Nutritional Prioritising Tool (RFH-NPT)—and to validate its use against the Royal Free Hospital Global Assessment (RFH-GA) which is the accepted gold standard for nutritional assessment of cirrhotic patients in the UK. Methods The RFH-NPT was devised and piloted; interobserver performance agreement was excellent. Validation was undertaken in a cohort of patients with cirrhosis on transplant units throughout the UK and Southern Ireland. Patients' nutritional status was determined and categorised, by nursing staff, using the RFH-NPT (completion time 3 min). The results were compared with the categorisation of nutritional status determined by dietitians using the RFH-GA (completion time 45 min). The analysis of descriptive data, cross-tabulation, performance variables, 95% CIs and κ values were calculated using standard methods. κ Values were interpreted according to Altman, 1999. Results The patient population comprised 133 patients with cirrhosis (98 men: 35 women; mean [range] age 56 [23–73] yr). Overall 49 (37 %) patients were classified, using the RFH-GA, as adequately nourished; 46 (35%) as moderately malnourished and 38 (29%) as severely malnourished. The RFH-NPT identified patients who were at high risk for malnutrition with a diagnostic sensitivity of 100% (95% CI 89 to 100) and specificity of 73% (95% CI 63 to 81) (κ value 0.41, 95% CI 29 to 53). Conclusion The RFH-NPT is a simple, quick and validated method for identifying patients with cirrhosis who at high risk for malnutrition. Further multicentre validation is warranted.Abstract PMO-040 Figure 1 Competing interests None declared.

Introduction Patients with cirrhosis are frequently malnourished and this has a detrimental effect on outcome in terms of complications, survival after liver transplantation and overall liver-related mortality. The detection of malnutrition and its active management is, therefore, pivotal to these patients9 well-being and survival. The aim of this study was to design and develop a simple, nutritional screening tool for use in patients with cirrhosis -The Royal Free Nutritional Prioritising Tool (RFH-NPT)—and to validate its use against the Royal Free Hospital Global Assessment (RFH-GA) which is the accepted gold standard for nutritional assessment of cirrhotic patients in the UK. Methods The RFH-NPT was devised and piloted; interobserver performance agreement was excellent. Validation was undertaken in a cohort of patients with cirrhosis on transplant units throughout the UK and Southern Ireland. Patients9 nutritional status was determined and categorised, by nursing staff, using the RFH-NPT (completion time 3 min). The results were compared with the categorisation of nutritional status determined by dietitians using the RFH-GA (completion time 45 min). The analysis of descriptive data, cross-tabulation, performance variables, 95% CIs and κ values were calculated using standard methods. κ Values were interpreted according to Altman, 1999. Results The patient population comprised 133 patients with cirrhosis (98 men: 35 women; mean [range] age 56 [23–73] yr). Overall 49 (37 %) patients were classified, using the RFH-GA, as adequately nourished; 46 (35%) as moderately malnourished and 38 (29%) as severely malnourished. The RFH-NPT identified patients who were at high risk for malnutrition with a diagnostic sensitivity of 100% (95% CI 89 to 100) and specificity of 73% (95% CI 63 to 81) (κ value 0.41, 95% CI 29 to 53). Conclusion The RFH-NPT is a simple, quick and validated method for identifying patients with cirrhosis who at high risk for malnutrition. Further multicentre validation is warranted. Competing interests None declared.

Introduction Patients with cirrhosis are frequently malnourished and this has a detrimental effect on outcome in terms of complications, survival after liver transplantation and overall liver-related mortality. The detection of malnutrition and its active management is, therefore, pivotal to these patients9 well-being and survival. The aim of this study was to design and develop a simple, nutritional screening tool for use in patients with cirrhosis -The Royal Free Nutritional Prioritising Tool (RFH-NPT)—and to validate its use against the Royal Free Hospital Global Assessment (RFH-GA) which is the accepted gold standard for nutritional assessment of cirrhotic patients in the UK. Methods The RFH-NPT was devised and piloted; interobserver performance agreement was excellent. Validation was undertaken in a cohort of patients with cirrhosis on transplant units throughout the UK and Southern Ireland. Patients9 nutritional status was determined and categorised, by nursing staff, using the RFH-NPT (completion time 3 min). The results were compared with the categorisation of nutritional status determined by dietitians using the RFH-GA (completion time 45 min). The analysis of descriptive data, cross-tabulation, performance variables, 95% CIs and κ values were calculated using standard methods. κ Values were interpreted according to Altman, 1999. Results The patient population comprised 133 patients with cirrhosis (98 men: 35 women; mean [range] age 56 [23–73] yr). Overall 49 (37 %) patients were classified, using the RFH-GA, as adequately nourished; 46 (35%) as moderately malnourished and 38 (29%) as severely malnourished. The RFH-NPT identified patients who were at high risk for malnutrition with a diagnostic sensitivity of 100% (95% CI 89 to 100) and specificity of 73% (95% CI 63 to 81) (κ value 0.41, 95% CI 29 to 53). Conclusion The RFH-NPT is a simple, quick and validated method for identifying patients with cirrhosis who at high risk for malnutrition. Further multicentre validation is warranted. Competing interests None declared.

Introduction Transjugular intrahepatic portosystemic shunt (TIPS) is indicated in the management of portal vein thrombosis or stenosis, portal hypertension and for veno-occlusive disease in post-liver transplant (LT) patients. Previous series have reported 1-year mortality rates of 14%–67%. A MELD score >15 at the time of insertion may indicate a poor long term prognosis. We aimed to evaluate the safety of TIPS after LT at a UK tertiary referral centre. Methods We retrospectively analysed data from the Royal Free Hospital TIPS database between 1st January 1991 and the 31st January 2011. All patients who had undergone TIPS following LT were included. Results During the period studied 629 patients received a TIPS. In the same period 1192 liver transplant operations were performed. 10 TIPS were inserted into patients following LT for recurrent cirrhosis with refractory ascites (4), veno-occlusive disease (3) and portal vein thrombosis (3). The original indications for transplantation were PSC (3), PBC (3), Hepatitis C (1), Autoimmune (1), Primary Oxalosis (1) and Acute Liver Failure (1). We noted a median survival of 38 months. Survival at 1 and 5 years was 100% and 60% respectively. The median MELD at the time of TIPS insertion was 12 (range 7–19). No correlation between the MELD score at the time of TIPS insertion and survival was demonstrated (p=0.62). Conclusion These results suggest that TIPS can be performed safely after LT and that survival rates better than those previously reported can be achieved. We suggest TIPS should be used in carefully selected candidates following LT as a definitive treatment for patients not suitable for re-transplant or as a bridge to re-transplantation. The alternative of re-transplantation should always be considered prior to TIPS insertion where indicated. Competing interests None declared. References 1. Saad WEA, et al. Transjugular intrahepatic portosystemic shunts in liver transplant recipients for management of refractory ascites: clinical outcome. J Vasc Interv Radiol 2010;21:218–23. 2. Kim JJ, et al. Transjugular intrahepatic portosystemic shunts in liver transplant recipients. Liver Int 2008;28:240–8. 3. Feyssa E, et al. MELD score less than 15 predicts prolonged survival after transjugular intrahepatic portosystemic shunt for refractory ascites after liver transplantation. Transplantation 2011;91:786–92.

Realistic modelling of mechanical interactions between tissues is an important part of surgical simulation, and may become a valuable asset in surgical computer guidance. Unfortunately, it is also computationally very demanding. Explicit matrix-free FEM solvers have been shown to be a good choice for fast tissue simulation, however little work has been done on contact algorithms for such FEM solvers. This work introduces such an algorithm that is capable of handling both deformable-deformable (soft-tissue interacting with soft-tissue) and deformable-rigid (e.g. soft-tissue interacting with surgical instruments) contacts. The proposed algorithm employs responses computed with a fully matrix-free, virtual node-based version of the model first used by Taylor and Flanagan in PRONTO3D. For contact detection, a bounding-volume hierarchy (BVH) capable of identifying self collisions is introduced. The proposed BVH generation and update strategies comprise novel heuristics to minimise the number of bounding volumes visited in hierarchy update and collision detection. Aside from speed, stability was a major objective in the development of the algorithm, hence a novel method for computation of response forces from C0-continuous normals, and a gradual application of response forces from rate constraints has been devised and incorporated in the scheme. The continuity of the surface normals has advantages particularly in applications such as sliding over irregular surfaces, which occurs, e.g., in simulated breathing. The effectiveness of the scheme is demonstrated on a number of meshes derived from medical image data and artificial test cases.