Oral (pre-recorded)
Liver 		  
OL01 Liver: Metastases (pre-recorded) 
Selection of Presentations from Abstract Submissions
OL01-01 Population-Based Study on Practice Variation Regarding Preoperative Systemic Chemotherapy in Patients with Colorectal Liver Metastases and Impact on Short-Term Outcomes
Arthur K.E. Elfrink, Netherlands

A.K.E. Elfrink1,2, N.F.M. Kok3, L.R. van der Werf1, E. Marra1, D.J. Grunhagen4, J.M. Klaase2, for the Dutch Hepato Biliary Audit Group & All Other Authors
1Scientific Bureau, DICA, Netherlands, 2Surgery, UMCG, Netherlands, 3Surgery, Netherlands Cancer Institute, Netherlands, 4Surgery, Erasmus MC, Netherlands

Introduction: Indications for preoperative chemotherapy for colorectal liver metastases (CRLM) vary. Use of preoperative chemotherapy may influence postoperative outcomes. This study assesed variation in use of preoperative chemotherapy for CRLM and related postoperative outcomes in the Netherlands.
Methods: All patients who underwent liver resection for CRLM in the Netherlands between 2014 and 2018 were included from a national auditing database. Case-mix factors contributing to use of preoperative chemotherapy, hospital variation and postoperative outcomes were assessed using multivariable logistic regression. Postoperative outcomes were postoperative complicated course (PCC), 30-day morbidity and 30-day mortality.
Results: In total, 4323 (69.6%) patients were included of whom 1314 (31.1%) patients received preoperative chemotherapy and 3009 patients did not. Patients receiving chemotherapy were younger (mean (+SD) 63.3 (10.2) versus 67.1 (10.3) p< 0.01) and had less comorbidity (Charlson scores 2+ (18% versus 27%, p< 0.01). Unadjusted hospital variation concerning administration of preoperative chemotherapy ranged between 4% and 100%. After adjusting for case-mix factors, three hospitals administered significantly more preoperative chemotherapy than expected and six administered significantly less preoperative chemotherapy than expected. PCC was 12.1%, 30-day morbidity was 8.8% and 30-day mortality was 1.5%. No association between preoperative chemotherapy and PCC (OR 1.22, 0.97 - 1.53, p = 0.09), 30-day morbidity
(OR 1.16, 0.90 - 1.48, p = 0.25) or 30-day mortality (OR 1.18, 0.68 - 2.01, p = 0.55) was observed.
Conclusions: Significant hospital variation in use of preoperative chemotherapy for CRLM was present in the Netherlands. No association between postoperative outcomes and preoperative chemotherapy was observed.
OL01-02 Novel Nomograms to Predict Lymph Node Metastasis and Liver Metastasis in Patients with Early Colon Carcinoma
Yongcong Yan, China

Y. Yan, K. Mao, J. Lin, R. Chen, J. Wang, Z. Xiao
Sun Yat-sen University, China

Background: Lymph node status and liver metastasis (LIM) are important in determining the prognosis of early colon carcinoma. We attempted to develop and validate nomograms to predict lymph node metastasis (LNM) and LIM in patients with early colon carcinoma.
Methods: A total of 32,819 patients who underwent surgery for pT1 or pT2 colon carcinoma were enrolled in the study based on their records in the SEER database. Risk factors for LNM and LIM were assessed based on univariate and multivariate binary logistic regression. The C-index and calibration plots were used to evaluate LNM and LIM model discrimination. The predictive accuracy and clinical values of the nomograms were measured by decision curve analysis. The predictive nomograms were further validated in the internal testing set.
Results: The LNM nomogram, consisting of seven features, achieved the same favorable prediction efficacy as the five-feature LIM nomogram. The calibration curves showed perfect agreement between nomogram predictions and actual observations. The decision curves indicated the clinical usefulness of the prediction nomograms. Receiver operating characteristic curves indicated good discrimination in the training set (area under the curve [AUC] = 0.667, 95% CI=0.661-0.673) and the testing set (AUC=0.658, 95% CI=0.649-0.667) for the LNM nomogram and encouraging performance in the training set (AUC=0.766, 95% CI=0.760-0.771) and the testing set (AUC=0.825, 95% CI=0.818-0.832) for the LIM nomogram.
Conclusion: Novel validated nomograms for patients with early colon carcinoma can effectively predict the individualized risk of LNM and LIM, and this predictive power may help doctors formulate suitable individual treatments.
OL01-03 Extended Indications for Laparoscopic Parenchyma-Sparing Resection of Posterosuperior Liver Segments in Patients with Colorectal Metastases. PSM Analysis of Outcomes in Comparison with Open Hepatectomy
Mikhail Efanov, Russian Federation

M. Efanov1, D. Granov2, R. Alikhanov1, I. Rutkin2, V. Tsvirkun1, I. Kazakov1, A. Vankovich1, A. Koroleva1, D. Kovalenko1
1Moscow Clinical Scientific Center, Russian Federation, 2Russian Research Center of Radiology and Surgical Technologies, Russian Federation

Background: Indications for laparoscopic liver resection (LLR) of posterosuperior segments (RSS) for colorectal liver metastases (CRLM) are commonly restricted to small, superficial lesions not adjacent to large vessels. There is no long-term survival comparative estimation after LRR and open liver resection (OLR). We aimed to compare outcomes after parenchyma-sparing LLR with extended indication and open liver resection (OLR) for CRLM located in PSS.
Methods: Two Russian centers took part in the study. Patients with hemihepatectomy, extrahepatic tumors were excluded. Logistic regression was used for 1:1 propensity score matching (PSM).
Results: PSS were resected in 77 patients. LLR were performed in 51 (66%) patients. Twenty pairs of patients were matched. Demographic and preoperative data are presented in the table. After matching, no differences were found in terms of blood loss, ICU and hospital stay, which were better for LLR group before PSM. No 90-day mortality was observed within both groups. After PSM, no difference was found in 5-year overall survival after LLR and OLR (78% and 63%, respectively) and 4-year disease-free survival (27% in both groups).
Conclusions: Expanding indication for laparoscopic parenchyma-sparing resection of PSS for CRLM are justified in majority of patients if performed in high volume specialized centers expertized in laparoscopic liver surgery.
FactorsBefore matching After matching 
 LLR (n=51)OLR (n=26)PLLR (n=20)OLR (n=20)P
Median size of largest metastatic tumor, mm40 (10-98)46 (15-100)0,38450 (15-98)44 (15-100)0,309
Rate of metastases >50 mm, n (%)17 (33%)10 (39%)0,6569 (45%)7 (35%)0,519
Proximity to the large hepatic vessels, n (%)12 (24%)8 (31%)0,4935 (25%)6 (30%)0,655
Anatomical resection, n (%)22 (43%)9 (35%)0,4717 (35%)8 (40%)0,744
Difficulty index, point7,0 (4,8-11,1)6,1 (2,9-10,7)0,0917,2 (4,8-11,8)6,4 (2,9-10,7)0,189
Positive response to neoadjuvant chemo (RECIST), n (%)27 (53%)16 (62%)0,47313 (65%)12 (60%)0,744
Positive response to adjuvant chemo (RECIST), n (%)36 (71%)18 (69%)0,90215 (75%)14 (70%)0,723
[Demographic data]
OL01-04 Renin-Angiotensin Inhibitor Treatment Attenuates Progression of Colorectal Liver Metastasis
Georgina Riddiough, Australia

G. Riddiough1, B. Tran2, T. Fifis1, K. Walsh1, V. Muralidharan1, E. Vincan2, C. Christophi1, M. Perini1
1Department of Surgery, University of Melbourne, Australia, 2Molecular Oncology, Doherty Institute, Australia

Introduction: Experimental and clinical data demonstrates that liver regeneration following hepatectomy drives tumour progression in the future liver remnant. We have previously shown using a murine model that treatment with the renin-angiotensin inhibitor (RASi) captopril attenuates tumour growth in the non-regenerating liver. This study aimed to assess the efficacy of RASi for attenuating CRLM progression in the regenerating liver post-hepatectomy in vivo and using patient-derived colorectal liver metastasis (CRLM) organoids ex vivo.
Methods: Male CBA mice underwent induction of colorectal liver metastases (CRLM) in conjunction with 70% partial hepactectomy. They were treated with either control or RASi, captopril 250mg/kg via intra-peritoneal injection. Mouse livers were collected for qRT-PCR.
Tumour and liver tissue tissue was obtained from consenting patients undergoing CRLM resection at Austin Health, Melbourne. Fresh human tissue samples were processed to facilitate the growth of organoids in vitro. Cellular proliferation assays (MTT) were used to assess the efficacy of RASi on tumour proliferation.
Results: In the regenerating liver RASi significantly reduced tumour burden (p< 0.01). RASi treatment was associated with a significant down-regulation of c-myc expression in tumour samples (p< 0.05). 10mM captopril significantly attenuated tumouroid proliferation in vitro compared to control (p< 0.001).
Conclusions: Tumour progression in the regenerating liver is significantly diminished by RASi and is accompanied by a significant reduction in c-myc expression in vivo.
Patient-derived CRLM tumoroids mimic in vivo disease and can be used for drug testing. High dose RASi attenuates patient-derived tumoroid proliferation in vitro.
OL01-05 Gut Microbiota Composition in CRC Patients with Liver Metastasis
Yunwei Wei, China

Y. Wei
Harbin Medical University, Harbin, China

Introduction: Liver metastases contribute to the high mortality rates of colorectal cancer (CRC).
Until now, no study has reported the gut microbiota composition in CRC patients with liver metastasis.
Method: We conducted a cross-sectional and case-control study, to analysis the gut microbiota of CRC with synchronous liver metastasis patients at initial diagnosis (sLM) and patients who develop metachronous liver metastasis (mLM) after surgery, compared to patients without liver metastasis (CG). Gut microbiota were analyzed by 16s RNA gene sequencing with mucosal samples.
Result: An analysis of the mucosal microbial diversity with two methodologies (Shannon and Simpson indices) showed that alpha-diversity of the mucosal microbiome was significantly higher in CG group compared with that of sLM or mLM groups.
Principal coordinate analysis (PCoA) on geuns level with Bary-Curtis distance showed a separated gut microbiota of both sLM and mLM from CG.
LEfSe analysis identified 5 predominated taxa between the sLM and CG groups, 12 predominated taxa between the mLM and CG groups on genus level with LAD score>3.0, It was noted that Escherichia_Shigella had higher relative abundance in the Both sLM and mLM groups compared to the CG group (P< 0.05)
Conclusion: CRC patients with synchronous or metachronous liver metastasis showed a different gut microbiota. Whether gut microbiota contribute to liver metastasis need further validation.
[Variance analysis of mucosal samples in mLM and CG、sLM and CG groups.]
OL01-06 The Effect of Adjuvant Chemotherapy after Hepatectomy for Colorectal Liver Metastasis Depends on RAS Mutation Status
Yoshinori Takeda, Japan

Y. Takeda1,2, A. Saiura1,2, R. Yoshioka1, H. Ichida1, T. Mizuno1, Y. Mise1, H. Imamura1, Y. Ono2, Y. Takahashi2
1Department of Hepatobiliary-Pancreatic Surgery, Juntendo University Hospital, Japan, 2Department of Hepatobiliary-Pancreatic Surgery, Cancer Institute Hospital, Japan

Background: Previous randomized controlled trial showed that adjuvant systemic chemotherapy after surgical resection did not prolong the overall survival (OS) of patients with colorectal liver metastasis (CLM), suggesting the selection for administration of adjuvant chemotherapy. Molecular analysis might help to select patients who obtain a survival benefit by adjuvant chemotherapy.
Method: Data from patients who underwent curative liver resection for CLM between 2010 and 2017 at two Japanese high-volume centers with known KRAS mutation status were analyzed retrospectively. Patients were divided into two groups according to their KRAS status and the relationship between adjuvant chemotherapy and OS was investigated.
Result: A total of 467 patients underwent curative resection, of whom 190 (40.1%) patients had mutant KRAS (KRAS-MT). Adjuvant chemotherapy was administrated to 274 (59%) patients.
Among patients with KRAS wild type (KRAS-WT), patients who received adjuvant chemotherapy after surgery had significantly prolonged OS compared to those who underwent surgery alone (5-year OS; 64.2% vs. 45.3%, p = 0.003), while among patients with KRAS-MT, OS were not different between patients who received adjuvant chemotherapy after surgery and those who underwent surgery alone (5-year OS; 49.2% vs. 36.7%, p = 0.146). Multivariate analysis showed adjuvant chemotherapy as prognostic factor among patients with KRAS-WT (Hazard ratio 0.529 [0.358 - 0781], p = 0.001).
Conclusion: Adjuvant chemotherapy after resection of CLM offers more survival benefit in patients with KRAS-WT compared to those with KRAS-MT.
[Overall survival of patients who received adjuvant chemotherapy and those who did not.]
OL01-07 Doubling Time of Residual Tumor Volume for Metastatic Neuroendocrine Tumors Determines Survival Outcomes after Hepatic Resection
Ek Khoon Tan, Singapore

E.K. Tan1,2, M.F. Shaheen2,3, H.S. Chen4, T. Taner2,5, R.L. Smoot6, D.M. Nagorney6
1Department of Hepatopancreatobliiary & Transplant Surgery, Singapore General Hospital, Singapore, 2Division of Transplantation Surgery, Mayo Clinic, United States, 3College of Medicine, King Saud bin Abdulaziz University for Health Sciences, Saudi Arabia, 4Division of Gastroenterologic and General Surgery, Mayo Clinic, United States, 5William J. Von Liebig Center for Transplantation and Clinical Regeneration, Mayo Clinic, United States, 6Division of Hepatobiliary and Pancreas Surgery, Mayo Clinic, United States

Introduction: Estimates of the percentage of resectable tumor to guide debulking surgery for metastatic gastro-entero-pancreatic neuroendocrine tumors (GEP-NET) does not account for tumor biology and absolute hepatic tumor burden. Thus, we studied the growth kinetics of residual metastatic disease after debulking surgery for GEP-NET to determine whether these tumor-specific factors correlated with survival.
Methods: Patients with measurable residual metastatic disease by imaging after debulking surgery for GEP-NET with or without resection of primary tumors between 2000-2017 were studied. Residual tumor volumes were measured over time to determine tumor growth kinetics. The primary outcome of interest was time from debulking surgery to adjuvant therapy or death.
Results: Twenty-eight patients had residual metastatic GEP-NET and half had an estimated ≥90% debulking resection. Median residual tumor volume (Y0) after debulking surgery was 3.0 (IQR 0.45-20.8) ml. Median tumor volume doubling time was 53.9 (IQR 29-159.2) weeks. Grade of tumor differentiation, Ki-67 index, and estimated percentage of debulking were not associated with outcomes. Multi-variate analysis showed that, after adjusting for Y0 and the use of somatostatin-analogues post-operatively, only tumor volume doubling time ≥52 weeks was associated with the time to adjuvant therapy or death (HR 0.17 95%CI 0.04-0.82). Patients with a tumor doubling time ≥52 weeks had improved overall survival at 1, 3, and 5 years when compared to a doubling time < 52 weeks (100.0%, 90.9%, 90.9% vs 92.3%, 68.4%, 68.4%; p=0.034).
Conclusions: Tumor volume doubling time, not the proportion of debulking performed, determines outcomes after debulking surgery for residual metastatic GEP-NET.
[Kaplan-Meier curve of time to adjuvant therapy or death, stratified by doubling time of residual tumor]
OL01-10 Response to Induction Chemotherapy with Antibodies on the Primary Tumor and the Liver Metastases in Synchronous Metastastic CRC Patients
Rebecca Lutz, Austria

R. Lutz1, P. Jonas1, R.A. Padickakudy1, I. Brandl2, M. Klimpfinger2, E. Bonner3, B. Weitmayr3, T. Grünberger1
1Surgery, Medical Center South, HPB Center, Austria, 2Pathology, Medical Center South, Austria, 3Pathology, Rudolfstiftung Hospital, Austria

Introduction: Treatment sequence and therapy combination are not standardized for metastatic colorectal cancer patients who present with synchronous disease. We investigated two different systemic therapy combinations to identify the best response both at the primary site and the liver metastases in potentially resectable patients.
Methods: Between 2014 and 2019 some 103 patients with liver metastases and a primary CRC were neoadjuvantly treated with Folfox/Xelox and either Bevacizumab or an EGFR-inhibitor for a total of 2 months. Thereafter liver resection was performed, followed by primary tumor resection 4 weeks later. The radiological response of both liver metastases and primary were evaluated and compared to pathological response (Rödel and Rubbia Brandt).
Results: Of 103 treated synchronously metastasized patients, 66 patients received Folfox/Xelox alone prior resection of liver metastases followed by primary resection, 63 with antibody-therapy. 24 received an EGFR antibody combined with chemotherapy and 38 Bevacizumab with Folfox/ Xelox. The histological evaluation of the primary tumor demonstrated a better response to the VEGF antibody-therapy, which was more pronounced clinically (Table 1). 96 patients received chemotherapy before liver resection, 87 with antibody-therapy. The radiological response and histological evaluation is shown in Table 1.
Conclusion: A combined induction chemotherapy with a VEGF antibody reveals clinical, radiological and histological better response both at the primary tumor and at the liver metastases compared to an EGFR-therapy and should be preferably used in patients with potential resectable synchronously metastasized colorectal cancer.

Histological Response of the primary tumorHistological Response of the liver metastases
Antibody-Therapyn=63TRG (Rödel)Antibody-Therapyn=87TRG (Rubbia Brandt)
VEGF391,96VEGF592,84
EGFR241,57EGFR283,25
Radiological Response on the liver metastases
Antibody-TherapyCR 1%PR 81%SD 16%PD 2% 
VEGF100%56%75%100% 
EGFR0%44%25%0% 
[Table 1 Histological and radiological response to induction chemotherapy with VEGF or EGFR antibody-therapy]
OL01-11 Histopathologic Growth Patterns of Colorectal Liver Metastases for Predicting Survival after Liver Resection: Data from the OSLO-COMET RCT
Davit Aghayan, Norway

D. Aghayan1, V.J. Dagenborg1, K. Grzyb1, Ĺ.A. Fretland1, A.M. Kazaryan2, S. Yaqub1, K. Flatmark1, B. Edwin1
1Oslo University Hospital, Norway, 2Řstfold Hospital Trust, Norway

Introduction: Based on clinico-pathological factors and biological markers, different score systems have been developed to stratify prognosis of patients with colorectal liver metastases (CRLM). We aimed to investigate the impact of tumor growth patterns (GP) on overall survival (OS) in patients undergoing liver resection for CRLM.
Methods: OSLO-COMET was a randomized controlled trial (NCT01516710) into laparoscopic vs open liver resection for CRLM recruiting 280 patients from Oslo University Hospital, Oslo, Norway from 2012 to 2016. A comprehensive biobank of all resected specimens was generated. Pathological specimens of 90 patients were evaluated with particular attention given to GP (desmoplastic n=43, replacement n=12, pushing n=14, and mixed n=21). Survival data for the comparison of GPs were analyzed with Kaplan-Meier plots, log-rank tests for equality of survival curves, and Cox proportional hazard regression. To identify predictors of survival, univariable and multivariable Cox-regression analyses were performed.
Results: Median follow-up was 85 months (95%CI, 83 - 86) and 5-year OS compiled 51% in this cohort. Patients with desmoplastic and replacement GPs had significantly better OS than patients with pushing and mixed GPs(p=0.01). Divided into two groups (desmoplastic and replacement GPs in Group 1, n=55; pushing and mixed GPs in Group 2, n=35), 5-year OS in Group 1 and Group 2 were 60% and 36%, respectively (p=0.001, HR 0.41). At multivariable analysis including other known clinicopathological prognostic variables, pushing and mixed GPs were independent predictors for poor OS (p=0,028, HR 1.95).
Conclusion: Histopathologic GP is a practical and reproducible factor of estimating prognosis after resection of CRLM.
OL01-12 Finding the Optimal Treatment for Stage IV Rectal Cancer: A Nationwide Comparison of Two Strategies
Myrtle Frederique Krul, Netherlands

M.F. Krul1, J.M. van Rees2, N.F.M. Kok1, E.N.D. Kok1, P.M.H. Nierop2, G.L. Beets1, C. Verhoef2, K.F.D. Kuhlmann1
1Surgical Oncology, Netherlands Cancer Institute, Netherlands, 2Surgical Oncology, Erasmus University Medical Centre, Netherlands

Introduction: The M1-schedule (short-course pelvic radiotherapy (5x5Gy) followed by systemic therapy and subsequent local treatment of tumour sites) was compared to the liver first approach (LFA: systemic therapy, local treatment of liver metastases with subsequent (chemo)radiotherapy and rectum surgery) for patients with stage IV rectal cancer.
Methods: Consecutive patients with stage IV rectal cancer and potentially resectable liver metastases who were treated with the M1-schedule or LFA between 2006 and 2018 were analysed in nine tertiary referral centres in the Netherlands. Survival and oncological outcome were assessed.
Results: Of the 281 patients, 117 of the 147 (79.6%) patients completed the M1-schedule and 94 of the 134 patients (70.1%) completed the LFA (p = 0.068). Most patients failed completion due to disease progression. The 3-year OS was 59.7% (95% CI: 51.6-69.1) and 53.9% (95% CI: 45.9-63.3, p=0.370) and the median PFS was 16.6 and 16.2 months for M1-schedule and LFA, respectively (p=0.753).
Radiologic complete response rates were higher in M1-schedule (10% vs. 4.3%), but did not reach significance (p=0.253). Pathologic complete responses were similar (13% vs. 11.5%, p=0.829). The duration of the M1-schedule was 7 weeks shorter than the LFA (37 vs 45 weeks, p=< 0.001). Surgical and oncological complication rates were comparable in both groups.
Conclusion: Both schedules have similar overall and progression free survival outcomes but the M1-schedule is 7 weeks shorter. Although significance was not reached, radiological complete responses were higher with M1-schedule. This schedule could lead to more rectum sparing treatment for stage IV rectal cancer patients.
OL01-13 Nomogram to Predict the Risk of Recurrence after Curative Liver Resection for Neuroendocrine Liver Metastasis
Diamantis Tsilimigras, United States

J.-X. Xiang1, X.-F. Zhang1, D. Tsilimigras2, A. Guglielmi3, L. Aldrighetti4, R. Fields5, G. Poultsides6, S. Maithel7, T. Pawlik8, The U S Neuroendocrine Tumor Study Group
1The First Affiliated Hospital of Xi'an Jiaotong University, China, 2The Ohio State University Wexner Medical Center, United States, 3University of Verona, Italy, 4Ospedale San Raffaele, Italy, 5Washington University School of Medicine, United States, 6Stanford University, United States, 7Emory University, United States, 8Department of Surgery, The Ohio State University Wexner Medical Center, United States

Introduction: We sought to develop and externally validate a novel nomogram to predict recurrence and long-term prognosisof patients undergoing curative liver resection for neuroendocrine liver metastasis (NELM).
Methods: Patients who underwent curative liver resection for NELM were identified from an international multicenter database between 1980-2015. A nomogram to predict recurrence was developed and validated externally.
Results: Among 279 patients with NELM who underwent resection, multivariable analysis identified primary tumor location(pancreatic vs. others, HR 2.1, 95% CI 1.3-3.4), tumor grade(moderate vs. well, HR 1.9, 95% CI 1.1-3.1; poor vs. well, HR 1.6, 95% CI 0.7-3.3), lymphatic metastasis (HR 2.6, 95% CI 1.4-4.6) and the type of hepatectomy(major resection vs. parenchymal-sparingresection, HR 0.3, 95% CI 0.1-0.6) to be independently associated with recurrence-free survival (RFS). A weighted nomogram was constructed based on the beta-coefficients of the final multivariable model. The nomogram demonstrated a good ability to predict risk of recurrence in both test and external validation cohorts (c-index; test cohort: 0.754; validation cohort: 0.748).Specifically, the calibrated nomogram predicted survival that closely corresponded to actual survival with a good net benefit for most threshold probabilities especially between 20% to 60% in both the development and validation cohorts.
Conclusions: A novel nomogram predicted recurrence-free survival rates for patients with NELM with excellent discrimination and calibration. The proposed nomogram may be helpful to discuss with patients their anticipated prognosis following resection of NELM.
OL01-14 Risk Factors of R1 Resection in Laparoscopic and Open Liver Surgery for Colorectal Liver Metastases: A European Multicentre Study
Andrea Benedetti Cacciaguerra, Italy

A. Benedetti Cacciaguerra1,2, I. Dagher3, D. Fuks4, F. Rotellar5, M. D’Hondt6, R.I. Troisi7, L. Aldrighetti8, B. Edwin9, M. Abu Hilal1
1Poliambulanza Foundation, Italy, 2Polytechnic University of Marche, Italy, 3Antoine Béclčre Hospital, France, 4Université Paris Descartes, France, 5Clinica Universidad de Navarra, Spain, 6Groeninge Hospital, Belgium, 7Ghent University Hospital, Belgium, 8San Raffaele Hospital, Italy, 9Oslo University Hospital, Norway

Objective: To assess the risk factors associated with R1 resection in patients undergoing open (OLS) and laparoscopic liver surgery (LLS) for CRLMs.
Background: The impact of R1 resection on the outcome of patients after liver resection for colorectal liver metastases (CRLMs) has been continuously appraised, but factors affecting R status have not been clearly defined.
Methods: A cohort study of patients who underwent OLS and LLS for CRLMs in nine European referral centres between January 1th 2004 and December 31th 2018 was performed. A multivariate analysis and the receiver operating characteristic (ROC) curves were used to investigate the risk factors for R1 resection.
Results: Overall, 3387 consecutive liver resections for CRLMs were included. OLS was performed in 1792 cases, LLS in 1595; the R1 resection rate was 14% and 14.2% respectively. The risk factors for R1 resection were: the type of resection (non-anatomic: p=0.001, 0.031 and anatomic/non-anatomic: p=< 0.001), the number of nodules (p=0.008, < 0.001) and the size of tumour (p=0.009, 0.007). In the LLS group only, blood loss was associated with higher risk of R1 (p=0.020), while the Pringle's manoeuvre had a protective effect (p=0.032). The predictive size of tumour for R1 resection was >45mm in OLS and >30mm in LLS, while having more than 2 lesions was significative in both groups.
Conclusions: The study describes the risk factors for R1 resection following liver surgery for CRLMs, which may be used to plan better the perioperative strategies to reduce the incidence of R1 resection during OLS and LLS.
OL01-16 Appropriate Surgical Selection in Patients with Breast Cancer Liver Metastases to Optimize Surgical Outcome: A Nomogram Based on a Multi-institutional Experience
Francesca Ratti, Italy

F. Ratti1, M. Serenari2, M. Zanello3, E. Prosperi2, F. Cipriani1, M. Ravaioli2, E. Jovine3, M. Cescon2, L. Aldrighetti1
1Hepatobiliary Surgery Division, IRCCS San Raffaele Hospital, Milano, Italy, 2Sant'Orsola Malpighi University Hospital, Bologna, Italy, 3Maggiore Hospital, Bologna, Italy

Background and aims: Favorable survivals and better outcomes compared with oncologic treatments only have been recently reported in patients with Breast Cancer Liver Metastases (BCLM). hence the need to focus on the factors determining a patient's prognosis following hepatectomy and to develop a nomogram to predict oncological outcome.
Methods: Using a multicentric database including patients who underwent hepatectomy for BCLM at three Italian HPB referral centers between 2000 and 2018, we sought to identify the predictors of survival and develop a clinical tool to predict patient's prognosis after liver resection for BCLM. The predictive ability of the nomogram to predict recurrence was assessed using the c-index.
Results: 154 patients undergoing surgery with curative intent were included in the analysis. Number of liver lesions (HR 1.66;95% CI, 1.01-2.67; p=0.035), tumor size (HR 1.71;95% CI, 1.22-2.97; p=0.019), triple negative status (HR = 1.07; 95% CI, 1.21 - 2.76; p=0.027), presence of extrahepatic metastases (HR=1.36; 95% CI, 1.19 - 2.91, p=0.032), response to preoperative chemotherapy (HR 1.97; 95% CI, 1.56 - 3.12; p=0.028) were the predictors of recurrence free survival. The nomogram demonstrated a good ability to predict risk of recurrence (c-index of 0.741). Patients were stratified by Predicted Recurrence Risk (PRR) groupings: PRR < 20%, PRR 20-40%, PRR 40-80%, PRR >80%, which correlated with disease free-survival at 3-year of 85%, 61%, 34%, and 0%, respectively.
Conclusion: The developed nomogram accurately (c-index >75 %) staged and predicted the prognosis of patients undergoing liver resection for BCLM, providing an accurate tool to select candidates to resection
OL01-17 The Impact of Surgical Approach on Oncological Outcomes and Efficiency of Treatment Sequencing for Synchronous Colorectal Liver Metastasis
Amelie Beauchamp, Canada

A. Beauchamp1, Y. Haddad1, R. Lapointe2, S. Turcotte2, Z. Rong2, R. Letourneau2,3, F. Vandenbroucke-Menu2, E. Simoneau2
1Surgery, University of Montreal, Canada, 2Hepato-Pancreato-Biliary and Liver Transplant Surgery, University of Montreal, Canada, 3Medical Oncology, University of Montreal, Canada

Introduction: Patients with synchronous colorectal liver metastasis (CRLM) can be managed by 3 surgical approaches: classical (primary-first (PF)), liver-first (LF) or simultaneous resection (SR). The effect of these strategies on oncological outcomes and overall efficiency of treatment sequencing is unclear.
Method: A retrospective analysis of prospectively collected data from a single institution was conducted. Perioperative data including return to adjuvant therapy and overall time for completion of intended systemic and surgical therapies were analysed. Long-term outcomes were compared using the log-rank test.
Results: Between 2010 and 2019, 499 patients with synchronous CRLM were resected: 60.3%(n=301) PF, 25.7%(n=128) LF and 14.0%(n=70) SR. There was no difference in baseline characteristics between the 3 groups. Hospital stay was longer in the SR group (median 9 days vs. 7 days for both PF and LF, p< 0.001). More patients underwent major hepatectomies in the LF group (56.3%(n=72) vs. 34.4%(n=24) SR and 44.5%(n=134) PF, p=0.019). Return to intended adjuvant treatment was similar for all groups (63 days (SR) vs. 54 days (LF) vs. 51.5 days (PF), p=0.6774), and total time to complete all curative therapies was significantly lower in the SR group (9.7 months vs. 11.3 months (LF group) vs. 12.1 months (PF group), p< 0.001), regardless of the type of hepatic resection. Overall survival and disease-free survival were similar between all groups (p>0.05).
Conclusions: Efficient treatment sequencing resulting in shorter time to complete all curative therapies such as seen with simultaneous resection in selected patients might improve overall patient experience without compromising oncological outcomes.
OL01-19 Safety and Feasibility of Adjuvant Hepatic Arterial Infusion Pump Chemotherapy in Patients with Resectable Colorectal Liver Metastasis; A Phase II Trial in the Netherlands
Wills Floris Filipe, Netherlands

W.F. Filipe1, F. Buisman1, D.J. Grünhagen1, C. Verhoef1, C. Grootscholten2, N. Kok3, K.F.D. Kuhlmann3, B. Groot Koerkamp1, M.Y.V. Homs4
1Surgery, Erasmus MC, Netherlands, 2Medical Oncology, Netherlands Cancer Institute, Netherlands, 3Surgery, Netherlands Cancer Institute, Netherlands, 4Medical Oncology, Erasmus MC, Netherlands

Introduction: Adjuvant hepatic arterial infusion pump (HAIP) chemotherapy after resection of colorectal liver metastasis (CRLM) was associated with a 2-year increased survival in Memorial Sloan Kettering Cancer Center. Due to the promising results published on HAIP chemotherapy we have started a HAIP program in the Netherlands. We now present our long-term safety and feasibility results.
Methods: A phase II study was performed in 2 centers in The Netherlands. Patients with resectable CRLM without extrahepatic disease were eligible. All patients underwent 6 cycles of adjuvant HAIP chemotherapy with floxuridine. Safety was determined by HAIP chemotherapy-related complications grade III or higher from surgery until 90 days after finishing the last cycle. Feasibility was determined by the number of completed cycles.
Results: Twenty patients were included. The median number of floxuridine cycles completed was 5 and all patients completed at least 3 cycles at full dose. Nine patients (45%) had 14 grade III or higher complications. Two patients (10%) required a reoperation for a malfunctioning or flipped pump. One patient (5%) developed biliary obstruction that resolved with temporary stent placement. One patient (5%) developed hepatic artery thrombosis after 4 cycles with subsequent pancreatitis, portal vein thrombosis, and died from liver failure. None of the patients developed arterial bleeding or dissection, pseudo-aneurysm, extrahepatic perfusion, or pump pocket infection.
Conclusion: Adjuvant HAIP chemotherapy was safely introduced in the Netherlands in a phase II trial. Long-term follow-up of this trial and an ongoing phase III trial are required to determine survival benefit.
OL01-20 Can the Presence of KRAS Mutation Be a Judge to Determine the Resection Type of Colorectal Liver Metastasis?
Munseok Choi, Korea, Republic of

M. Choi, D.H. Han, J.S. Choi, G.H. Choi
Department of Surgery, Yonsei University College of Medicine, Korea, Republic of

Introduction: Colorectal cancer harboring KRAS mutations are known to exhibit a high rate of vascular invasion and hematogenous metastases. Although it is generally accepted that non-anatomical resection (NAR) in colorectal liver metastasis (CRLM) is comparable to safety and efficacy compared to anatomical resection (AR) and doesn't affect the oncologic outcome, there are reports that AR has oncologic benefit in KRAS-mutated CRLM. This study aimed to compare survival outcomes between AR versus NAR resection for CRLM, according to KRAS mutational status.
Methods: 344 patients who underwent hepatic resection of CRLM with known KRAS mutational status were reviewed from 2007 to 2018. Among them, 117 KRAS mutated CRLMs and 227 KRAS wild type CRLMs were compared.
Results: Both KRAS wild type and mutated type, the size of the resected tumor, tended to be larger in the AR group than in the NAR group (p< 0.001, respectively). In the short term surgical outcome, hospital stay was significantly longer in the AR group (15.65 vs. 11.27 days, p < 0.001), and there was no difference between the two groups in EBL, postoperative complications, and 30-day mortality. There was no difference in disease-free survival (DFS) between AR and NAR in both the KRAS wild type and the mutated type (p=0.326, p=0.954, respectively). In addition, there was no difference in intrahepatic DFS between AR and NAR in both the KRAS wild type and the mutated type (p= 0.165, p=0.516, respectively).
Conclusions: The presence of KRAS mutation cannot determine the resection type of colorectal liver metastasis.
OL01-21 Perioperative Radiofrequency Ablation or Microwave Ablation of Colorectal Liver Metastases: A Single Centre Experience
Myrtle Frederique Krul, Netherlands

M.F. Krul1, S.L. Gerritsen2, F.L. Vissers3, E.G. Klompenhouwer4, T.J.M. Ruers1, K.F.D. Kuhlmann5, N.F.M. Kok1
1Surgical Oncology, Netherlands Cancer Institute, Netherlands, 2Surgery, Antonius Ziekenhuis Nieuwegein, Netherlands, 3Surgical Oncology, Amsterdam UMC, University of Amsterdam, Netherlands, 4Radiology, Netherlands Cancer Institute, Netherlands, 5Netherlands Cancer Institute, Netherlands

Introduction: Radiofrequency ablation (RFA) and microwave ablation (MWA) are generally accepted to treat colorectal liver metastases (CRLM) with a curative intent. MWA is considered to have better outcomes than RFA but true comparing data are scarce.
Methods: This single centre retrospective cohort study analysed 121 consecutive patients who underwent open surgery and ultrasound guided ablation between January 2013 and August 2018. The main objective was to compare RFA with MWA ablation outcomes, including local tumor progression per lesion (LTP), complete ablation rates per lesion, overall survival (OS), progression-free survival (PFS) and postoperative complication rates. Logistic regression models were used for univariable and multivariable analysis to identify predictors of LTP.
Results: Forty-two patients underwent RFA of 96 lesions between 2013 and 2015 and 79 patients MWA of 192 lesions between 2015 and 2018. The median number of ablated CRLM was 2 (range 1-12) with a median diameter of 11 mm (range 1-31). Incomplete ablations were observed in 2 CRLM (2.1%) for RFA and 5 CLRM (2.6%) for MWA (p=0.787). LTP after complete ablation within one year occurred in 6 CRLM (6.3%) treated with RFA and in 11 CRLM (5.7%) treated with MWA (p=0.860). Complications after RFA and MWA requiring re-intervention (e.g. abscess and biloma drainage) within 30 days were observed in 2 (4.8%) and 6 (7.6%) patients, respectively (p=0.550). Ablation technique, operator (surgeon or interventional radiologist), size of ablated lesion and neoadjuvant chemotherapy were not associated with LTP.
Conclusion: Perioperative RFA seems equally effective as MWA for CRLM up to 31mm.
OL01-23 Morphology of Tumor-Associated Macrophages Dictates the Prognosis of Patients with Colorectal Liver Metastases
Matteo Donadon, Italy

M. Donadon1, F. Marchesi2, L. Di Tommaso3, C. Soldani1, N. Cortese2, B. Franceschini1, R. Carriero4, A. Mantovani2, G. Torzilli1
1Department of Hepatobiliary and General Surgery, Humanitas Research Hospital, Rozzano, Milan, Italy, 2Department of Immunology and Inflammation, Humanitas Research Hospital, Rozzano, Milan, Italy, 3Department of Pathology, Humanitas Research Hospital, Rozzano, Milan, Italy, 4Department of Bioinformatics, Humanitas Research Hospital, Rozzano, Milan, Italy

Background: In the era of precision medicine there is a need of reliable prognostic markers to cope with the clinical heterogeneity of patients with colorectal liver metastases (CLM) and, for their predominance in metastatic tissues, tumor-associated macrophages (TAMs) emerge as promising candidates. The aim of this study was to test the presence of discrete TAMs in CLM patients, and to test their role on recurrence-free survival (RFS).
Methods: NCT038888638 is a single-center study that examined a cohort of CLM patients that underwent hepatectomy between 2005 and 2017. TAMs cell density, cell area and cell perimeter were quantified in 3 non-contiguous and non-overlapping areas of by means of immuno-reactive area of CD163+ macrophages. The association of TAMs metrics and RFS was tested by using receiver operating characteristics (ROC) curves, and multivariate Cox regression analysis.
Results: 101 patients resected between 2005 and 2017 were considered. Among density (AUC=0.555;95%CI=0.410-0.701;P=0.449), perimeter (AUC=0.526;95%CI=0.383-0.671;P=0.708) and area (AUC=0.791;95%CI=0.572-0.841;P=0.006) of CD163+ TAMs, only the latter was significantly associated with differences in survival time. Small and large TAMs, as defined by using the best cutoff value extrapolated from the ROC curve (area:60.39µm2;Se=0.79;Sp=0.44), were clearly associated with significantly different 5-year RFS rates of 27.8% and 0.2% respectively (P< 0.001). At the multivariate analysis, including TAMs area and several prognostic factors, only TAMs area was found to be independently statistically associated with RFS (HR=3.41;95%CI=1.13-5.43;P=0.001).
Conclusions: Macrophage morphology is associated with CLM patients' prognosis. Quantitative morphometric characterization of TAMs can serve as an easily quantifiable correlate of functional diversity with prognostic significance.
OL01-22 Novel Therapeutic Targets in the Cancer Associated Stroma Define Poor Prognosis Colorectal Cancer and Colorectal Liver Metastases
Kai M Brown, Australia

K.M. Brown1,2,3, A. Xue3, A.J. Gill2,4,5, T.J. Hugh1,2,3
1Upper GI Surgery Unit, Department of Gastrointestinal Surgery, Royal North Shore Hospital, Australia, 2Northern Clinical School, University of Sydney, Australia, 3Cancer Surgery and Metabolism Research Group, Kolling Institute of Medical Research, Australia, 4Department of Anatomical Pathology, Royal North Shore Hospital, Australia, 5Cancer Diagnosis and Pathology Group, Kolling Institute of Medical Research, Australia

Introduction: Cancer associated stroma (CAS) is emerging as a key determinant of metastasis in colorectal cancer (CRC), however very little is known about the CAS in colorectal liver metastases (CRLM). This study aimed to characterise the transcriptome of CAS in CRLM and identify novel stromal biomarkers or targetable pathways of metastasis.
Method: A nested case-control study design from a prospectively maintained database of paired primary CRCs and CRLMs was adopted. Differential gene expression (DGE), followed by pathway enrichment and sparse partial least square discriminant analysis (sPLS-DA) were performed after isolating epithelial tumour and CAS RNA by laser capture microdissection. The transcriptomes of epithelial tumour versus the CAS between case and control CRC as well as paired primary CRCs and CRLMs were compared.
Results: Isolated tumour and CAS from ten primary CRCs, ten paired CRLMs, and ten control stage-matched primary CRCs that had not developed metastases were included (60 RNA samples in total). Median follow up was 62, 63 and 45 months for case primary, control primary and CRLM groups, respectively. DGE and sPLS-DA identified a number of novel targetable stromal genes (shown below) and pathways including phospholipases, BMP signalling and MAPK pathways that defined poor prognosis CRC and the CAS of CRLM.
Conclusion: This study is the first to describe key differences in CAS gene expression between paired primary CRC and CRLM, as well as identifying a number of targetable genes and pathways whose prognostic relevance specifically in the CAS of CRC and CRLM have not been previously described.
[sPLS-DA fitted onto two components, and corresponding ten topmost predictive genes]
OL01-24 Expression Pattern of CXCR4/CXCL12 Is Related to Survival after Hepatic Resection for Colorectal Liver Metastases
Hiroyuki Yoshidome, Japan

H. Yoshidome1, N. Sakai2, H. Yoshitomi2, K. Furukawa2, T. Takayashiki2, S. Kuboki2, S. Takano2, T. Konishi2, M. Ohtsuka2
1Surgery, Oami Municipal Hospital, Japan, 2General Surgery, Chiba University, Japan

Introduction: The chemokine network such as interaction between CXCR4 and CXCL12 plays a role in the induction of organ-specific metastases. The present study evaluated CXCR4/CXCL12 expression pattern in colorectal liver metastases (CRLM) and determined whether the expression patterns affect tumor progression.
Methods: We examined pattern of CXCR4 and CXCL12 immunohistochemistry in 92 CRLM patients. CXCL12/CD133 immunoreactivity was also evaluated. The median follow-up time of these patients was 38 months. Clinicopathological data of these patients were evaluated. Overall survival rates were evaluated by the Kaplan-Meier method. The expression profile of CXCR4 in the colorectal cancer cell line was determined by fluorescence microscopy.
Results: The cytoplasmic CXCR4 expression was greater in 36 patients than that indicated by CXCR4 staining intensity of hepatocytes. CXCL12 was also expressed in hepatocytes surrounding the tumors at high and low levels in 68 (74%) and 24 (26%), respectively. High levels of nuclear CXCR4 expression were observed in 23 patients which significantly correlated with CXCL12 expression in hepatocytes. The nuclear CXCR4 expression in the cancer cell line increased after exposure to CXCL12. The univariate and multivariate analyses demonstrated that high levels of nuclear CXCR4 and the increased CXCL12 expression in hepatocytes were significantly better prognostic factors for overall and hepatic disease-free survival in patients with CRLM.
Conclusion: The CXCR4 expression in CRLM together with the upregulation of CXCL12 in hepatocytes may help to predict the clinical outcomes of patients with CRLM after hepatic resection and to determine whether adjuvant chemotherapy may be required.