Management of perioperative iron deficiency anemia as part of patient blood management in France: A budget impact model-based analysis based on real world data.

OBJECTIVES: Patient Blood Management (PBM) is defined as a patient-centered, systematic, evidence-based approach to improve patient outcomes by managing and preserving a patient's own blood, while promoting patient safety and empowerment. As a corollary, it also reduces the utilization of allogeneic blood components. However, demonstrating cost-effectiveness depends on the health insurance system considered. This analysis aims to estimate the one-year budget impact of PBM in four elective surgical areas, from French National Health Insurance and hospital perspectives. METHODS: A budget impact model was developed to estimate the difference in the cost of care between scenarios with and without PBM. The impact of hematopoiesis optimization (first pillar of PBM) was studied throughout the management of preoperative anemia and iron deficiency in four types of surgeries: orthopedic, cardiac & cardiovascular, vascular & thoracic, and urologic & visceral surgery. Estimation of model's parameters was based on data collected in 10 French hospitals, literature, and on data from the French national medico-administrative database. RESULTS: A total of 980,125 patients were modeled for all four therapeutic areas. Results shows that implementation of a PBM program could generate annual savings up to €1079 M from the French National Health Insurance perspective (€1018 M from the hospital perspective), and the sparing of 181,451 red blood cells units per year. The deterministic sensitivity analysis showed that PBM generates savings for both perspectives in most parameters tested. CONCLUSION: Implementing PBM programs could result in important savings for the health care system in France.

Laparoscopic versus open resection of intrahepatic cholangiocarcinoma: nationwide analysis.

BACKGROUND: The relevance of laparoscopic resection of intrahepatic cholangiocarcinoma (ICC) remains debated. The aim of this study was to compare laparoscopic (LLR) and open (OLR) liver resection for ICC, with specific focus on textbook outcome and lymph node dissection (LND). METHODS: Patients undergoing LLR or OLR for ICC were included from two French, nationwide hepatopancreatobiliary surveys undertaken between 2000 and 2017. Patients with negative margins, and without transfusion, severe complications, prolonged hospital stay, readmission or death were considered to have a textbook outcome. Patients who achieved both a textbook outcome and LND were deemed to have an adjusted textbook outcome. OLR and LLR were compared after propensity score matching. RESULTS: In total, 548 patients with ICC (127 LLR, 421 OLR) were included. Textbook-outcome and LND completion rates were 22.1 and 48.2 per cent respectively. LLR was independently associated with a decreased rate of LND (odds ratio 0.37, 95 per cent c.i. 0.20 to 0.69). After matching, 109 patients remained in each group. LLR was associated with a decreased rate of transfusion (7.3 versus 21.1 per cent; P = 0.001) and shorter hospital stay (median 7 versus 14 days; P = 0.001), but lower rate of LND (33.9 versus 73.4 per cent; P = 0.001). Patients who underwent LLR had lower rate of adjusted TO completion than patients who had OLR (6.5 versus 17.4 per cent; P = 0.012). CONCLUSION: The laparoscopic approach did not substantially improve quality of care of patients with resectable ICC.

Results of laparoscopic intragastric surgery.

BACKGROUND: Technological and operative advancements have allowed laparoscopic intragastric surgery (LIGS) to be applied in the treatment of superficial gastric or submucosal lesions. The aim of this study was to evaluate short- and long-term outcomes following LIGS. METHODS: From 2000 to 2013, 25 LIGSs were performed for superficial gastric lesions. Clinical records were reviewed retrospectively for peri-operative course and long-term outcomes with particular attention to the oncological follow-up for patients with malignant lesions. RESULTS: Nineteen (76%) lesions were located close to the EGJ, three (12%) in the lesser curvature, two (8%) in the posterior wall and one (4%) in the prepyloric-antral region. A multiport technique was used in 15 (60%) patients and a single-access approach in 10 (40%) patients. The median operative time was 140 (50-210) minutes. No conversion to open or conventional laparoscopic surgery was needed. Mortality was nil, and severe morbidity occurred in one (4%) patient. The median length of stay was 6 (3-10) days. Indications of LIGS were adenocarcinoma in 11 (44%) patients, gastrointestinal stromal tumors (GISTs) in 6 (24%) patients and benign lesions in eight (32%) patients. En bloc resection was obtained in 24 (96%) patients with R0 margins in 23 (92%) patients. After a median follow-up of 76 (26-171) months, recurrence was detected in 4 (36%) patients with advanced malignant adenocarcinoma. CONCLUSION: LIGS provides an interesting alternative to major gastric and EGJ resection when endoscopic resection is not suitable for highly selected patients with superficial gastric lesions.

Predictors for Survival in an International Cohort of Patients Undergoing Distal Pancreatectomy for Pancreatic Ductal Adenocarcinoma.

BACKGROUND: Surgical factors, including resection of Gerota's fascia, R0-resection, and lymph node yield, may be associated with survival after distal pancreatectomy (DP) for pancreatic ductal adenocarcinoma (PDAC), but evidence from large multicenter studies is lacking. This study aimed to identify predictors for overall survival after DP for PDAC, especially those related to surgical technique. PATIENTS AND METHODS: Data from an international retrospective cohort including patients from 11 European countries and the USA who underwent DP for PDAC (2007-2015) were analyzed. Cox proportional hazard analyses were performed and included Gerota's fascia resection, R0 resection, lymph node ratio, extended resection, and a minimally invasive approach. RESULTS: Overall, 1200 patients from 34 centers with median follow-up of 15 months [interquartile range (IQR) 5-31 months] and median survival period of 30 months [95% confidence interval (CI), 27-33 months] were included. Gerota's fascia resection [hazard ratio (HR) 0.74; p = 0.019], R0 resection (HR 0.70; p = 0.006), and decreased lymph node ratio (HR 0.28; p < 0.001) were associated with improved overall survival, whereas extended resection (HR 1.75; p < 0.001) was associated with worse overall survival. A minimally invasive approach did not improve survival as compared with an open approach (HR 1.14; p = 0.350). Adjuvant chemotherapy (HR 0.67; p = 0.003) was also associated with improved overall survival. CONCLUSIONS: This international cohort identified Gerota's fascia resection, R0 resection, and decreased lymph node ratio as factors associated with improved overall survival during DP for PDAC. Surgeons should strive for R0 resection and adequate lymphadenectomy and could also consider Gerota's fascia resection in their routine surgical approach.

Impact of cirrhosis in patients undergoing laparoscopic liver resection in a nationwide multicentre survey.

BACKGROUND: The aim was to analyse the impact of cirrhosis on short-term outcomes after laparoscopic liver resection (LLR) in a multicentre national cohort study. METHODS: This retrospective study included all patients undergoing LLR in 27 centres between 2000 and 2017. Cirrhosis was defined as F4 fibrosis on pathological examination. Short-term outcomes of patients with and without liver cirrhosis were compared after propensity score matching by centre volume, demographic and tumour characteristics, and extent of resection. RESULTS: Among 3150 patients included, LLR was performed in 774 patients with (24·6 per cent) and 2376 (75·4 per cent) without cirrhosis. Severe complication and mortality rates in patients with cirrhosis were 10·6 and 2·6 per cent respectively. Posthepatectomy liver failure (PHLF) developed in 3·6 per cent of patients with cirrhosis and was the major cause of death (11 of 20 patients). After matching, patients with cirrhosis tended to have higher rates of severe complications (odds ratio (OR) 1·74, 95 per cent c.i. 0·92 to 3·41; P = 0·096) and PHLF (OR 7·13, 0·91 to 323·10; P = 0·068) than those without cirrhosis. They also had a higher risk of death (OR 5·13, 1·08 to 48·61; P = 0·039). Rates of cardiorespiratory complications (P = 0·338), bile leakage (P = 0·286) and reoperation (P = 0·352) were similar in the two groups. Patients with cirrhosis had a longer hospital stay than those without (11 versus 8 days; P = 0·018). Centre expertise was an independent protective factor against PHLF in patients with cirrhosis (OR 0·33, 0·14 to 0·76; P = 0·010). CONCLUSION: Underlying cirrhosis remains an independent risk factor for impaired outcomes in patients undergoing LLR, even in expert centres.

ANTECEDENTES: El objetivo de este estudio fue analizar el impacto de la cirrosis en los resultados a corto plazo después de la resección hepática laparoscópica (laparoscopic liver resection, LLR) en un estudio de cohortes multicéntrico nacional. MÉTODOS: Este estudio retrospectivo incluyó todos los pacientes sometidos a LLR en 27 centros entre 2000 y 2017. La cirrosis se definió como fibrosis F4 en el examen histopatológico. Los resultados a corto plazo de los pacientes con hígado cirrótico (cirrhotic liver CL) (pacientes CL) y los pacientes con hígado no cirrótico (non-cirrhotic liver, NCL) (pacientes NCL) se compararon después de realizar un emparejamiento por puntaje de propension del volumen del centro, las características demográficas y del tumor, y la extensión de la resección. RESULTADOS: Del total de 3.150 pacientes incluidos, se realizó LLR en 774 (24,6%) pacientes CL y en 2.376 (75,4%) pacientes NCL. Las tasas de complicaciones graves y mortalidad en el grupo de pacientes CL fueron del 10,6% y 2,6%, respectivamente. La insuficiencia hepática posterior a la hepatectomía (post-hepatectomy liver failure, PHLF) fue la principal causa de mortalidad (55% de los casos) y se produjo en el 3,6% de los casos en pacientes CL. Después del emparejamiento, los pacientes CL tendieron a tener tasas más altas de complicaciones graves (razón de oportunidades, odds ratio, OR 1,74; i.c. del 95% 0,92-0,41; P = 0,096) y de PHLF (OR 7,13; i.c. del 95% 0,91-323,10; P = 0,068) en comparación con los pacientes NCL. Los pacientes CL estuvieron expuestos a un mayor riesgo de mortalidad (OR 5,13; i.c. del 95% 1,08-48,6; P = 0,039) en comparación con los pacientes NCL. Los pacientes CL presentaron tasas similares de complicaciones cardiorrespiratorias graves (P = 0,338), de fuga biliar (P = 0,286) y de reintervenciones (P = 0,352) que los pacientes NCL. Los pacientes CL tuvieron una estancia hospitalaria más larga (11 versus 8 días; P = 0,018) que los pacientes NCL. La experiencia del centro fue un factor protector independiente de PHLF (OR 0,33; i.c. del 95% 0,14-0,76; P = 0,010) pacientes CL. CONCLUSIÓN: La presencia de cirrosis subyacente sigue siendo un factor de riesgo independiente de peores resultados en pacientes sometidos a resección hepática laparoscópica, incluso en centros con experiencia.

Appendix intussusception in a young patient with Crohn's disease.

Appendix intussusception is rare, can be responsible of chronic unexplained abdominal pain and is mostly associated with inflammatory changes or endometriosis. A 23-year-old woman with Crohn's disease was diagnosed invaginated appendix by CT-scan and colonoscopy after several acute atypical abdominal pain episodes. These episodes were suspected to be independent of the Crohn's disease and to be attributed to the intussusception itself. A resection combining ligation of the base by colonoscopy and cecectomy by laparoscopy was performed successfully. However, no argument for Crohn's disease was found on the pathologic examination.

Highlights, limitations and future challenges of laparoscopic resection for colorectal liver metastases.

The liver is the most common site for metastatic colorectal cancer (CRLM). Despite advances in oncologic treatment, resection of metastases is still the only curative option. Although laparoscopic surgery for primary colorectal cancer is well documented and widely used, laparoscopic surgery for liver metastases has developed more slowly. However, in spite of some difficulties, laparoscopic approach demonstrated strong advantages including minimal parietal damage, decreased morbidity (reduced blood loss and need for transfusion, fewer pulmonary complications), and simplification of subsequent iterative hepatectomy. Up to now, more than 9 000 laparoscopic procedures have been reported worldwide and long-term results in colorectal liver metastases seem comparable to the open approach. Only one recent randomized controlled trial has compared the laparoscopic and the open approach. The purpose of the present update was to identify the barriers limiting widespread acceptance of laparoscopic approach, the benefits and the limits of laparoscopic hepatectomies in CRLM.

Multicentre propensity score-matched study of laparoscopic versus open repeat liver resection for colorectal liver metastases.

BACKGROUND: Repeat liver resection is often the best treatment option for patients with recurrent colorectal liver metastases (CRLM). Repeat resections can be complex, however, owing to adhesions and altered liver anatomy. It remains uncertain whether the advantages of a laparoscopic approach are upheld in this setting. The aim of this retrospective, propensity score-matched study was to compare the short-term outcome of laparoscopic (LRLR) and open (ORLR) repeat liver resection. METHODS: A multicentre retrospective propensity score-matched study was performed including all patients who underwent LRLRs and ORLRs for CRLM performed in nine high-volume centres from seven European countries between 2000 and 2016. Patients were matched based on propensity scores in a 1 : 1 ratio. Propensity scores were calculated based on 12 preoperative variables, including the approach to, and extent of, the previous liver resection. Operative outcomes were compared using paired tests. RESULTS: Overall, 425 repeat liver resections were included. Of 271 LRLRs, 105 were matched with an ORLR. Baseline characteristics were comparable after matching. LRLR was associated with a shorter duration of operation (median 200 (i.q.r. 123-273) versus 256 (199-320) min; P < 0·001), less intraoperative blood loss (200 (50-450) versus 300 (100-600) ml; P = 0·077) and a shorter postoperative hospital stay (5 (3-8) versus 6 (5-8) days; P = 0·028). Postoperative morbidity and mortality rates were similar after LRLR and ORLR. CONCLUSION: LRLR for CRLM is feasible in selected patients and may offer advantages over an open approach.

Early removal of intraperitoneal drainage after pancreatoduodenectomy in patients without postoperative fistula at POD3: Results of a randomized clinical trial.

BACKGROUND: To determine whether the timing of removal of abdominal drainage (AD) after pancreatoduodenectomy (PD) influences the 30-day surgical site infection (30-day SSI) rate. METHODS: A multicenter randomized, intention-to-treat trial with two parallel arms (superiority of early vs. standard AD removal on SSI) was performed between 2011 and 2015 in patients with no pancreatic fistula (PF) on POD3 after PD (NCT01368094). The primary endpoint was the 30-day SSI rate. The secondary endpoints were specific post-PD complications (grade BC PF), postoperative morbidity and risk factor of SSI, reoperation rate, 30-day mortality, length of drainage, length of stay and postoperative infectious complications. RESULTS: One hundred and forty-one patients were randomized: 71 in the early arm, 70 in the standard arm (70.2% of pancreatic adenocarcinomas; 91.5% of pancreatojejunostomies; 66.0% of bilateral drainages; feasibility: 39.9%). Early removal of drains was not associated with a significant decrease of 30-day SSI (14.1% vs. 24.3%, P=0.12). A lower rate of deep SSI was observed in the early arm (2.8% vs. 17.1%, P=0.03), leading to a shorter length of stay (17.8±6.8 vs. 21.0±6.1, P=0.01). Grade BC PF rate (5.6%), severe morbidity (17.7%), reoperation rate (7.8%), 30-day mortality (1.4%) and wound-SSI rate (7.8%) were similar between arms. After multivariate analysis, the timing of AD removal was not associated with an increase of 30-day SSI (OR=0.74 [95% CI 0.35-1.13, P=0.38]). CONCLUSION: In selected patients with no PF on POD3, early removal of abdominal drainage does not seem to increase or decrease surgical site infection's occurrence.

The effect of metabolic syndrome on postoperative outcomes following laparoscopic colectomy.

BACKGROUND: Among the criteria used to diagnose metabolic syndrome (MS), obesity and diabetes mellitus (DM) are associated with poor postoperative outcomes following colectomy. MS is also associated with colorectal cancer (CRC) and diverticulosis, both of which may be treated with colectomy. However, the effect of MS on postoperative outcomes following laparoscopic colectomy has yet to be clarified. METHODS: In an academic tertiary hospital, data from all consecutive patients undergoing laparoscopic colectomy from 2005 to 2014 were prospectively recorded and analysed. Patients presenting with MS [defined by the presence of three or more of the following criteria: elevated blood pressure, body mass index > 28 kg/m2, dyslipidemia (decreased serum HDL cholesterol, increased serum triglycerides) and increased fasting glucose/DM] were compared with patients without MS regarding peri-operative outcome [mainly anastomotic leaks, severe postoperative complications (Clavien-Dindo III and IV)] and mortality. RESULTS: Overall, 1236 patients were included: 508 (41.1%) right colectomies and 728 (58.9%) left colectomies. Seven hundred seventy-two (62.4%) of these procedures were performed for CRC. MS was diagnosed in 85 (6.9%) patients, who were significantly older than the others (70 vs. 64.2 years, p < 0.001), and presented with more cardiac comorbidities (p < 0.001). MS was associated with increased blood loss (122.5 vs. 79.9 mL p = 0.001) and blood transfusion requirement (5.9 vs. 1.7%, p = 0.021). The anastomotic leak rate was 6.6% (with 2.2% of anastomotic leaks requiring surgical treatment), and the overall reoperation rate was 6.9%. The incidence of severe postoperative complications was 11.5%, and the overall mortality rate 0.6%. No differences were found between the groups in overall postoperative morbidity and mortality. Median length of stay was similar in both groups (7 days). CONCLUSIONS: MS does not jeopardize postoperative outcomes following laparoscopic colectomy.

Comparison between open and laparoscopic gastrectomy for gastric cancer: A monocentric retrospective study from a western country.

BACKGROUND: The majority of laparoscopic gastrectomy (LG) reports arise from Asia and the benefit of this approach in western countries remains unclear. The objective of this study was to compare the postoperative outcomes between LG and open gastrectomy (OG) for gastric cancer in a western center. METHODS: Between 2005 and 2015, all consecutive patients with gastric cancer who underwent either LG or OG were enrolled. Postoperative morbimortality was evaluated according to Dindo-Clavien classification. RESULTS: Over 164 patients, 60 had LG and 104 OG with a mean age of 62 and 65 years, respectively. Total gastrectomy represented 58% of LG and 54% of OG (P=0.749). Operative time was not different in the two groups (160.8 vs. 174.2min, P=0.780) so as intraoperative blood loss (111 vs. 173mL, P=0.057). The rate of severe complications (including postoperative bleeding) was significantly higher in the LG group (40% vs. 23%, P=0.012) so as reoperation rate (27% vs. 6%, P<0.001). There was no statistical difference in terms of postoperative mortality (0 vs. 3%, P=0.252) or length of hospital stay (20 vs. 16 days, P=0.116). CONCLUSION: Laparoscopic gastrectomy for the treatment of gastric cancer in western countries appears to be feasible but with a higher rate of severe complications compared to open gastrectomy.

Preoperative imaging and prediction of oesophageal conduit necrosis after oesophagectomy for cancer.

BACKGROUND: Oesophageal conduit necrosis following oesophagectomy is a rare but life-threatening complication. The present study aimed to assess the impact of coeliac axis stenosis on outcomes after oesophagectomy for cancer. METHODS: The study included consecutive patients who had an Ivor Lewis procedure with curative intent for middle- and lower-third oesophageal cancer at two tertiary referral centres. All patients underwent preoperative multidetector CT with arterial phase to detect coeliac axis stenosis. The coeliac artery was classified as normal, with extrinsic stenosis due to a median arcuate ligament or with intrinsic stenosis caused by atherosclerosis. RESULTS: Some 481 patients underwent an Ivor Lewis procedure. Of these, ten (2·1 per cent) developed oesophageal conduit necrosis after surgery. Coeliac artery evaluation revealed a completely normal artery in 431 patients (91·5 per cent) in the group without conduit necrosis and in one (10 per cent) with necrosis (P < 0·001). Extrinsic stenosis of the coeliac artery due to a median arcuate ligament was found in two patients (0·4 per cent) without conduit necrosis and five (50 per cent) with necrosis (P < 0·001). Intrinsic stenosis of the coeliac artery was found in 11 (2·3 per cent) and eight (80 per cent) patients respectively (P < 0·001). Eight patients without (1·7 per cent) and five (50 per cent) with conduit necrosis had a single and thin left gastric artery (P < 0·001). CONCLUSION: This study suggests that oesophageal conduit necrosis after oesophagectomy for cancer may be due to pre-existing coeliac axis stenosis.

Multicentre study of the impact of morbidity on long-term survival following hepatectomy for intrahepatic cholangiocarcinoma.

BACKGROUND: The impact of morbidity on long-term outcomes following liver resection for intrahepatic cholangiocarcinoma is currently unclear. METHODS: This was a retrospective analysis of all consecutive patients who underwent liver resection for intrahepatic cholangiocarcinoma with curative intent in 24 university hospitals between 1989 and 2009. Severe morbidity was defined as any complication of Dindo-Clavien grade III or IV. Patients with severe morbidity were compared with those without in terms of demographics, pathology, management, morbidity, overall survival, disease-free survival and time to recurrence. Independent predictors of severe morbidity were identified by multivariable analysis. RESULTS: A total of 522 patients were enrolled. Severe morbidity occurred in 113 patients (21·6 per cent) and was an independent predictor of overall survival (hazard ratio 1·64, 95 per cent c.i. 1·21 to 2·23), as were age at resection, multifocal disease, positive lymph node status and R0 resection margin. Severe morbidity did not emerge as an independent predictor of disease-free survival. Independent predictors of time to recurrence included severe morbidity, tumour size, multifocal disease, vascular invasion and R0 resection margin. Major hepatectomy and intraoperative transfusion were independent predictors of severe morbidity. CONCLUSION: Severe morbidity adversely affects overall survival following liver resection for intrahepatic cholangiocarcinoma.

Accuracy of staging laparoscopy in detecting peritoneal dissemination in patients with gastroesophageal adenocarcinoma.

Despite staging laparoscopy (SL) with peritoneal lavage is recommended in US Guidelines in patients with potentially resectable gastroesophageal adenocarcinoma, this procedure is not systematically proposed in French Guidelines. Therefore, we decided to analyze the results of systematic SL in patients considered for preoperative chemotherapy. From 2005 to 2011, 116 consecutive patients with distal esophagus, esogastric junction, and gastric adenocarcinoma ≥T3 or N+ without detectable metastatic dissemination by computed tomography (CT) scan imaging underwent SL before neoadjuvant chemotherapy. Positive and negative SLs were compared according to tumor characteristics. SL was positive in 15 cases (12.9%) including 14 with peritoneal seeding (localized in five, diffuse in nine). SL was positive in 7 (24.1%) of 29 patients with poorly differentiated tumor, in 9 (32.1%) of 28 patients with signet ring cells, in 7 (50%) of 14 patients with gastric linitis tumor, and in 15 (16.3%) of 92 patients with T3 or T4 tumor. All the lesions of distal esophagus extending to the cardia had a negative SL. Among the 14 patients with peritoneal carcinomatosis at SL, nine (65%) had signs of peritoneal seeding on initial CT scan. One (0.8%) patient had a small bowel perforation closed laparoscopically. If systematic SL before preoperative chemotherapy does not seem justified because of its low accuracy, it should be performed in patients with poorly differentiated tumor, signet ring cell, and gastric linitis plastica components on biopsy and when CT scan is suggestive of T4 tumor, ascites, or peritoneal nodule.

Minimally-invasive surgery for liver metastases.

The practice of laparoscopic liver surgery has developed gradually, and most liver resections seem currently feasible and safe for selected patients in centers where surgeons are experienced both in liver surgery and in laparoscopic surgery. However, further studies would be required, especially for long-term oncological results and for major hepatectomy follow-up, before laparoscopic hepatectomy becomes a common alternative to open liver surgeries. Laparoscopic major hepatectomy is a recent technique, due to its significant complexity. In this review, we showed that patients undergoing laparoscopic hepatectomy for liver metastases experienced decreased rates of overall and major postoperative complications without any compromise on long-term outcomes compared to patients treated by open resection.

Laparoscopic two-stage hepatectomy for bilobar colorectal liver metastases.

BACKGROUND: Despite the gradual diffusion of laparoscopic liver resection, the feasibility and results of laparoscopic two-stage hepatectomy (TSH) for bilobar colorectal liver metastases (CRLM) have not been described frequently. This study aimed to evaluate the feasibility, safety and oncological outcomes of laparoscopic TSH for bilobar CRLM. METHODS: All patients eligible for laparoscopic TSH among those treated for bilobar CRLM from 2000 to 2013 were included. Demographics, tumour characteristics, surgical procedures, and short- and long-term outcomes were analysed. RESULTS: Laparoscopic TSH was planned in 34 patients with bilobar CRLM, representing 17·2 per cent of all 198 patients treated for bilobar CRLM. Thirty patients received preoperative chemotherapy, and 20 had portal vein occlusion to increase the volume of the remnant liver. Laparoscopic resection of the primary colorectal tumour was integrated within the first-stage hepatectomy in 11 patients. After a median interval of 3·1 months, 26 patients subsequently had a successful laparoscopic second-stage hepatectomy, including 18 laparoscopic right or extended right hepatectomies. The mortality rate for both stages was 3 per cent (1 of 34), and the overall morbidity rate for the first and second stages was 50 per cent (17 of 34) and 54 per cent (14 of 26) respectively. Mean length of hospital stay was 6·1 and 9·0 days respectively. With a median follow-up of 37·8 (range 6-129) months, 3- and 5-year overall survival rates in patients who completed TSH were 78 and 41 per cent respectively. The 3- and 5-year disease-free survival rates were 26 and 13 per cent respectively. CONCLUSION: Laparoscopic TSH for bilobar CRLM is safe and does not jeopardize long-term outcomes in selected patients.

Risk factors and consequences of conversion in laparoscopic major liver resection.

BACKGROUND: Although recent reports have suggested potential benefits of the laparoscopic approach in patients requiring major hepatectomy, it remains unclear whether conversion to open surgery could offset these advantages. This study aimed to determine the risk factors for and postoperative consequences of conversion in patients undergoing laparoscopic major hepatectomy (LMH). METHODS: Data for all patients undergoing LMH between 2000 and 2013 at two tertiary referral centres were reviewed retrospectively. Risk factors for conversion were determined using multivariable analysis. After propensity score matching, the outcomes of patients who underwent conversion were compared with those of matched patients undergoing laparoscopic hepatectomy who did not have conversion, operated on at the same centres, and also with matched patients operated on at another tertiary centre during the same period by an open laparotomy approach. RESULTS: Conversion was needed in 30 (13·5 per cent) of the 223 patients undergoing LMH. The most frequent reasons for conversion were bleeding and failure to progress, in 14 (47 per cent) and nine (30 per cent) patients respectively. On multivariable analysis, risk factors for conversion were patient age above 75 years (hazard ratio (HR) 7·72, 95 per cent c.i. 1·67 to 35·70; P = 0·009), diabetes (HR 4·51, 1·16 to 17·57; P = 0·030), body mass index (BMI) above 28 kg/m(2) (HR 6·41, 1·56 to 26·37; P = 0·010), tumour diameter greater than 10 cm (HR 8·91, 1·57 to 50·79; P = 0·014) and biliary reconstruction (HR 13·99, 1·82 to 238·13; P = 0·048). After propensity score matching, the complication rate in patients who had conversion was higher than in patients who did not (75 versus 47·3 per cent respectively; P = 0·038), but was not significantly different from the rate in patients treated by planned laparotomy (79 versus 67·9 per cent respectively; P = 0·438). CONCLUSION: Conversion during LMH should be anticipated in patients with raised BMI, large lesions and biliary reconstruction. Conversion does not lead to increased morbidity compared with planned laparotomy.

Learning curve for laparoscopic major hepatectomy.

BACKGROUND: Laparoscopic major hepatectomy (LMH) is evolving as an important surgical approach in hepatopancreatobiliary surgery. The present study aimed to evaluate the learning curve for LMH at a single centre. METHODS: Data for all patients undergoing LMH between January 1998 and September 2013 were recorded in a prospective database and analysed. The learning curve for operating time (OT) was evaluated using the cumulative sum (CUSUM) method. RESULTS: Of 173 patients undergoing major hepatectomy, left hepatectomy was performed in 28 (16·2 per cent), left trisectionectomy in nine (5·2 per cent), right hepatectomy in 115 (66·5 per cent), right trisectionectomy in 13 (7·5 per cent) and central hepatectomy in eight (4·6 per cent). Median duration of surgery was 270 (range 100-540) min and median blood loss was 300 (10-4500) ml. There were 20 conversions to an open procedure (11·6 per cent). Vascular clamping was independently associated with conversion on multivariable analysis (hazard ratio 5·95, 95 per cent c.i. 1·24 to 28·56; P = 0·026). The CUSUMOT learning curve was modelled as a parabola (CUSUMOT = 0·2149 × patient number(2) - 30·586 × patient number - 1118·3; R(2) = 0·7356). The learning curve comprised three phases: phase 1 (45 initial patients), phase 2 (30 intermediate patients) and phase 3 (the subsequent 98 patients). Although right hepatectomy was most common in phase 1, a significant decrease was observed from phase 1 to 3 (P = 0·007) in favour of more complex procedures. CONCLUSION: The learning curve for LMH consisted of three characteristic phases identified by CUSUM analysis. The data suggest that the learning phase of LMH included 45 to 75 patients.