Pancreas-sparing tumor resection for peripancreatic paraganglioma: a case series of six patients.

Paragangliomas (PGLs) located around the pancreas are rare and challenging to diagnose preoperatively. Tumor resection with pancreatectomy is often performed for peripancreatic PGL. However, pancreas-sparing tumor resection can be indicated with an accurate preoperative diagnosis. Six patients with pathologically diagnosed peripancreatic PGL were included. The clinical data were retrospectively collected from medical records. Five of them were suspected of peripancreatic PGL on imaging studies due to the fat plane identified between the tumor and pancreas, and subsequently diagnosed with PGL preoperatively based on elevated urinary catecholamine levels and/or metaiodobenzylguanidine scintigraphy without biopsy. All patients underwent pancreas-sparing tumor resection with negative surgical margins, and they did not develop postoperative complications related to potential damage to the pancreas. A fat plane between the tumor and pancreas on imaging studies and hormone levels are key findings for obtaining an accurate preoperative diagnosis of peripancreatic PGL, which can be managed with pancreas-sparing tumor resection.

Predictive risk factors for resection surface-related complications after laparoscopic simultaneous resection of primary colorectal tumor and synchronous liver metastases: a single-center retrospective study.

BACKGROUND: Laparoscopic simultaneous resection (LSR) of primary colorectal tumors and synchronous colorectal liver metastases (sCRLM) has been recently performed. This study aimed to evaluate the postoperative outcomes after LSR and determine the risk factors for resection surface-related complications (RSRC), such as postoperative biliary fistula and liver-transection surface abscess. METHODS: Between 2009 and 2022, consecutive patients with sCRLM who underwent LSR were included. We retrospectively analyzed clinicopathological data, including intraoperative factors and postoperative outcomes. The difficulty level of all liver resections was classified according to the IWATE difficulty scoring system (DSS). We then performed univariate and multivariate analyses to identify the risk factors for RSRC. RESULTS: Of the 112 patients, 94 (83.9%) underwent partial hepatectomy and colorectal surgery. The median DSS score was 5 points (1-11), with 12 (10.7%) patients scoring ≥ 7 points. Postoperative complications were observed in 41 (36.6%) patients, of whom 16 (14.3%) experienced severe complications classified as Clavien-Dindo grade IIIa or higher. There was no postoperative mortality. The most common complication was RSRC (19 patients, 17.0%). Multivariate analysis identified American Society of Anesthesiologists (ASA) classification ≥ 3 [odds ratio (OR) 10.3, 95% confidence interval (CI) 1.37-77.8; P = 0.023], DSS score ≥ 7 points (OR 5.08, 95% CI 1.17-20.0; P = 0.030), and right-sided colectomy (OR 4.67, 95% CI 1.46-15.0; P = 0.009) as independent risk factors for RSRC. Postoperative hospital stays were significantly longer for patients with RSRC than for those without RSRC (22 days vs. 11 days; P < 0.001). CONCLUSION: Short-term outcomes of LSR for patients with sCRLM were acceptable in an experienced center. RSRC was the most common complication, and high-difficulty hepatectomy, right-sided colectomy, and ASA classification ≥ 3 were independent risk factors for RSRC.

Pathological Complete Response in Patients With Resected Pancreatic Adenocarcinoma After Preoperative Chemotherapy.

Importance: Preoperative chemo(radio)therapy is increasingly used in patients with localized pancreatic adenocarcinoma, leading to pathological complete response (pCR) in a small subset of patients. However, multicenter studies with in-depth data about pCR are lacking. Objective: To investigate the incidence, outcome, and risk factors of pCR after preoperative chemo(radio)therapy. Design, Setting, and Participants: This observational, international, multicenter cohort study assessed all consecutive patients with pathology-proven localized pancreatic adenocarcinoma who underwent resection after 2 or more cycles of chemotherapy (with or without radiotherapy) in 19 centers from 8 countries (January 1, 2010, to December 31, 2018). Data collection was performed from February 1, 2020, to April 30, 2022, and analyses from January 1, 2022, to December 31, 2023. Median follow-up was 19 months. Exposures: Preoperative chemotherapy (with or without radiotherapy) followed by resection. Main Outcomes and Measures: The incidence of pCR (defined as absence of vital tumor cells in the sampled pancreas specimen after resection), its association with OS from surgery, and factors associated with pCR. Factors associated with overall survival (OS) and pCR were investigated with Cox proportional hazards and logistic regression models, respectively. Results: Overall, 1758 patients (mean [SD] age, 64 [9] years; 879 [50.0%] male) were studied. The rate of pCR was 4.8% (n = 85), and pCR was associated with OS (hazard ratio, 0.46; 95% CI, 0.26-0.83). The 1-, 3-, and 5-year OS rates were 95%, 82%, and 63% in patients with pCR vs 80%, 46%, and 30% in patients without pCR, respectively (P < .001). Factors associated with pCR included preoperative multiagent chemotherapy other than (m)FOLFIRINOX ([modified] leucovorin calcium [folinic acid], fluorouracil, irinotecan hydrochloride, and oxaliplatin) (odds ratio [OR], 0.48; 95% CI, 0.26-0.87), preoperative conventional radiotherapy (OR, 2.03; 95% CI, 1.00-4.10), preoperative stereotactic body radiotherapy (OR, 8.91; 95% CI, 4.17-19.05), radiologic response (OR, 13.00; 95% CI, 7.02-24.08), and normal(ized) serum carbohydrate antigen 19-9 after preoperative therapy (OR, 3.76; 95% CI, 1.79-7.89). Conclusions and Relevance: This international, retrospective cohort study found that pCR occurred in 4.8% of patients with resected localized pancreatic adenocarcinoma after preoperative chemo(radio)therapy. Although pCR does not reflect cure, it is associated with improved OS, with a doubled 5-year OS of 63% compared with 30% in patients without pCR. Factors associated with pCR related to preoperative chemo(radio)therapy regimens and anatomical and biological disease response features may have implications for treatment strategies that require validation in prospective studies because they may not universally apply to all patients with pancreatic adenocarcinoma.

Factors Affecting Nonfunctioning Small Pancreatic Neuroendocrine Neoplasms and Proposed New Treatment Strategies.

BACKGROUND & AIMS: Despite previously reported treatment strategies for nonfunctioning small (≤20 mm) pancreatic neuroendocrine neoplasms (pNENs), uncertainties persist. We aimed to evaluate the surgically resected cases of nonfunctioning small pNENs (NF-spNENs) in a large Japanese cohort to elucidate an optimal treatment strategy for NF-spNENs. METHODS: In this Japanese multicenter study, data were retrospectively collected from patients who underwent pancreatectomy between January 1996 and December 2019, were pathologically diagnosed with pNEN, and were treated according to the World Health Organization 2019 classification. Overall, 1490 patients met the eligibility criteria, and 1014 were included in the analysis cohort. RESULTS: In the analysis cohort, 606 patients (59.8%) had NF-spNENs, with 82% classified as grade 1 (NET-G1) and 18% as grade 2 (NET-G2) or higher. The incidence of lymph node metastasis (N1) by grade was significantly higher in NET-G2 (G1: 3.1% vs G2: 15.0%). Independent factors contributing to N1 were NET-G2 or higher and tumor diameter ≥15 mm. The predictive ability of tumor size for N1 was high. Independent factors contributing to recurrence included multiple lesions, NET-G2 or higher, tumor diameter ≥15 mm, and N1. However, the independent factor contributing to survival was tumor grade (NET-G2 or higher). The appropriate timing for surgical resection of NET-G1 and NET-G2 or higher was when tumors were >20 and >10 mm, respectively. For neoplasms with unknown preoperative grades, tumor size >15 mm was considered appropriate. CONCLUSIONS: NF-spNENs are heterogeneous with varying levels of malignancy. Therefore, treatment strategies based on tumor size alone can be unreliable; personalized treatment strategies that consider tumor grading are preferable.

Explainable artificial intelligence prediction-based model in laparoscopic liver surgery for segments 7 and 8: an international multicenter study.

BACKGROUND: Artificial intelligence (AI) is becoming more useful as a decision-making and outcomes predictor tool. We have developed AI models to predict surgical complexity and the postoperative course in laparoscopic liver surgery for segments 7 and 8. METHODS: We included patients with lesions located in segments 7 and 8 operated by minimally invasive liver surgery from an international multi-institutional database. We have employed AI models to predict surgical complexity and postoperative outcomes. Furthermore, we have applied SHapley Additive exPlanations (SHAP) to make the AI models interpretable. Finally, we analyzed the surgeries not converted to open versus those converted to open. RESULTS: Overall, 585 patients and 22 variables were included. Multi-layer Perceptron (MLP) showed the highest performance for predicting surgery complexity and Random Forest (RF) for predicting postoperative outcomes. SHAP detected that MLP and RF gave the highest relevance to the variables "resection type" and "largest tumor size" for predicting surgery complexity and postoperative outcomes. In addition, we explored between surgeries converted to open and non-converted, finding statistically significant differences in the variables "tumor location," "blood loss," "complications," and "operation time." CONCLUSION: We have observed how the application of SHAP allows us to understand the predictions of AI models in surgical complexity and the postoperative outcomes of laparoscopic liver surgery in segments 7 and 8.

A multi-institutional randomized phase III study comparing minimally invasive distal pancreatectomy versus open distal pancreatectomy for pancreatic cancer; Japan Clinical Oncology Group study JCOG2202 (LAPAN study).

BACKGROUND: Minimally invasive distal pancreatectomy (MIDP), including laparoscopic and robotic distal pancreatectomy, has gained widespread acceptance over the last decade owing to its favorable short-term outcomes. However, evidence regarding its oncologic safety is insufficient. In March 2023, a randomized phase III study was launched in Japan to confirm the non-inferiority of overall survival in patients with resectable pancreatic cancer undergoing MIDP compared with that of patients undergoing open distal pancreatectomy (ODP). METHODS: This is a multi-institutional, randomized, phase III study. A total of 370 patients will be enrolled from 40 institutions within 4 years. The primary endpoint of this study is overall survival, and the secondary endpoints include relapse-free survival, proportion of patients undergoing radical resection, proportion of patients undergoing complete laparoscopic surgery, incidence of adverse surgical events, and length of postoperative hospital stay. Only a credentialed surgeon is eligible to perform both ODP and MIDP. All ODP and MIDP procedures will undergo centralized review using intraoperative photographs. The non-inferiority of MIDP to ODP in terms of overall survival will be statistically analyzed. Only if non-inferiority is confirmed will the analysis assess the superiority of MIDP over ODP. DISCUSSION: If our study demonstrates the non-inferiority of MIDP in terms of overall survival, it would validate its short-term advantages and establish its long-term clinical efficacy. TRIAL REGISTRATION: This trial is registered with the Japan Registry of Clinical Trials as jRCT 1,031,220,705 [ https://jrct.niph.go.jp/en-latest-detail/jRCT1031220705 ].

Intraoperative artificial intelligence system identifying liver vessels in laparoscopic liver resection: a retrospective experimental study.

BACKGROUND: The precise recognition of liver vessels during liver parenchymal dissection is the crucial technique for laparoscopic liver resection (LLR). This retrospective feasibility study aimed to develop artificial intelligence (AI) models to recognize liver vessels in LLR, and to evaluate their accuracy and real-time performance. METHODS: Images from LLR videos were extracted, and the hepatic veins and Glissonean pedicles were labeled separately. Two AI models were developed to recognize liver vessels: the "2-class model" which recognized both hepatic veins and Glissonean pedicles as equivalent vessels and distinguished them from the background class, and the "3-class model" which recognized them all separately. The Feature Pyramid Network was used as a neural network architecture for both models in their semantic segmentation tasks. The models were evaluated using fivefold cross-validation tests, and the Dice coefficient (DC) was used as an evaluation metric. Ten gastroenterological surgeons also evaluated the models qualitatively through rubric. RESULTS: In total, 2421 frames from 48 video clips were extracted. The mean DC value of the 2-class model was 0.789, with a processing speed of 0.094 s. The mean DC values for the hepatic vein and the Glissonean pedicle in the 3-class model were 0.631 and 0.482, respectively. The average processing time for the 3-class model was 0.097 s. Qualitative evaluation by surgeons revealed that false-negative and false-positive ratings in the 2-class model averaged 4.40 and 3.46, respectively, on a five-point scale, while the false-negative, false-positive, and vessel differentiation ratings in the 3-class model averaged 4.36, 3.44, and 3.28, respectively, on a five-point scale. CONCLUSION: We successfully developed deep-learning models that recognize liver vessels in LLR with high accuracy and sufficient processing speed. These findings suggest the potential of a new real-time automated navigation system for LLR.

Survival outcome of patient with pT1N0 biliary tract cancer treated with surgery alone.

INTRODUCTION: Adjuvant chemotherapy (AC) with S-1 or capecitabine monotherapy is now the standard of care for resected biliary tract cancer (BTC) according to the Adjuvant S-1 for Cholangiocarcinoma Trial (ASCOT) and the BILCAP study. Patients selection criteria, especially regarding pT1N0 BTC, differed in both trials. We aimed to clarify the survival outcomes regarding resected pT1N0 BTC without AC. METHODS: Among patients with macroscopically complete resection for BTC treated without AC between September 1992 and December 2020, the survival outcomes of those with pT1N0 BTC, except for intrahepatic cholangiocarcinoma, according to the Union for International Cancer Control 7th and 8th edition (TNM7 and 8), were investigated. RESULTS: Of 749 patients who underwent curative resection for BTC, 69 were identified as having pT1N0 BTC according to TNM8. Six patients (9 %) developed recurrence during the median follow-up period of 53 months (range: 14-263 months) with only one patient (2 %) being pT1N0 according to TNM7. Based on TNM8, the 5-year recurrence-free survival, disease-specific survival, and overall survival reached 90.7 % (95 % confidence interval [CI]: 80.3-95.7 %), 96.4 % (95 % CI: 86.1-99.1 %), and 85.3 % (95 % CI: 71.2-92.8 %), respectively. Perineural invasion (PNI) was significantly associated with recurrence, and the recurrence rate in patients with PNI reached as high as 40 %. CONCLUSIONS: The survival outcomes regarding resected pT1N0 BTC according to TNM7 were excellent without AC; however, those of TNM8 were not, with PNI being associated with recurrence risk.

FBXW7 regulates the sensitivity of imatinib in gastrointestinal stromal tumors by targeting MCL1.

BACKGROUND: Imatinib contributes to improving prognosis of high-risk or unresectable gastrointestinal stromal tumors (GISTs). As therapeutic efficacy is limited by imatinib resistance and toxicity, the exploration of predictive markers of imatinib therapeutic efficacy that enables patients to utilize more effective therapeutic strategies remains urgent. METHODS: The correlation between FBXW7 and imatinib resistance via FBXW7-MCL1 axis was evaluated in vitro and in vivo experiments. The significance of FBXW7 as a predictor of imatinib treatment efficacy was examined in 140 high-risk patients with GISTs. RESULTS: The ability of FBXW7 to predict therapeutic efficacy of adjuvant imatinib in high-risk GIST patients was determined through 5-year recurrence-free survival (RFS) rates analysis and multivariate analysis. FBXW7 affects imatinib sensitivity by regulating apoptosis in GIST-T1 cells. FBXW7 targets MCL1 to regulate apoptosis. MCL1 involves in the regulation of imatinib sensitivity through inhibiting apoptosis in GIST-T1 cells. FBXW7 regulates imatinib sensitivity by down-regulating MCL1 to enhance imatinib-induced apoptosis in vitro. FBXW7 regulates imatinib sensitivity of GIST cells by targeting MCL1 to predict efficacy of imatinib treatment in vivo. CONCLUSIONS: FBXW7 regulates imatinib sensitivity by inhibiting MCL1 to enhance imatinib-induced apoptosis in GIST, and predicts efficacy of imatinib treatment in high-risk GIST patients treated with imatinib.

Feasibility of S-1 adjuvant chemotherapy after major hepatectomy for biliary tract cancers: An exploratory subset analysis of JCOG1202.

INTRODUCTION: Major hepatectomy (MH) may produce the impaired liver function and affect the feasibility of adjuvant chemotherapy in terms of early period after the surgery, but there have not been detailed investigations. JCOG1202 (UMIN000011688) is a randomized phase III trial demonstrating the superiority of adjuvant S-1 chemotherapy for biliary tract cancer (BTC). The aim of this study is to examine the influence of MH for BTC on adjuvant S-1. MATERIALS AND METHODS: Of the total 424 patients, 207 received S-1 (S-1 arm) while the remaining 217 were not. We compared MH with non-major hepatectomy (NMH) for BTC. RESULTS: In the S-1 arm, 42 had undergone MH, and 165 had undergone NMH. MH had similar pretreatment features to NMH, including the proportion of biliary reconstruction, to NMH, except for a lower platelet count (17.7 vs. 23.4 × 104/mm3, p < 0.0001) and lower serum albumin level (3.5 vs. 3.8 g/dL, p < 0.0001). The treatment completion proportion tended to be lower for MH than for NMH (59.5 % vs. 75.8 %; risk ratio, 0.786 [95 % confidence interval, 0.603-1.023], p = 0.0733), and the median dose intensity was lower as well (88.7 % vs. 99.6 %, p = 0.0358). The major reasons for discontinuation were biliary tract infections and gastrointestinal disorders after MH. The frequency of grade 3-4 biliary tract infection was 19.0 % in MH vs. 4.2 % in NMH. CONCLUSION: The treatment completion proportion and dose intensity were lower in MH than in NMH. Caution should be exercised against biliary tract infections and gastrointestinal disorders during adjuvant S-1 after MH for BTC.

Re-appraisal of the universal definition of tumor rupture among patients with high-risk gastrointestinal stromal tumors.

Aim: Tumor rupture has been indicated as a risk factor for recurrence of gastrointestinal stromal tumors (GISTs). The universal definition of tumor rupture was proposed. This study evaluated whether the universal definition was more accurate in identification of GISTs with high recurrent risk than subjective judgment. Methods: The study included 507 patients with high-risk GISTs who underwent complete resection between December 2012 and December 2015. We conducted a questionnaire survey in participating institutes to re-diagnose tumor rupture based on the universal definition according to their surgical and pathological findings. We compared the clinical outcomes of tumor rupture based on the definition to those based on the surgeon's judgment and clarified the clinical importance of the rupture. Results: Sixty-four patients were initially registered to have tumor rupture by surgeon's judgment, and it became 90 patients who had tumor rupture after reevaluation. Although there were significant differences in recurrence-free survival (RFS) between no rupture and rupture for both initial registration and reevaluation (p = 0.002, <0.001, respectively), a significant difference in overall survival was only observed after reevaluation (p = 0.011). Tumor rupture was significantly associated with large tumor size, mixed cell type in histology, R1 resection, frequent adjuvant therapy and recurrence, but not with location, mitosis, and genotype. Adjuvant therapy more than 3 years improved RFS of patients with tumor rupture. Conclusion: This study suggested that tumor rupture based on the universal definition more accurately identified GISTs with poor prognostic outcomes than the subjective judgment.

Common clinicopathological and immunological features of sarcomatoid carcinoma across organs: A histomorphology-based cross-organ study.

Sarcomatoid carcinoma (SC), which can occur in any organ, is a rare disease. To elucidate common characteristics of SC beyond organs, we evaluated clinicopathological and immunological features of SC defined by the single histological criterion beyond organs compared to randomly matched conventional carcinoma (non-SC) adjusted for the disease stage. Immunological features were assessed by multiplex immunohistochemistry, comparing immune cell density in tumor tissues and tumor programmed death-ligand 1 (PD-L1) expression. A total of 101 patients with SC or non-SC (31 lung, 19 esophagus, 22 pancreas, 15 liver, 4 bile duct, 6 kidney, 2 uterus and 2 ovary) were identified among 7197 patients who underwent surgery at our institute (1997-2020). SC was significantly associated with worse survival (HR: 1.571; 95% CI: 1.084-2.277; P = .017). The frequency of postoperative progression within 6 months was significantly higher for SC patients (54% vs 28%; P = .002). The immune profiling revealed the densities of CD8+ T cells (130 vs 72 cells/mm2 ; P = .004) and tumor-associated macrophages (566 vs 413 cells/mm2 ; P < .0001) and the tumor PD-L1 expression score (40% vs 5%; P < .0001) were significantly higher in SCs than in non-SCs. Among 73 SC patients with postoperative progression, multivariate Cox regression analysis showed that immunotherapy tended to be associated with favorable survival (HR: 0.256; 95% CI: 0.062-1.057; P = .060). Collectively, SCs shared clinicopathological and immunological features across organs. Our study can initiate to standardize the pathological definition of SC and provide a rationale for the investigation and development for this rare disease in a cross-organ manner.

Adjuvant and neoadjuvant chemotherapy for biliary tract cancer: a review of randomized controlled trials.

The first randomized controlled trial of adjuvant chemotherapy for biliary tract cancer was reported in 2002. Since then, studies have continued, with efficacy reported for capecitabine in 2018 and S-1 in 2023. Oral fluoropyrimidines have become established as the standard of care. This article reviews the evidence from the randomized controlled trials reported to date and those that are ongoing or from which results have not yet been reported.

Liver functional assessment using time-associated change in the liver-to-spleen signal intensity ratio on enhanced magnetic resonance imaging: a retrospective study.

BACKGROUND: Liver-to-spleen signal intensity ratio (LSR) is evaluated by magnetic resonance imaging (MRI) in the hepatobiliary phase and has been reported as a useful radiological assessment of regional liver function. However, LSR is a passive (non-time-associated) assessment of liver function, not a dynamic (time-associated) assessment. Moreover, LSR shows limitations such as a dose bias of contrast medium and a timing bias of imaging. Previous studies have reported the advantages of time-associated liver functional assessment as a precise assessment of liver function. For instance, the indocyanine green (ICG) disappearance rate, which is calculated from serum ICG concentrations at multiple time points, reflects a precise preoperative liver function for predicting post-hepatectomy liver failure without the dose bias of ICG or the timing bias of blood sampling. The aim of this study was to develop a novel time-associated radiological liver functional assessment and verify its correlation with traditional liver functional parameters. METHODS: A total of 279 pancreatic cancer patients were evaluated to clarify fundamental time-associated changes to LSR in normal liver. We defined the time-associated radiological assessment of liver function, calculated using information on LSR from four time points, as the "LSR increasing rate" (LSRi). We then investigated correlations between LSRi and previous liver functional parameters. Furthermore, we evaluated how timing bias and protocol bias affect LSRi. RESULTS: Significant correlations were observed between LSRi and previous liver functional parameters such as total bilirubin, Child-Pugh grade, and albumin-bilirubin grade (P < 0.001 each). Moreover, considerably high correlations were observed between LSRi calculated using four time points and that calculated using three time points (r > 0.973 each), indicating that the timing bias of imaging was minimal. CONCLUSIONS: This study propose a novel time-associated radiological assessment, and revealed that the LSRi correlated significantly with traditional liver functional parameters. Changes in LSR over time may provide a superior preoperative assessment of regional liver function that is better for predicting post-hepatectomy liver failure than LSR using the hepatobiliary phase alone.

Relationships between physical and immunological tumor microenvironment in pancreatic ductal adenocarcinoma.

Pancreatic ductal adenocarcinoma (PDAC) is physically palpated as a hard tumor with an unfavorable prognosis. Assessing physical features and their association with pathological features could help to elucidate the mechanism of physical abnormalities in cancer tissues. A total of 93 patients who underwent radical surgery for pancreatic and bile duct cancers at a single center hospital during a 28-month period were recruited for this study that aimed to estimate the stiffness of PDAC tissues compared to the other neoplasms and assess relationships between tumor stiffness and pathological features. Physical alterations and pathological features of PDAC, with or without preoperative therapy, were analyzed. The immunological tumor microenvironment was evaluated using multiplexed fluorescent immunohistochemistry. The stiffness of PDAC correlated with the ratio of Azan-Mallory staining, α-smooth muscle actin, and collagen I-positive areas of the tumors. Densities of CD8+ T cells and CD204+ macrophages were associated with tumor stiffness in cases without preoperative therapy. Pancreatic ductal adenocarcinoma treated with preoperative therapy was softer than that without, and the association between tumor stiffness and immune cell infiltration was not shown after preoperative therapy. We observed the relationship between tumor stiffness and immunological features in human PDAC for the first time. Immune cell densities in the tumor center were smaller in hard tumors than in soft tumors without preoperative therapies. Preoperative therapy could alter physical and immunological aspects, warranting further study. Understanding of the correlations between physical and immunological aspects could lead to the development of new therapies.

A novel predictive model for postoperative delirium using multiple geriatric screening factors.

OBJECTIVES: The purpose of this study was to develop a new composite score to accurately predict postoperative delirium (POD) after major urological cancer surgery. METHODS: Our retrospective analysis included, in total, 449 consecutive patients who experienced major urological cancer surgery and a preoperative geriatric functional assessment at our institution (development cohort). Geriatric functional assessments included Geriatric 8, Instrumental Activities of Daily Living, and mini-cognitive assessment instrument (Mini-Cog). Multivariate analysis was used to identify factors related to POD and combined to create a predictive score. The composite score was externally validated using a cohort of 92 consecutive pancreatic cancer patients who underwent pancreaticoduodenectomy and a preoperative geriatric functional assessment (validation cohort). The predictive accuracy and performance of the composite score were evaluated using the area under the receiver operating characteristic curves (AUC) and calibration plots. RESULTS: In multivariate analysis of a development cohort, the following factors were significantly associated with POD: a Mini-Cog score of <3 (odds ratio [OR] = 9.5; p < 0.001), disability in the responsibility for medication (OR = 4.1; p = 0.03), and the preoperative use of benzodiazepine (OR = 6.4; p < 0.001). The composite score of these three factors showed excellent discrimination in predicting POD, with AUC values of 0.819 and 0.804 in development and validation cohorts, respectively. Calibration plots showing predicted probability and actual observation in both cohorts showed good agreement. CONCLUSIONS: A combined model of Mini-Cog, a disability in the responsibility for medication, and preoperative benzodiazepine use showed excellent discriminative power in predicting POD.

Machine learning-based Automatic Evaluation of Tissue Handling Skills in Laparoscopic Colorectal Surgery: A Retrospective Experimental Study.

OBJECTIVE: To develop a machine learning model that automatically quantifies the spread of blood in the surgical field using intraoperative videos of laparoscopic colorectal surgery and evaluate whether the index measured with the developed model can be used to assess tissue handling skill. BACKGROUND: Although skill evaluation is crucial in laparoscopic surgery, existing evaluation systems suffer from evaluator subjectivity and are labor-intensive. Therefore, automatic evaluation using machine learning is potentially useful. MATERIALS AND METHODS: In this retrospective experimental study, we used training data with annotated labels of blood or non-blood pixels on intraoperative images to develop a machine learning model to classify pixel RGB values into blood and non-blood. The blood pixel count per frame (the total number of blood pixels throughout a surgery divided by the number of frames) was compared among groups of surgeons with different tissue handling skills. RESULTS: The overall accuracy of the machine learning model for the blood classification task was 85.7%. The high tissue handling skill group had the lowest blood pixel count per frame, and the novice surgeon group had the highest count (mean [SD]: high tissue handling skill group 20972.23 [19287.05] vs. low tissue handling skill group 34473.42 [28144.29] vs. novice surgeon group 50630.04 [42427.76], P <0.01). The difference between any 2 groups was significant. CONCLUSIONS: We developed a machine learning model to measure blood pixels in laparoscopic colorectal surgery images using RGB information. The blood pixel count per frame measured with this model significantly correlated with surgeons' tissue handling skills.

Early postoperative decrease of skeletal muscle mass predicts recurrence and poor survival after surgical resection for perihilar cholangiocarcinoma.

BACKGROUND: Preoperative sarcopenia is a predictor of poor survival in cancer patients. We hypothesized that sarcopenia could progress as occult metastasis arose, especially after highly invasive surgery for highly aggressive malignancy. This study aimed to evaluate the associations of postoperative changes in skeletal muscle mass volume with survival outcomes in patients who underwent surgery for perihilar cholangiocarcinoma. METHODS: Fifty-six patients who underwent major hepatectomy with extrahepatic bile duct resection for perihilar cholangiocarcinoma were studied. The skeletal muscle index (SMI) at the third lumbar vertebra was calculated from axial computed tomography images taken preoperatively and 3-6 months postoperatively (early postoperative period). The associations of clinicopathological variables, including changes of SMI after surgery, with overall survival and recurrence-free survival were evaluated. Moreover, the associations of decreased SMI and elevated serum carbohydrate antigen 19-9 level with early recurrence and poor survival was compared. RESULTS: Among 56 patients, 26 (46%) had sarcopenia preoperatively and SMI decreased in 29 (52%) in the early postoperative period. During the median follow-up of 57.9 months, 35 patients (63%) developed recurrence and 29 (50%) died. Decreased SMI in the early postoperative period was independently associated with a shorter overall survival (hazard ratio, 2.39; 95% confidence interval, 1.00-6.18; P = 0.049) and a shorter recurrence-free survival (hazard ratio, 2.14; 95% confidence interval, 1.04-4.57; P = 0.039), whereas elevated carbohydrate antigen 19-9 level was not. CONCLUSIONS: Decreased SMI in the early postoperative period may be used as a predictor for recurrence and poor survival in patients undergoing surgery for perihilar cholangiocarcinoma.

Automated surgical workflow identification by artificial intelligence in laparoscopic hepatectomy: Experimental research.

BACKGROUND: To perform accurate laparoscopic hepatectomy (LH) without injury, novel intraoperative systems of computer-assisted surgery (CAS) for LH are expected. Automated surgical workflow identification is a key component for developing CAS systems. This study aimed to develop a deep-learning model for automated surgical step identification in LH. MATERIALS AND METHODS: We constructed a dataset comprising 40 cases of pure LH videos; 30 and 10 cases were used for the training and testing datasets, respectively. Each video was divided into 30 frames per second as static images. LH was divided into nine surgical steps (Steps 0-8), and each frame was annotated as being within one of these steps in the training set. After extracorporeal actions (Step 0) were excluded from the video, two deep-learning models of automated surgical step identification for 8-step and 6-step models were developed using a convolutional neural network (Models 1 & 2). Each frame in the testing dataset was classified using the constructed model performed in real-time. RESULTS: Above 8 million frames were annotated for surgical step identification from the pure LH videos. The overall accuracy of Model 1 was 0.891, which was increased to 0.947 in Model 2. Median and average accuracy for each case in Model 2 was 0.927 (range, 0.884-0.997) and 0.937 ± 0.04 (standardized difference), respectively. Real-time automated surgical step identification was performed at 21 frames per second. CONCLUSIONS: We developed a highly accurate deep-learning model for surgical step identification in pure LH. Our model could be applied to intraoperative systems of CAS.