Evaluating surgical expertise with AI-based automated instrument recognition for robotic distal gastrectomy.

Introduction: Complexities of robotic distal gastrectomy (RDG) give reason to assess physician's surgical skill. Varying levels in surgical skill affect patient outcomes. We aim to investigate how a novel artificial intelligence (AI) model can be used to evaluate surgical skill in RDG by recognizing surgical instruments. Methods: Fifty-five consecutive robotic surgical videos of RDG for gastric cancer were analyzed. We used Deeplab, a multi-stage temporal convolutional network, and it trained on 1234 manually annotated images. The model was then tested on 149 annotated images for accuracy. Deep learning metrics such as Intersection over Union (IoU) and accuracy were assessed, and the comparison between experienced and non-experienced surgeons based on usage of instruments during infrapyloric lymph node dissection was performed. Results: We annotated 540 Cadiere forceps, 898 Fenestrated bipolars, 359 Suction tubes, 307 Maryland bipolars, 688 Harmonic scalpels, 400 Staplers, and 59 Large clips. The average IoU and accuracy were 0.82 ± 0.12 and 87.2 ± 11.9% respectively. Moreover, the percentage of each instrument's usage to overall infrapyloric lymphadenectomy duration predicted by AI were compared. The use of Stapler and Large clip were significantly shorter in the experienced group compared to the non-experienced group. Conclusions: This study is the first to report that surgical skill can be successfully and accurately determined by an AI model for RDG. Our AI gives us a way to recognize and automatically generate instance segmentation of the surgical instruments present in this procedure. Use of this technology allows unbiased, more accessible RDG surgical skill.

MUC1-C Dependence for the Progression of Pancreatic Neuroendocrine Tumors Identifies a Druggable Target for the Treatment of This Rare Cancer.

Patients with pancreatic neuroendocrine tumors (pNETs) have limited access to effective targeted agents and invariably succumb to progressive disease. MUC1-C is a druggable oncogenic protein linked to driving pan-cancers. There is no known involvement of MUC1-C in pNET progression. The present work was performed to determine if MUC1-C represents a potential target for advancing pNET treatment. We demonstrate that the MUC1 gene is upregulated in primary pNETs that progress with metastatic disease. In pNET cells, MUC1-C drives E2F- and MYC-signaling pathways necessary for survival. Targeting MUC1-C genetically and pharmacologically also inhibits self-renewal capacity and tumorigenicity. Studies of primary pNET tissues further demonstrate that MUC1-C expression is associated with (i) an advanced NET grade and pathological stage, (ii) metastatic disease, and (iii) decreased disease-free survival. These findings demonstrate that MUC1-C is necessary for pNET progression and is a novel target for treating these rare cancers with anti-MUC1-C agents under clinical development.

Identifying the Trends of Urinary microRNAs within Extracellular Vesicles for Esophageal Cancer.

Background: The advancement of multidisciplinary treatment has increased the need to develop tests to monitor tumor burden during treatment. We herein analyzed urinary microRNAs within extracellular vesicles from patients with esophageal squamous cell carcinoma (ESCC) and normal individuals using a microarray. Methods: Patients with advanced ESCC who underwent esophagectomy (A), endoscopic submucosal resection (ESD) (B), and healthy donors (C) were included. Based on microRNA expression among the groups (Analysis 1), microRNAs with significant differences between groups A and C were selected (Analysis 2). Of these candidates, microRNAs in which the change between A and C was consistent with the change between B and C were selected for downstream analysis (Analysis 3). Finally, microRNA expression was validated in patients with recurrence from A (exploratory analysis). Results: For analysis 1, 205 microRNAs were selected. For Analyses 2 and 3, the changes in 18 microRNAs were consistent with changes in tumor burden as determined by clinical imaging and pathological findings. The AUC for the detection of ESCC using 18 microRNAs was 0.72. In exploratory analysis, three of eighteen microRNAs exhibited a concordant trend with recurrence. Conclusions: The current study identified the urinary microRNAs which were significantly expressed in ESCC patients. Validation study is warranted to evaluate whether these microRNAs could reflect tumor burden during multidisciplinary treatment for ESCC.

Automated artificial intelligence-based phase-recognition system for esophageal endoscopic submucosal dissection (with video).

BACKGROUND AND AIMS: Endoscopic submucosal dissection (ESD) for superficial esophageal cancer is a multistep treatment involving several endoscopic processes. Although analyzing each phase separately is worthwhile, it is not realistic in practice owing to the need for considerable manpower. To solve this problem, we aimed to establish a state-of-the-art artificial intelligence (AI)-based system, specifically, an automated phase-recognition system that can automatically identify each endoscopic phase based on video images. METHODS: Ninety-four videos of ESD procedures for superficial esophageal cancer were evaluated in this single-center study. A deep neural network-based phase-recognition system was developed in an automated manner to recognize each of the endoscopic phases. The system was trained with the use of videos that were annotated and verified by 2 GI endoscopists. RESULTS: The overall accuracy of the AI model for automated phase recognition was 90%, and the average precision, recall, and F value rates were 91%, 90%, and 90%, respectively. Two representative ESD videos predicted by the model indicated the usability of AI in clinical practice. CONCLUSIONS: We demonstrated that an AI-based automated phase-recognition system for esophageal ESD can be established with high accuracy. To the best of our knowledge, this is the first report on automated recognition of ESD treatment phases. Because this system enabled a detailed analysis of phases, collecting large volumes of data in the future may help to identify quality indicators for treatment techniques and uncover unmet medical needs that necessitate the creation of new treatment methods and devices.

Sarcoidosis Associated with Enlarged Mediastinal Lymph Nodes with the Detection of Streptococcus gordonii and Cutibacterium acnes Using a Clone Library Method.

A 77-year-old Japanese woman with mediastinal lymphadenopathy and uveitis was diagnosed with sarcoidosis. The bacterial flora in biopsied samples from mediastinal lymph nodes was analyzed using a clone library method with Sanger sequencing of the 16S rRNA gene, and Streptococcus gordonii (52 of 71 clones) and Cutibacterium acnes (19 of 71 clones) were detected. No previous study has conducted a bacterial floral analysis using the Sanger method for the mediastinal lymph node in sarcoidosis, making this case report the first to document the presence of S. gordonii and C. acnes in the mediastinal lymph node of a patient with sarcoidosis.

Disseminated Mycobacterium genavense Infection Mimicking Sarcoidosis: A Case Report and Review of Literature on Japanese Patients.

Sarcoidosis is a systemic inflammatory disease characterized by noncaseating epithelioid cell granulomas. However, certain infections can exhibit similar histological findings. We present a case of a 69-year-old man who was initially diagnosed with sarcoidosis and later was confirmed, through 16S rRNA sequencing, to have disseminated Mycobacterium genavense infection. Acid-fast bacteria were detected in the bone marrow biopsy using Ziehl-Neelsen staining, but routine clinical tests did not provide a definitive diagnosis. The patient tested negative for HIV, anti-interferon-gamma antibodies, and genetic immunodeficiency disorders. He was treated with multiple drugs, including aminoglycosides and macrolides, but showed no improvement in fever and pancytopenia. However, these clinical signs responded favorably to steroid therapy. We reviewed 17 Japanese cases of M. genavense infection. All cases were in males; 7/17 (41%) were HIV-negative; and 12/17 (71%) had a decreased CD4 count. Genetic analysis confirmed M. genavense isolation, and macrolides were used universally. Mycobacterium genavense infection is challenging to identify and mimics other systemic inflammatory diseases such as sarcoidosis. There are no standard treatment protocols. Our case report and Japanese case review contribute to understanding this rare disease.

The impact of thoracic duct resection on the long-term body composition of patients who underwent esophagectomy for esophageal cancer and survived without recurrence.

BACKGROUND: We have reported the possible benefits of radical esophagectomy with thoracic duct (TD) resection in elective esophageal cancer surgery. However, the effect of TD resection on the long-term nutrition status remains unclear. METHODS: Patients who underwent esophagectomy at Keio University between January 2006 and December 2018 were included, and those who had no recurrence for more than three years were evaluated. Changes in each body composition (muscle mass and body fat) were comparatively assessed between those who underwent TD resection or not, before and at, one, three and five years after surgery. Computed tomography images were analyzed on postoperative year 1, 3 and 5. RESULTS: This study included 217 patients categorized in the TD-resected (TD-R) (156 patients) and TD-preserved (TD-P) (61 patients) groups. The loss of muscle mass was comparable between the groups. On the other hand, the loss of adipose tissues was significantly greater in the TD-R group than in the TD-P group at one and three years after surgery, while there was no statistical difference five years after surgery. Additionally, among patients with cT1N0M0 disease in whom survival advantage of TD resection has been reported previously, the loss of muscle mass did not differ between each group. CONCLUSIONS: The change of muscle mass between the two groups was comparable. Although body fat mass was reduced by TD resection, it eventually recovered in the long term. In patients with esophageal cancer, TD resection may be acceptable without significant impact on body composition in the long term.

Clinical predictors of early postoperative recurrence after radical esophagectomy for thoracic esophageal cancer.

PURPOSE: Esophagectomy for esophageal cancer has a high incidence rate of early postoperative recurrence and death. This study aimed to identify the clinical and pathological features in early recurrence cases and to confirm the usefulness of prediction using these factors for effective adjuvant therapy and postoperative surveillance. METHODS: One hundred and twenty five patients who developed postoperative recurrence after undergoing radical esophagectomy for thoracic esophageal cancer were classified into two groups as follows: those with early recurrence at ≤ 6 months and those with nonearly recurrence at > 6 months after surgery. After identifying related factors of early recurrence, usefulness of these factors for prediction were examined in all patients with and without recurrence. RESULTS: The analysis cohort consisted of 43 and 82 patients in the early and nonearly recurrence groups, respectively. In multivariate analysis, factors associated with early recurrence were higher initial levels of tumor markers (squamous cell carcinoma [SCC] ≥ 1.5 ng/ml in tumors, except for adenocarcinoma, and carcinoembryonic antigen [CEA] ≥ 5.0 ng/ml in adenocarcinoma) and higher venous invasion (v), i.e., ≥ 2 (p = 0.040 and p = 0.004, respectively). The usefulness of these two factors for recurrence prediction was confirmed in 378 patients, including 253 patients without recurrence. Patients with at least one of the two factors had significantly higher early recurrence rates than those without any factors in pStages II and III (odds ratio [OR], 6.333; p = 0016 and OR, 4.346; p = 0.008, respectively). CONCLUSIONS: Early recurrence of thoracic esophageal cancer (i.e., during ≤ 6 months after esophagectomy) was associated with higher initial tumor marker levels and pathological findings of v ≥ 2. The combination of these two factors is useful as a simple and critical predictor of early postoperative recurrence.

Evaluation of Endoscopic Response Using Deep Neural Network in Esophageal Cancer Patients Who Received Neoadjuvant Chemotherapy.

BACKGROUND: We previously reported that endoscopic response evaluation can preoperatively predict the prognosis and distribution of residual tumors after neoadjuvant chemotherapy (NAC). In this study, we developed artificial intelligence (AI)-guided endoscopic response evaluation using a deep neural network to discriminate endoscopic responders (ERs) in patients with esophageal squamous cell carcinoma (ESCC) after NAC. METHOD: Surgically resectable ESCC patients who underwent esophagectomy following NAC were retrospectively analyzed in this study. Endoscopic images of the tumors were analyzed using a deep neural network. The model was validated with a test data set using 10 newly collected ERs and 10 newly collected non-ER images. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the endoscopic response evaluation by AI and endoscopists were calculated and compared. RESULTS: Of 193 patients, 40 (21%) were diagnosed as ERs. The median sensitivity, specificity, PPV, and NPV values for ER detection in 10 models were 60%, 100%, 100%, and 71%, respectively. Similarly, the median values by the endoscopist were 80%, 80%, 81%, and 81%, respectively. CONCLUSION: This proof-of-concept study using a deep learning algorithm demonstrated that the constructed AI-guided endoscopic response evaluation after NAC could identify ER with high specificity and PPV. It would appropriately guide an individualized treatment strategy that includes an organ preservation approach in ESCC patients.

Tumor Burden Monitoring with Circulating Tumor DNA During Treatment in Patients with Esophageal Squamous Cell Carcinoma.

BACKGROUND: To guide appropriate treatment strategy, an accurate tumor monitoring modality that reflects tumor burden during neoadjuvant treatment is required for esophageal squamous cell carcinoma (ESCC). We aimed to investigate the clinical utility of circulating tumor DNA (ctDNA) in plasma in patients who received neoadjuvant chemotherapy (NAC) followed by esophagectomy. PATIENTS AND METHODS: Longitudinally collected plasma samples for ctDNA combined with genomic DNA from primary lesions were obtained from patients with histologically confirmed ESCC who underwent NAC followed by subtotal esophagectomy. Next-generation sequencing was performed to identify mutations from the plasma and the primary tumor. The relationships between changes in ctDNA and the pathological response and recurrence were assessed in patients with locally advanced ESCC. RESULTS: In pretreatment samples from 13 patients, multiple concordant mutations in ctDNA and primary tumors were observed in 11 patients (85%), who were classified as ctDNA positive before treatment. The ctDNA positive rate after NAC correlated with the pathological response (responders, 25%; nonresponders, 100%; p = 0.007). The risk of recurrence increased significantly in patients with positive ctDNA after surgery in analysis of 16 patients; the 1-year recurrence-free survival rates were 90 and 0% in ctDNA-negative and ctDNA-positive groups, respectively (p = 0.0008). In two patients with postoperative recurrence, ctDNA was detected approximately 5.5 months earlier than the diagnosis using radiographical imaging. CONCLUSIONS: ctDNA is a promising biomarker for predicting pathological response and postoperative recurrence in ESCC. To demonstrate the external validity, we are currently preparing a multicenter prospective study.

Evaluation of surgical complexity by automated surgical process recognition in robotic distal gastrectomy using artificial intelligence.

BACKGROUND: Although radical gastrectomy with lymph node dissection is the standard treatment for gastric cancer, the complication rate remains high. Thus, estimation of surgical complexity is required for safety. We aim to investigate the association between the surgical process and complexity, such as a risk of complications in robotic distal gastrectomy (RDG), to establish an artificial intelligence (AI)-based automated surgical phase recognition by analyzing robotic surgical videos, and to investigate the predictability of surgical complexity by AI. METHOD: This study assessed clinical data and robotic surgical videos for 56 patients who underwent RDG for gastric cancer. We investigated (1) the relationship between surgical complexity and perioperative factors (patient characteristics, surgical process); (2) AI training for automated phase recognition and model performance was assessed by comparing predictions to the surgeon-annotated reference; (3) AI model predictability for surgical complexity was calculated by the area under the curve. RESULT: Surgical complexity score comprised extended total surgical duration, bleeding, and complications and was strongly associated with the intraoperative surgical process, especially in the beginning phases (area under the curve 0.913). We established an AI model that can recognize surgical phases from video with 87% accuracy; AI can determine intraoperative surgical complexity by calculating the duration of beginning phases from phases 1-3 (area under the curve 0.859). CONCLUSION: Surgical complexity, as a surrogate of short-term outcomes, can be predicted by the surgical process, especially in the extended duration of beginning phases. Surgical complexity can also be evaluated with automation using our artificial intelligence-based model.

Inhibitory effect of aspirin on inflammation-induced lung metastasis of cancer cells associated with neutrophil infiltration.

PURPOSE: Systematic inflammation has been reported to contribute to cancer progression through various mechanisms; however, the exact mechanism is still the subject of research. In this study, we evaluated the influence of systematic inflammation on lung metastasis, using a murine abdominal sepsis model, and assessed its relationship with pneumonia after curative esophagectomy in patients with esophageal cancer. METHODS: We used a murine abdominal sepsis model given highly metastatic osteosarcoma, to reveal the mechanism of systematic inflammation and its potential for lung metastasis. The therapeutic effect of aspirin (ASA) in preventing distant metastasis was also investigated. Subsequently, we analyzed, retrospectively, the relationship between pneumonia and lung metastasis after esophagectomy in patients who underwent esophagectomy at Keio University between January, 2007 and October, 2020. RESULTS: Abdominal sepsis provoked lung injury in the acute phase. ASA inhibited the recruitment of neutrophils triggered by the lung injury, and it also suppressed lung metastasis. Our retrospective study revealed that lung metastasis was more frequent in patients with postoperative pneumonia. CONCLUSIONS: Postoperative acute lung injury is associated with a higher risk of lung metastasis. ASA may be a potential preoperative treatment for inhibiting lung metastasis by preventing the recruitment of neutrophils.

Comparison of changes in body-fat mass and reflux esophagitis among reconstruction methods for proximal gastrectomy.

BACKGROUND: Although proximal gastrectomy (PG) is a function-preserving surgical option, it remains unclear as to which reconstruction method can prevent reflux and maintain body composition. METHODS: Patients who underwent PG at Keio University between April 2011 and November 2018 were analyzed. Changes in the subcutaneous and visceral adipose tissues were comparatively assessed before and after a year of surgery for three common reconstruction methods. We also compared the endoscopic findings of reflux esophagitis and the number of patients prescribed with proton-pump inhibitor after a year of surgery. RESULTS: This study included 76 patients, of which 33 patients underwent esophagogastrostomy with a circular stapler (CS), 35 under double flap (DF) reconstruction, and 8 underwent double tract (DT) reconstruction. Comparing esophagogastrostomy (CS and DF) and DT showed that esophagogastrostomy could significantly preserve both subcutaneous and visceral adipose tissues (P < 0.001 and P = 0.04, respectively). However, the change in the subcutaneous and visceral adipose tissues was comparable between CS and DF. As for reflux esophagitis, DF showed the lowest incidence rate for esophagitis and the least number of patients who were prescribed a proton-pump inhibitor. CONCLUSION: DF is a relatively better reconstruction method for preserving fat mass and preventing reflux among the three common reconstruction methods.

Long-term variation in psoas muscle mass index is affected by short-term loss after esophagectomy in survivors of esophageal cancer.

Changes in muscle mass may be an objective approach toward measuring the quality of life after surgery, but long-term changes due to surgery without the effect of cachexia remain unclear. Patients with esophageal cancer who underwent esophagectomy and did not experience cancer recurrence for 3 years were analyzed. The psoas muscle mass index (PMI) was assessed before surgery and 7 days, 1 year, and 3 years after surgery. Patients with no change or increased PMI within 7 days after surgery were categorized into the Nondecreasing-PMI group, whereas those with decreased PMI were categorized into the Decreasing-PMI group. Eighty-four and 51 patients were categorized into the Nondecreasing- and Decreasing-PMI groups, respectively. The Decreasing-PMI group had a higher incidence rate of anastomotic leakage than the Nondecreasing-PMI group (25% vs. 12%, respectively; P = 0.042). Moreover, the Decreasing-PMI group showed a significantly greater decrease in the PMI 1 year after surgery than the Nondecreasing-PMI group (-9.2% vs. -4.0%, respectively; P = 0.048). However, although the Decreasing-PMI group had a greater decrease in the PMI than the Nondecreasing-PMI group, no significant difference was observed 3 years after surgery (-9.8% vs. -5.3%, respectively; P = 0.115). A decrease in PMI in the acute phase after esophagectomy may contribute to a long-term decrease in the PMI. Intensive interventions may be beneficial for these patients to improve their long-term quality of life.

Specificity of Presepsin as a Biomarker of Bacterial Infection in Mouse Sepsis Models.

INTRODUCTION: Since its discovery in 2002, presepsin (P-SEP) has been reported to be useful in the early diagnosis of sepsis and has been evaluated in many clinical studies. However, as antibodies that bind to mouse P-SEP were previously unavailable, serum P-SEP levels in mice are limited. This study used a P-SEP enzyme-linked immunosorbent assay kit to evaluate the changes in serum P-SEP levels in mouse sepsis models compared with changes in other inflammatory markers and determine whether P-SEP can function as a biomarker specific to bacterial infections. METHODS: Sepsis was induced in mice via cecal ligation and puncture (CLP), induction with lipopolysaccharide (LPS), and cecal ligation (CL) model was created as a control for the CLP model, following which clinical biomarkers (P-SEP, C-reactive protein, and procalcitonin) were evaluated. RESULTS: The 48-h survival rates in the CLP, CL, and LPS-induced sepsis models were 67%, 89%, and 57%, respectively. Serum C-reactive protein levels did not increase in the CLP and CL models within 24 h but significantly increased in the LPS-induced sepsis model. Serum procalcitonin levels increased in the CLP and CL models and especially increased in the LPS-induced sepsis model. In contrast, an increase in serum P-SEP level was found in the CLP model at 6 h compared with those at baseline, the CL, and LPS-induced sepsis models. CONCLUSIONS: Mouse P-SEP is elevated early in infection and more specific to bacterial infection compared with other biomarkers.