High incidence of lung cancer death after curative endoscopic submucosal dissection for superficial esophageal squamous cell carcinoma.

BACKGROUND AND AIM: Following treatment of superficial esophageal squamous cell carcinoma (ESCC), surveillance for a second primary malignancy (SPM) is necessary. However, detailed evidence regarding the timing and prognosis of SPMs is insufficient. We aimed to clarify the details of SPMs and their effects on patient outcomes. METHODS: This retrospective, multicenter study involved 11 hospitals. Patients with superficial ESCC curatively resected using endoscopic submucosal dissection between May 2005 and December 2012, were included in this study. RESULTS: The 5-year survival rate of 187 patients was 92.6% during a median follow-up duration of 96.8 months. Thirty-one patients died, 14 of whom died of SPMs. Compared to patients with SPMs detectable by esophagogastroduodenoscopy (EGD), patients with SPMs detectable only by modalities other than EGD had a significantly higher mortality rate (p < 0.001). Patients with second primary lung cancer (LC) had a high mortality rate (56.3%). Univariate and multivariate analyses showed that multiple Lugol-voiding lesions (LVLs) tended to be associated with SPMs (p = 0.077, hazard ratio [HR] 4.43, 95% confidence interval [CI]: 0.91-6.50), and metachronous ESCC was an independent risk factor for the incidence of second primary LC (p = 0.037, HR 3.51, 95% CI: 1.08-11.41). CONCLUSIONS: SPMs that cannot be detected by EGD, such as LC, must be considered after the curative resection of ESCC. We suggest strict screening by both EGD and computed tomography for patients with multiple LVLs or metachronous ESCC to detect SPMs in their early stages.

A machine learning model for predicting the lymph node metastasis of early gastric cancer not meeting the endoscopic curability criteria.

BACKGROUND: We developed a machine learning (ML) model to predict the risk of lymph node metastasis (LNM) in patients with early gastric cancer (EGC) who did not meet the existing Japanese endoscopic curability criteria and compared its performance with that of the most common clinical risk scoring system, the eCura system. METHODS: We used data from 4,042 consecutive patients with EGC from 21 institutions who underwent endoscopic submucosal dissection (ESD) and/or surgery between 2010 and 2021. All resected EGCs were histologically confirmed not to satisfy the current Japanese endoscopic curability criteria. Of all patients, 3,506 constituted the training cohort to develop the neural network-based ML model, and 536 constituted the validation cohort. The performance of our ML model, as measured by the area under the receiver operating characteristic curve (AUC), was compared with that of the eCura system in the validation cohort. RESULTS: LNM rates were 14% (503/3,506) and 7% (39/536) in the training and validation cohorts, respectively. The ML model identified patients with LNM with an AUC of 0.83 (95% confidence interval, 0.76-0.89) in the validation cohort, while the eCura system identified patients with LNM with an AUC of 0.77 (95% confidence interval, 0.70-0.85) (P = 0.006, DeLong's test). CONCLUSIONS: Our ML model performed better than the eCura system for predicting LNM risk in patients with EGC who did not meet the existing Japanese endoscopic curability criteria. We developed a neural network-based machine learning model that predicts the risk of lymph node metastasis in patients with early gastric cancer who did not meet the endoscopic curability criteria.

A novel artificial intelligence-based endoscopic ultrasonography diagnostic system for diagnosing the invasion depth of early gastric cancer.

BACKGROUND: We developed an artificial intelligence (AI)-based endoscopic ultrasonography (EUS) system for diagnosing the invasion depth of early gastric cancer (EGC), and we evaluated the performance of this system. METHODS: A total of 8280 EUS images from 559 EGC cases were collected from 11 institutions. Within this dataset, 3451 images (285 cases) from one institution were used as a development dataset. The AI model consisted of segmentation and classification steps, followed by the CycleGAN method to bridge differences in EUS images captured by different equipment. AI model performance was evaluated using an internal validation dataset collected from the same institution as the development dataset (1726 images, 135 cases). External validation was conducted using images collected from the other 10 institutions (3103 images, 139 cases). RESULTS: The area under the curve (AUC) of the AI model in the internal validation dataset was 0.870 (95% CI: 0.796-0.944). Regarding diagnostic performance, the accuracy/sensitivity/specificity values of the AI model, experts (n = 6), and nonexperts (n = 8) were 82.2/63.4/90.4%, 81.9/66.3/88.7%, and 68.3/60.9/71.5%, respectively. The AUC of the AI model in the external validation dataset was 0.815 (95% CI: 0.743-0.886). The accuracy/sensitivity/specificity values of the AI model (74.1/73.1/75.0%) and the real-time diagnoses of experts (75.5/79.1/72.2%) in the external validation dataset were comparable. CONCLUSIONS: Our AI model demonstrated a diagnostic performance equivalent to that of experts.

Recent steroid use and the relapse risk in ulcerative colitis patients with endoscopic healing.

BACKGROUND: Endoscopic healing (EH) is a therapeutic target in ulcerative colitis (UC). However, even patients who have achieved EH relapse frequently. AIMS: To investigate the association between recent steroid use and relapse risk in UC patients with EH. METHODS: This multi-centre cohort study included 1212 UC patients with confirmed EH (Mayo endoscopic subscore ≤1). We excluded patients with current systemic steroid use or history of advanced therapy. We divided patients into a recent steroid group (last systemic steroid use within 1 year; n = 59) and a non-recent or steroid-naïve group (n = 1153). We followed the patients for 2 years to evaluate relapse, defined as induction of systemic steroids or advanced therapy. We used logistic regression to estimate the odds ratio (OR) of relapse. RESULTS: Relapse occurred in 28.8% of the recent steroid group and 5.6% of the non-recent/steroid-naïve group (multi-variable-adjusted OR 5.53 [95% CI 2.85-10.7]). The risk of relapse decreased with time since the last steroid use: 28.8% for less than 1 year after steroid therapy, 22.9% for 1 year, 16.0% for 2 years and 7.9% beyond 3 years, approaching 4.0% in steroid-naïve patients. (ptrend <0.001). CONCLUSIONS: Even for patients with UC who achieved EH, the risk of relapse remains high following recent steroid therapy. Physicians need to consider the duration since last steroid use to stratify the relapse risk in UC patients with EH.

Real-world NUDT15 genotyping and thiopurine treatment optimization in inflammatory bowel disease: a multicenter study.

BACKGROUND: This study evaluated the effectiveness of NUDT15 codon 139 genotyping in optimizing thiopurine treatment for inflammatory bowel disease (IBD) in Japan, using real-world data, and aimed to establish genotype-based treatment strategies. METHODS: A retrospective analysis of 4628 IBD patients who underwent NUDT15 codon 139 genotyping was conducted. This study assessed the purpose of the genotyping test and subsequent prescriptions following the obtained results. Outcomes were compared between the Genotyping group (thiopurine with genotyping test) and Non-genotyping group (thiopurine without genotyping test). Risk factors for adverse events (AEs) were analyzed by genotype and prior genotyping status. RESULTS: Genotyping test for medical purposes showed no significant difference in thiopurine induction rates between Arg/Arg and Arg/Cys genotypes, but nine Arg/Cys patients opted out of thiopurine treatment. In the Genotyping group, Arg/Arg patients received higher initial doses than the Non-genotyping group, while Arg/Cys patients received lower ones (median 25 mg/day). Fewer AEs occurred in the Genotyping group because of their lower incidence in Arg/Cys cases. Starting with < 25 mg/day of AZA reduced AEs in Arg/Cys patients, while Arg/Arg patients had better retention rates when maintaining ≥ 75 mg AZA. Nausea and liver injury correlated with thiopurine formulation but not dosage. pH-dependent mesalamine reduced leukopenia risk in mesalamine users. CONCLUSIONS: NUDT15 codon 139 genotyping effectively reduces thiopurine-induced AEs and improves treatment retention rates in IBD patients after genotype-based dose adjustments. This study provides data-driven treatment strategies based on genotype and identifies risk factors for specific AEs, contributing to a refined thiopurine treatment approach.