Deep learning-based diagnostic model for predicting complications after gastrectomy.

BACKGROUND: Gastric cancer is one of the leading causes of cancer deaths, and gastrectomy with lymph node dissection is the mainstay of treatment. Despite clinician efforts and advances in surgical methods, the incidence of complications after gastrectomy remains 10%-20% including fatalities. To the best of our knowledge, this is the first report on utilization of a deep learning method to build a new artificial intelligence model that could help surgeons diagnose these complications. METHODS: A neural network was constructed with a total of 4000 variables. Clinical, surgical, and pathological data of patients who underwent radical gastrectomy at our institute were collected to maintain a deep learning model. We optimized the parameters of the neural network to diagnose whether these patients would develop complications after gastrectomy or not. RESULTS: Seventy percent of the data was used to optimize the neural network parameters, and the rest was used to validate the model. A model that maximized the receiver operating characteristics (ROC) area under the curve (AUC) for validation of the data was extracted. The ROC-AUC, sensitivity, and specificity of the model to diagnose all complications were 0.8 vs 0.7, 81% vs 50%, and 69% vs 75%, for the teaching and validation data, respectively. CONCLUSIONS: A predictive model for postoperative complications after radical gastrectomy was successfully constructed using the deep learning method. This model can help surgeons accurately predict the incidence of complications on postoperative day 3.

Impact of Preoperative Time Interval on Survival in Patients With Gastric Cancer.

BACKGROUND: A time interval between diagnosis and surgery for gastric cancer is necessary, although its impact on survival remains controversial. We evaluated the impact of preoperative time interval on survival in gastric cancer patients. METHODS: We enrolled 332 patients who underwent curative gastrectomy for clinical stage (cStage) I-III gastric cancer between 2012 and 2015. We separately analyzed early- (cStage I) and advanced-stage (cStages II and III) patients. Early-stage patients were divided according to preoperative time interval: short (≤ 42 days) and long (> 42 days) groups. Advanced-stage patients were also divided into short (≤ 21 days) and long (> 21 days) groups. We compared the survival between the short and long groups in early- and advanced-stage patients. RESULTS: The median preoperative time interval was 29 days, and no significant differences were found in patient characteristics between the short and long groups in early- and advanced-stage patients. In early-stage patients, the 5-year survival rates of the short and long groups were 86.5% and 88.4%, respectively (P = 0.917). In advanced-stage patients, the 5-year survival rates were 72.1% and 70.0%, respectively (P = 0.552). In multivariate analysis, a longer time interval was not selected as an independent prognostic factor in early- and advanced-stage patients. CONCLUSIONS: In this study, survival difference was not found based upon preoperative time interval. The results do not affirm the delay of treatment without reason, however, imperative extension of preoperative time interval may be justified from the standpoint of long-term survival.

Preoperative lymphocyte-to-monocyte ratio is the most predictive inflammatory response marker of survival in gastric cancer.

PURPOSE: Systemic inflammatory responses play a key role in cancer progression, and detecting the predictive inflammatory response markers is needed. The present study explored inflammatory response markers capable of predicting survival in patients with gastric cancer. METHODS: We enrolled 264 patients, who underwent curative gastrectomy for clinical stage (cStage) I-III gastric cancer between 2012 and 2015. The cut-off point of eight preoperative inflammatory response markers was determined by receiver operating characteristic (ROC) curve analysis. The marker with the highest Harrell's concordance index (C-index) was adopted for subsequent univariate and multivariate analyses using the Cox proportional-hazards model. RESULTS: Among eight representative inflammatory response markers, lymphocyte-to-monocyte ratio (LMR; cut-off point, 4.60) achieved the highest C-index (0.633). The 5-year survival rate was significantly worse in patients with LMR < 4.60 than in those with LMR ≥ 4.60 (67.5% versus 89.0%, P < 0.001). In multivariate analysis, LMR < 4.60 was identified as an independent prognostic factor (hazard ratio: 2.372; 95% confidence interval: 1.266-4.442; P = 0.007). CONCLUSION: In this study, LMR had the strongest ability to predict the survival of patients with gastric cancer among other inflammatory response markers, with lower LMRs being associated with poor survival following curative gastrectomy.

Surgical and nutritional outcomes of laparoscopic proximal gastrectomy versus total gastrectomy: a meta-analysis.

BACKGROUND: Laparoscopic proximal gastrectomy (LPG) is regarded as a less invasive surgery than laparoscopic total gastrectomy (LTG) for early gastric cancer located on the proximal side of the stomach. However, whether LPG is more effective than LTG remains unclear. METHODS: A systematic literature search of studies assessing short-term surgical and nutritional outcomes after LPG and LTG was conducted. A meta-analysis of surgical outcomes (operative time, intraoperative estimated blood loss, postoperative complications, and length of hospital stay) and nutritional outcomes (decrease in body weight, albumin, hemoglobin, total protein, and lymphocyte count) was then performed. All of 11 papers are a retrospective cohort study. RESULTS: Eleven studies reported assessments of the above-mentioned outcomes in 883 patients. There was a trend towards shorter operative time and lower blood loss for LPG compared to LTG though not reaching statistical significance. Other surgical outcomes showed no significant differences. Patients who underwent LTG had a significantly lower body weight (95% confidence interval, 3.01-6.05, [Formula: see text] = 4.53, p < 0.01) and hemoglobin level (95% confidence interval, 1.88-5.87, [Formula: see text] = 3.87, p < 0.01) than patients who underwent LPG at 1 year after surgery. There were no significant differences in other nutritional outcomes. CONCLUSIONS: These results indicate LPG had some advantages in postoperative nutrition. However, no significant differences in short-term surgical outcomes were noted between the two operations. Our analysis suggests that LPG may be more beneficial compared with LTG in terms of perioperative and nutritional outcomes for early-stage gastric cancer.

Indolent T cell lymphoproliferative disorder with villous atrophy in small intestine diagnosed by single-balloon enteroscopy.

Chronic diarrhea is one of the major symptoms in gastroenterology. However, this may be caused by pathologic conditions for which the diagnosis is critical. Villous atrophy, as an endoscopic lesion, accompanied by chronic diarrhea can occasionally be observed in the patients with inflammatory diseases of the gastrointestinal (GI) tract. Herein, we present a case with persistent diarrhea accompanied by intestinal wall thickening without any other significant endoscopic features other than villous atrophy in the jejunum and the ileum, where we diagnosed as an indolent T cell lymphoproliferative disorder (T-LPD) of the GI tract, defined in the 2016-2017 revised World Health Organization classification, via single-balloon enteroscopy (SBE). Interestingly, we found the same lymphocyte infiltration from the distal third portion of the duodenum, where gastroscopy could not reach, via SBE, even though no endoscopic findings were observed such as villous atrophy. Since infiltrating cells in the intestinal tissues were CCR4+, mogamulizumab was administered with resulting durable symptomatic remission for more than 2 years. Patients with persistent diarrhea may have serious small intestinal disorder including not only chronic inflammatory diseases but also lymphoid neoplasmic conditions including T-LPD of GI tract.