Machine learning-based model for prediction and feature analysis of recurrence in pancreatic neuroendocrine tumors G1/G2.

BACKGROUND: Pancreatic neuroendocrine neoplasms (PanNENs) are a heterogeneous group of tumors. Although the prognosis of resected PanNENs is generally considered to be good, a relatively high recurrence rate has been reported. Given the scarcity of large-scale reports about PanNEN recurrence due to their rarity, we aimed to identify the predictors for recurrence in patients with resected PanNENs to improve prognosis. METHODS: We established a multicenter database of 573 patients with PanNENs, who underwent resection between January 1987 and July 2020 at 22 Japanese centers, mainly in the Kyushu region. We evaluated the clinical characteristics of 371 patients with localized non-functioning pancreatic neuroendocrine tumors (G1/G2). We also constructed a machine learning-based prediction model to analyze the important features to determine recurrence. RESULTS: Fifty-two patients experienced recurrence (14.0%) during the follow-up period, with the median time of recurrence being 33.7 months. The random survival forest (RSF) model showed better predictive performance than the Cox proportional hazards regression model in terms of the Harrell's C-index (0.841 vs. 0.820). The Ki-67 index, residual tumor, WHO grade, tumor size, and lymph node metastasis were the top five predictors in the RSF model; tumor size above 20 mm was the watershed with increased recurrence probability, whereas the 5-year disease-free survival rate decreased linearly as the Ki-67 index increased. CONCLUSIONS: Our study revealed the characteristics of resected PanNENs in real-world clinical practice. Machine learning techniques can be powerful analytical tools that provide new insights into the relationship between the Ki-67 index or tumor size and recurrence.

Efficacy of Combination Therapy with Epinephrine Local Injection and Hemostatic Clips on Active Diverticular Bleeding.

Epinephrine local injection is a hemostatic procedure used in active diverticular bleeding that elicits vasoconstriction and tamponade effects. We compared the additional benefit of combination therapy with HSE-C (hypertonic saline epinephrine injection with clipping) to clipping monotherapy. Retrospective data on diverticular bleeding between 2011 and 2016 was reviewed. Those with an active bleeding source confirmed by colonoscopy (excluding non-bleeding vessels and adherent clots) who received either HSE-C or clipping were evaluated. Endpoints were rates of successful primary hemostasis, recurrent bleeding, and surgical intervention during hospitalization. A total of 320 patients with diverticular bleeding were evaluated, on which either HSE-C (n = 35) or clipping monotherapy (n = 18) was performed. Rates of successful primary hemostasis (91.4% vs. 66.7%, p = 0.048) and direct placement of endoclips (60.0% vs. 16.7%, p = 0.004) were significantly higher in the HSE-C group. Although not statistically significant, the HSE-C group had a higher rate of early rebleeding (18.8% vs. 8.3%, p = 0.653), while no difference was seen in the number of patients requiring surgery (11.4% vs. 5.5%, p = 0.651). HSE-C is associated with a higher rate of successful primary hemostasis for severe active diverticular bleeding but has no significant difference in reducing early recurrent bleeding or the number of patients requiring surgery, suggesting that hemostatic effects may be temporary.

Performance of novel tissue harmonic echo imaging using endoscopic ultrasound for pancreatic diseases.

BACKGROUND AND STUDY AIMS: Recently, tissue harmonic echo (THE) imaging has advanced with the development of a new endoscopic ultrasound (EUS) monitor/processing unit. With this new technology, penetration (THE-P) and resolution (THE-R) images can be obtained. The aim of this study was to investigate the performance of this novel THE imaging using a new processing unit for pancreatic diseases. PATIENTS AND METHODS: Fifty patients with pancreatic lesions (38 cystic, 12 solid) were retrospectively analyzed. At each examination, 3 EUS images of the same pancreatic lesion were obtained using B-mode, THE-P mode, and THE-R mode imaging. Each set of EUS images was randomly arranged and evaluated independently by 4 physicians blinded to the imaging technique. Images were compared using a Likert scale 5-point grading system for each parameter. RESULTS: For cystic lesions, THE-P mode images were significantly superior to conventional B-mode images for visualizing the boundary, septum, nodules, and total image quality (P < 0.05). THE-R mode images were significantly superior to conventional B-mode images for visualizing the boundary, septum, and total image quality (P < 0.05). However, for solid lesions, there was no significant difference in all the evaluation points between THE-P and conventional B-mode images. THE-R mode images were inferior to conventional B-mode images for visualizing the boundary, internal structure, and total image quality (P < 0.05). CONCLUSIONS: For pancreatic cystic lesions, THE mode images provided better lesion characterization than conventional B-mode images. Further research is required to determine if this improvement will result in improved EUS diagnostics.

Diagnostic ability of EUS-FNA for pancreatic solid lesions with conventional 22-gauge needle using the slow pull technique: a prospective study.

OBJECTIVE: Endoscopic ultrasonography-guided fine needle aspiration (EUS-FNA) using the slow pull technique (SP-FNA) has recently attracted attention as an effective tissue acquisition technique. However, efficacy of SP-FNA with a 22-gauge conventional needle remains unclear. The aim of this study is to evaluate the diagnostic ability of SP-FNA with a 22-gauge needle. MATERIAL AND METHODS: Patients with a pancreatic solid lesion were prospectively enrolled in this study. SP-FNA was performed at two needle passes with a 22-gauge needle. One dedicated pathologist evaluated the obtained samples in terms of quantity (Grade 0: scant; Grade 1: inadequate; Grade 2: adequate), quality (Grade 0: poor; Grade 1: moderate; Grade 2: good), and blood contamination (Grade 0: significant; Grade 1: moderate; Grade 2: low), and provided a pathological diagnosis. Additional EUS-FNA was performed by applying suction (SA-FNA). The evaluation points were as follows: diagnostic accuracy of SP-FNA compared with that of SA-FNA, and the quantity, quality, and blood contamination level of SP-FNA-obtained samples. RESULTS: We enrolled 40 cases. The diagnostic accuracy of SP-FNA was 90% (36/40). There was no significant difference in the accuracy between SP-FNA and SA-FNA (90% vs. 90%, p = 1.000). The samples obtained using SP-FNA were assessed as Grade 2 for quantity in 29 cases (73%), quality in 31 (78%), and blood contamination in 25 (63%). CONCLUSIONS: Adequate, high-quality, and unsubstantially blood-contaminated samples could be obtained using SP-FNA. The diagnostic ability of SP-FNA was 90%, which appeared to be similar to that of SA-FNA.

Feasibility, efficacy, and predictive factors for the technical success of endoscopic nasogallbladder drainage: a prospective study.

BACKGROUND/AIMS: Several studies have shown the useful-ness of endoscopic nasogallbladder drainage (ENGBD) in patients with acute cholecystitis. However, the procedure is difficult, and factors that affect technical success have not yet been clarified. We conducted a prospective study to eval-uate the technical feasibility, efficacy, and predictive factors for the technical success of ENGBD in patients with acute cholecystitis. METHODS: All patients with moderate or severe acute cholecystitis who were enrolled underwent ENGBD between April 2009 and April 2011. Patients with surgically altered anatomy or pancreatobiliary malignancies were ex-cluded. The primary outcomes included technical success, clinical success, and complications. Factors that could affect the technical success were also examined. RESULTS: Of the 27 patients who underwent ENGBD during the study period, technical success was achieved in 21 (78%) and clinical improvement was achieved in 20 (95%). Early complications were encountered in four patients (15%). Gallbladder wall thickness (odds ratio [OR], 1.64; 95% confidence interval [CI], 1.08 to 2.47) and age (OR, 1.16; 95% CI, 1.00 to 1.35) were effective predictors of technical failure. CONCLUSIONS: ENGBD was effective in resolving acute cholecystitis; however, this modality was technically challenging and had a limited suc-cess rate. Because of technical difficulties, ENGBD should be reserved for limited indications. (Gut Liver, 2015;9239-246).

A novel dry model for practicable sphincterotomy and precut needle knife sphincterotomy.

Aim. We aimed to develop a simulation dry model for endoscopic sphincterotomy (ES) and needle knife precut sphincterotomy (NKP) and to evaluate its usefulness as a training simulator. Materials and Methods. An endoscopic retrograde cholangiopancreatography trainer was used as a duodenum, bile duct, and papilla simulator. A simulated papilla was created with a piece of rolled uncured ham, and ES and NKP were performed. Hands-on training was carried out using this model, and success and failure of the procedures were evaluated. A questionnaire survey was conducted among the participants to assess the performance and usefulness of the dry model for ES and NKP training. Results. Twenty-two endoscopists participated in the hands-on training using this dry model. ES was successful in 33 out of 34 attempts (97%) whereas NKP was successful in all 7 attempts (100%). Based on the results of the questionnaire survey, the median score for realism was 7 (range: 2-9) for ES and 8 for NKP on a scale of 1 to 10. Conclusions. The dry model using an uncured ham provides a condition closely similar to actual clinical practice and is useful as a training model for ES and NKP.

Efficacy, Safety, and Long-Term Follow-Up Results of EUS-Guided Transmural Drainage for Pancreatic Pseudocyst.

Background and Aim. EUS-guided transmural drainage (EUS-GTD) is now considered a minimally invasive and effective alternative to surgery for drainage of symptomatic pancreatic pseudocysts. However, the technique is rather difficult, and sometimes serious complications occur to patients undergoing this procedure. We retrospectively evaluated efficacy, safety, and long-term follow-up results of EUS-GTD for pancreatic pseudocyst. Methods. Sixty-seven patients with pancreatic pseudocyst who underwent EUS-GTD from April 2000 to March 2011 were enrolled. We retrospectively evaluated (1) technical success, (2) clinical success, (3) adverse event of procedure, and (4) long-term follow-up results. Results. Total technical success rate was 88%. Ninety-one percent of external drainage, 79% of internal drainage, and 66% of puncture and aspiration only achieved clinical success. There was only one case with an adverse event, perforation (1.5%). The case required emergency operation. Total recurrence rate was 23.9%. Median follow-up period was 33.9 months. The recurrence rates in the cases of stent remaining, spontaneously dislodged, removed on schedule, external tube removal, and aspiration only were 10.0%, 12.5%, 42.9%, 50%, and 0%, respectively. Conclusion. EUS-GTD is a relatively safe and effective therapeutic method. However, further analysis should be done by larger series to determine the method of EUS-GTD for pancreatic pseudocyst.

[A case of 5-fluorouracil-induced hyperammmonia after chemotherapy for metastatic colon cancer].

A 79-year-old woman with colon cancer and multiple liver metastases was admitted to our hospital for systemic chemotherapy. She underwent first cycle of modified FOLFOX6 chemotherapy. She was confused on treatment day 5. Blood test revealed her serum ammonia level to be 121 microg/dl. We diagnosed 5-fluorouracil (5FU)-induced hyperammonemia. Conservative treatment resulted in improvement of metal status. The reason for hyperammonemia after administration of 5FU was the excess production of ammonium from metabolites of 5FU.

Whipple's disease: the first Japanese case diagnosed by electron microscopy and polymerase chain reaction.

A 52-year-old man presented with diarrhea and 20 kg weight loss in one year. Enteroscopy showed diffuse yellow-white shaggy mucosa in the duodenum and jejunum. Biopsies of the duodenal mucosa revealed massive infiltration within the lamina propria by foamy macrophages strongly positive for periodic acid-Schiff stain. Electron microscopy demonstrated numerous bacilli within macrophages of the lamina propria. Tropheryma whipplei DNA was detected by polymerase chain reaction. The definitive diagnosis of Whipple's disease was made. Antibiotic therapy dramatically improved his clinical picture. This is the first Japanese case with Whipple's disease diagnosed by electron microscopy and polymerase chain reaction.