Machine learning prediction for COVID-19 disease severity at hospital admission.

IMPORTANCE: Early prognostication of patients hospitalized with COVID-19 who may require mechanical ventilation and have worse outcomes within 30 days of admission is useful for delivering appropriate clinical care and optimizing resource allocation. OBJECTIVE: To develop machine learning models to predict COVID-19 severity at the time of the hospital admission based on a single institution data. DESIGN, SETTING, AND PARTICIPANTS: We established a retrospective cohort of patients with COVID-19 from University of Texas Southwestern Medical Center from May 2020 to March 2022. Easily accessible objective markers including basic laboratory variables and initial respiratory status were assessed using Random Forest's feature importance score to create a predictive risk score. Twenty-five significant variables were identified to be used in classification models. The best predictive models were selected with repeated tenfold cross-validation methods. MAIN OUTCOMES AND MEASURES: Among patients with COVID-19 admitted to the hospital, severity was defined by 30-day mortality (30DM) rates and need for mechanical ventilation. RESULTS: This was a large, single institution COVID-19 cohort including total of 1795 patients. The average age was 59.7 years old with diverse heterogeneity. 236 (13%) required mechanical ventilation and 156 patients (8.6%) died within 30 days of hospitalization. Predictive accuracy of each predictive model was validated with the 10-CV method. Random Forest classifier for 30DM model had 192 sub-trees, and obtained 0.72 sensitivity and 0.78 specificity, and 0.82 AUC. The model used to predict MV has 64 sub-trees and returned obtained 0.75 sensitivity and 0.75 specificity, and 0.81 AUC. Our scoring tool can be accessed at https://faculty.tamuc.edu/mmete/covid-risk.html . CONCLUSIONS AND RELEVANCE: In this study, we developed a risk score based on objective variables of COVID-19 patients within six hours of admission to the hospital, therefore helping predict a patient's risk of developing critical illness secondary to COVID-19.

Ex vivo human bile duct radiofrequency ablation with a bipolar catheter.

BACKGROUND: Management of the primary and secondary tumors of the bile ducts still remains as a major clinical challenge. Radiofrequency (RF) ablation (RFA) of these tumors is feasible but the effect of RF energy on the human common bile duct (CBD) and surrounding tissues has not been investigated. This pilot study aimed to determine the relationship between RF energy and the depth of ablation in the normal human CBD. METHODS: The study was performed on fresh ex vivo human biliary-pancreatic tissue which had been resected for a pancreatic cyst or mass. The study was conducted within 15 min after resection. A bipolar Habib RFA catheter was placed into the middle of the intact CBD, and three different (5, 7, 10 W) power settings were applied over a 90-s period by an RF generator. Gross and histological examinations were performed. The depth of coagulation necrosis in CBD and the effect of RFA on CBD wall and surrounding pancreas tissue were determined by microscopic examination. RESULTS: The study included eight tissue samples. 5 W power was applied to three sites and RFA caused only focal epithelial necrosis limited to the CBD mucosa. 7 and 10 W were applied to five sites and coagulation necrosis occurred in all cases. Microscopically, necrosis was transmural, involved accessory bile duct glands, and extended to the surrounding pancreatic tissue in four of these cases. Macroscopically, RFA resulted in circumferential white-yellowish color change extending approximately 2 cm of the CBD. CONCLUSION: Bipolar RF energy application with 5 W resulted in limited ablation on CBD wall. However, 7 and 10 W generated tissue necrosis which extended through the CBD wall and into surrounding pancreas tissue. Endoscopic biliary RFA is an effective technique for local biliary tissue ablation but the use of high energy may injure surrounding tissue.

Needle-Based Confocal Laser Endomicroscopy for Evaluation of Cystic Neoplasms of the Pancreas.

BACKGROUND: The accurate diagnosis of cystic neoplasms of the pancreas (CNP) with current diagnostic methods is limited. Endoscopic ultrasound (EUS)-guided needle-based confocal laser endomicroscopy (nCLE) is a new technique which can obtain images from the cyst wall during EUS-fine needle aspiration (EUS-FNA). The aim of this study was to assess the safety, feasibility, and diagnostic value of nCLE for CNP. METHODS: Patients who underwent EUS-FNA to evaluate a CNP larger than 2 cm were enrolled. The cyst was punctured with 19-G FNA needle preloaded with an nCLE probe. The images from different areas of the cyst wall were recorded. Using the final diagnosis defined by surgery or EUS-FNA cyst fluid analysis, the accuracy of the confocal images was defined. RESULTS: The procedure and image acquisition was successful in 18 of the 20 patients. Predefined typical structures for mucinous cysts were visualized in 8 of 12 (66%) cysts but none of the non-mucinous cysts. The superficial vascular network which is a typical finding of serous cysts was observed in 2 of 3 patients. The sensitivity, specificity, and diagnostic accuracy of the findings of epithelial structures by nCLE were 66, 100, and 80%, respectively, for a mucinous cyst diagnosis. All patients tolerated the procedure well, and no adverse effects were determined. CONCLUSION: nCLE was found to be safe and feasible with high technical success, in this pilot study. With an overall accuracy of 80%, it has the potential to contribute to the diagnosis of CNP with specific imaging.

Imaging of pancreatic cystic lesions with confocal laser endomicroscopy: an ex vivo pilot study.

BACKGROUND: The differential diagnosis of pancreatic cystic lesions (PCLs) is an increasingly common clinical challenge. Confocal laser endomicroscopy (CLE) may differentiate PCLs by imaging of the cyst wall. However, clinical experience is still limited, and better image definition and characterization of the cyst wall in a spectrum of cysts are needed. This experimental study aimed to expose detailed imaging characteristics of PCLs by CLE. METHODS: Patients who underwent surgery of a PCL were enrolled. During surgery, intravenous fluorescein (2.5 ml of 10%) was injected just prior to the ligation of blood vessels supplying the pancreas. The freshly excised specimens were transected along the long axis to fully expose the luminal surface. A Gastroflex-UHD CLE probe (pCLE) was used manually to acquire images directly from the surface of cyst wall. The specimen subsequently underwent cross-sectional histology. All recorded data were analyzed by two investigators for predefined and original image findings of PCLs. RESULTS: Ten cases were recruited into the study. All patients underwent surgery because of a mucinous cyst with worrisome features or a symptomatic PCL. Imaging was successful in all patients and differently shaped papillary projections (PP) were visualized in eight patients. Pathological examination of those patients confirmed 6 cases with Intraductal Papillary Mucinous Neoplasm (IPMN) and 2 cases with Mucinous Cystic Neoplasm (MCN). In two patients with serous cystadenoma, typical vascular network was visualized in one patient, and microcystic structures in the other. Three of the IPMNs were malignant. The loss of papillary margin integrity and significant fragmentation together with irregularity was detected in malignant IPMNs by CLE. CONCLUSIONS: Pancreatic cyst epithelial wall can be visualized successfully by pCLE in ex vivo surgical specimens. Different papillary projections have been seen in all cases of IPMNs and MCNs. CLE has potential for identifying IPMN subtypes and for grading dysplasia.

Value of adding GNAS testing to pancreatic cyst fluid KRAS and carcinoembryonic antigen analysis for the diagnosis of intraductal papillary mucinous neoplasms.

BACKGROUND AND AIM: Molecular analysis of pancreatic cyst fluid (PCF) has been proposed as a novel method for differentiating pancreatic cystic lesions (PCL). The present study aimed to investigate the value of GNAS testing when added to KRAS and carcinoembryonic antigen (CEA) testing of PCF for the diagnosis of intraductal papillary mucinous neoplasms (IPMN). METHODS: Prospectively collected endoscopic ultrasonography fine-needle aspiration (EUS-FNA) data were analyzed retrospectively for GNAS and KRAS mutations and CEA results. IPMN were histologically confirmed or supported by imaging and EUS-FNA findings (KRAS, CEA, cytology). Performance characteristics of GNAS added to KRAS and CEA for the diagnosis of IPMN were calculated. RESULTS: The study population consisted of 197 patients with cyst fluid test results. Cysts were histologically classified in 33 patients and by clinical criteria in 164 patients. The IPMN group included 108 patients and the non-IPMN group included 89 patients. GNAS was positive in 51 patients (47.2%) with IPMN. Forty-two of these patients (82.3%) also had a KRAS mutation. Adding GNAS to KRAS increased the diagnostic accuracy from 76.6% to 79.1% (P > 0.05). Adding GNAS to CEA increased the diagnostic accuracy from 66.4% to 80.7 % (P < 0.05), but did not achieve a diagnostic superiority to KRAS testing alone (80.7% vs 76.6%, P > 0.05). The diagnostic accuracy of the triple combination was significantly better than all single tests (P < 0.05). CONCLUSION: GNAS mutation is a highly specific test for IPMN. When GNAS testing is added to CEA and KRAS, a significantly greater overall accuracy (86.2%) is achieved.

Radiofrequency ablation for the management of occluded biliary metal stents.

Background and study aim: Obstruction of biliary self-expandable metal stents (SEMSs) is seen frequently. Radiofrequency ablation (RFA) causes tissue necrosis. This study aimed to assess the efficacy of RFA for management of occluded SEMS. Patients and methods: Patients with biliary malignancy and treated for an occluded SEMS were retrospectively reviewed. The study group comprised patients treated with RFA using an Habib endoprobe inside the SEMS. The control group comprised patients treated only with insertion of a plastic stent into an occluded SEMS. The end points were; 90-day stent patency rate, time to stent reocclusion, 30-day mortality, and 3 - and 6-month survival rates. Results: During the 5-year study period 25 patients with an occluded SEMS underwent RFA and 25 patients underwent a plastic stent placement only. Both groups were matched for age and diagnosis. Immediate biliary drainage was restored in all patients. In the RFA group, the stenosis was ablated successfully in 14 patients (56 %); ablation failed in 11 and a plastic stent was then inserted into the SEMS for these. The control group had only placement of a plastic stent across the stenosis. Stent patency rate at 90 days was 56 % and 24 % in the RFA and control groups, respectively (P = 0.04). The mean stent patency time was significantly longer in the RFA group compared to the control group (119.5 vs. 65.3 days, P = 0.03). Conclusion: The application of RFA for occluded SEMS improves stent patency. RFA is an alternative treatment of tissue ingrowth in malignant biliary obstruction.

Successful Treatment of Recurrent Pericarditis With Rilonacept in Acute Promyelocytic Leukemia.

A 40-year-old woman presented with recurrent pericarditis and pericardial effusion while receiving treatment with all-trans retinoic acid and arsenic trioxide for recently diagnosed acute promyelocytic leukemia. She was successfully treated with the interleukin-1 inhibitor rilonacept after experiencing multiple recurrences with triple therapy with aspirin, colchicine, and steroids. (Level of Difficulty: Advanced.).

Thermal ablation of pancreatic cyst with a prototype endoscopic ultrasound capable radiofrequency needle device: A pilot feasibility study.

BACKGROUND AND OBJECTIVES: Pancreatic cysts are evaluated by endoscopic ultrasound and fine needle aspiration (EUS). The only accepted treatment is pancreatectomy, which is associated with morbidity and mortality. This study evaluated the optimal thermal dosimetry of a novel radiofrequency ablation device using a standard electrosurgical unit in ex vivo cyst models. METHODS: A modified EUS 22-gauge monopolar needle prototype with a tip electrode connected to a standard electrosurgical unit (Erbe USA, Marietta, GA, USA) was used to induce a subboiling point temperature. A cyst model was created using 2-cm sections of porcine small intestine ligated and filled with saline. After ablation, the cyst models were prepared for pathological evaluation. The epithelial layers were measured in at least two different sites with a micrometer and compared with the corresponding control sample. RESULTS: Thirty-two cyst models were ablated with maximum temperatures of 50°C, 60°C, 90°C, and 97°C in 8, 11, 11, and 2 cysts, respectively. Longer ablation times were required to induce higher temperatures. A trend in the reduction in thickness of the measured layers was observed after exposure to higher temperatures. A temperature over 50°C was required for the ablation of the muscularis, submucosa, and villi, and over 60°C was required to ablate the mucosal crypts. CONCLUSIONS: In a preclinical model, a novel radiofrequency EUS-capable needle connected to a standard electrosurgical unit using standard low-voltage coagulation provided ablation in a temperature-dependent fashion with a threshold of at least 60°C and a safe cyst margin below 97°C. This potentially will allow low-cost, convenient cyst ablation.

The value of KRAS mutation testing with CEA for the diagnosis of pancreatic mucinous cysts.

BACKGROUND AND AIMS: Pancreatic cyst fluid (PCF) CEA has been shown to be the most accurate preoperative test for detection of cystic mucinous neoplasms (CMNs). This study aimed to assess the added value of PCF KRAS mutational analysis to CEA for diagnosis of CMNs. PATIENTS AND METHODS: This is a retrospective study of prospectively collected endoscopic ultrasonography (EUS) fine-needle aspiration (FNA) data. KRAS mutation was determined by direct sequencing or equivalent methods. Cysts were classified histologically (surgical cohort) or by clinical (EUS or FNA) findings (clinical cohort). Performance characteristics of KRAS, CEA and their combination for detection of a cystic mucinous neoplasm (CMN) and malignancy were calculated. RESULTS: The study cohort consisted of 943 patients: 147 in the surgical cohort and 796 in the clinical cohort. Overall, KRAS and CEA each had high specificity (100 % and 93.2 %), but low sensitivity (48.3 % and 56.3 %) for the diagnosis of a CMN. The positivity of KRAS or CEA increased the diagnostic accuracy (80.8 %) and AUC (0.84) significantly compared to KRAS (65.3 % and 0.74) or CEA (65.8 % and 0.74) alone, but only in the clinical cohort (P < 0.0001 for both). KRAS mutation was significantly more frequent in malignant CMNs compared to histologically confirmed non-malignant CMNs (73 % vs. 37 %, P = 0.001). The negative predictive value of KRAS mutation was 77.6 % in differentiating non-malignant cysts. CONCLUSIONS: The detection of a KRAS mutation in PCF is a highly specific test for mucinous cysts. It outperforms CEA for sensitivity in mucinous cyst diagnosis, but the data does not support its routine use.

Transnasal endoscopy: Technical considerations, advantages and limitations.

Transnasal endoscopy (TNE) is an upper endoscopy method which is performed by the nasal route using a thin endoscope less than 6 mm in diameter. The primary goal of this method is to improve patient tolerance and convenience of the procedure. TNE can be performed without sedation and thus eliminates the risks associated with general anesthesia. In this way, TNE decreases the cost and total duration of endoscopic procedures, while maintaining the image quality of standard caliber endoscopes, providing good results for diagnostic purposes. However, the small working channel of the ultra-thin endoscope used for TNE makes it difficult to use for therapeutic procedures except in certain conditions which require a thinner endoscope. Biopsy is possible with special forceps less than 2 mm in diameter. Recently, TNE has been used for screening endoscopy in Far East Asia, including Japan. In most controlled studies, TNE was found to have better patient tolerance when compared to unsedated endoscopy. Nasal pain is the most significant symptom associated with endoscopic procedures but can be reduced with nasal pretreatment. Despite the potential advantage of TNE, it is not common in Western countries, usually due to a lack of training in the technique and a lack of awareness of its potential advantages. This paper briefly reviews the technical considerations as well as the potential advantages and limitations of TNE with ultra-thin scopes.

Transnasal endoscopy is preferred by transoral endoscopy experienced patients.

BACKGROUND & AIMS: Both unsedated transoral endoscopy (TOE) and sedated TOE have some drawbacks in clinical practice. Unsedated transnasal endoscopy (TNE) has been suggested as an alternative to both methods. This study aimed to determine the advantages of TNE in patients who have previously undergone unsedated conventional TOE. METHODS: Patients who had received an unsedated TOE in the last 12 months and were scheduled for a second upper endoscopy were included. They were randomized to undergo either unsedated TOE, using a standard endoscope, or unsedated TNE, using an ultrathin endoscope. Post-procedure, patients were asked to complete a questionnaire to assess pain, discomfort and acceptability of the procedure, and to compare the current procedure with their previous unsedated TOE. Endoscope insertion rate, procedure duration, and side-effects were recorded. RESULTS: Each group included 50 patients. With the exception of nasal pain, the tolerability and acceptance were significantly greater in the unsedated TNE group. Significantly more TNE patients (82%) found the current endoscopic procedure to be better than their previous TOE when compared with patients who had received a second TOE (12%). A repeat procedure was significantly more acceptable for TNE patients when compared to the TOE group (68% vs.16%). The duration of endoscopy was significantly shorter in TOE than in TNE (p<0.05). Endoscope insertion failed in 4% and mild epistaxis was observed in 4% of TNE patients. CONCLUSION: Unsedated TNE was better tolerated in endoscopy experienced patients when compared with unsedated TOE. The majority of patients found TNE more acceptable and preferable to TOE, suggesting that TNE should become a more common practice in clinics when applicable.