Introduction

Pancreatic Ductal Adenocarcinoma (PDAC) is a highly aggressive and lethal malignancy of the pancreas due to limited treatment response and lack of early diagnosis. PDAC accounts for more than 90% of all pancreatic neoplasms and early PDAC may be missed on Computed Tomography (CT) scan due to similar attenuation as healthy pancreas, small size and hidden location.

Aims & Methods

The current study aims at improving detection of PDAC with CT scans using a deep learning model. It uses three different Convolutional Neural Network (CNN) models namely, Visual Geometry Group 19 (VGG19), Residual Neural Network 50 (ResNet-50) and Inception-v3. A compendium of 129,910 abdomen CT images (107 patients) of PDAC patients and 18,942 abdomen CT images (80 patients) of healthy controls were considered from The Cancer Imaging Archive. Additionally, several pre-processing and image augmentation methods were used to prevent overfitting and improve accuracies.

Results

The current model showed high construct validity with markedly lowered computational power. The Inception-v3 model showed the highest accuracy in the detection of PDAC with an accuracy of 97.93%. Specificity of the VGG-19 model was the highest at 98.77 whereas sensitivity was the highest for the Inception-v3 model at 98.75%. Data on the CNN models’ Performance Metrics (accuracy, sensitivity, specificity and area under ROC) ​were all compared.

Conclusion

The deep learning enabled tool accurately detected PDAC on CT scans versus healthy controls. The models were also reasonably sensitive for tumours below 2 cm.

References

1. Ushio J, Kanno A, Ikeda E, et al. Pancreatic Ductal Adenocarcinoma: Epidemiology and Risk Factors. Diagnostics (Basel). 2021;11(3):562. doi:10.3390/diagnostics11030562
2. Qureshi TA, Gaddam S, Wachsman AM, et al. Predicting pancreatic ductal adenocarcinoma using artificial intelligence analysis of pre-diagnostic computed tomography images. Cancer Biomark. 2022;33(2):211-217. doi:10.3233/CBM-210273
3. Ahmed HS. BEYOND TRADITIONAL TOOLS: EXPLORING CONVOLUTIONAL NEURAL NETWORKS AS INNOVATIVE PROGNOSTIC MODELS IN PANCREATIC DUCTAL ADENOCARCINOMA. Arq Gastroenterol. 2024;61:e23107. doi:10.1590/S0004-2803.24612023-117

Introduction

Despite advances with chemotherapy regimens, Pancreatic adenocarcinoma (PDAC) is becoming the 2^nd cause of cancer-related death. Personalized medicine will be the game changer, but molecular characterization of pancreatic neoplasms is yet extremely cumbersome and has been performed on surgical samples or few EUS-tissue acquisition (TA) cases, mostly for DNA analysis. Our aim was to evaluate DNA and RNA extracted for research purposes from pancreatic neoplasms employing Endoscopic Ultrasound (EUS) Fine needle aspiration (FNA) and biopsy (FNB) and adequacy to molecular analyses.

Aims & Methods

From Oct 2022, patients with solid pancreatic neoplasms suggestive of either PDAC or neuroendocrine neoplasm (NEN) undergoing diagnostic EUS-TA were considered for enrollment. Employed needles were either FNA, or Franseen, Fork-tip or Three-prong, either 22G or 25G based on operator’s preference, with Rapid on Site Evaluation and, in case of adequacy within the first 2 passes, 1 additional pass with the same needle was performed for molecular analyses. DNA and RNA were then extracted with a new technique and nucleic acids were quantified and evaluated for adequacy for molecular analysis. Mann-Whitney was employed for comparison of continuous variables, and Kruskal-Wallis test for comparison between multiple groups; Logistic regression was applied to determine factors associated with adequacy to molecular tests.

Results

351 patients underwent sampling for molecular analyses, 309 with a definitive diagnosis of PDAC and 42 with NEN. 92,9% of patients were sampled with 25G FNA needles. Median quantity of DNA extracted was significantly lower in PDAC 2048 ng (95%CI: 1470 -3651) compared to NEN 6752 ng (95%CI: 3502-9655)(p<0.0144). Mean quantity of RNA extracted was similar, 312 ng for PDAC, and 444 ng for NEN (p=0.07), with a higher median RIN of 6,4 (95%CI:6,2-6,5) for PDAC versus 4,7 (95%CI:3,9-5) for NEN (p<0.0001). In terms of needle type, concerning DNA extraction, 25G FNA needles retrieved a median of 3384 ng, 25G Franseen needles (4312 ng), 22G Franseen (73864 ng), 25G Fork-tip or Three-prong (3104 ng), with no significant difference. In terms of needle type, concerning RNA extraction, 25G FNA needles retrieved a median of 310 ng (RIN 6,3), 25G Franseen needles 460 ng (RIN 5,5), 22G Franseen 884 ng (RIN 5,9), 25G Fork-tip or Three-prong 301 ng (RIN 5,15), with no significant difference. 295 (95,5%) of PDAC samples and 34 (80,9%) of NEN samples were considered adequate for sequencing. The only factor associated to adequacy for molecular analysis was tumor type, with PDAC having an OR of 5,44 (95%CI 2,06-14,37)(p=0.0006), while the type of employed needle was not

Conclusion

Molecular analysis on EUS-acquired tissue of pancreatic neoplasms will be the mainstay of personalized therapies but how to best perform tissue acquisition and sample handling for this specific purpose is yet to be defined. This is the largest study ever performed on DNA and RNA extraction from EUS-acquired samples, demonstrating superior adequacy for molecular analysis in PDAC compared to NEN. All needle types result similar in terms of performance despite the low number of needles other than 25G FNA needles.

Introduction

In patients with chronic pancreatitis (CP), pancreatic exocrine insufficiency (PEI) and poor nutritional intake may cause nutritional deficiencies. Guidelines advise monitoring key nutrients and supplementing is case of deficiency to prevent malnutrition-related complications. However, adherence remains low, possibly due to limited evidence on the clinical relevance of single nutrient deficiencies and the benefit of targeted supplementation. The aim of this study is to evaluate the prevalence of nutritional deficiencies in patients with CP and to identify predictors of those at highest risk. Secondary, we seek to explore association between nutritional deficiencies and malnutrition-related complications, and to provide insight into the natural course of nutritional deficiencies.

Aims & Methods

We analyzed data from the COMBO-trial, a nationwide stepped-wedged cluster randomized trial to evaluate the effect of an evidence-based algorithm for the management of CP. Measurements of fat-soluble vitamins, minerals, and albumin conducted in the outpatient clinic during the 35-month study period were included. PEI was defined as fecal elastase-1 concentration <200 µg/g or the use of pancreatic enzyme replacement therapy (PERT). The Pancreatic Exocrine Insufficiency Questionnaire (PEI-Q), designed to evaluate the severity of PEI symptoms, was administered at four distinct time points throughout the study. Supplement use was collected from health records and patient questionnaires. Multivariate logistic regression analysis was performed for the presence of multiple deficiencies. Potential predictors included patient and disease characteristics at baseline and the mean score of the PEI-Q during follow-up.

Results

In total 409 patients were included in this study. At baseline 327 (80.0%) had PEI and PERT was used by 297 (72.6%). Nutritional deficiencies were present in 195 patients (47.7%) of which 81 (19.8%) had multiple deficiencies. The prevalence of deficiencies included vitamin A in 13/172 (7.6%) patients, vitamin D 111/289 (38.4%), vitamin E 15/168 (8.9%), vitamin K 52/88 (59.1%), iron 44/137 (32.1%) calcium 4/198 (2.0%), magnesium 24/132 (18.2%) zinc 13/67 (19.4%), selenium 4/42 (9.5%), and albumin in 41/288 (14.2%) patients. No abnormal coagulation was observed in the 20 patients with vitamin K deficiency and available INR measurements. Osteopenia was diagnosed in 110/236 (46.6%) and osteoporosis in 56/236 (23.7%). Deficiencies in vitamin K and vitamin D, micronutrients critical to bone metabolism, were not associated with osteopenia or osteoporosis. Of 44 patients with an iron deficiency, 22 (50%) were anemic. Deficiency resolved without single nutrient supplementation in 71.4% (vitamin A), 75.0% (vitamin D), 50.0% (vitamin E) , 33.3% (vitamin K), 63.6% (iron), 100% (magnesium), 80.0% (zinc) of cases. Independent predictors for multiple deficiencies were more severe PEI symptoms (OR 1.86, 1.09 – 3.17) and daily alcohol consumption (OR 2.80, 1.27 – 6.18).

Conclusion

Nutritional deficiencies are present in approximately half of the patients with CP. Most nutritional deficiencies resolve without single nutrient supplementation. Predictors for multiple nutritional deficiencies were severity of PEI symptoms and daily alcohol consumption. Therefore, nutritional deficiencies should be considered a marker of malnutrition, with treatment focused on addressing the underlying cause rather than isolated correction through single nutrient supplementation.

Introduction

The recommended treatment for severe complications of portal hypertension (PH) is transjugular intrahepatic portosystemic shunt (TIPS) implantation. Despite alcohol-related liver disease (ALD) representing the most frequent cause of advanced chronic liver disease (ACLD) and PH, literature on the effects of ongoing alcohol use (AU) after TIPS placement currently remains scarce.

Aims & Methods

The study population consisted of patients with ALD-related ACLD who had undergone TIPS implantation at two Viennese tertiary care centers (pooled cohort from the prospective AUTIPS study, NCT03409263: 01/2017-11/2022; and a retrospective cohort: 09/2000-09/2019). Through the electronic patient management system, individual demographic, laboratory, and clinical data were retrieved. AU after TIPS placement was investigated according to clinical documentation, serum ethanol quantification if available, and patients' reporting.

Results

Overall, the study population consisted of 276 patients (median follow-up 29 [IQR 54] months), 196 (71.0%) of which had received TIPS for refractory ascites while in 80 (29.0%) the indication for TIPS implantation had been variceal hemorrhage. The median age was 56 (IQR 13) years and 215 (77.9%) of the patients were men. At the time of the TIPS insertion, the median MELD was 11 (IQR 9) and 38 (13.8%) individuals were classified as Child-Pugh C. AU after TIPS placement was documented in 85 (30.8%) study participants. Baseline characteristics at TIPS implantation were similar in patients with and without post-TIPS AU, specifically regarding the prevalence of ascites (p=0.51), variceal hemorrhage (p=0.87), and hepatic encephalopathy (p=0.87).
Bacterial infections were more frequently detected among patients with ongoing AU (log-rank p=0.002). Importantly, patients with ongoing AU showed a higher incidence of acute-on-chronic liver failure (ACLF; log-rank p<0.001) and higher overall mortality (log-rank p=0.005; including liver-related death [log-rank p<0.001]).
Furthermore, cumulative incidence of ACLF (p=0.014) and liver-related death (p=0.004) were significantly higher in patients with ongoing AU. Of note, ongoing AU was independently associated with ACLF (asHR: 1.66; 95%CI: 1.13-2.44; p=0.010) and liver-related death (asHR: 2.09; 95%CI: 1.32-3.33; p=0.002) in multivariable competing risks regression analysis adjusted for age, MELD, and portal pressure gradient after TIPS implantation.

Conclusion

AU represents an independent risk factor for outcome deterioration after TIPS placement. These findings emphasize the paramount role of psychiatric care and harm reduction measures in addition to hepatological optimization after TIPS implantation.

Disclosure

C.S. received travel support from Gilead, Abbvie, Galapagos and Gebro; speaking honoraria from Abbvie and Gilead; and payments for consulting from Gilead.
M.S. received travel support from MSD, Sandoz, BMS, Abbvie and Gilead; and speaking honoraria from BMS.
M.J. received speaking honoraria from Abbvie and Gilead.
T.R. received grant support from Abbvie, Boehringer- Ingelheim, Gilead, MSD, Philips Healthcare, Gore; speaking honoraria from Abbvie, Gilead, Gore, Intercept, Roche, MSD; consulting/advisory board fee from Abbvie, Bayer, Boehringer-Ingelheim, Gilead, Intercept, MSD, Siemens; and travel support from Boehringer-Ingelheim, Gilead, and Roche.
M.M. served as a speaker and/or consultant and/or advisory board member for AbbVie, Bristol-Myers Squibb, Collective Acumen, Gilead, and W. L. Gore & Associates and received travel support from AbbVie, Bristol-Myers Squibb, and Gilead.
B.S. received travel support from AbbVie and Gilead.