A deep learning approach for Direct Immunofluorescence pattern recognition of Autoimmune Bullous Diseases.

BACKGROUND: Artificial intelligence (AI) is reshaping healthcare, using machine and deep learning to enhance disease management. Dermatology has seen improved diagnostics, particularly in skin cancer detection, through the integration of AI. However, the potential of AI in automating immunofluorescence imaging for autoimmune bullous skin diseases remains untapped. While direct immunofluorescence (DIF) supports diagnosis, its manual interpretation can hinder efficiency. The use of deep learning to automatically classify DIF patterns, including the Intercellular Pattern (ICP) and the Linear Pattern (LP), holds promise for improving the diagnosis of autoimmune bullous skin diseases. OBJECTIVES: The objectives of this study are to develop AI algorithms for automated classification of autoimmune bullous skin disease DIF patterns, such as ICP and LP. This aims to enhance diagnostic accuracy, streamline disease management, and improve patient outcomes through deep learning-driven immunofluorescence interpretation. METHODS: We collected immunofluorescence images from skin biopsies of patients suspected of AIBD between January 2022 and January 2024. Skin tissue was obtained via 5-mm punch biopsy, prepared for direct immunofluorescence. Experienced dermatologists classified the images into three classes: ICP, LP, and negative. To evaluate our deep learning approach, we divided the images into training (436) and test sets (93). We employed transfer learning with pre-trained deep neural networks and conducted 5-fold cross-validation to assess model performance. Our dataset's class imbalance was addressed using weighted loss and data augmentation strategies. The models were trained for 50 epochs using Pytorch, achieving an image size of 224x224 for both CNNs and the Swin Transformer. RESULTS: Our study compared six CNNs and the Swin transformer for AIBDs image classification, with the Swin transformer achieving the highest average validation accuracy of 98.5%. On a separate test set, the best model attained an accuracy of 94.6%, demonstrating 95.3% sensitivity and 97.5% specificity across AIBDs classes. Visualization with Grad-CAM highlighted the model's reliance on characteristic patterns for accurate classification. CONCLUSIONS: The study highlighted CNN's accuracy in identifying DIF features. This approach aids automated analysis and reporting, offering reproducibility, speed, data handling, and cost-efficiency. Integrating deep learning in skin immunofluorescence promises precise diagnostics and streamlined reporting in this branch of dermatology.

Very Early Recurrence After Curative Resection for Pancreatic Ductal Adenocarcinoma: Proof of Concept for a "Biological R2 Definition".

PURPOSE: Very early recurrence after radical surgery for pancreatic ductal adenocarcinoma (PDAC) has been poorly investigated. This study was designed to evaluate this group of patients who developed recurrence, within 12 weeks after surgery, defined as "biological R2 resections (bR2)." METHODS: Data from patients who underwent surgical resection as upfront procedure or after neoadjuvant treatment for PDAC between 2015 and 2019 were analyzed. Disease-free, disease-specific survival, and independent predictors of early recurrence were examined. The same analysis was performed separately for upfront and neoadjuvant treated patients. RESULTS: Of the 573 patients included in the study, 63 (11%) were classified as bR2. The rate of neoadjuvant treatment was similar in bR2 and in the remaining patients (44 vs. 42%, p = 0.78). After a median follow-up of 27 months, median DFS and DSS for the entire cohort were 17 and 43 months, respectively. Median DSS of bR2 group was 13 months. The only preoperative identifiable independent predictor of very early recurrence was body-tail site lesion, whereas all other were pathological: higher pT (8th classification), G3 differentiation, and high lymph node ratio. These predictors were confirmed for patients undergoing upfront surgery, whereas in the neoadjuvant group the only independent predictor was pT. CONCLUSIONS: One of ten patients with "radical" resected PDAC relapses very early after surgery (bR2); hence, imaging must be routinely repeated within 12 weeks. Despite higher biological aggressiveness and worse pathology, this bR2 cluster eludes our preoperative examinations.

Exploring the optimal therapeutic management of stage ypIA pancreatic ductal adenocarcinoma patients in the era of primary chemotherapy.

BACKGROUND: Data on the proper post-surgical chemotherapy (PSC) in pancreatic ductal adenocarcinoma (PDAC) patients already treated with neoadjuvant therapy (NAT) are lacking, especially for stage ypIA. AIM AND METHODS: We retrospectively analyzed ypT1N0M0 (ypIA) PDAC patients resected after NAT between 2015 and 2020 at our Institution. Primary endpoint was median disease free-survival (DFS) according to PSC treatment. RESULTS: Seventy-five out of 363 patients achieved a pathological ypIA after NAT (20.6%) and 72 were analyzed. Among the study population 34 patients (47%) were treated with NAT ≤4 months and 38 (53%) >4 months. After surgery, 10 patients (14%) received PSC using the same multidrug NAT regimen (Group A); 35 (49%) received PSC with a different regimen (Group B), with either single agents in 24 patients (68.5%) or combination schedules in 11 (31.5%); 27 patients (14%) did not receive any PSC (Group C). DFS was longer in group A and C as opposed to group B (p = 0.006). CONCLUSION: Patients affected by ypIA PDAC treated with a proper multi-agent chemotherapy for more than 4 months show an improved DFS, regardless of the peri­operative or totally pre-surgical administration of treatment.

Environmental factors in autoimmune bullous diseases with a focus on seasonality: new insights.

Autoimmune bullous diseases are a heterogeneous group of rare conditions clinically characterized by the presence of blisters and/or erosions on the skin and the mucous membranes. Practically, they can be divided into two large groups: the pemphigoid group and the pemphigus group, depending on the depth of the autoimmune process on the skin. A family history of autoimmune diseases can often be found, demonstrating that genetic predisposition is crucial for their development. Moreover, numerous environmental risk factors, such as solar radiation, drugs, and infections, are known. This study aimed to evaluate how seasonality can affect the trend of bullous pemphigoid and pemphigus vulgaris, especially considering the number of hospitalizations recorded over the course of individual months. The total number of hospitalizations in the twelve months of the year was evaluated. Moreover, blood chemistry assay and, for some patients, enzyme-linked immunosorbent assay were executed to evaluate antibodies. Regarding the severity of the disease, the bullous pemphigoid area index and the pemphigus disease area index score systems were used. Results showed a complex interplay between environmental factors such as seasons and autoimmune conditions.

Improvement in quality of life and sexual function in patients affected by vulvar lichen sclerosus treated with combined autologous platelet-rich plasma and fat grafting.

BACKGROUND: Vulvar lichen sclerosus (LS) severely impairs patients' quality of life. OBJECTIVES: To evaluate the impact of a combined application of autologous platelet-rich plasma (PRP) and fat grafting as treatment for vulvar LS on patient quality of life. MATERIALS & METHODS: We reviewed the clinical charts of 72 patients affected by LS, who underwent regenerative surgery. The patients' quality of life was assessed using: the Dermatology Life Quality Index (DLQI), the Skindex-29, the Female Sexual Function Index (FSFI) and the patient-administered - Clinical Scoring System (CSS). RESULTS: After reconstructive surgery, all scores improved: Skindex-29 (-31.8 [IQR: 42.1, -21.8] points; p<0.001), FSFI (7.6 [IQR: 2.7, 14.7)] points; p<0.001), Patient-administered CSS (-24 [IQR: -30, -15] points; p<0.001), DLQI (-9 [IQR: -17, -7] points; p<0.001), Physician-administered CSS (-5 [IQR: -7, -5] points; p<0.001), and IGA (median ΔIGA: -4, IQR: -4, -3; p<0.001). CONCLUSION: Combined treatment with PRP and fat grafting proved to be effective in improving the quality of life of patients with vulvar LS.

Impact of powered circular stapler on anastomotic leak after anastomosis to the rectum: a propensity score matched study.

PURPOSE: The aim of the present study is to assess the impact of Echelon Circular™ powered stapler (PCS) on left-sided colorectal anastomotic leaks and to compare results to conventional circular staplers (CCS). METHODS: A single center cohort study was carried out on 552 consecutive patients, who underwent laparoscopic colorectal resection and anastomosis to the rectum between December 2017 and September 2022. Patients who underwent powered circular anastomosis to the rectum were matched to those who had a conventional stapled anastomosis using a propensity score matching. Main outcomes were anastomotic leak (AL) rate, anastomotic bleeding, and postoperative outcomes. RESULTS: After adjusting cases with propensity score matching, two new groups of patients were generated: 145 patients in the PCS and 145 in the CCS. The two groups were homogeneous with respect to demographics and comorbidities on admission. Overall, AL occurred in 21 (7.3%) patients. No significant differences were observed with respect to AL (5.5% in PCS vs 9% in CCS; p = 0.66), fistula severity (p = 0.60) or reoperation rate (p = 0.65) in the two groups in study. A higher rate of anastomotic bleeding was observed in the CCS vs PCS (5.5% vs 0.7%, p = 0.03). At univariate analysis performed after propensity score matching, stapler diameter ≥ 31mm and age ≥ 70 years were the only variable significantly associated with anastomotic leak (p = 0.001 and p = 0.031; respectively). CONCLUSIONS: The powered circular stapler has no impact on AL, while it could affect bleeding rate at the anastomotic site.

Autoimmune bullous dermatoses in cancer patients treated by immunotherapy: a literature review and Italian multicentric experience.

Cutaneous immune-related adverse events are frequently associated with immune checkpoint inhibitors (ICIs) administration in cancer patients. In fact, these monoclonal antibodies bind the cytotoxic T-lymphocyte antigen-4 and programmed cell death-1/ligand 1 leading to a non-specific activation of the immune system against both tumoral cells and self-antigens. The skin is the most frequently affected organ system appearing involved especially by inflammatory manifestations such as maculopapular, lichenoid, psoriatic, and eczematous eruptions. Although less common, ICI-induced autoimmune blistering diseases have also been reported, with an estimated overall incidence of less than 5%. Bullous pemphigoid-like eruption is the predominant phenotype, while lichen planus pemphigoides, pemphigus vulgaris, and mucous membrane pemphigoid have been described anecdotally. Overall, they have a wide range of clinical presentations and often overlap with each other leading to a delayed diagnosis. Achieving adequate control of skin toxicity in these cases often requires immunosuppressive systemic therapies and/or interruption of ICI treatment, presenting a therapeutic challenge in the context of cancer management. In this study, we present a case series from Italy based on a multicenter, retrospective, observational study, which included 45 patients treated with ICIs who developed ICI-induced bullous pemphigoid. In addition, we performed a comprehensive review to identify the cases reported in the literature on ICI-induced autoimmune bullous diseases. Several theories seeking their underlying pathogenesis have been reported and this work aims to better understand what is known so far on this issue.

Pathological staging in postneoadjuvant pancreatectomy for pancreatic cancer: implications for adjuvant therapy.

BACKGROUND: It is unclear whether pathological staging is significant prognostically and can inform the delivery of adjuvant therapy after pancreatectomy preceded by neoadjuvant therapy. METHODS: This multicentre retrospective study included patients who underwent pancreatectomy for pancreatic ductal adenocarcinoma after neoadjuvant treatment at two Italian centres between 2013 and 2017. T and N status were assigned in accordance with the seventh and eighth editions of the AJCC staging system, as well as according to a modified system with T status definition combining extrapancreatic invasion and tumour size. Patients were then stratified by receipt of adjuvant therapy. Survival analysis and multivariable interaction analysis of adjuvant therapy with pathological parameters were performed. The results were validated in an external cohort from the USA. RESULTS: The developmental set consisted of 389 patients, with a median survival of 34.6 months. The modified staging system displayed the best prognostic stratification and the highest discrimination (C-index 0.763; 1-, 2- and 3-year time-dependent area under the curve (AUC) 0.746, 0.722, and 0.705; Uno's AUC 0.710). Overall, 67.0 per cent of patients received adjuvant therapy. There was no survival difference by receipt of adjuvant therapy (35.0 versus 36.0 months; P = 0.772). After multivariable adjustment, interaction analysis suggested a benefit of adjuvant therapy for patients with nodal metastases or with tumours larger than 2 cm with extrapancreatic extension, regardless of nodal status. These results were confirmed in the external cohort of 216 patients. CONCLUSION: Modified staging with a T status definition combining extrapancreatic invasion and tumour size is associated with better prognostic segregation after postneoadjuvant pancreatectomy. This system allows identification of patients who might benefit from adjuvant therapy.

Quantification of perineural invasion in pancreatic ductal adenocarcinoma: proposal of a severity score system.

Perineural invasion (PNI) is a common feature in pancreatic ductal adenocarcinoma (PDAC) and correlates with an aggressive tumor behavior already at early stages of disease. PNI is currently considered as a "present vs. absent" feature, and a severity score system has not yet been established. The aim of the present study was thus to develop and validate a score system for PNI and to correlate it with other prognostic features. In this monocentric retrospective study, 356 consecutive PDAC patients (61.8% upfront surgery patients, 38.2% received neoadjuvant therapy) were analyzed. PNI was scored as follows: 0: absent; 1: the presence of neoplasia along nerves < 3 mm in caliber; and 2: neoplastic infiltration of nerve fibers ≥ 3 mm and/or massive perineural infiltration and/or the presence of necrosis of the infiltrated nerve bundle. For every PNI grade, the correlation with other pathological features, disease-free survival (DFS), and disease-specific survival (DSS) were analyzed. Uni- and multivariate analysis for DFS and DSS were also performed. PNI was found in 72.5% of the patients. Relevant trends between PNI score and tumor differentiation grade, lymph node metastases, vascular invasion, and surgical margins status were found. The latter was the only parameter statistically correlated with the proposed score. The agreement between pathologists was substantial (Cohen's K 0.61). PNI severity score significantly correlated also with decreased DFS and DSS at univariate analysis (p < 0.001). At multivariate analysis, only the presence of lymph node metastases was an independent predictor of DFS (HR 2.235 p < 0.001). Lymph node metastases (HR 2.902, p < 0.001) and tumor differentiation grade (HR 1.677, p = 0.002) were independent predictors of DSS. Our newly developed PNI score correlates with other features of PDAC aggressiveness and proved to have a prognostic role though less robust than lymph nodes metastases and tumor differentiation grade. A prospective validation is needed.

Immunosuppressive treatment for autoimmune blistering diseases does not impair production of neutralizing antibodies after "booster" dose of COVID-19 vaccine.

Since vaccines against COVID-19 are available, it has been debated if immunosuppressed patients with autoimmune blistering diseases (AIBDs) should be advised to interrupt the immunosuppressive therapy before receiving the vaccine, with consequent risk to experience a flare of disease. In the present study, we measured the neutralizing antibodies production after anti-SARS-CoV-2 vaccination in patients with AIBDs on immunosuppressive treatment, compared to healthy controls. Our results give strength to the hypothesis that these patients do not need to discontinue their therapy to produce effective levels of neutralizing antibodies, in other words to achieve successful protection.

The role of biliary events in treatment and survival of patients with advanced pancreatic ductal adenocarcinoma.

BACKGROUND: Systemic chemotherapy (CT) is the treatment of choice for advanced pancreatic ductal adenocarcinoma (PDAC). Biliary obstruction is common in this setting and may interfere with CT administration due to jaundice or cholangitis related to biliary stent malfunction. AIMS: To evaluate the impact of biliary events on CT administration and survival in patients with stage III-IV PDAC. METHODS: Patients enrolled in a randomized trial of nab-paclitaxel plus gemcitabine with/without capecitabine and cisplatin in advanced PDAC were included. Data on management of jaundice, biliary stents/complications and CT were prospectively collected and retrospectively analyzed. Modified overall (mOS) and progression-free (mPFS) survival were evaluated. RESULTS: Eighty-eight patients met the inclusion criteria (50% females; median age 65years). Seven of eight (87.5%) patients who placed plastic stents developed biliary complications versus 14/30 (46.7%) with metallic stents (p = 0.071). Patients without biliary complications completed planned CT in 64.2% versus 47.6% of cases (p = 0.207). CT completion was related to longer mOS (17 vs 12 months, p = 0.005) and mPFS (9 vs 6 months, p = 0.011). mOS was shorter when biliary complications occurred (12 vs 17 months, p = 0.937), as was mPFS (6 vs 8 months, p = 0.438). CONCLUSION: Complications related to biliary obstruction influence chemotherapy completion and survival in patients with advanced PDAC.

Pancreatectomy with venous vascular resection for pancreatic cancer: Impact of types of vein resection on timing and pattern of recurrence.

INTRODUCTION: Few studies analysed the impact of different venous resection techniques on recurrence in patients with pancreatic ductal adenocarcinoma (PDAC). Primary aim was to compare local recurrence rate and disease-free survival (DFS) between patients who underwent pancreatectomy with tangential versus segmental resection of portal vein/superior mesenteric vein. MATERIALS AND METHODS: All consecutive patients who underwent pancreatectomy with venous resection for PDAC between 2009 and 2019 were included. A propensity score matching (PSM) was used to reduce the effect of treatment selection bias. RESULTS: Overall, 120 patients (68%) underwent pancreatectomy with tangential venous resection and 57 patients (32%) were submitted to pancreatectomy with segmental venous resection. After a median follow-up of 24 months, local recurrence was comparable between the two groups (tangential: n = 32/120, 26.7% versus segmental: n = 10/57, 17.5%; p = 0.58). The median DFS was 17 months (IQR 9-31) in patients who underwent tangential resection, as compared to 12 months (IQR 5-21) in those who underwent segmental resection (p = 0.049). After PSM (n = 106), the median DFS was 18 months (IQR 9-26) in the tangential resection group, and 12 months (IQR 5-21) in the segmental resection group (p = 0.17). In the PSM population, lymph node ratio (HR 4.83; p = 0.028) and tumor size >25 mm (HR 3.26; p = 0.007) were identified as determinants of local recurrence. CONCLUSION: Tangential venous resections are not associated with a higher rate of local recurrence. Long-term outcomes are more related to tumors characteristics than to venous resection techniques. A step-up approach to vein resection, with tangential resection being performed whenever technically feasible, should be strongly encouraged.

Prognostic value of preoperative CT scan derived body composition measures in resected pancreatic cancer.

BACKGROUND: It remains unclear whether preoperative body composition may affect the prognosis of pancreatic cancer patients undergoing surgery. The aim of the present study was to assess the extent to which preoperative body composition impacts on postoperative complication severity and survival in patients undergoing pancreatoduodenectomy for pancreatic ductal adenocarcinoma (PDAC). METHODS: A retrospective cohort study was performed on consecutive patients who underwent pancreatoduodenectomy with preoperative CT scan imaging available. Body composition parameters including total abdominal muscle area (TAMA), visceral fat area (VFA), subcutaneous fat area and liver steatosis (LS) were assessed. Sarcopenic obesity was defined as a high VFA/TAMA ratio. Postoperative complication burden was evaluated with the comprehensive complication index (CCI). RESULTS: Overall, 371 patients were included in the study. At 90 days after surgery, 80 patients (22%) experienced severe complications. The median CCI was 20.9 (IQR 0-30). At multivariate linear regression analysis, preoperative biliary drainage, ASA score ≥3, fistula risk score and sarcopenic obesity (37% increase; 95%CI 0.06-0.74; p = 0.046) were associated to an increase in CCI. Patient characteristics associated to sarcopenic obesity were older age, male gender and preoperative LS. At a median follow-up of 25 months (IQR 18-49), median disease-free survival (DFS) was 19 months (IQR 15-22). At cox-regression analysis, only pathological features were associated with DFS, while LS and other body composition measures did not show any prognostic role. CONCLUSION: The combination of sarcopenia and visceral obesity was significantly associated with increased complication severity after pancreatoduodenectomy for cancer. Patients' body composition did not affect disease free survival after pancreatic cancer surgery.