St George Hospital

Colorectal cancer with peritoneal metastases (PM) presents a significant therapeutic challenge. Cytoreductive surgery (CRS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) is one option to prolong survival. While completeness of cytoreduction (CC) score 0 is associated with improved outcomes, the clinical value of near-complete CC-score 1 versus open-close laparotomy (CC-3) remains unclear. This retrospective study evaluates overall survival (OS) in patients with colorectal cancer PM scheduled for CRS and HIPEC from 23 global peritoneal-surface oncology centers from 2006 to 2023. A propensity score matching was performed using tumor location (colon/rectum), lymph node status, liver metastases, signet-ring histology, preoperative chemotherapy, peritoneal cancer index, and treatment year. Matching was performed using the nearest neighbor method with a caliper of 0.1, chosen after several iterations to optimize intergroup balance. Balance was assessed using standardized mean differences. Sensitivity analyses with alternative calipers and multivariable Cox regression in the unmatched cohort were considered to test the robustness of the findings. The study time-period was divided into 4 equal quartiles for analysis. In the unmatched cohort (n=284), patients with CC-1 had significantly longer median OS compared to those with CC-3 (22.2 vs. 9.4 months, p<0.001). After 1:1 matching (n=172), the OS advantage of CC-1 persisted, with a median OS of 18.9 months (95% CI: 14.2-24.7) versus 10.5 months (95% CI: 9.4-12.3) for CC-3, p<0.0001, HR 0.4 (95% CI:0.27-0.56). Multivariable Cox regression confirmed CC-1 as a significant predictor of survival (HR: 0.15, 95% CI: 0.08-0.26). The CC-1 proportion went from 55 to 65% in time-periods 1 & 2-39% in period 3, to 11% in period 4; leading to significantly reduced survival rates in the latter time-periods 3 & 4. Near complete cytoreduction is associated with improved overall survival compared to open-close laparotomy. Prospective or standardized multicenter analyses will be required to confirm the clinical value of a near complete cytoreduction.

Autoimmune blistering diseases (AIBDs) in children and adolescents present a major challenge in diagnosis and management because of their rarity and the presence of other differential diagnoses with blistering. They are potentially serious, and timely diagnosis and optimal management can help prevent morbidity related to both the diseases and their treatments. The Autoimmune Blistering Diseases Task Force of the European Academy of Dermatology and Venereology (EADV) has initiated evidence-based, expert-informed clinical guidelines for the diagnosis, treatment and monitoring of paediatric AIBDs. A writing group conducted a systematic review of the available literature on AIBDs in children and adolescents, and since there is limited strong evidence in the literature, recommendations were formulated by consensus over three rounds of structured feedback and revision by the Core Group and Expert Group of the task force (32 voting experts; >75% agreement threshold). Although all AIBDs-including intra-epidermal and subepidermal blistering disorders-are reported in children, linear IgA dermatosis (LAD) is the most common (42-83%). The diagnostic approach in the paediatric age group does not differ from that in adults. Treatment recommendations are tailored to paediatric patients, taking into account developmental pharmacokinetics, potential adverse effects and the impact of disease and therapy on quality of life. These guidelines provide the first comprehensive, paediatric-focused framework for the diagnosis and management of AIBDs. Implementing the recommendations should harmonize care, reduce treatment-related toxicity and improve quality of life for affected children while steering future research towards critical knowledge gaps.

Zero-shot learning (ZSL) offers tremendous potential for histopathology image analysis, enabling models to generalize to unseen classes without extensive labeled data. Recent vision-language model (VLM) advancements have expanded ZSL capabilities, allowing task performance without task-specific fine-tuning. However, applying VLMs to histopathology presents considerable challenges due to the complexity of histopathological imagery and the nuanced nature of diagnostic tasks. We propose Multi-Resolution Prompt-guided Hybrid Embedding (MR-PHE), a novel framework for zero-shot histopathology image classification. MR-PHE mimics pathologists' workflow through multiresolution patch extraction to capture key cellular and tissue features. It introduces a hybrid embedding strategy that integrates global image embeddings with weighted patch embeddings, effectively combining local and global contextual information. Additionally, we develop a comprehensive prompt generation and selection framework, enriching class descriptions with domain-specific synonyms and clinically relevant features to enhance semantic understanding. A similarity-based patch weighting mechanism assigns attention-like weights to patches based on their relevance to class embeddings, emphasizing diagnostically important regions during classification. Experimental results demonstrate MR-PHE significantly improves zero-shot classification performance on histopathology datasets, often surpassing fully supervised models, showing its effectiveness and potential to advance computational pathology.