This Week in Pathology — Jun 23, 2026

Welcome to This Week in Pathology. This week we are covering nine notable papers spanning diagnostic reproducibility and staging refinements, novel molecular signatures in solid tumors and hematopathology, and the emerging role of high-throughput data and artificial intelligence in medicine. Let us dive in.

We begin with a series of major studies in gastrointestinal and gynecologic pathology that focus on improving diagnostic precision, prognostic stratification, and clinical decision-making. In a nationwide, real-world cohort study published in Histopathology, researchers from the Dutch Esophageal Pathology Panel evaluated the diagnostic impact and reproducibility of p53 immunohistochemistry in Barrett's esophagus [5]. Analyzing over eleven hundred patient consultation cases and more than seventeen hundred biopsy levels, the expert panel demonstrated substantial interobserver agreement for p53 interpretation, with a Fleiss' kappa of 0.65. Furthermore, the inclusion of p53 immunohistochemistry directly influenced diagnostic decisions, prompting the reclassification of roughly nine percent of non-dysplastic biopsies with aberrant p53 to indefinite for dysplasia, and nearly five percent to low-grade dysplasia. For biopsies initially classified as indefinite, p53 staining shifted seventeen percent toward non-dysplastic and thirty-seven percent toward low-grade dysplasia. This robust, real-world evidence highlights that p53 immunohistochemistry is highly reliable and clinically meaningful, supporting its routine implementation to resolve diagnostic ambiguity in Barrett's esophagus. Moving from the esophagus to the colon, a study in The Journal of Pathology introduces the Tumor Invasive Border Index, or TIBI, as a reproducible method to quantify colorectal cancer morphology [7]. By measuring the proportion of tumor stroma and adipose tissue within a hotspot at the deepest point of invasion across two large cohorts of over eighteen hundred patients, the researchers found that a high TIBI, reflecting an infiltrative border, independently predicted significantly higher colorectal cancer-specific mortality. Specifically, patients with high TIBI had roughly one-and-a-half to two-and-a-half times the risk of death compared to those with pushing borders. On a molecular level, high TIBI was associated with mismatch repair proficiency, TP53 and KRAS mutations, and epithelial-mesenchymal transition upregulation, including high expression of L1CAM and DSG3. This establishes TIBI as a highly relevant, reproducible morphologic biomarker that bridges histopathology with underlying aggressive tumor biology. In gynecologic pathology, a study published in Modern Pathology explores whether staging and risk stratification can be refined for patients with co-existent endometrial and fallopian tube carcinoma [4]. Currently, fallopian tube involvement automatically classifies endometrial carcinoma as pathologic T3 or FIGO stage III, often leading to intensive therapy. The authors adapted the FIGO 2023 stage IA3 criteria, which are traditionally used for low-risk co-existent endometrial and ovarian carcinomas, to a cohort of seventy-one patients with fallopian tube involvement. They found that eleven percent of these carcinomas met the adapted low-risk criteria and experienced zero recurrences. By further refining these criteria to exclude adverse molecular features such as mismatch repair deficiency, p53 abnormalities, and ER-low or negative No Specific Molecular Profile, and expanding the criteria to include bilateral adnexal involvement, they identified fifteen percent of patients who had excellent, recurrence-free outcomes. This suggests that combining refined histopathological criteria with molecular profiling can successfully identify a subset of patients with indolent, adnexa-confined disease who may be safely considered for treatment de-escalation.

Our next theme highlights the ongoing challenges of diagnostic reproducibility, the discovery of novel molecular drivers, and the critical need for standardized reporting. Writing in Histopathology, researchers evaluated the reproducibility of the three-tiered World Health Organization fifth edition classification system for grading appendiceal goblet cell adenocarcinoma [1]. Seven pathologists with an interest in appendiceal pathology reviewed fifty-eight whole-slide images from twenty cases. Strikingly, all seven observers agreed on the final grade in only twenty percent of the cases. The overall interobserver agreement was fair at best, with a Fleiss' kappa of 0.29. When looking at individual histopathological features, agreement was worst for low-grade features like mild architectural disarray or tubular fusion, and high-grade features like necrosis. The best agreement was seen for neutral features like extracellular mucin and high-grade tumor sheets. These findings indicate that the current three-tiered World Health Organization grading system suffers from poor reproducibility in clinical practice, and the authors suggest that a validated two-tiered system of low- versus high-grade may be far more reliable for patient management. Meanwhile, a study in Modern Pathology introduces a highly specific molecular marker for salivary gland keratocystoma, a rare and benign parotid gland tumor that is frequently difficult to distinguish from malignant mimics [6]. Analyzing nineteen cases, the researchers identified RUNX2 gene rearrangements in one hundred percent of the tumors, with targeted RNA sequencing revealing a recurrent IRF2BP2::RUNX2 fusion gene. While immunohistochemistry showed high RUNX2 protein expression in keratocystomas compared to squamous cell carcinoma, epidermoid cysts, and branchial cleft cysts, the authors noted that immunohistochemistry alone was not sufficiently specific due to background staining. Therefore, molecular testing for RUNX2 rearrangements is recommended as a highly specific diagnostic tool to confirm this rare entity and avoid misdiagnosis. To address the clinical need for standardized reporting across diverse tumor types, the College of American Pathologists Biomarker Project Team published a comprehensive guidebook in the Archives of Pathology & Laboratory Medicine for navigating HER2 assessment across different organ systems [2]. With seven of eleven active College of American Pathologists biomarker templates now including HER2 status, this guide provides a critical framework for interpreting immunohistochemistry and in situ hybridization scoring criteria, addressing the challenges of designing unified reporting formats that accommodate alternative scoring systems across various primary solid tumors.

Our final theme explores how high-throughput technologies, comprehensive proteomic mapping, and artificial intelligence are shaping the future of pathology and clinical medicine. In Modern Pathology, researchers utilized optical genome mapping to characterize chromoanagenesis in three hundred and thirty-two patients with myelodysplastic syndromes [3]. Chromoanagenesis, a catastrophic genomic event marked by massive chromosomal rearrangements, was identified in approximately sixteen percent of cases. Among newly diagnosed patients, chromoanagenesis was almost exclusively associated with highly complex karyotypes and multi-hit TP53 alterations, both occurring in ninety-seven percent of cases, while canonical driver mutations were rare. Clinically, these patients had an exceptionally poor prognosis, with a median overall survival of only nine.point-nine months, which was significantly worse than non-chromoanagenesis patients, even those in the very high-risk IPSS-M category. This study demonstrates that optical genome mapping is an efficient tool for detecting chromoanagenesis, defining a biologically distinct, ultra-high-risk patient subset that requires aggressive therapeutic strategies. On a broader scale, a landmark study in Nature presents a massive, spatially resolved quantitative map of the human proteome [8]. Utilizing data-independent acquisition mass spectrometry, researchers profiled over thirteen thousand proteins across more than twenty-eight hundred samples, spanning fifty-eight major tissue types, two hundred and fifty-one sub-types, and twenty-five distinct carcinomas. This comprehensive proteomic landscape provides unprecedented insights into developmental biology, organ-specific drug toxicity, and oncogenic progression, offering a highly valuable quantitative resource for identifying novel therapeutic targets and drug repurposing opportunities. Finally, the integration of artificial intelligence into clinical workflows takes a major step forward in another Nature publication introducing MIRA, an autonomous medical artificial intelligence agent [9]. Operating in a sandboxed electronic health record environment, MIRA was designed to navigate complex clinical actions, including obtaining patient histories, ordering and interpreting laboratory and imaging tests, and formulating treatment plans. In simulations using real patient cases, MIRA outperformed physicians in diagnostic accuracy and made guideline-concordant, medication-safe decisions. While prospective, real-world studies are required to establish safety and governance, this work demonstrates the potential of electronic health record-integrated artificial intelligence agents to serve as highly effective clinical decision-support partners.

If you only have time for one paper this week, make it the nationwide evaluation of p53 immunohistochemistry in Barrett's esophagus by Weeda and colleagues in Histopathology [5]. This study provides robust, real-world evidence that standardized p53 immunohistochemistry is highly reproducible and directly resolves diagnostic ambiguity, supporting its routine implementation in clinical practice guidelines.

Here are the key takeaways from this week in Pathology. First, routine p53 immunohistochemistry in Barrett's esophagus is highly reproducible and significantly aids in reclassifying cases that are indefinite for dysplasia. Second, the three-tiered World Health Organization fifth edition grading system for appendiceal goblet cell adenocarcinoma suffers from high interobserver variability, suggesting a two-tiered system may be more reliable. Third, adapting the FIGO 2023 endometrial carcinoma stage IA3 criteria, combined with molecular profiling, can identify low-risk patients with co-existent endometrial and fallopian tube carcinoma who may benefit from treatment de-escalation. Fourth, optical genome mapping identifies chromoanagenesis in roughly sixteen percent of myelodysplastic syndrome cases, defining an ultra-high-risk patient subset with a dismal median survival of under ten months. And finally, RUNX2 gene rearrangements, specifically the IRF2BP2::RUNX2 fusion, serve as a highly specific diagnostic marker for salivary gland keratocystoma.

That's your roundup for This Week in Pathology. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And for audio briefings on your own clinical questions and papers, visit audioscholar dot C C.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.