This Week in Pathology — Jul 4, 2026

Welcome to This Week in Pathology. This week we're covering ten notable papers spanning diagnostic quality and immunohistochemical specificity, advancements in thoracic and gastrointestinal oncology, and critical management thresholds in breast and pediatric pathology. Let's dive in.

We begin with diagnostic quality and the critical role of subspecialty review in genitourinary pathology. Accurate evaluation of bladder biopsies and transurethral resections of bladder tumors is essential for clinical management, yet diagnostic variability remains a significant challenge. A large-scale study published in Human Pathology evaluated over thirty-eight hundred consultations and institutional reviews, comparing original diagnoses with expert second opinions [2]. The researchers discovered that diagnostic discrepancies were not only common but frequently had major clinical implications. Specifically, tumor grade was discordant in seven percent of patients, and morphologic subtypes or variants differed in over ten percent. Remarkably, roughly twelve percent of cases required reclassification between benign and malignant diagnoses. Discrepancies in the depth of invasion resulted in upstaging in nearly eight percent of tumors, including critical transitions from carcinoma in situ to lamina propria invasion, and from lamina propria to muscularis propria invasion. Furthermore, two percent of cases were completely reclassified as non-urothelial tumors, including tumors of prostatic origin, pure squamous cell carcinoma, and inflammatory myofibroblastic tumors. This high rate of discrepancy underscores the clinical necessity of subspecialty consultation in challenging cases. This issue of diagnostic precision is also highly relevant when evaluating therapeutic biomarkers in bladder cancer. A study in the Archives of Pathology & Laboratory Medicine investigated human epidermal growth factor receptor 2, or HER2, in urothelial carcinoma [3]. While HER2-targeted therapies like trastuzumab deruxtecan show promise, the optimal scoring criteria remain controversial. The investigators compared the breast and upper gastrointestinal scoring criteria from the College of American Pathologists across one hundred thirty-nine specimens. They found that the two scoring systems yielded divergent results in a significant subset of cases. Using the upper gastrointestinal criteria, forty-one percent of cases were classified as HER2 three-positive, compared to only thirty-four percent when using the breast criteria. This means nearly one in ten patients could have different eligibility for targeted therapies depending solely on the scoring system applied. The study also noted that HER2 high expression was particularly enriched in micropapillary and plasmacytoid subtypes, and in brain metastases. Beyond scoring criteria, diagnostic specificity often hinges on the structural properties of our immunohistochemical antibodies. Writing in the Archives of Pathology & Laboratory Medicine, researchers investigated a common diagnostic pitfall involving the TPIT antibody clone OTI2G1, which is widely used to classify pituitary corticotroph tumors [4]. They discovered that this specific clone exhibits widespread, diffuse nuclear cross-reactivity in chordomas, staining over three-quarters of chordoma cells. By performing structural modeling, the authors demonstrated that the T-box domains of TBX19 and Brachyury share marked structural similarity, and the immunogen for this specific antibody clone directly overlaps with this conserved region. To avoid misdiagnosing a poorly differentiated chordoma as a pituitary neuroendocrine tumor in the sellar region, pathologists should utilize more specific clones, such as CL6251, which do not show this cross-reactivity. A similar need for careful marker interpretation is highlighted in a study from Virchows Archiv, which evaluated the diagnostic utility of the second-generation neuroendocrine marker INSM1 in pancreatic acinar cell carcinomas [5]. These rare tumors can show divergent neuroendocrine differentiation in about thirty-five percent of cases, creating a difficult differential diagnosis with pancreatic neuroendocrine tumors. Testing fifty-eight acinar cell carcinomas, the researchers found that INSM1 expression closely overlapped with traditional markers like synaptophysin and chromogranin A, but it offered no additional diagnostic advantage. Therefore, INSM1 expression in pancreatic lesions must always be interpreted in conjunction with overall morphology and a broad immunohistochemical panel. Finally, identifying the primary site of poorly differentiated tumors remains a classic pathological challenge, especially for rare, high-grade BRG1-deficient malignant neoplasms. A study in Human Pathology investigated the utility of TRPS1 and GATA3 in identifying the breast as the primary site for these aggressive tumors [7]. Although BRG1 loss is exceptionally rare in breast cancer, occurring in only one out of over four hundred cases evaluated on a tissue microarray, the researchers found that all five BRG1-deficient breast carcinomas in their cohort expressed TRPS1, and three co-expressed GATA3. In contrast, none of the seventeen BRG1-deficient tumors originating in the thorax, gastrointestinal tract, or gynecologic tract showed co-expression of these two markers. This makes the combination of TRPS1 and GATA3 highly specific for confirming a breast origin in these challenging cases.

Let's turn our attention to the thoracic cavity and gastrointestinal tract, where advanced biomarkers, computational tools, and precursor pathways are shaping our understanding of tumor biology. In thoracic oncology, distinguishing primary lung cancer from pulmonary metastasis during surgery is critical for determining the extent of surgical resection. A study published in Modern Pathology developed a deep learning model using over sixteen hundred frozen section slides to assist with this intraoperative decision [6]. The model achieved an overall accuracy of eighty percent and an area under the curve of zero point eight eight eight. It performed exceptionally well for lung adenocarcinomas, reaching over ninety-five percent accuracy for lepidic patterns, but was less reliable for squamous cell carcinomas, where accuracy dropped to sixty-two percent. This suggests that while artificial intelligence can provide reasonable, real-time support during intraoperative consultations, its performance varies significantly by histological subtype. For rare and aggressive lung cancers like large cell neuroendocrine carcinoma, identifying novel therapeutic targets is vital. A study in Histopathology investigated TROP2 immunoreactivity in fifty-eight pure cases and twenty-five mixed cases [10]. TROP2 expression was found in sixty percent of pure cases, with a quarter showing high expression in at least half of the tumor cells. Interestingly, high TROP2 expression was significantly associated with retained RB1 staining, though it had no impact on recurrence-free or disease-specific survival. These findings suggest that TROP2 antibody-drug conjugates could be a viable treatment option for a subset of these aggressive tumors. Moving down to the gastrointestinal tract, the American Journal of Surgical Pathology published a large series of two hundred seventy-one colorectal serrated lesions exploring the precursor pathways of traditional serrated adenomas, or TSAs [9]. The researchers investigated the newly proposed subtype, superficially serrated adenoma, or SuSA, alongside sessile serrated lesions and hyperplastic polyps. They discovered that RSPO fusions and KRAS mutations were highly prevalent in SuSAs and TSAs with SuSA-like precursor components. Conversely, TSAs with BRAF mutations were strongly associated with sessile serrated lesions and microvesicular hyperplastic polyps. This suggests two distinct evolutionary pathways: one where SuSA serves as a precursor to KRAS-mutated, RSPO-fusion-positive TSAs, and another where traditional sessile serrated lesions or hyperplastic polyps progress to BRAF-mutated TSAs.

Finally, we examine critical management thresholds and diagnostic signatures in breast surgical margins and pediatric hematopathology. In breast pathology, determining the optimal surgical margin for phyllodes tumors has long been a source of debate. A structured critical interpretative synthesis published in Modern Pathology reviewed forty cohorts to evaluate whether evidence supports specific margin widths [1]. The meta-analysis revealed that current guidelines do not recommend re-excision for positive or close margins in benign phyllodes tumors. For borderline and malignant tumors, although guidelines recommend negative margins, there is no established association between wider margins and lower rates of local recurrence. The estimated local recurrence rates are about twelve point five percent for borderline and sixteen point five percent for malignant tumors, but a mandatory ten-millimeter threshold is not supported by the data, suggesting narrower negative margins of at least one millimeter may be perfectly adequate in selected cases. Transitioning to pediatric hematopathology, NUP98 rearrangements represent a high-risk and often cytogenetically cryptic subgroup of myeloid neoplasms. A retrospective study of sixty-two patients in Modern Pathology characterized their immunophenotypic, cytogenetic, and molecular patterns based on their fusion partners [8]. The NUP98-NSD1 fusion, accounting for over half of the cases, occurred in older children with a median age of thirteen years and showed an immature myeloid phenotype with uniform CD123 expression and frequent FLT3-ITD alterations. In contrast, the NUP98-KDM5A fusion occurred in much younger children, with a median age of less than two years, and was associated with erythroid or megakaryocytic differentiation, complex karyotypes, and RB1 alterations. These distinct immunophenotypic and cytogenetic signatures can help pathologists prioritize targeted RNA-based fusion testing.

If you only have time for one paper this week, make it the multi-center reappraisal of surgical margins in breast phyllodes tumors published in Modern Pathology [1]. This comprehensive synthesis directly challenges the traditional, dogmatic adherence to wide surgical margins, demonstrating that a mandatory ten-millimeter threshold is unsupported and that a narrower negative margin of at least one millimeter is often sufficient for borderline and malignant cases, thereby sparing patients from unnecessary, disfiguring re-excisions.

Here are the key takeaways from this week in Pathology: First, second-opinion consultations for bladder biopsies and transurethral resections frequently result in clinically significant changes, with roughly one in eight cases reclassified between benign and malignant diagnoses, underlining the value of expert review. Second, when evaluating HER2 expression in urothelial carcinoma, the choice of scoring system is critical, as upper gastrointestinal criteria yield significantly more positive results than breast criteria, directly impacting patient eligibility for targeted therapies. Third, be aware of antibody cross-reactivity in the sellar region; the TPIT clone OTI2G1 cross-reacts with Brachyury in chordomas, meaning a more specific clone like CL6251 should be used to avoid diagnostic errors. Fourth, for BRG1-deficient malignant neoplasms of unknown primary, the co-expression of TRPS1 and GATA3 is highly specific for confirming a breast origin. And finally, in colorectal pathology, superficially serrated adenomas represent a distinct precursor pathway characterized by RSPO fusions and KRAS mutations, separate from the traditional BRAF-mutated sessile serrated pathway.

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.