Histopathology最新文献

Background: Breast phyllodes tumours (PT) are rare biphasic neoplasms consisting of epithelial and stromal components. They are classified into benign, borderline and malignant categories. Diagnosing and grading PTs present a challenge for pathologists, as there are multiple histological parameters and their own tiers. We aim to investigate the potential role of artificial intelligence (AI) as a diagnostic aid in determining PT grade.

Materials and methods: We investigated 15 PT whole slide images (WSIs), comprising 5 benign, 5 borderline and 5 malignant cases. We sought to classify and retrieve the most relevant WSIs by matching histological features at different patch sizes, using the Yottixel framework for WSI processing. Patches were extracted at a 20× magnification level. We then used the KimiaNet to extract feature vectors and transformed these into barcode representations. Barcodes of a query WSI were compared with those of others in the archive, allowing us to identify the most similar WSI and determine the PT grades from histological similarities.

Results: We utilized 'majority-n accuracy' as a measure of correctness, that is when the majority of the top-n search results have the correct diagnosis as the query patient. We achieved a maximum reported accuracy of 67%, with a 3000 × 3000 patch size when grading PT at majority voting with n = 4.

Conclusion: Despite the small sample size and absence of fine-tuning, our study demonstrated the potential of AI-based PT grade stratification using various patch sizes through histological matching. This serves as a preliminary proof of concept, with the prospect of refinement for potential routine clinical application.

Aim: Whole slide imaging (WSI) scanners are fundamental to digital pathology, providing high-resolution digitized histology slides. While scanner models vary in throughput, usability and image quality, the impact of these differences on pathologists' diagnostic performance remains insufficiently understood. This study evaluates the perceived image quality and diagnostic sufficiency of four commonly used high-throughput WSI scanners.

Methods and results: We selected four bright-field WSI scanners: Leica's Aperio GT 450 DX, Hamamatsu's NanoZoomer S360, 3DHISTECH's Pannoramic 1000 DX and Philips' IntelliSite Ultra Fast Scanner (first generation). A set of 601 slides was digitized on each scanner, and 96 representative slides were selected for evaluation. Fourteen pathologists from the University Hospital of Augsburg's Institute of Pathology assessed sharpness, brightness, contrast, colour realism and artefact frequency. To evaluate diagnostic sufficiency, pathologists were asked whether each scanned image was adequate for making a precise diagnosis when considering brightness/contrast/colour realism, sharpness and artefact frequency. Significant differences in perceived quality were observed across scanners, yet diagnostic sufficiency was rated more positively overall. Here, the Aperio GT 450 DX and NanoZoomer S360 scanners were ranked equally, each achieving the highest median rating for all image quality aspects. The Pannoramic 1000 DX and IntelliSite UFS scanners closely followed with deficits in colour realism and sharpness, respectively. Depending on the scanner model, 4.2% to 22.2% of images were deemed insufficient for diagnosis, with performance varying by tissue type and staining method.

Conclusion: Scanner differences in diagnostic sufficiency ratings were less pronounced than differences in perceived image quality. These findings highlight the importance of selecting appropriate scanning solutions based on workload and tissue characteristics.

Aim: Sessile serrated lesions with dysplasia (SSLd) are direct precursors of colorectal carcinomas (CRCs) but pose a diagnostic challenge. We aim to explore contemporary practice leading to recommendations to improve detection of dysplasia.

Methods and results: (i) The frequency of dysplasia in SSLs was estimated through a nationwide study of individuals undergoing colonoscopy in the Netherlands (2014-2022). Out of 186 427 SSLs, 17 456 showed dysplasia, yielding a frequency of 9.4%. (ii) A national audit evaluating diagnostic practices and interobserver variability revealed diagnostic discrepancies in 11% of SSLd cases, while ancillary immunohistochemistry (IHC) was used by only 20% of participating laboratories. (iii) The additional value of biomarkers (MLH1, p16, p53, beta-catenin and c-myc) was assessed using retrospective (n = 213) and prospective (n = 348) SSL cohorts. MLH1 emerged as the only useful biomarker for identifying dysplasia among SSLs, increasing SSLd diagnoses from 33 to 41 cases in the retrospective cohort (P = 0.008) and from 35 to 43 cases in the prospective cohort (P = 0.013). (iv) Misdiagnosis of SSLd among conventional adenomas was investigated using BRAF IHC in a cohort of 1572 advanced adenomas and was found to be very rare, occurring in only 2 cases.

Conclusions: The diagnosis of SSLd appears to have good reproducibility among pathologists and positive diagnostic trends that are observed in recent years. MLH1 is the only IHC marker with robust clinical utility to identify SSLd that do not meet the morphologic criteria for overt dysplasia but should be used only in selected cases due to its low prevalence.

Aims: Accurate diagnosis of lung adenocarcinoma (LUAD), particularly in small biopsy specimens, can be challenging due to limited tissue availability and the presence of histological mimickers. Claudin-18 (Cldn18), a tight junction protein with a lung-specific isoform (Cldn18.1), is normally expressed in alveolar epithelium and is implicated in maintaining alveolar integrity. Loss of Cldn18 has been linked to LUAD development through dysregulated YAP signalling. This study evaluated Cldn18 expression across lung cancer subtypes, with particular focus on LUAD.

Methods and results: Cldn18 immunohistochemical expression was retrospectively assessed in 391 lung resection specimens and 53 small biopsy samples. Normal alveoli and reactive pneumocytes consistently expressed Cldn18, whereas LUADs, especially non-mucinous LUADs (NM-LUADs), typically showed marked loss. Complete absence of Cldn18 was observed in 83% of NM-LUAD resections and 84.8% of NM-LUAD biopsies. ROC analysis demonstrated excellent diagnostic accuracy (AUC 0.992), with sensitivity of 98.4% and specificity of 100%. Lepidic NM-LUAD components, a frequent diagnostic pitfall especially in small specimens, showed low expression in most cases (>50% of tumour cells in 24/26). Similarly, 6/7 atypical adenomatous hyperplasia (AAH) cases exhibited low expression, while all reactive pneumocyte proliferations retained staining. By contrast, mucinous LUADs (M-LUADs) preserved Cldn18 expression in 35/54 cases, providing a clear distinction from NM-LUADs with mucin production.

Conclusions: Cldn18 loss is a promising diagnostic biomarker for NM-LUAD. Complete loss of Cldn18 expression could serve as an ancillary tool for distinguishing malignant lesions from NM-LUAD mimickers and for differentiating NM-LUADs with mucin production from true M-LUADs, particularly in small or challenging biopsies.

Introduction and aims: Lymphovascular space invasion (LVSI) is associated with increased nodal involvement and disease recurrence in cervical squamous cell carcinoma (SCC). While LVSI is currently included in some clinical risk assessment algorithms, the reproducibility of diagnosing LVSI remains understudied in cervical SCC. We aimed to assess the inter-observer agreement for LVSI diagnosis using digital whole slide images (WSIs) of haematoxylin and eosin-stained (H&E) slides with presumed true LVSI (n = 20) and LVSI mimics (n = 20) from early-stage cervical SCC identified in an international cohort.

Methods and results: The reference diagnosis for LVSI status was established by 2 pathologists through consensus review. Ten independent pathologists from North and South America, Europe, Asia and South Africa were included for blinded scoring of true LVSI versus LVSI mimics. Complete concordance was seen in 57% of cases. Overall percent agreement was 88% with the κ value being 0.76, reflecting substantial agreement on LVSI diagnosis based on H&E WSIs.

Conclusions: Our findings support the diagnostic feasibility and reliability of including LVSI as a parameter in routine histologic evaluation for risk stratification of cervical SCC. While the inter-observer agreement in our study is acceptable, there remains a need for standardised diagnostic criteria and guidance on the use of immunohistochemical studies for the detection or confirmation of LVSI in daily practice.

Aims: CIC-rearranged sarcoma (CRS), a highly aggressive sarcoma, is characterized by CIC fusion, with DUX4 being the most common partner. CRS is among the most difficult tumours to diagnose. Besides its relatively nonspecific small round or epithelioid morphology, genetic analyses often fail to detect CIC rearrangements. Although several diagnostic markers for CRS, namely WT1, MUC5AC, ETV4 and DUX4, have been introduced, their accessibility and interpretability are limited. Prompted by previous RNA-seq studies, we investigated the diagnostic utility of HMGA2 immunohistochemistry.

Methods: HMGA2 immunohistochemistry was evaluated in institutional archival cases from 2000 to 2025. Nine CRS and 119 other small round/epithelioid cell tumours were included. GATA3 and FOSL1 were also tested in selected CRS cases.

Results: HMGA2 was frequently positive for CRS, with a sensitivity of 78% (strong expression) or 100% (any expression). It was positive in other small round epithelioid cell tumours, including a subset of Ewing sarcoma, EWSR1::NFATC2 sarcoma, sclerosing epithelioid fibrosarcoma, mesenchymal tumours with ACTB::GLI1 fusion, and malignant melanoma. Overall, the specificity was 82% (strong expression) and 70% (any expression). HMGA2 was more readily interpretable than WT1 and MUC5AC, and double-negative CRS cases were strongly positive for HMGA2. Three cases were negative for GATA3 and FOSL1.

Conclusions: HMGA2 may be more readily implemented than ETV4 and DUX4, even in non-specialized hospitals. Thus, HMGA2 immunohistochemistry is a useful adjunct for CRS diagnosis. HMGA2 expression in CRS and other small round or epithelioid cell tumours should be tested in a larger series, particularly in non-DUX4 CRS and ATXN1/ATXN1L-rearranged sarcomas.