A recent Pathology digest,
as sent to subscribers

Study Type: Narrative review/commentary

Key Question: How should pathologists recognise and diagnostically approach cutaneous involvement by myeloid leukaemias and precursor conditions?

Key Findings:

• Reviews clinical and histopathological features of myelodysplasia cutis, myeloid sarcoma (cutaneous AML), and blastic plasmacytoid dendritic cell neoplasm, plus brief coverage of chronic myelomonocytic leukaemia skin involvement.
• Highlights key genetic markers relevant to diagnosis and discusses differential diagnoses for each entity, given overlapping clinical/histological presentations.
• No original data, statistics, or outcome measures presented; content is expert-based educational guidance.

Clinical Relevance: UK dermatopathologists and haematopathologists need clear diagnostic frameworks for these rare but prognostically significant skin infiltrates, as accurate recognition directly informs haematology referral, staging, and treatment pathways within NHS multidisciplinary teams.

Limitations: As a narrative review, it lacks systematic methodology and primary data, so recommendations reflect expert opinion rather than validated evidence.

Study Type: Narrative/scoping review (evidence mapping) of commercial and non-commercial AI tools for PD-L1 TPS scoring.

Key Question: Does published evidence support ranking or routine adoption of AI tools for PD-L1 tumour proportion scoring in NSCLC?

Key Findings:

• 20 tools reviewed (7 commercial, 13 non-commercial); commercial sources emphasised workflow integration and certification, while non-commercial studies detailed model architecture and code availability.
• Marked heterogeneity in cohorts, assay-scanner combinations, reference standards, and reported decision thresholds precluded direct performance comparison.
• Prospective multi-centre validation and reporting at clinically relevant TPS cut-offs remain scarce.

Clinical Relevance: UK pathology labs considering AI-assisted PD-L1 scoring should locally verify tool performance against their own assay-scanner workflow and TPS thresholds rather than relying on vendor claims or cross-study comparisons.

Limitations: The review relies solely on publicly available evidence, which is inconsistently reported and largely lacks independent external or prospective validation.

Study Type: Case report with molecular and immunohistochemical correlation

Key Question: Does Wagner-Meissner (W-M) schwannoma have a distinct molecular driver linking it to other HTRA1-rearranged schwannoma subtypes?

Key Findings:

• A plexiform W-M schwannoma from the volar pulp of a 37-year-old woman's finger harboured a SH3PXD2A::HTRA1 fusion, detected by RNA-based NGS.
• Tumour cells expressed S100, SOX10, Calretinin, and CD56, with NF1 loss; stroma showed CD34 and patchy EMA positivity.
• Findings suggest W-M body formation correlates histologically with the "serpentine" palisading pattern previously reported in HTRA1-fusion schwannomas.

Clinical Relevance: Recognising SH3PXD2A::HTRA1 fusions in W-M-predominant schwannomas may refine molecular classification and diagnostic algorithms for peripheral nerve sheath tumours in UK histopathology practice, particularly where NGS fusion panels are available.

Limitations: Single-case observation limits generalisability; further cases are needed to confirm this fusion as a consistent molecular feature of W-M schwannoma.

Study Type: Narrative review / classification update (non-systematic)

Key Question: How have recent WHO/ICC classification updates changed the diagnostic approach to primary cutaneous lymphomas and lymphoproliferative disorders?

Key Findings:

• Several previously provisional entities (e.g., CD4+ small/medium T-cell LPD, gamma/delta T-cell lymphoma, CD8+ aggressive epidermotropic cytotoxic T-cell lymphoma, EBV+ mucocutaneous ulcer) are now formally recognised diagnostic categories.
• Primary cutaneous CD8+ acral T-cell lymphoma has been reclassified as a lymphoproliferative disorder (renamed "LPD") reflecting its indolent, excellent-prognosis behaviour.
• Emerging tools highlighted include TCR sequencing, novel immunohistochemical/molecular markers, and AI-assisted diagnostics to improve discrimination between indolent, aggressive, and reactive lymphoid processes.

Clinical Relevance: Accurate synoptic diagnosis—integrating histology, immunophenotype, and molecular data—is essential for UK pathologists to prevent overtreatment of indolent LPDs or missed aggressive lymphomas, directly affecting staging and referral pathways within NHS haemato-oncology MDTs.

Limitations: As a narrative review, it synthesises expert opinion and classification consensus rather than new primary data, limiting evidence grading.

Study Type: Retrospective method-development study (deep-learning model validation)

Key Question: Can a deep-learning segmentation model detect occult cervical lymph node metastases from H&E sections alone, reducing the need for cytokeratin (CK) immunohistochemistry?

Key Findings:

• After fine-tuning on lymph node metastases, the model achieved precision 0.8778, recall 0.8607, accuracy 0.9193, and specificity 0.9458.
• Case-level analysis showed reduced Dice scores in harder cases, reflecting difficulty with micrometastatic/occult disease.
• AI-predicted tumour regions showed substantial spatial overlap with CK-positive areas, suggesting reliable localisation of small metastatic deposits using H&E alone.

Clinical Relevance: For NHS head and neck pathology services facing pressure on IHC turnaround times and staffing, this approach could streamline nodal staging in OSCC by triaging cases needing CK confirmation, improving efficiency without necessarily compromising detection of occult metastases.

Limitations: Single-institution, expert-annotated dataset with CK used selectively to guide annotation, and no independent multi-centre validation, limiting generalisability.

Study Type: Systematic literature review (narrative synthesis of 39 studies/84 liver samples)

Key Question: What are the histopathological features of vitamin A-induced liver toxicity in humans?

Key Findings:

• Hepatic stellate cell hypertrophy/hyperplasia was the near-universal finding (90.1%), representing the primary adaptive storage response.
• Progressive injury patterns included sinusoidal fibrosis (44.1%), chronic hepatitis (23.8%), and cirrhosis (25.0%), with severity linked to cumulative dose and duration (mean intake 112,006 IU/day over 6.8 years).
• Most cases arose from medical therapy (67.9%) or supplements (27.4%), affecting a wide age range (3–78 years) with balanced sex distribution.

Clinical Relevance: Provides UK liver pathologists with a histological reference framework for recognising VitA-related hepatotoxicity—relevant given rising supplement use and retinoid-based therapies—supporting differentiation from other steatotic/fibrotic liver diseases.

Limitations: Small, heterogeneous case-level dataset spanning six decades limits standardisation of histological criteria and generalisability.

Study Type: Retrospective cohort study

Key Question: Do radiologic, pathologic, and deep learning (DL)-derived response metrics predict clinical outcomes in RCC patients receiving ICI therapy before nephrectomy?

Key Findings:

• Radiologic shrinkage ≥30% was significantly associated with longer freedom from next systemic therapy (FFNT) (p=0.0036).
• Greater pathologic regression correlated with prolonged FFNT (HR 0.97, 95% CI 0.95–0.99; p=0.0023), with concordant DL-based quantification (HR 0.96, 95% CI 0.93–0.99; p=0.0041).
• Coagulative tumour necrosis remained associated with worse outcomes; DL-derived immune infiltrate and tumour size independently predicted FFNT on multivariable analysis.

Clinical Relevance: Establishes reproducible pathologic and AI-assisted criteria for assessing ICI treatment response in nephrectomy specimens, potentially informing standardised reporting and postoperative treatment stratification within UK renal cancer MDTs.

Limitations: Single-centre, retrospective design with modest sample size (n=99) limits generalisability pending prospective validation.

Study Type: Translational research study (integrated bulk and spatial proteomic analysis of lymph node tissue)

Key Question: What proteomic signatures distinguish Castleman disease subtypes, and what spatially compartmentalised molecular pathways drive iMCD-TAFRO pathogenesis?

Key Findings:

• Bulk proteomics differentiated hyaline vascular vs plasmacytic CD variants, showed systemic complement activation distinguishing iMCD from UCD, and supported iMCD-IPL as a molecularly distinct entity.
• iMCD-TAFRO showed upregulated angiogenesis (PDGFRβ, NOTCH3), interferon signalling (STAT1/ISG15), and fibrosis markers (COL3A1/LOXL1).
• Spatial proteomics revealed compartment-specific pathology: intrafollicular vessels enriched for myofibroblast/pyroptosis markers (ACTA2, GSDMD, TFRC); interfollicular regions showing TGF-β/SMAD3-driven fibrosis (LOXL1); follicular zones showing complement activation (CFHR1/2) and M2 macrophage infiltration.

Clinical Relevance: These findings refine molecular classification of Castleman disease subtypes and identify candidate diagnostic biomarkers and therapeutic targets relevant to UK haematopathology services managing this rare, diagnostically challenging condition.

Limitations: Likely limited by small sample size given tissue scarcity in this rare disease, restricting generalisability and statistical power.