Lung Cancer Pathology

What it does

Lung Cancer Pathology reads the H&E slides of a lung biopsy or resection, together with a PD-L1 immunostain when one is uploaded. Lung biopsies are small, and every section cut for IHC or PCR is tissue lost for next-generation sequencing; the read is designed to get the most out of the slides already made.

On H&E the run separates adenocarcinoma, squamous cell carcinoma and other carcinomas. For invasive non-mucinous adenocarcinoma it gives the percentage of each growth pattern (lepidic, acinar, papillary, micropapillary, solid, cribriform) and the IASLC grade. On the PD-L1 slide it computes the tumour proportion score with an overlay of positive tumour cells. An EGFR-mutation probability from H&E ranks the case for rapid molecular testing.

Intended use

Decision support for subtyping and biomarker work-up in non-small cell lung cancer. The EGFR probability helps decide which cases go to rapid PCR or NGS first, and the TPS is reported with its clone and assay.

Who it is for

Thoracic and general surgical pathologists, residents, and molecular tumour boards that plan testing on scant tissue. Developers call the run to add structured lung-cancer results to a LIS or a trial pre-screening pipeline.

Accepted input

• H&E: whole-slide images of a small biopsy or of a resection.
• PD-L1 IHC: one slide of the same tumour, with the antibody clone stated.
• Formats: DICOM VL Whole Slide Microscopy Image, or vendor SVS, NDPI, MRXS, iSyntax, SCN, BIF, CZI and OME-TIFF files.

Requirements

Requirement	Why
Stain and marker per slide	H&E and PD-L1 slides go to different parts of the read
PD-L1 clone in metadata or upload	TPS bands follow the 22C3 cut-offs, and the score is reported with its assay
Pixel spacing in µm/px	Pattern areas and cell densities are computed per mm²
Enough viable tumour	Small crushed biopsies give low-confidence subtype and TPS, flagged in quality

Optional context

Clinical history and the question to answer (diagnosis, immunotherapy eligibility, molecular planning) can go in clinical_context to shape the guidance and draft report.

The result

One JSON result for the case, organised in sections.

Section	Content
Subtype and IASLC grade	Adenocarcinoma, squamous or other carcinoma; adenocarcinoma pattern percentages and IASLC grade
PD-L1 TPS	Tumour proportion score in the band <1 %, 1–49 % or ≥50 %, with a positive tumour-cell overlay
EGFR mutation probability	EGFR mutation probability from H&E in adenocarcinoma
Draft report	Opt-in (options.report): an English narrative that interprets the result for the reader (findings, impression, limitations); every number is checked against the findings

Standards

• Classification: WHO Classification of Thoracic Tumours, 5th edition; IASLC 2020 grading of invasive non-mucinous adenocarcinoma.
• PD-L1: 22C3 TPS cut-offs <1 %, 1–49 % and ≥50 %.
• Reporting: CAP lung cancer protocol.
• DICOM and HL7: pattern and tumour maps as Segmentation, scored cells as Microscopy Bulk Simple Annotations, scores in SR TID 1500; FHIR R4 DiagnosticReport with Observation entries coded in LOINC.

IASLC grade

The IASLC grade applies to invasive non-mucinous adenocarcinoma. It combines the predominant pattern with the amount of high-grade patterns (solid, micropapillary, complex glandular or cribriform): grade 1 is lepidic-predominant and grade 2 acinar- or papillary-predominant, each with less than 20 % high-grade pattern; grade 3 is any tumour with 20 % or more high-grade pattern.