# Extracting Histologic Features to Distinguish Primary and Metastatic Squamous Cell Carcinoma of the Lung

**Authors:** Yuri Tachibana, Andrey Bychkov, Kris Lami, Jijgee Munkhdelger, Hoa Pham, Thiyaphat Laohawetwanit, Zun Pwint Oo, Izumi Sato, Luka Brcic, Junya Fukuoka

PMC · DOI: 10.1111/pin.70084 · 2026-01-14

## TL;DR

This study develops a histological scoring system to distinguish primary from metastatic squamous cell carcinoma in the lung, improving diagnostic accuracy.

## Contribution

The study introduces a novel scoring formula combining histologic features to differentiate primary and metastatic lung squamous cell carcinoma.

## Key findings

- A multivariate formula with interstitial fibrosis, squamous dysplasia, and tumor size achieved an AUC of 0.86.
- The formula showed 89.5% sensitivity and 70.2% specificity in distinguishing primary from metastatic cases.
- Seven histologic features were evaluated, with six associated with primary and one with metastatic tumors.

## Abstract

Lung is a common site of metastasis for squamous cell carcinoma (SqCC), and distinguishing primary lung SqCC from pulmonary metastatic SqCC is critical for clinical decision‐making, including treatment planning. However, no practical histological criteria have been established for routine diagnosis. This study aimed to develop histopathological criteria to differentiate lung SqCC and pulmonary metastatic SqCC. A total of 85 surgical cases (48 primary and 37 metastatic) were collected with clinical background data. Seven histological features were evaluated. Six [morphological heterogeneity, dilated airway, interstitial fibrosis (IF), squamous dysplasia (SD), emphysema, RB macrophage] were presented as primary‐associated features, while stromal infarction was proposed as metastatic. Seven pathologists scored these features. Primary‐associated findings showed significantly higher score in primary cases. Tumor size (size) was significantly larger in the primary group (median: primary, 30 mm; metastatic, 14 mm; p < 0.001). A multivariate analysis incorporating size produced the “pathological primary formula” based on parameter estimates: 0.70・IF + 0.36・SD + 0.09・size, with an AUC of 0.86, sensitivity of 89.5%, and specificity of 70.2%. These results suggest that the extracted histologic features may provide reproducible criteria for distinguishing primary lung SqCC from pulmonary metastatic SqCC, offering insight into potential diagnostic applications.

A scoring‐based formula was developed to distinguish primary from metastatic lung squamous cell carcinoma. Six features suggestive of primary and one of metastasis were scored from 0 to 2 by seven pathologists. The formula derived from multivariate analysis showed high AUC and sensitivity in both internal (n = 85) and external (n = 20) cohorts. Upper right: interstitial fibrosis; lower right: squamous dysplasia.

## Linked entities

- **Diseases:** squamous cell carcinoma (MONDO:0005096), lung squamous cell carcinoma (MONDO:0005097)

## Full-text entities

- **Genes:** CDKN2A (cyclin dependent kinase inhibitor 2A) [NCBI Gene 1029] {aka ARF, CAI2, CDK4I, CDKN2, CMM2, INK4}, TP53 (tumor protein p53) [NCBI Gene 7157] {aka BCC7, BMFS5, LFS1, P53, TRP53}
- **Diseases:** metastases (MESH:D009362), emphysema (MESH:D004646), lung lesion (MESH:D008171), CPFE (MESH:D011656), esophageal cancer (MESH:D004938), RB (MESH:D001988), necrosis (MESH:D009336), combined pulmonary fibrosis (MESH:D011658), Primary Tumors (MESH:D001932), SD (MESH:D002294), infarction (MESH:D007238), interstitial pneumonia (MESH:D017563), precancerous (MESH:D011230), Stage IV (MESH:D062706), head and neck (MESH:D006258), MH (MESH:C566911), dilated airway (MESH:D002311), Tumor (MESH:D009369), IF (MESH:D005355), Primary (MESH:D010538)
- **Chemicals:** eosin (MESH:D004801), H&amp;E (MESH:D006371), alcohol (MESH:D000438), hematoxylin (MESH:D006416), DA (-)
- **Species:** Homo sapiens (human, species) [taxon 9606]

## Figures

5 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12835960/full.md

---
Source: https://tomesphere.com/paper/PMC12835960