Quantification of Intratumoral Heterogeneity: Distinguishing Histological Subtypes in Clinical T1 Stage Lung Adenocarcinoma Presenting as Pure Ground-Glass Nodules on Computed Tomography

Rationale and Objectives

To quantify intratumor heterogeneity (ITH) in clinical T1 stage lung adenocarcinoma presenting as pure ground-glass nodules (pGGN) on computed tomography, assessing its value in distinguishing histological subtypes.

Materials and Methods

An ITH score was developed for quantitative measurement by integrating local radiomics features and global pixel distribution patterns. Diagnostic efficacy in distinguishing histological subtypes was evaluated using receiver operating characteristic curve analysis and area under the curve (AUC) values. The ITH score's performance was compared to those of conventional radiomics (C-radiomics), and radiological assessments conducted by experienced radiologists.

Results

The ITH score demonstrated excellent performance in distinguishing lepidic-predominant adenocarcinoma (LPA) from other histological subtypes of clinical T1 stage lung adenocarcinoma presenting as pGGN. It outperformed both C-radiomics and radiological findings, exhibiting higher AUCs of 0.784 (95% confidence interval [CI]: 0.742–0.826) and 0.801 (95% CI: 0.739–0.863) in the training and validation cohorts, respectively. The AUCs of C-radiomics were 0.764 (95% CI: 0.718–0.810, DeLong test, p = 0.025) and 0.760 (95% CI: 0.692–0.829, p = 0.023) and those of radiological findings were 0.722 (95% CI: 0.673–0.771, p = 0.003) and 0.754 (95% CI: 0.684–0.823, p = 0.016) in the training and validation cohorts, respectively. Subgroup analysis revealed varying diagnostic efficacy across clinical T1 stages, with the highest efficacy in the T1a stage, followed by the T1b stage, and lowest in the T1c stage.

Conclusion

The ITH score presents a superior method for evaluating histological subtypes and distinguishing LPA from other subtypes in clinical T1 stage lung adenocarcinoma presenting as pGGN.