Discrepancy Between Preoperative CT and Pathological Tumor Diameters in Stage I Lung Adenocarcinoma and Its Association with Postoperative Recurrence.
Authors
Affiliations (4)
Affiliations (4)
- Radiology Department, Ningbo No. 2 Hospital, Ningbo 315010, China (M.L., Q.D., J.Z., H.C., J.Z.); School of Medicine, Shaoxing University, Shaoxing, Zhejiang, China (M.L., M.D., H.Z.).
- School of Medicine, Shaoxing University, Shaoxing, Zhejiang, China (M.L., M.D., H.Z.).
- Radiology Department, Ningbo No. 2 Hospital, Ningbo 315010, China (M.L., Q.D., J.Z., H.C., J.Z.); Ningbo Key Laboratory of Digital Imaging and Medical-Engineering Interdisciplinarity, Ningbo 315010, China (Q.D., J.Z., H.C., J.Z.).
- Radiology Department, Ningbo No. 2 Hospital, Ningbo 315010, China (M.L., Q.D., J.Z., H.C., J.Z.); Ningbo Key Laboratory of Digital Imaging and Medical-Engineering Interdisciplinarity, Ningbo 315010, China (Q.D., J.Z., H.C., J.Z.). Electronic address: [email protected].
Abstract
To assess the concordance between preoperative CT-measured and postoperative pathological tumor diameters in stage I lung adenocarcinoma, and to examine whether their discrepancy is associated with postoperative recurrence. We retrospectively analyzed patients with stage I lung adenocarcinoma who underwent curative resection from 2019-2021. CT tumor diameter was obtained using AI-assisted 3D segmentation and reviewed by experienced thoracic radiologists. Pathological size was measured on the maximal tumor section of formalin-fixed specimens. Correlation was assessed using Pearson's r, and agreement examined with Bland-Altman analysis. Patients were grouped by CT-pathology discrepancy (<-3 mm, -3 to 3Â mm, >3Â mm). Recurrence-free survival (RFS) was compared using Kaplan-Meier analysis, and independent predictors of recurrence were identified via multivariable Cox proportional hazards regression. An integrated prognostic model incorporating significant variables was internally validated. A total of 239 patients were included in the study and of those, 48 recurred. CT and pathological tumor diameters were moderately to highly correlated (r = 0.876, P < .001), with a mean positive bias of 2.9Â mm. RFS differed significantly across groups (P < .001). Multivariable Cox analysis identified imaging overestimation (CT-pathology discrepancy >3Â mm), higher pathological T stage, older age and male as independent predictors of recurrence. The integrated model showed acceptable discrimination (optimism-corrected C-index 0.818) and calibration, enabling risk stratification. CT measurements tend to overestimate tumor size, and excessive overestimation (difference >3Â mm) was independently associated with higher postoperative recurrence risk. Incorporating CT-pathology discrepancies into prognostic models may improve postoperative risk stratification.