Diagnostic sensitivity of solid volume measurement for pathological invasion in non-solid lung adenocarcinoma.

Background:In the current tumor-node-metastasis (TNM) classification, the clinical T descriptor is defined by solid size (SS) on a computed tomography (CT) slice and the pathological one is done by invasive size (IS) in microscopic evaluations. We sometimes experience discrepancies in diagnosis of both descriptors. A volume analyzing application enables semi-automatic measurement of three-dimensional (3D) parameters in cases where there are discrepancies in diagnosing tumors' solid size and IS. In this study, we aimed to evaluate the association between 3D parameters and pathological invasion in non-solid small-sized lung adenocarcinomas. Methods:We enrolled 246 consecutive patients who underwent pulmonary resection at Shizuoka Cancer Center. Patients with lung adenocarcinomas that were radiologically non-solid, node-negative and sized ≤3 cm were eligible. We used a volume analyzing application to retrospectively measure 3D parameters of max and mean Hounsfield units (HUs) and solid volume (SV). The cut-off value of these parameters for diagnosing invasive adenocarcinoma (IAD) was set by describing receiver operating characteristic (ROC) curves. The correlation of IAD with these parameters was compared to its correlation with the SS. This study was not registered. Results:Of 246 patients with adenocarcinoma, 183 (74.4%) had IADs. In multivariate analyses, the total size (TS) and SS were significantly associated with IAD (P=0.006, 0.001, respectively), whereas 3D parameters including SV were not (P=0.80). In radiological adenocarcinoma (2.1-3.0 cm), SV >300 mm3 diagnosed IAD with a higher sensitivity than that of the SS (0.93 and 0.83, respectively). Conclusions:TS >20 mm and SS >5 mm were well-correlated with IAD. SV measurement may complement the current computed tomographic diagnosis of IAD based on the SS (2.1-3.0 cm).