The value of relative CT attenuation in predicting invasiveness in patients with T1-stage lung adenocarcinoma.

Background: Computed tomography (CT) values of the lungs vary among different populations. Relative CT values, which are derived by comparing the CT differences between pulmonary nodules and normal lung tissue, can theoretically reduce the interference of factors such as emphysema and improve diagnostic accuracy. This study aims to investigate the performance of relative CT values in predicting the invasiveness of T1 stage lung adenocarcinoma (LUAD) by comparing them with absolute CT values of pulmonary nodules, using the mean CT value of the lung lobe where the nodule is located as a reference. Methods: This retrospective study analyzed 1,259 patients with T1 stage LUAD [including atypical adenomatous hyperplasia (AAH), adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and invasive adenocarcinoma (IAC)] who underwent surgical resection and had pathological confirmation from three centers. Patients were divided into non-invasive (AAH, AIS, MIA) and invasive (IAC) groups based on pathological invasiveness. Original imaging data were imported into the Shukun artificial intelligence workstation to quantify three-dimensional CT features of the pulmonary nodules, including maximum CT value, mean CT value, and median CT value. The mean CT value of the lung lobe where the nodule was located was obtained using the software's whole-lung analysis function, and relative CT differences and ratios were calculated. Statistical analysis was performed to compare the differences in relative and absolute CT values between the two groups to evaluate their performance in predicting the invasiveness of T1 stage LUAD. Patients were categorized into four groups based on the mean CT value of the lung lobe: >-750 Hounsfield units (HU), -750 to -800 HU, -800 to -850 HU, and ≤-850 HU. Subgroup analysis was conducted on factors with statistical differences to explore the differences in predictive performance between relative and absolute CT values as the degree of emphysema in the lung lobe where the nodule is located increased. Statistical analysis and data visualization were performed using R 4.2.1, with P<0.05 indicating statistical significance. Results: A total of 1,259 pulmonary nodules were enrolled, with 828 cases (65.77%) assigned to the non-invasive group and 431 cases (34.23%) to the invasive group. The non-invasive group had a mean age of 54.08 years, significantly lower than the invasive group's mean age of 60.48 years (P<0.001). Nodule distribution across lung lobes was: right upper lobe [418], right middle lobe [91], right lower lobe [235], left upper lobe [322], and left lower lobe [193]. Both relative CT attenuation and absolute CT attenuation metrics (mean, median, maximum CT values) demonstrated consistent predictive value for invasiveness in T1 adenocarcinoma (P<0.05 for all). Subgroup analyses showed no significant differences in predictive performance between relative and absolute CT values, irrespective of the mean CT attenuation of the nodule-containing lobe (P>0.05). Conclusions: In T1-stage LUAD, both lobe-normalized relative CT attenuation and nodule-specific absolute CT attenuation provide reliable predictors of invasiveness. However, relative CT values did not outperform absolute CT values in this study. Thus, we advocate for continuing to use absolute CT attenuation values as the key imaging biomarker for predicting adenocarcinoma invasiveness, due to their simplicity and equivalent predictive validity.