Clinical and radiological characteristics of intermediate and high-risk cases in the Brazilian early lung cancer screening trial (BRELT3): Insights into Lung-RADS categories 3 and 4 and biopsy decision-making factors.

Abstract

10540 Background: Lung cancer is the leading cause of cancer death worldwide, and low-dose computed tomography (LDCT) effectively reduces mortality through early detection. BRELT1 and BRELT2, from the Propulmão initiative, highlighted the feasibility of lung cancer screening (LCS) programs in Brazil. This report examines demographic and radiological characteristics of patients with Lung-RADS (LR) 3 and 4 nodules in BRELT3, a mobile LDCT-based LCS initiative, and factors influencing biopsy indication. Methods: This prospective cohort study included current or former smokers (cessation ≤15 years), with a smoking history of ≥20 pack-years, aged 50–80 years. Those with LDCT classified as LR 3 or 4 were analyzed, and clinical and tomographic data were collected. The Brock malignancy probability model (PMB, 10% threshold) was applied retrospectively to assess its association with biopsy indication. Statistical methods included logistic regression, t-test, Mann-Whitney, chi-square, and Fisher’s exact test. Ethical Committee approval was obtained (SENAI-CIMATEC / Santa Izabel Hospital; n° 67431523.6.0000.9287 / 67431523.6.3001.5520). Results: Among 2018 screened patients, 223 (11.1%) had findings for lung cancer risk, Of these, 44.4% classified as LR3 and 55.6% as LR4. Median age was 64 years, with 87.4% self-identified as Black, 63.2% as current smokers, and 75.8% reported no family history of lung cancer. Nodules were predominantly single (82.5%) and solid (77.1%), with a mean size of 13.12 mm. LR3 nodules were smaller (9.53 mm) and exhibited lower PMB (7.43%) compared to LR 4 (15.99 mm; PMB: 18.93%). Biopsy was indicated for 45 participants (98% LR4). Nodules requiring biopsy were larger (22.97 vs 10.63 mm) and had higher PMB (26.44 vs 10.64%). Predictors for biopsy included irregular or spiculated contours (OR 5.83; p < 0.05), LR4B/4X vs.4A classification (OR 5.00; p < 0.05), PMB > 10% (OR 5.36; p < 0.05) and nodule size (OR 1.10; p < 0.05). Of the 45 nodules indicated for biopsy, 24 underwent the procedure, and 5 progressed to surgery. Conclusions: Mobile LDCT-based screening programs showed potential in identifying high-risk nodules among underserved populations. Factors influencing biopsy decisions included nodule size, irregular or spiculated contours, LR 4B/4X classification, and PMB > 10%. Integrating PMB into LCS may improve diagnostic accuracy and biopsy decision-making, enhancing early lung cancer detection. Association between nodules’ characteristics and biopsy indication. Characteristic (n = 45) OR (CI 95%; p-value)* Irregular or spiculated contours 5.83 (2.86 - 11.88; p < 0.05) Lung-RADS 4B/4X vs.4A 5.00 (1.61 - 15.54; p < 0.05) PMB - % 1.05 (1.03 - 1.07; p < 0.05) Nodule size - mm 1.10 (1.06 - 1.13; p < 0.05) *Univariate logistic regression. OR = Odds Ratio; CI = Confidence Interval.