Association of preoperative SUVmax with histological high-grade pattern in resected clinical stage IA lung adenocarcinoma: Derivation and temporal validation.

BACKGROUND: IASLC grading defines histologic high-grade lung adenocarcinoma as Grade 3 (≥20% high-grade pattern) and is associated with poor outcomes even in clinical stage IA disease. Reliable preoperative identification remains challenging. We evaluated whether primary-tumor SUVmax on FDG-PET/CT is associated with IASLC Grade 3 and whether a high-sensitivity cutoff retains performance in a temporally distinct cohort. PATIENTS AND METHODS: We retrospectively analyzed 224 patients with resected clinical stage IA lung adenocarcinoma who underwent nodal evaluation and preoperative FDG-PET/CT performed at our institution. The 2019-2022 cohort (n = 118) was used for derivation and the 2023-2025 cohort (n = 106) for temporal validation. In the derivation cohort, ROC analysis with a prespecified high-sensitivity strategy (sensitivity ≥0.85, maximal specificity) selected an SUVmax cutoff. Multivariable logistic regression entered tumor size, consolidation-to-tumor ratio (CTR; per 0.1), and SUVmax simultaneously. The fixed cutoff was then applied to the validation cohort. RESULTS: In derivation, SUVmax yielded an AUC of 0.733 (95% CI, 0.63-0.83), and a cutoff of 2.40 was selected. In multivariable logistic regression, SUVmax was the only variable independently associated with IASLC Grade 3 (OR, 1.12; 95% CI, 1.02-1.24; P = .022). In validation, SUVmax ≥2.40 achieved 92.9% sensitivity, 47.4% specificity, 94.9% negative predictive value, and 38.8% positive predictive value. CONCLUSION: In this selected cohort of resected clinical stage IA invasive lung adenocarcinoma, a fixed SUVmax cutoff of 2.40 demonstrated reproducible sensitivity and negative predictive value for IASLC Grade 3. SUVmax may serve as an adjunctive high-sensitivity marker for preoperative risk stratification.