Murat Kiliç, Merve Biyikli, Ayşenur Akatli
SUVmax shows a graded association with the predominant histological pattern and with LVI, but discrimination is modest. Our thresholds are two to three times higher than those from screen-detected series, indicating that published cut-offs are not transferable across populations, and SUVmax complements rather than replaces clinical and radiological assessment.
OBJECTIVES: We investigated whether preoperative maximum standardized uptake value (SUVmax) on PET-CT tracks the predominant histological pattern of resected lung adenocarcinoma and lymphovascular invasion (LVI).
METHODS: We retrospectively reviewed 166 consecutive patients resected between 2014 and 2026. Tumours were classified by predominant pattern (IASLC/ATS/ERS 2011) using descriptive labels, not grade terms: lepidic-predominant (n = 49), acinar/papillary-predominant (n = 59), solid/micropapillary/cribriform-predominant (n = 47) and invasive mucinous adenocarcinoma (n = 8). PET-CT was available in 94 (56.6%), who were compared with the 72 without. Pairwise comparisons were Holm-corrected and logistic models included age, sex and pleural invasion.
RESULTS: Imaged and non-imaged patients did not differ in any recorded variable (p ≥ 0.10). Median SUVmax rose stepwise across the three non-mucinous groups (8.80, 12.60, 15.40; Kruskal-Wallis p = 0.034, ε²=0.080); after Holm correction only the lepidic- versus solid/micropapillary/cribriform-predominant contrast remained significant (p = 0.042). Mucinous tumours showed low uptake (median 2.70; five imaged cases, descriptive only) and SUVmax was higher when LVI was present (13.10 vs 7.35; p = 0.001). Discrimination was modest: area under the curve 0.648 (95% CI 0.513-0.783) for an aggressive pattern and 0.698 (0.593-0.803) for LVI, with specificity below 50%. Adjusted for tumour size, age, sex and pleural invasion, SUVmax remained associated with LVI (odds ratio 1.12; 95% CI 1.05-1.21; p = 0.002).
CONCLUSIONS: SUVmax shows a graded association with the predominant histological pattern and with LVI, but discrimination is modest. Our thresholds are two to three times higher than those from screen-detected series, indicating that published cut-offs are not transferable across populations, and SUVmax complements rather than replaces clinical and radiological assessment.