Co-Occurrence Patterns of Pathological High-Risk Factors and Predictive Value of PET Metabolic Features in Lung Adenocarcinoma: A Retrospective Cohort Study of 1292 Cases

Background/Objectives: Co-occurrence patterns among pathological high-risk factors in lung adenocarcinoma remain poorly characterized, and whether preoperative 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) predicts them is unsettled. We mapped their interrelationships and evaluated the predictive value of preoperative maximum standardized uptake value (SUVmax). Methods: We retrospectively enrolled 1292 consecutive patients with resected invasive lung adenocarcinoma (May 2020–September 2023) and extracted six high-risk factors: spread through air spaces (STAS), visceral pleural invasion (VPI), lymphovascular invasion (LVI), occult lymph node metastasis (OLNM), and solid and micropapillary components. Pairwise co-occurrence was quantified by the Phi coefficient with Bonferroni correction; logistic regression and receiver operating characteristic (ROC) analysis assessed the predictive value of SUVmax. High-risk burden was defined as ≥3 concurrent factors. Results: STAS (29.8%) and the micropapillary component (24.4%) were the most prevalent. All 15 pairwise associations were positive and significant after Bonferroni correction; the strongest enrichment was between LVI and OLNM (observed-to-expected [O/E] ratio 5.85, Phi = 0.41, p < 0.001), with no mutual exclusivity. SUVmax independently predicted all six factors (e.g., solid component, odds ratio [OR] 1.283; STAS, OR 1.225; micropapillary component, OR 1.141; all p < 0.001). For high-risk burden, SUVmax achieved an area under the curve (AUC) of 0.838, which was numerically higher than that for any individual factor. At a cut-off of 3.0, sensitivity was 86.8% and the negative predictive value was 96.8% (bootstrap 95% confidence interval [CI] 0.807–0.865). Conclusions: High-risk pathological factors co-occur rather than segregate. Preoperative SUVmax is independently associated with all six factors but is best positioned to identify high-risk burden rather than individual features, although a clinical benefit remains to be demonstrated.