Digital Quantification of Stromal Tumor-Infiltrating Lymphocytes in Lung Adenocarcinoma: Associations With Histologic Grade and PET-Derived Metabolic Parameters

Background/Aim Tumor-infiltrating lymphocytes (TILs) are important components of the tumor immune microenvironment in lung adenocarcinoma. However, their relationship with fluorodeoxyglucose (FDG) positron emission tomography (PET)/computed tomography (CT)-derived metabolic parameters remains unclear. This study investigated the association between digitally quantified stromal TIL density and clinicopathologic as well as metabolic imaging parameters in resected lung adenocarcinoma. Materials and Methods This retrospective study included 57 patients with surgically resected lung adenocarcinoma who underwent preoperative FDG PET/CT. Stromal TIL density was quantified digitally using QuPath (version 0.6.0) in selected stromal hotspot regions and expressed as cells/mm 2 . PET parameters included SUVmax, SUVmean, SULpeak, metabolic tumor volume (MTV), and total lesion glycolysis (TLG). Correlation and multivariable regression analyses were performed. Results Stromal TIL density showed a weak inverse correlation with MTV ( ρ = −0.287, nominal P = .039); however, this association did not remain statistically significant after FDR correction ( q = .109). No significant associations were observed with SUV-based parameters or TLG. Histologic grade demonstrated a moderate inverse correlation with TIL density ( ρ = −0.36, P = .007) and remained independently associated in multivariable analysis, although the association was borderline ( P = .046). Higher-grade tumors exhibited significantly increased metabolic activity. Lower stromal TIL density was also associated with advanced clinical stage and N1 lymph node metastasis ( P = .008). Tumor size correlated positively with all PET-derived metabolic parameters, most strongly with TLG ( ρ = 0.91, P < .001). Conclusion Digitally quantified stromal TIL density was associated with histologic grade and more advanced disease in resected lung adenocarcinoma, whereas its relationship with PET-derived metabolic parameters was limited. These exploratory findings require validation in larger cohorts using spatially representative immune assessment methods.