Combining volume based metabolic parameters on preoperative ¹⁸F-FDG PET/CT and clinical features could predict lymphovascular invasion in patients with non-small-cell lung cancer

Objectives: Our study intended to explore the association combining ¹⁸F-FDG PET/CT metabolic parameters and clinical features and lymphovascular invasion status in non-small-cell lung cancer (NSCLC).

Methods: This retrospective study included 161 patients (median age, 63 years; interquartile range, 55.5 - 67.0 years) from January 2018 to November 2020 with NSCLC who underwent ¹⁸F-FDG PET/CT scan within one month before surgery in our hospital. The PET/CT metabolic parameters including maximum and mean standard uptake values (SUVmax and SUVmean), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) of primary lesion were measured. The associations between PET/CT parameters and clinical characteristics (age, sex, smoking, serum tumor marker, histological type, T-stage, N-stage, TNM stage, pleural invasion) and lymphovascular invasion status were analyzed. Multivariate logistic regression analysis was performed to explore independent predictive factors for lymphovascular invasion.

Results: A total of 161 eligible patients with NSCLC enrolled in this study, including 111 males and 50 females, were divided into two groups, lymphovascular invasion-positive group (n=38, 23.6%) and lymphovascular invasion-negative group (n=123, 76.4%). Histopathological findings included adenocarcinoma (n=119), squamous cell carcinoma (n=38), undefined NSCLC (n=1) and adenosquamous carcinoma (n=3). The average of SUVmax and SUVmean of primary lesion were 15.43 ± 5.91 vs. 11.73 ± 7.53 (p = 0.006) and 6.02 ± 1.58 vs. 5.07 ± 2.03 (p = 0.009), in the lymphovascular invasion-positive group and lymphovascular group-negative group, respectively. The median MTV and TLG were 15.80 (8.22, 31.07) vs. 5.47 (1.47, 20.17) (p < 0.001), 88.78 (42.68, 230.00) vs. 24.47 (5.95, 136.08) (p < 0.001) in the two groups, respectively. Multivariate analysis revealed that high MTV ≥ 6.385 (p = 0.002) and N-stage (p = 0.003) correlated with lymphovascular invasion in NSCLC patients. The receiver operating characteristic (ROC) curve yielded the area under curve (AUC) values of 0.704 and 0.854 for the high MTV and the combination of the abovementioned three factors, respectively. Conclusion: This study indicated that combining MTV on preoperative ¹⁸F-FDG PET/CT and N-stage yield a favorable predictive performance for lymphovascular invasion in NSCLC patients.