Value of Patlak-Ki from ultra-high sensitivity dynamic total body [ 18 F]FDG PET/CT for evaluation of treatment response to induction immuno-chemotherapy in locally advanced non-small cell lung cancer (LA-NSCLC) patients.
DaQuan Wang,Bo Qiu,Qianwen Liu,Liangping Xia,Liu Songran,ChaoJie Zheng,Hui Liu,Yiwen Mo,Xu Zhang,Ying-Ying Hu,ShiYang Zheng,Yin Zhou,Jia Fu,NaiBin Chen,FangJie Liu,Rui Zhou,JinYu Guo,Wei Fan,Hui Liu
DOI: https://doi.org/10.1200/jco.2023.41.16_suppl.e20508
IF: 45.3
2023-06-01
Journal of Clinical Oncology
Abstract:e20508 Background: This study aimed to explore the value of metabolic features in predicting the response to induction immuno-chemotherapy in locally advanced NSCLC, using dynamic total body [ 18 F]FDG PET/CT. Methods: The LA-NSCLC patients who received two cycles of induction immuno-chemotherapy were analyzed in the study. The 60-minute dynamic total body [ 18 F]FDG PET/CT scan was administered before treatment. The primary tumors (PTs) were manually delineated. The metabolic features including the Patlak-Ki, Patlak-Intercept, the SUV max , metabolic tumor volume (MTV) and total lesion glycolysis (TLG) of PTs were evaluated. Using the Patlak graphical analysis, the Patlak-Ki of PTs was calculated from the 20-60 minutes frames. The Laplacian feature importance scores was used to select the best feature and an unsupervised K-Means method was applied to cluster patients. ROC curve was used to examine the effect of selected metabolic feature in predicting tumor response to treatment. The targeted next generation sequencing on 1021 genes was conducted. The expressions of CD68, CD86, CD163, CD206, CD33, CD34, Ki67 and VEGFA were assayed by immunohistochemistry. The independent samples t test and the Mann-Whitney U test were applied in the intergroup comparison. Results: Thirty-seven LA-NSCLC patients were analyzed between September 2020 and November 2021. All patients received two cycles of induction chemotherapy combined with Nivolumab/ Camrelizumab. The Laplacian scores showed that the Patlak-Ki of PTs had the highest importance for patient clustering, and the unsupervised K-Means derived decision boundary of Patlak-Ki was 2.779 ml/min/100g. Then Patients were divided into a high FDG Patlak-Ki (H-FDG-Ki, Patlak-Ki > 2.779 ml/min/100g) group (n = 23) and a low FDG Patlak-Ki (L-FDG-Ki, Patlak-Ki ≤ 2.779 ml/min/100g) group (n = 14). The ORR to induction immuno-chemotherapy was 67.6% (25/37) in the whole cohort, with 87% (20/23) in H-FDG-Ki group and 35.7% (5/14) in the L-FDG-Ki group (P = 0.001). The sensitivity and specificity of Patlak-Ki in predicting the treatment response were 80% and 75%, respectively [AUC = 0.775 (95%CI 0.605-0.945)]. The expression of CD3 + /CD8 + T cells and CD86 + /CD163 + /CD206 + macrophages were higher in the H-FDG-Ki group, while Ki67, CD33 + myeloid cells, CD34 + micro-vessel density (MVD) and tumor mutation burden (TMB) were comparable between the two groups. Conclusions: The total body [ 18 F]FDG PET/CT scanner performed a dynamic acquisition of the entire body and clustered LA-NSCLC patients into H-FDG-Ki and L-FDG-Ki groups based on the Patlak-Ki. H-FDG-Ki patients had better response to induction immuno-chemotherapy and higher immune cells infiltrations in the PTs than L-FDG-Ki patients. Further studies in a large patient cohort are warranted.
oncology