RT Journal Article SR Electronic T1 Prediction of NSCLC and SCLC using ki67 and the intratumoral metabolic heterogeneity assessed by 18F-FDG PET/CT JF Journal of Nuclear Medicine JO J Nucl Med FD Society of Nuclear Medicine SP 1047 OP 1047 VO 62 IS supplement 1 A1 Wu, Zhifang A1 Li, Xiaomeng A1 Hou, Mingxia A1 Li, Bingbing A1 Wang, Xinchao A1 Yang, Shuai A1 Zhao, Jingxu A1 Li, Yayuan YR 2021 UL http://jnm.snmjournals.org/content/62/supplement_1/1047.abstract AB 1047Objectives: To evaluate intratumoral matabolic heterogeneity and identify whether 18F-FDG PET/CT-derived metabolic parameters of the primary tumor and ki67 could predict different pathological type of primary lung cancer.Methods:Primary tumor regions of all patients were delineated, SUVmax, SUVpeak, SUVmean, metabolic tumor volume (MTV) and total lesion glycolysis (TLG) were obtained under different SUVmax thresholds. The intratumoral metabolic heterogeneity is represented by heterogeneity factor (HF), which is obtained by calculating the derivative of the volume threshold function of the 40%-90% threshold value of SUVmax. Mann-Whitney U test analyse was used to identify whether these PET/CT parameters and ki67 significantly different in NSCLC and SCLC. The optimal cutoff value for PET/CT parameters was assessed by the receiver operating characteristic (ROC) curve to discriminate NSCLC or SCLC.Results: A total of 59 consecutive patients who were performed pretreatment 18F-FDG PET/CT and confirmed different pathological type of primary lung cancer with tracheoscope or percutaneous lung puncture biopsy, of which 21(35.6%) squamous cell carcinoma(SQCC); 25(42.4%) adenocarcinoma(ADC) and 13(22%) small cell lung cancer(SCLC), and immunohistochemistry was performed. HF (12.556±2.337; P<0.01) and ki67 (51.95%±25.98%; P<0.001) were potential factors for the discrimination of NSCLC or SCLC. The ROC curve demonstrated HF and ki67 for discriminating NSCLC or SCLC showed the best performance compared with other 18F-FDG PET/CT metabolic parameters. The sensitivity, specifcity and AUC for HF and ki67 were 84.6%, 69.6%, 0.737; 84.6%,78.3%, 0.860, respectively. The optimal cutoff values were 7.891 and 65% for HF and ki67 were determined by ROC.Conclusions: HF and ki67 may predict different pathological types of primary lung cancer , especially for NSCLC and SCLC, and ki67 was better factor compared with HF. These fndings would be helpful in distinguishing patients with NSCLC or SCLC. More well-designed studies to confirm the current results and whether these parameters can further predict different types of NSCLC are needed.