TY - JOUR T1 - Influence of counts statistic variation on metabolic response assessment in solid tumor lesions using new PET/CT technologies JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 716 LP - 716 VL - 58 IS - supplement 1 AU - Chiara Spadavecchia AU - Carlotta Dolci AU - Luca Guerra AU - Elena De Ponti AU - Sabrina Morzenti AU - Elia Anna Turolla AU - Sergio Todde AU - Claudio Landoni AU - Maurizio Arosio Y1 - 2017/05/01 UR - http://jnm.snmjournals.org/content/58/supplement_1/716.abstract N2 - 716Objectives: Semiquantitative PET data such as SUV (Standardized Uptake Value) peak can provide relevant clinical information for treatment response assessment in oncology. Aim of this study was to investigate if the counts statistic variation in terms of time/FOV can influence quantification and consequently metabolic response evaluation after treatment with new PET/CT technologies (large FOV -field of view- and high sensitivity) and reconstruction algorithms (RR -regularized reconstruction-).Methods: 19 target lesions (range 1.1 - 5.2 cm, median 2.4 cm) were selected in 16 patients with solid tumors (mean age 71 years, range 59-81). All lesions were evaluated with a baseline PET scan (1.5 minutes/FOV, standard iterative algorithm OSEM and RR reconstruction) before chemotherapy and/or radiotherapy and with a follow up scan after treatment on 5 rings Discovery IQ PET/CT scanner (GE Healthcare, Milwaukee, WI, USA). Follow up scans were acquired at 2 min/FOV and LIST data were reconstructed also at 1.5 and 1 min/FOV both using OSEM and RR algorithms. SUL (lean body mass SUV) peak was measured for each target lesion in baseline and follow up PET datasets with dedicated software (PETVCAR, AW 4.6 workstation-GE Healthcare, Milwaukee, USA). Statistical analysis was performed to evaluate the difference in quantification obtained from the PET datasets with different time/FOV, both for OSEM and RR datasets. For any pair of PET datasets, metabolic response for each lesion was assessed based on SUL peak variation, with a threshold of 30% considered as significant according to PERCIST criteria.Results: Median SUL peak of target lesions in baseline exams were 5.7 and 6.0 for OSEM and RR algorithms respectively (p=0.0032, paired t-test). Based on different time/FOV, SUL peak variations in follow up scans ranged from 0.2% to 21.2% and from 1.0% to 15.9% for OSEM and RR algorithms respectively. No statistically significant difference between SUL values associated to different time/FOV was found using the rank sum test. Metabolic response of target lesions was classified as follows: 2 (11%) progression of disease (PD), 9 (47%) stable disease (SD), 5 (26%) partial response (PR), 3 (16%) complete response (CR) and 2 (11%) PD, 8 (42%) SD, 6 (32%) PR, 3 (16%) CR for OSEM and RR algorithms respectively. No difference was observed comparing metabolic responses relative to different time/FOV in the follow up exams. Just in one case (5%) classification was different according to reconstruction algorithm (SD with OSEM, PR with RR reconstruction).Conclusion: SUL peak values at baseline exams showed a dependence on the reconstruction algorithm (with RR algorithm proving better) while no dependence was found comparing different counts statistics (time/FOV). Metabolic response assessment with new PET/CT technologies turned out to be highly stable and independent on counts statistic, thus on acquisition time, which can be very useful in routine clinical practice when patients disease or clinical workflow could require more flexible protocols, including lower scan duration. Research Support: none ER -