The aim of this study was to analyse early and delayed acquisition on copper-64 diacetyl-bisN4-methylthiosemicarbazone (64Cu-ATSM) PET/CT in a small cohort of patients by comparing semiquantitative and computer-aided fractal geometry analyses.Patients and methods
Five cancer patients, including non-small-cell lung cancer and head and neck cancer, were investigated with 64Cu-ATSM PET/CT. Participants received an intravenous injection of 64Cu-ATSM according to body size and were imaged 60 min (early) and 16 h (delayed) later on hybrid PET/CT. Reconstructed images were visualized on advanced workstations for the definition of semiquantitative parameters: standardized uptake value (SUV)max, SUVratio-to-muscle, SUVmean, hypoxic volume (HV) and hypoxic burden (HB=HV×SUVmean). DICOM data retrieved from both scans were analysed using an ad-hoc computer program to determine the mean intensity value, SD, relative dispersion, three-dimensional histogram fractal dimension and three-dimensional fractal dimension.Results
All tumour lesions showed increased uptake of 64Cu-ATSM at early evaluation, with a median SUVratio-to-muscle of 4.42 (range: 1.58–5.62), a median SUVmax of 5.3 (range: 1.9–7.3), a median SUVmean of 2.8 (range: 1.5–3.9), a median HV of 41.6 cm3 (range: 2.8–453.7) and a median HB of 161.5 cm3 (range: 4.4–1112.5). All semiquantitative data obtained at 1 h were consistent with the parameters obtained on delayed imaging (P>0.05). A borderline statistically significant difference was found only for SUVmax of the muscle (P=0.045). Fractal geometry analysis on DICOM images showed that all parameters at early imaging showed no statistically significant difference with late acquisition (P>0.05).Conclusion
Our findings support the consistency of 64Cu-ATSM PET/CT images obtained at early and delayed acquisition for the assessment of tumour lesions.