Clin Nucl Med 2016 Feb;41(2):e87-92
From the *Department of Nuclear Medicine, Humanitas Clinical and Research Center, Rozzano, Milan; †Alma Mater Studiorum University of Bologna, Doctoral Course in Specialized Medical Sciences, and ‡Department of Nuclear Medicine, University Hospital S. Orsola-Malpighi, Bologna; §Department of Nuclear Medicine, Ospedale Santa Maria della Misericordia, Rovigo, Italy; ∥Department of Radiology, University of Southern California, Los Angeles, CA; ¶Department of Medical Oncology, Ospedale degli Infermi, Faenza; #Department of Radiation Oncology, Bellaria Hospital, Bologna; **Department of Medical Oncology, Istituto Oncologico Romagnolo, Ravenna; and ††Division of Thoracic Surgery, University Hospital S. Orsola-Malpighi, Bologna, Italy.
Purpose: Cu-ATSM is a very promising PET radiopharmaceutical for tumor imaging of hypoxia. One of the advantages of this compound compared with other hypoxia-avid tracers is the high tumor-to-background signal offered, which guaranties facilitated tumor delineation. This study analyzes optimal semiquantitative and quantitative parameters obtained by Cu-ATSM PET/CT in the same cohort of patients with special focus on their correlation to disease outcome.
Patients And Methods: A prospective recruitment of 18 consecutive patients (M:F, 13:5; mean age, 60.7 years) with locally advanced non-small cell lung cancer (n = 7) or head and neck cancer (HNC) was performed. Each participant received 105 to 500 MBq of tracer according to body size and was scanned in a 3-dimensional mode PET/CT 60 minutes after tracer injection. PET images were reconstructed and visualized on a GE Advanced 4.6 workstation for the definition of semiquantitative and quantitative parameters: SUVmax, SUVratio-to-muscle, hypoxic tumor volume (HTV), and hypoxic burden (HB = HTV × SUVmean). These data were subsequently correlated to disease outcome, expressed in terms of progression-free survival calculated on a follow-up period with a median of 14.6 months.
Results: All patients showed a moderately to highly increased uptake of Cu-ATSM in tumor lesions, with a mean SUVmax of 5.2 (range, 1.9-8.3) and mean SUVratio of 4.4 (range, 1.6-6.8). In addition, a broad range of HTV and HB was defined as mean values of 99.3 cm (range, 2.5-453.7 cm) and 301 (4.2-1134), respectively. Receiver operating characteristic analysis identified as reference cutoffs with respect to disease outcome with the following values: SUVmax >2.5 (AUC, 0.57; sensitivity, 88.9%; specificity, 50%), SUVratio ≤4.4 (AUC, 0.60; sensitivity, 50; specificity, 83.3%), HTV >160.7 cm (AUC, 0.61; sensitivity, 55.6%; specificity, 75%), and HB >160.7 (AUC, 0.67; sensitivity, 58.3%; specificity, 83.3%). In our cohort, HB showed a statistically significant difference in terms of mean values on the analysis of variance test with respect to disease progression (P = 0.04). On univariate analysis, Cox regression confirmed these findings and showed a significant correlation to progression-free survival for HB (P = 0.05) and HTV (P = 0.02).
Conclusions: In our cohort, the definition of optimal semiquantitative and quantitative parameters on Cu-ATSM PET/CT seems feasible and in line with previously published data. However, when considering the prognostic role with respect to disease outcome, the more robust parameters are represented by HTV and HB.