Standardization of image quality across multiple centers by optimization of acquisition and reconstruction parameters with interim FDG-PET/CT for evaluating diffuse large B cell lymphoma

Objective: A multicenter trial is currently underway using FDG-PET/CT to evaluate diffuse large B cell lymphoma in Japan (JSCT NHL10). Standardization of the image quality between the participating centers is a fundamental aspect of the study. Within the framework of JSCT NHL10, standardization of the image quality was attempted by optimizing the acquisition and reconstruction conditions using mid-therapy FDG-PET/CT for diffuse large B cell lymphoma. This report describes the procedures and results of this attempt. Methods: The acquisition protocols and imaging quality were initially determined at each center and again after modification. The image quality was based on performance with an ¹⁸F-filled National Electrical Manufacturers Association standards body phantom. We determined that the acquisition duration and reconstruction parameters of each scanner evaluated were in compliance with the Japanese guideline for the oncology FDG-PET/CT data acquisition protocol: synopsis of Version 1.0 (the Guideline) based on the results of the phantom experiments performed by the Core Laboratory. Results: A total of 18 centers (19 scanners) participated in this trial. The center's default protocol was unchanged for 9 scanners (47.4 %) and changed for 10 scanners (52.6 %). Both acquisition duration and reconstruction parameters were changed in 3 (15.8 %) of 10 scanners and the acquisition duration alone was changed in 7 (36.8 %) scanners. Also, the accuracy of the standardized uptake value (SUV) was evaluated with the acceptable level 1.0 ± 0.1, resulting in readjustment and recalibration in 2 scanners (10.5 %), which were confirmed to attain the acceptable accuracy after the required readjustment. As of August 2012, 21 patients have undergone an FDG-PET/CT examination under the acquisition protocols determined by the Core Laboratory. Evaluation of the image quality using several physical parameters confirmed that the accumulated data were of sufficient quality. Conclusions: Optimization of the acquisition protocol, in compliance with the guideline, was successfully achieved by the Core Laboratory in the framework of JSCT NHL10 to accumulate equivalent quality data across multiple centers. The progress of the trial was greatly facilitated by support from the Japan Society of Nuclear Medicine Working Group for Investigation of Response Evaluation Criteria in Malignant Tumors Using Standardized PET/CT (Principal Investigator: Ukihide Tateishi, MD., PhD).