Abstract

Prospective randomised trials are required to evaluate response adapted treatment algorithms in lymphoma featuring decision making based on positron emission tomography (PET). Trial recruitment is likely to be optimised if patients can be scanned close to home but consistency in quality control (QC), scan acquisition and interpretation is needed for results to be comparable when scans are performed at different locations and to validate the eventual clinical results of the trial. We describe the arrangements for QC of PET imaging in an ongoing randomised controlled trial (RAPID) involving patients with clinical stages IA/IIA HL and 13 PET Centres across the UK. The aim of the trial is to test whether 18F-fluorodeoxyglucose PET can identify patients cured by chemotherapy who can therefore avoid consolidation radiotherapy (RT) and involves a randomisation between involved field RT and no further treatment in pts who are PET ‘negative’ (-ve) after 3 cycles ABVD. When 400 PET -ve patients have been randomised the trial is powered to reliably exclude a 7% difference in progression free survival between the trial arms. All PET Centres wishing to take part in RAPID had to have a fixed-installation PET scanner and were visited by a physicist from the ‘core lab’ in London who performed a phantom acquisition, validated data anonymisation and transfer, and confirmed adherence to a specific QC protocol. Phantom data were assessed to ensure consistency of image quality and reconstruction across all scanners. Scans are reviewed at the core lab by the same two nuclear medicine specialists using visual assessment and scored according to a 5 point scale (1, no disease; 2, probably no disease; 3, possibly disease; 4, probably disease; 5, definitely disease). For the purposes of the trial, centrally reviewed scores of 1 and 2 are considered ‘negative’ and these patients are randomised. Scores of 3, 4 and 5 are considered ‘positive’ and these patients receive a 4th cycle of ABVD and involved field RT. 331 patients who have completed initial chemotherapy have had PET scans centrally reviewed with central review scores returned to the clinical centre and trial office within 72 hours of receipt of images at the core lab. Scans have been acquired 7–15 days after day 15 of cycle 3 ABVD in 69.5% pts, 1–6 days after in 14.5% and 16 or more days after in 16%. One scan was unsatisfactory and had to be repeated because a site of original disease in the olecranon fossa was not included. The proportion of scans scored as ‘negative’ has been consistent over successive 6 month periods, varying between 72% and 85%. We conclude that central review of PET imaging can be successfully co-ordinated across multiple centres in real time using a central Core Lab featuring agreed methods for QC and image interpretation. This methodology, which can easily be applied to other clinical trial settings, provides quality assured PET data and ultimately allows robust conclusions to be drawn from trial results.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author