Research Article: Conspicuity of Malignant Lesions on PET/CT and Simultaneous Time-Of-Flight PET/MRI

Date Published: January 19, 2017

Publisher: Public Library of Science

Author(s): Ryogo Minamimoto, Andrei Iagaru, Mehran Jamali, Dawn Holley, Amir Barkhodari, Shreyas Vasanawala, Greg Zaharchuk, Gayle E. Woloschak.


To compare the conspicuity of malignant lesions between FDG PET/CT and a new simultaneous, time-of-flight (TOF) enabled PET/MRI scanner.

All patients underwent a single-injection of FDG, followed by a dual imaging protocol consisting of PET/CT followed by TOF PET/MRI. PET/CT and PET/MRI images were evaluated by two readers independently for areas of FDG uptake compatible with malignancy, and then categorized into 5 groups (1: PET/MRI and PET/CT positive; 2: PET/MRI positive, PET/CT positive in retrospect; 3: PET/CT positive, PET/MRI positive in retrospect; 4: PET/MRI positive, PET/CT negative; 5: PET/MRI negative, PET/CT positive) by consensus. Patients with no lesions on either study or greater than 10 lesions based on either modality were excluded from the study.

Fifty-two patients (mean±SD age: 58±14 years) underwent the dual imaging protocol; of these, 29 patients with a total of 93 FDG-avid lesions met the inclusion criteria. The majority of lesions (56%) were recorded prospectively in the same location on PET/CT and PET/MRI. About an equal small fraction of lesions were seen on PET/CT but only retrospectively on PET/MRI (9%) and vice versa (12%). More lesions were identified only on PET/MRI but not on PET/CT, even in retrospect (96% vs. 81%, respectively; p = 0.003). Discrepant lesions had lower maximum standardized uptake value (SUVmax) than concordant lesions on both modalities (p<0.001). While most lesions were identified prospectively on both modalities, significantly more lesions were identified with PET/MRI than with PET/CT.

Partial Text

Hybrid positron emission tomography / magnetic resonance imaging (PET/MRI) is one of the latest advances in multimodality technologies, and provides both biological and morphological information of malignant lesions [1]. Compared to PET/CT, the general advantages of PET/MRI are reduction of radiation exposure, use of MRI to image organ function, and improvement of diagnostic ability due to the better contrast of MRI imaging [2, 3]. Several studies have used combined data from separate PET and MRI examinations; however, these studies had limitations in terms of the time interval between studies and the potential for misregistration [2, 4]. Recent studies of simultaneous, non-time of flight (TOF) enabled PET/MRI scanners in clinical practice have shown promising initial results for several clinical indications [5–7].

Between January 2014 and February 2015, 52 patients were recruited consecutively in this prospective study (mean age, 58±14 years [range 27–86 years]; 25 male, 27 female). 11.5% of the participants were referred for initial treatment strategy, the rest were referred for subsequent treatment strategy (treatment monitoring, restaging and detection of suspected recurrence, etc.) [10]. More than 10 lesions were identified either by PET/MRI or PET/CT in 9 patients; these patients were excluded from analysis. In 14 patients, no focal FDG uptake suggesting malignancy was detected on either PET/CT or PET/MRI (Fig 1).

This study directly compares the sensitivity of PET/CT and TOF-enabled PET/MRI for detection of malignant lesions on FDG examinations obtained for clinical purposes in oncology patients. TOF PET/MRI provided comparable diagnostic ability with PET/CT, despite decreased FDG activity for imaging at a later time point. Significantly more lesions were identified with PET/MRI than with PET/CT. The reasons for this are not entirely clear, but may relate to improved lesion-to-background at later imaging times, the increased sensitivity of the TOF PET detectors in PET/MRI, longer imaging times that are possible due to the need to acquire MR information, and the superior soft tissue contrast afforded by the simultaneous MR imaging. Given the structure of the study, it is difficult to disentangle these factors and, in any case, it is likely to be multifactorial. Perhaps it is not surprising that PET/MRI performed better than PET/CT given that the PET/CT was performed without a diagnostic CT, while the PET/MRI, in contrast, included multiple diagnostic MR sequences. However, each method of review is consistent with the most commonly used approach for that modality at the current time, so we believe this simulates the clinical situation accurately.

While most lesions were identified prospectively on both modalities, significantly more lesions were identified with PET/MRI than with PET/CT.