Diagnostic performance of simultaneous PET-MR versus PET-CT in oncology with an overview on clinical utility and referral pattern of PET-MR: a single institutional study.

Background: PET-Magnetic Resonance (PET-MR) imaging is an upcoming investigative modality with a few installations in Asia and only three in India. PET-Computed Tomography (PET-CT) is an established diagnostic cornerstone for oncological indications but with limited resolution for small lesions due to low soft-tissue contrast and additional radiation exposure.

Objective: Our primary objective was to evaluate the diagnostic performance of simultaneous PET-MR and PET-CT in lesion detection in oncological practice. Secondly to assess the referral pattern and study the clinical utility of PET-MR in a university hospital practice.

Materials and methods: A total of 100 consecutive biopsy-proven cancer patients (breast or lung malignancy with suspected metastases) underwent 18 F Fluorodeoxyglucose (FDG) PET-MR and PET-CT for staging as a single injection, double examination protocol. Morphological lesion detection on correlative imaging/histopathology was used as the gold standard. Analysing the referral pattern, a total of 9366 patients underwent simultaneous PET-MR imaging for various indications in the past 5 years since installation.

Results: 18 F FDG PET-MR detected 100% of primary tumours in breast/lung carcinoma patients while PET-CT was positive in 96%. Overall accuracy of nodal metastases detection for PET-MR and PET-CT was 96 and 88%, while for distant metastases the accuracy was 100 and 93%, respectively. FDG PET-MR proved more sensitive and specific than PET-CT for regional nodal ( P  = 0.011 and P  < 0.001) and distant metastases detection ( P  = 0.017 and P < 0.001, respectively). Analysing the general referral pattern for PET-MR, the majority were oncology referrals when compared to nononcological indications (66.5, 33.5%). About 66.24% were FDG based, followed by 68 Ga Prostate-Specific Membrane Antigen (PSMA) and dodecane tetraacetic acid (DOTA). The general utility of PET-MR was found incremental in better delineation of small lesions especially in head, neck, liver, brain and gynaecological malignancies.

Conclusion: In our past 5 years of PET-MR practice, we found that simultaneous PET-MR is a highly recommended ideal imaging technique for oncological and nononcological indications. It has excellent diagnostic performance with high sensitivity, specificity and accuracy when compared to PET-CT in primary tumour, nodal and distant metastases (TNM) staging in specific subgroup of breast and lung malignancy patients.