Purpose: The advent of hybrid positron emission tomography/magnetic resonance (PET/MR) systems has opened the gate to integrated multiparametric evaluation of patients with suspected cardiac sarcoidosis. The aim of the present paper is to systematically review all available literature to define the role of hybrid PET/MR in patients with confirmed or suspected cardiac sarcoidosis. Methods: The review strategy was developed following the PRISMA 2009 statement. An electronic database (Medline, Google Scholar, Clinical Trials.gov) search was conducted to identify articles on hybrid PET/MR in cardiac sarcoidosis. Only reports on hybrid PET/MR systems were considered. All PET tracers were considered. No date limitation was applied. Following duplicate removal, abstracts or full articles in case of case reports were screened for inclusion criteria. The full text of the remaining studies was reviewed. Results: Eleven studies were found, including seven case reports, two case series, and two diagnostic accuracy studies. All studies reported feasibility and interpretability of cardiac 18F-FDG PET/MR in cardiac sarcoidosis. The two case series demonstrated non-inferiority of PET/MR compared to PET/CT and, in fact, suggested greater sensitivity. Diagnostic accuracy studies showed satisfactory sensitivity and specificity of PET/MR to diagnose cardiac sarcoidosis. Two reports also showed the usefulness of cardiac PET/MR to assess response to treatment. Conclusion: Hybrid cardiac PET/MR is a valuable tool for the diagnosis of cardiac sarcoidosis. Moreover, as it offers activity and tissue characterisation, it can be used to tailor and monitor treatment.

Original languageEnglish
Pages (from-to)317-326
Number of pages10
JournalClinical and Translational Imaging
Issue number5
StatePublished - 1 Oct 2019


  • Cardiac inflammation
  • Cardiac sarcoidosis
  • Fusion imaging
  • PET/MR


Dive into the research topics of 'Hybrid cardiac PET/MR: the value of multiparametric assessment in cardiac sarcoidosis'. Together they form a unique fingerprint.

Cite this