Objectives
MRI is indispensable for staging of nasopharyngeal carcinoma (NPC) as it offers superior soft-tissue contrast. PET/CT and MRI are complementary in accurate staging of NPC. The combination of MRI and functional imaging from PET in PET/MR is promising in NPC management. We compared the diagnostic performance of PET/CT and PET/MR in 46 patients with NPC under different disease scenarios, including primary nonmetastatic cases, primary metastatic cases, recurrence and/or metastasis after treatment, and post-treatment follow-up cases.
Methods
Forty-six patients (37 males and 9 females) underwent both PET/CT and PET/MR within one day (median age: 54.5 years). Primary tumor extension into risk-stratified anatomic structures, retropharyngeal and cervical lymph node metastasis, distant metastasis and post-treatment follow-up results, as well as maximum of standardized uptake value (SUVmax) were evaluated and compared. Bland-Altman analysis was conducted to assess reproducibility of SUVmax between the two modalities. P < 0.05 was considered statistically significant.
Results
For high-risk structures, PET/MR detected two more sides of tensor/levator veli palatine muscle involvement, one more case of clivus involvement, and ruled out 12 false-positive sides of prevertebral muscle involvement by PET/CT. For medium-risk structures, PET/MR detected four more sides of medial pterygoid muscle involvement. For low-risk structures, abnormal signal on massa lateralis atlantis was detected by PET/MR. PET/MR detected 14 more positive retropharyngeal lymph nodes and more liver micrometastases than PET/CT. Overall, PET/MR changed two patients’ T staging. Furthermore, SUVmax showed high reproducibility between PET/CT and PET/MR (P < 0.001).
Conclusions
PET/MR outperforms PET/CT in delineating muscle, skull-base bone, and nodal involvement, and identifying liver micrometastases, may serve as a single-step staging modality for NPC.