The value of 18F-FDG PET / MR in skull-base bone invasion of nasopharyngeal carcinoma

Objectives: To explore the value of 18F-FDG PET/MR in evaluating skull-base bone invasion of nasopharyngeal carcinoma.

Methods: One hundred and twenty-four patients with NPC who underwent PET/MR and enhanced MR scan before or after treatment were reviewed retrospectively. All images were analyzed by an experienced PET/MR doctor and a radiotherapy doctors. Difference of metabolism between the bone invasion region and tumor body was compared and the metabolic differences of different types of bone invasion was also compared. Difference between GTV_PET(gross tumor volume), a tumor target region delineated with 30% of SUVmax as a threshold, and GTV_MR, a tumor target region delineated with fat suppression enhanced T2 sequence was compared.

Results: 68 patients showed skull-base bone erosion, of the 68 patients, 24 showed osteolytic bone destruction and 44 showed bone marrow infiltration. The osteolytic bone destruction area is characterized by soft tissue mass signals. The bone marrow infiltration type is characterized by low signal intensity replaced high signal intensity in the bone marrow on the T1-weighted images or when hyperintensity was observed on the fat-suppressed T2WI, and a high signal was seen on DWI. All enhanced scans of the bone invasion area showed abnormal enhancement. The SUVmax [3.31(1.64,4.95)]of the skull base lesion area was significantly lower than that [9.77(6.47,11.54)]of the main body of the nasopharyngeal lesion (Z=-7.143,P<0.01). The SUVmax [3.22(2.44,3.60)]of bone lesions in bone marrow infiltrating cases was significantly lower than that[6.18(4.72,7.45)] of osteolytic bone destruction (Z=-6.776,P<0.01). Among the 44 cases of bone marrow infiltration, FDG metabolism in the visible lesion area was higher than that in the contralateral or surrounding normal bone in 23 cases, and no intense high metabolic region was observed in 21 cases. The area under the ROC curve (AUC) for the evaluation of skull base invasion by SUVmax was 0.866, the SUVmax corresponding to the most approximate index was 1.5, and the sensitivity and specificity were 79.4% and 46.0%, respectively. In osteolytic bone destruction cases, 4.17% (1/24) of the cases of skull-base bone lesions were located outside the GTV_PET area, and 61.36% (27/44) of bone marrow infiltration cases were located outside the GTV_PET area,there were statistical differences(χ²=20.97,P<0.01). GTV_PET [35.46(25.27,43.56)] is smaller than GTV_MR [41.27(30.45,29.58)] (Z=-5.316,P<0.01).

Conclusions: Although PET / MR is of great value for staging of nasopharyngeal carcinoma, FDG uptake in the area of &#8203;&#8203;skull base invasion is lower than that of nasopharyngeal lesions, and false negatives often occur in bone marrow infiltrating types. Evaluation of skull base bone invasion and radiotherapy target region with PET metabolic range may not be accurate, and enhanced MR is still a necessary detection method for pre-treatment evaluation and efficacy evaluation.Female, 40 years old. Diagnosis of nasopharyngeal cancer (blue arrow). 1) pre-treatment; 2) post- treatment. The clivus signal behind the nasopharyngeal mass was abnormal before treatment. It showed a low signal on T1WI, an asymmetric hyperintense on T2WI, a high signal on DWI ,no significant increase in FDG uptake and enhancement was visible and significantly enhanced. Post-treatment, the nasopharyngeal mass subsided, and the clivus lesions still showed a low signal on T1WI, lower signal on T2WI than pre-treatment, no significant increase in FDG uptake, no high signal on DWI and no significant enhancement was observed.

• Download figure
• Open in new tab
• Download powerpoint