The role of 13 N -ammonia in the differential diagnosis of gliomas and brain inflammatory lesions

Abstract

Objective

To investigate the utility of ¹³N-ammonia PET/CT imaging in the differential diagnosis of gliomas and brain inflammations.

Methods

¹³N-ammonia PET/CT imaging data of 77 patients with gliomas and 34 patients with brain inflammations were retrospectively analyzed. No patients received any treatment before ¹³N-ammonia imaging. All the patients were diagnosed by stereotactic biopsy or clinical follow-up. Visual and semi-quantitative analysis was performed to analyze the results of ¹³N-ammonia imaging. Finally, the uptake ratios of each lesion were calculated and its differences among different groups were tested with one-way ANOVA.

Results

29.4% inflammations, 51.6% low-grade gliomas and 91.3% high-grade gliomas were positive by visual analysis in ¹³N-ammonia imaging. The sensitivity, specificity and accuracy for the diagnosis of gliomas were 75.3%, 55.8% and 67.8%, respectively. As for semi-quantitative analysis, the T/G ratios of inflammatory lesions, low-grade gliomas and high-grade gliomas were 0.88 ± 0.24, 1.04 ± 0.43 and 1.43 ± 0.49, respectively. One-way ANOVA revealed that the T/G ratios of high-grade gliomas were significantly higher than those of low-grade gliomas and inflammations (P < 0.05), but there was no statistical difference between low-grade gliomas and inflammations (P = 0.118). Among the inflammatory lesions, T/G ratios were not statistically different between infectious and demyelinating lesions (P > 0.05). ROC curve analysis showed that the optimal cut-off value of T/G ratio in distinguishing gliomas from inflammations was 1.21 with the AUC 0.78. The sensitivity, specificity, accuracy, PPV and NPV were 52.9%, 94.4%, 65.3%, 95.7% and 45.9%, respectively. ROC curve analysis showed that the optimal cut-off value of T/G ratio in distinguishing high-grade gliomas from low-grade gliomas was 1.06 with the AUC 0.78. The sensitivity, specificity, accuracy, PPV and NPV were 81.5%, 67.7%, 76.5%, 81.5% and 67.7%, respectively. ROC curve analysis showed that the optimal cut-off value of T/G ratio in distinguishing high-grade gliomas from low-grade gliomas and inflammations was 1.19 with the AUC 0.84. The sensitivity, specificity, accuracy, PPV and NPV were 70.4%, 85.1%, 78.5%, 79.2% and 78.1%, respectively.

Conclusions

¹³N-ammonia imaging is effective in distinguishing high-grade gliomas from low-grade gliomas and inflammations, but its role in the differential diagnosis of low-grade gliomas and brain inflammatory lesions is limited, and the accuracy needs to be improved.