Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging

Abstract

Background and purpose: Sensitivity, positive predictive value (PPV), and negative predictive value (NPV) of conventional MR imaging in predicting glioma grade are not high. Relative cerebral blood volume (rCBV) measurements derived from perfusion MR imaging and metabolite ratios from proton MR spectroscopy are useful in predicting glioma grade. We evaluated the sensitivity, specificity, PPV, and NPV of perfusion MR imaging and MR spectroscopy compared with conventional MR imaging in grading primary gliomas.

Methods: One hundred sixty patients with a primary cerebral glioma underwent conventional MR imaging, dynamic contrast-enhanced T2*-weighted perfusion MR imaging, and proton MR spectroscopy. Gliomas were graded as low or high based on conventional MR imaging findings. The rCBV measurements were obtained from regions of maximum perfusion. Metabolite ratios (choline [Cho]/creatine [Cr], Cho/N-acetylaspartate [NAA], and NAA/Cr) were measured at a TE of 144 ms. Tumor grade determined with the three methods was then compared with that from histopathologic grading. Logistic regression and receiver operating characteristic analyses were performed to determine optimum thresholds for tumor grading. Sensitivity, specificity, PPV, and NPV for identifying high-grade gliomas were also calculated.

Results: Sensitivity, specificity, PPV, and NPV for determining a high-grade glioma with conventional MR imaging were 72.5%, 65.0%, 86.1%, and 44.1%, respectively. Statistical analysis demonstrated a threshold value of 1.75 for rCBV to provide sensitivity, specificity, PPV, and NPV of 95.0%, 57.5%, 87.0%, and 79.3%, respectively. Threshold values of 1.08 and 1.56 for Cho/Cr and 0.75 and 1.60 for Cho/NAA provided the minimum C2 and C1 errors, respectively, for determining a high-grade glioma. The combination of rCBV, Cho/Cr, and Cho/NAA resulted in sensitivity, specificity, PPV, and NPV of 93.3%, 60.0%, 87.5%, and 75.0%, respectively. Significant differences were noted in the rCBV and Cho/Cr, Cho/NAA, and NAA/Cr ratios between low- and high-grade gliomas (P <.0001,.0121,.001, and.0038, respectively).

Conclusion: The rCBV measurements and metabolite ratios both individually and in combination can increase the sensitivity and PPV when compared with conventional MR imaging alone in determining glioma grade. The rCBV measurements had the most superior diagnostic performance (either with or without metabolite ratios) in predicting glioma grade. Threshold values can be used in a clinical setting to evaluate tumors preoperatively for histologic grade and provide a means for guiding treatment and predicting postoperative patient outcome.

Fig 1.

20-year-old woman with biopsy-proved high-grade glioma. A, Contrast-enhanced axial T1-weighted image (600/14/1 [TR/TE/NEX]) demonstrates an ill-defined nonenhancing mass (arrow) in the right frontal region. The lack of enhancement on the conventional MR image suggests a low-grade glioma. B, Axial T2-weighted image (3400/119/1) shows increased signal intensity in the mass, with minimal peritumoral edema. This mass was graded as a low-grade glioma with conventional MR imaging because of lack of enhancement, minimal edema, no necrosis, and no mass effect. C, Gradient-echo (1000/54) axial perfusion MR image with rCBV color overlay map shows increased perfusion with a high rCBV of 7.72, in keeping with a high-grade glioma. D, Spectrum from proton MR spectroscopy with the PRESS sequence (1500/144) demonstrates markedly elevated Cho and decreased NAA with a Cho/NAA ratio of 2.60, as well as increased lactate (Lac), in keeping with a high-grade glioma.

Fig 2.

43-year-old man with biopsy-proved low-grade glioma. A, Contrast-enhanced axial T1-weighted image (600/14/1) demonstrates a peripherally enhancing mass (arrow) in the right frontal region. The presence of contrast material enhancement on the conventional MR image would suggest a high-grade glioma. B, Axial T2-weighted image (3400/119/1) shows marked peritumoral edema with possible necrosis and blood products. This mass was graded as a high-grade glioma with conventional MR imaging because of the contrast material enhancement, heterogeneity, blood products, possible necrosis, and degree of edema. C, Gradient-echo (1000/54) axial perfusion MR image with rCBV color overlay map shows a low rCBV of 1.70, in keeping with a low-grade glioma. D, Spectrum from proton MR spectroscopy with the PRESS sequence (1500/144) demonstrates elevated Cho and slightly decreased NAA with a Cho/NAA ratio of 0.90, which is more in keeping with a low-grade glioma.

Fig 3.

ROC curves for rCBV plus metabolites, rCBV alone, Cho/Cr, and Cho/NAA demonstrate superior sensitivity and specificity of rCBV plus metabolites and rCBV alone compared with conventional MR imaging (cMRI, green asterisk) for glioma grading.