Using MRS and DWI to differentiate between glioblastoma recurrence and pseudoprogression
Glioblastoma, the most serious and frequent primary brain tumor in adults, progresses in most patients after initial treatment. Rapid diagnosis of relapse and prompt initiation of salvage treatment may extend a patient’s life. However, diagnosis can be challenging as glioblastoma recurrence can exhibit similar radiologic characteristics on advanced imaging as therapy-related changes like pseudoprogression.
MR spectroscopy (MRS) is a complementary method to increase the diagnostic accuracy of the brain tumor imaging protocol of structural MRI. A multi-specialty team of physicians from three institutions in Bruno, Czech Republic, developed standardized imaging protocols and established thresholds for the main metabolite concentrations obtained by MRS and the values of the apparent diffusion coefficient (ADC). In an article published in Neuroimage: Clinical, lead author Radim Jancalek, MD, of the Department of Neurosurgery at St. Anne’s University Hospital, and colleagues explain how establishment of thresholds and cutoffs can help radiologists better distinguish tumor recurrence from pseudoprogression.
The authors analyzed data from the metabolites measured by MRS and diffusion-weighted imaging (DWI) in 39 patients suspected of having disease progression. Tumor recurrence was diagnosed in 75% of the patients, with the remaining 25% having pseudoprogression.
MRS provides a measure of brain chemistry. When elevated, choline (Cho), the marker of cell membrane integrity and turnover, is associated with the presence of an increased tumor cell proliferation. N-acetylaspartate (NAA) is the marker of density and visibility of neurons. It is a neuronal marker, and decreases with any disease that affects neuronal integrity. The authors use a mutual ratio.
They determined that a significant difference in the total Cho/total NAA and the tNAA/tCr ratios was found between patients who had a tumor relapse and those with pseudoprogression. Specifically, the glioblastoma relapse was characterized by the tCho/tNAA ratio greater than 1.3 with a sensitivity of 100% and a specificity of 94.7%.
DWI describes changes in water diffusivity, and diffusion changes can be quantified by the ADC. Enhanced tumor cell proliferation is a cause for decreased diffusivity, which is reflected by the water diffusion restriction that lowers ADC values.
Not surprisingly, the calculated ADC mean values were significantly lower in patients who relapsed. A mean value higher than 1313 x 10-6 mm 2/s was associated with pseudoprogression, with a sensitivity of 98.3% and a specificity of 100%. When these thresholds were applied to a validation cohort of 16 patients, 75% of whom had been diagnosed with disease progression, the tCho/tNAA ratio and the ADC mean value correctly diagnosed 15 patients of the 16 patients.
The authors believe that the combination of ADC values and metabolite concentrates measured by MRS can enable radiologists to more accurately distinguish and diagnose glioblastoma progression from pseudoprogression. Because the experiences of individual radiologists can influence spectrum quality, the authors recommend institutional-specific establishment of protocols. They also encourage other researchers to perform similar studies with larger patient cohorts to further validate their findings.
REFERENCE
- Kazda T, Burlik M, Pospisil P, et al. Advanced MRI increases the diagnostic accuracy of recurrent glioblastoma: Single institution thresholds and validation of MR spectroscopy and diffusion weighted MR imaging. Neuroimage: Clin 2016 11: 316-321.