Differentiation of tumor progression and radiation-induced effects after intracranial radiosurgery.

A number of intracranial tumors demonstrate some degree of enlargement after stereotactic radiosurgery (SRS). It necessitates differentiation of their regrowth and various treatment-induced effects. Introduction of low-dose standards for SRS of benign neoplasms significantly decreased the risk of the radiation-induced necrosis after -management of schwannomas and meningiomas. Although in such cases a transient increase of the mass volume within several months after irradiation is rather common, it usually followed by spontaneous shrinkage. Nevertheless, distinguishing tumor recurrence from radiation injury is often required in cases of malignant parenchymal brain neoplasms, such as metastases and gliomas. The diagnosis is frequently complicated by histopathological heterogeneity of the lesion with coexistent viable tumor and treatment-related changes. Several neuroimaging modalities, namely structural magnetic resonance imaging (MRI), diffusion-weighted imaging, diffusion tensor imaging, perfusion computed tomography (CT) and MRI, single-voxel and multivoxel proton magnetic resonance spectroscopy as well as single photon emission CT and positron emission tomography with various radioisotope tracers, may provide valuable diagnostic information. Each of these methods has advantages and limitations that may influence its usefulness and accuracy. Therefore, use of a multimodal radiological approach seems reasonable. Addition of functional and metabolic neuroimaging to regular structural MRI investigations during follow-up after SRS of parenchymal brain neoplasms may permit detailed evaluation of the treatment effects and early prediction of the response. If tissue sampling of irradiated intracranial lesions is required, it is preferably performed with the use of metabolic guidance. In conclusion, differentiation of tumor progression and radiation-induced effects after intracranial SRS is challenging. It should be based on a complex evaluation of the multiple clinical, radiosurgical, and radiological factors.

Spectroscopy-supported frame-based image-guided stereotactic biopsy of parenchymal brain lesions: comparative evaluation of diagnostic yield and diagnostic accuracy.

OBJECTIVE: Comparative evaluation of diagnostic efficacy of stereotactic brain biopsy performed with and without additional use of spectroscopic imaging ((1)H-MRS) for target selection was done. METHODS: From 2002 to 2006, 30 patients with parenchymal brain lesions underwent (1)H-MRS-supported frame-based stereotactic biopsy, whereas in 39 others MRI-guided technique was used. Comparison of diagnostic yield of the procedure in these two groups was performed. Additionally, the diagnostic accuracy was evaluated in 37 lesions, which were surgically resected within 1 month thereafter. RESULTS: Stereotactic biopsy permitted establishment of a definitive histopathological diagnosis in 57 cases and diagnosis of low-grade glioma without specific tumor typing in 8 cases. In 4 cases tissue sampling was non-diagnostic. In 5 out of 8 cases with incomplete diagnosis and in all non-diagnostic cases target selection was performed without the use of (1)H-MRS (P=0.2073). The diagnostic yields of (1)H-MRS-supported and MRI-guided procedures were 100% and 90%, respectively (P=0.1268). Comparison of the histopathological diagnoses after stereotactic biopsy and surgical resection revealed complete diagnostic agreement in 13 cases, minor disagreement in 14 cases, and major disagreement in 10 cases. Among these last 10 cases, initial undergrading of non-enhancing WHO grade III gliomas was the most common (7 cases). The diagnostic accuracy of (1)H-MRS-supported and MRI-guided procedures was 67% and 79%, respectively (P=0.4756). CONCLUSION: While in the present study the diagnostic yield of (1)H-MRS-supported frame-based stereotactic brain biopsy was 100%, its statistically significant diagnostic advantages over MRI-guided technique were not proved. Optimal selection of the spectroscopic target for tissue sampling remains unclear.

Oral administration of choline does not affect metabolic characteristics of gliomas and normal-appearing white matter, as detected with single-voxel (1)H-MRS at 1.5 T.

INTRODUCTION: The present study was done for evaluation of the possible influence of the oral administration of choline on metabolic characteristics of gliomas detected with proton magnetic resonance spectroscopy ((1)H-MRS). MATERIALS AND METHODS: Thirty patients (22 men and eight women; mean age 38 +/- 15 years) with suspicious intracranial gliomas underwent single-voxel long-echo (TR 2,000 ms, TE 136 ms, 128-256 acquisitions) (1)H-MRS of the tumor, peritumoral brain tissue, and distant normal-appearing white matter before and several hours (median, 3 h; range, 1.2-3.7 h) after ingestion of choline with prescribed dose of 50 mg/kg (median actual dose, 52 mg/kg; range, 48-78 mg/kg). Investigations were done using 1.5 T clinical magnetic resonance imager. The volume of the rectangular (1)H-MRS voxel was either 3.4 or 8 cm(3). At the time of both spectroscopic examinations, similar voxels' positioning and size and technical parameters of (1)H-MRS were used. Surgery was done in 27 patients within 1 to 68 days thereafter. In all cases, more than 80% resection of the neoplasm was attained. RESULTS: There were 12 low-grade gliomas and 15 high-grade gliomas. MIB-1 index varied from 0% to 51.7% (median, 13.8%). Statistical analysis did not disclose significant differences of any investigated metabolic parameter of the tumor, peritumoral brain tissue and distant normal-appearing white matter between two spectroscopic examinations. CONCLUSION: Single-voxel (1)H-MRS at 1.5 T could not detect significant changes of the metabolic characteristics of gliomas, peritumoral brain tissue, and distant normal-appearing white matter after oral administration of choline.

Multivoxel proton MRS for differentiation of radiation-induced necrosis and tumor recurrence after gamma knife radiosurgery for brain metastases.

Multivoxel proton magnetic resonance spectroscopy (MRS) was used for differentiation of radiation-induced necrosis and tumor recurrence after gamma knife radiosurgery for intracranial metastases in 33 consecutive cases. All patients presented with enlargement of the treated lesion, increase of perilesional brain edema, and aggravation or appearance of neurological signs and symptoms on average 9.3 +/- 4.9 months after primary treatment. Metabolic imaging defined four types of lesions: pure tumor recurrence (11 cases), partial tumor recurrence (11 cases), radiation-induced tumor necrosis (10 cases), and radiation-induced necrosis of the peritumoral brain (1 case). In 1 patient, radiation-induced tumor necrosis was diagnosed 9 months after radiosurgery; however, partial tumor recurrence was identified 6 months later. With the exception of midline shift, which was found to be more typical for radiation-induced necrosis (P < 0.01), no one clinical, radiologic, or radiosurgical parameter either at the time of primary treatment or at the time of deterioration showed a statistically significant association with the type of the lesion. Proton MRS-based diagnosis was confirmed histologically in all surgically treated patients (7 cases) and corresponded well to the clinical course in others. In conclusion, multivoxel proton MRS is an effective diagnostic modality for identification of radiation-induced necrosis and tumor recurrence that can be used for monitoring of metabolic changes in intracranial neoplasms after radiosurgical treatment. It can be also helpful for differentiation of radiation-induced necrosis of the tumor and that of the peritumoral brain, which may have important clinical and medicolegal implications.

Proton MRS of the peritumoral brain.

Long-echo (TR: 2000 ms, TE: 136 ms) proton MRS of the cerebral tissue in the vicinity to intracranial lesion was done in 15 patients, mainly with parenchymal brain tumors. Significant decrease of N-acetylaspartate (NAA) (P<0.001) and more frequent presence of lactate (P<0.01) comparing with distant normal white matter were found in the perilesional brain tissue. The level of NAA in the perilesional brain tissue had negative associations with presence of lactate in the lesion (P<0.05), excess of lactate in the lesion compared to perilesional brain (P<0.01), grade of the perilesional edema (P<0.01) and patient's age (P<0.05). Multivariate analysis disclosed that identification of lactate in the lesion is associated with lower relative NAA content in the perilesional brain tissue, independently on the presence or absence of any other factor, including brain edema (P<0.001). In patients with lobar lesions who had at least one epileptic seizure during course of their disease the relative NAA content in the perilesional brain was significantly lower, comparing with those who were seizure-free (P<0.05). Therefore, lactate diffused from the tumor, or other metabolites secreted by lactate-producing neoplasm, should be considered as important contributors to the neuronal dysfunction in the surrounding brain. Decrease of NAA in the vicinity to intracranial lesions may reflect neuronal alteration responsible for associated epilepsy.

Gamma knife surgery for choroidal neovascularization in age-related macular degeneration. Technical note.

The authors conducted a study to determine a way of overcoming the poor-quality demonstration of choroidal neovascularization (CNV) associated with age-related macular degeneration (AMD) on conventional magnetic resonance (MR) imaging studies. The poor MR imaging demonstration of CNV in patients with AMD makes the use of gamma knife surgery more difficult. This difficulty, however, can be overcome by use of a modified time-of-flight MR imaging,sequence with Gd enhancement and coronal reconstruction.

Early metabolic changes in metastatic brain tumors after Gamma Knife radiosurgery: 1H-MRS study.

Evaluation of early metabolic changes in metastatic brain tumors after Gamma Knife radiosurgery was performed by long-echo (TR, 2000ms; TE, 136ms; 128-236 acquisitions) volume-selected single-voxel proton magnetic resonance spectroscopy (MRS). Eighty-five brain metastases in 81 patients were investigated before treatment and 16-18h thereafter. Standard metabolic ratios, namely N-acetylaspartate (NAA)/creatine (Cr), phosphorylcholine/glycerophosphorylcholine (Cho)/Cr, NAA/Cho, lactate (Lac)/Cr, and mobile lipids (Lip)/Cr, were calculated, and comparison of their values before and after irradiation was done. No volumetric changes of any neoplasm were found in any case on the next day after treatment. At the same time, significant reduction of Cho/Cr (P < 0.001) and NAA/Cr (P < 0.01) ratios on the proton MRS of the tumor was disclosed. Reduction of Cho/Cr ratio was significantly more prominent in neoplasms with higher pretreatment Cho/Cr ratios (P < 0.001) and heterogeneous contrast enhancement (P < 0.01). Reduction of NAA/Cr ratio was predominantly determined by its pretreatment value (P < 0.001). The observed decrease of Cho/Cr ratio probably reflects inhibition of proliferative activity and early apoptotic cell loss, whereas reduction of NAA/Cr may result from radiation-induced modulation of neuronal activity in the peritumoral brain tissue. Serial proton MRS represents a valuable diagnostic tool for evaluation of metabolic changes in intracranial neoplasms after radiosurgical treatment.