Jonathan L. Brisman

Eosinophilic granuloma of the clivus: case report, follow-up of two previously reported cases, and review of the literature on cranial base eosinophilic granuloma.

OBJECTIVE AND IMPORTANCE To our knowledge, this is the first reported case of the use of stereotactic radiotherapy for an eosinophilic granuloma (EG) of the clivus. We report follow-up information on two previously reported cases and suggest a management plan for this rare lesion. CLINICAL PRESENTATION We report the case of a 4.5-year-old boy who presented with a complete abducens palsy on the right with an associated head turn. A computed tomographic scan of his head revealed a lytic lesion on that side, and magnetic resonance imaging showed the mass to be of low intensity on T1-weighted images and of high intensity on T2-weighted images with heterogeneous enhancement. INTERVENTION A transnasal stereotactic biopsy was performed, revealing an EG. The patient was treated with stereotactic radiotherapy, and he became symptom-free with radiographic resolution of his lesion. Reviewing the literature, we found 13 series with 87 cases of EG in the petrous portion of the temporal bone. EG in the cranial base occurring outside of the temporal bone or in the temporal bone and extending intracranially is, however, quite rare, with only nine other cases reported, two of them clival. CONCLUSION These findings suggest a classification schema in which cranial base EG lesions be grouped with either the more common extracranial petrous temporal bone lesions or the very rare intracranial lesions. Although there are few cases in the literature, treatment results indicate that clival EG, and perhaps all intracranial cranial base EGs, be treated by a biopsy alone, followed by surgery or stereotactic radiotherapy if there is an incomplete resolution of the symptoms or if there is a recurrence.

Extramedullary Hemangioblastoma of the Conus Medullaris

Summary¶ Background. We report the case of an extramedullary pathologically proven hemangioblastoma of the conus medullaris. As spinal dural arteriovenous fistulas most commonly present with a conus medullaris syndrome, our presentation of the MRI, myelographic, and angiographic findings of this unique lesion may be useful in differentiating these two entities. Clinical Material. We report the case of a 57 year old woman with a two year history of progressive low back and right lower extremity pain and weakness. Spinal MRI and myelography demonstrated serpiginous vasculature on the dorsum of the spinal cord consistent with either a vascular tumor or malformation. Selective spinal angiography was thus undertaken by the neuroendovascular team which revealed a tumor nodule consistent with vascular tumor. T12-L1 laminectomy was performed and a 6 mm vascularized tumor was found in the intradural extramedullary compartment adjacent to the conus medullaris. The tumor was completely removed and pathological analysis was consistent with hemangioblastoma. Conclusion. This report documents a unique location for extramedullary spinal hemangioblastomas. Although both MRI and myelography are helpful in studying these lesions, angiography remains the gold standard in differentiating between vascular tumor and malformation. We suggest that the angiography be performed by a neurointerventional team to facilitate embolization, should this be warranted.

Radiosurgery of the rat hippocampus: magnetic resonance imaging, neurophysiological, histological, and behavioral studies.

OBJECTIVETo explore the histological, electrophysiological, radiological, and behavioral effects of radiosurgery using a new model of proton beam radiosurgery (PBR) of the rodent hippocampus. METHODSForty-one rats underwent PBR of the right hippocampus with nominal doses of 5 to 130 cobalt Gray equivalents (CGE). Three control animals were untreated. Three months after PBR, 41 animals were evaluated with the Morris water maze, 23 with T2-weighted magnetic resonance imaging, and 22 with intrahippocampal microelectrode recordings. Animals that were studied physiologically were killed, and their brains were examined with Nissl staining and immunocytochemical staining for glutamic acid decarboxylase, heat shock protein 72 (HSP-72), parvalbumin, calmodulin, calretinin, calbindin, and somatostatin. RESULTSNinety and 130 CGE resulted in decreased performance in the Morris water maze, increased signal on T2-weighted magnetic resonance imaging, diminished granule cell field potentials, and tissue necrosis, which was restricted to the irradiated side. These doses also resulted in ipsilateral up-regulation of calbindin and HSP-72. Parvalbumin was down-regulated at 130 CGE. The 30 and 60 CGE animals displayed a marked increase in HSP-72 staining on the irradiated side but no demonstrable cell loss. No asymmetries were noted in somatostatin, calretinin, and glutamic acid decarboxylase staining. Normal physiology was found in rats receiving up to 60 CGE. CONCLUSIONThis study expands our understanding of the effects of radiosurgery on the mammalian brain. Three months after PBR, the irradiated rat hippocampus demonstrates necrosis at 90 CGE, but not at 60 CGE, with associated abnormalities in magnetic resonance imaging, physiology, and memory testing. HSP-72 was up-regulated at nonnecrotic doses.

Phosphorylation of P42/P44 MAP kinase and DNA fragmentation in the rat perforant pathway stimulation model of limbic epilepsy

The intracellular signaling pathways associated with neuronal injury after perforant pathway stimulation of the rodent hippocampus have not been examined. To determine whether activation of the p42/p44 (Erk1/2) MAP kinase (MAPK) phosphorylation cascade is linked to neuronal injury after perforant pathway stimulation (PPS), we stained for phosphorylated Erk1/2 (P-Erk1/2) and for DNA fragmentation, a marker of cell death after PPS. Eighteen Sprague-Dawley rats underwent PPS for 6 (n=6), 12 (n=6), or 24 (n=6) h and were sacrificed either immediately (n=9) or 48 h (n=9) after stimulation. Sham-operated non-stimulated control animals (n=2) and control animals receiving low frequency stimulation only (n=4) were also examined. Brain sections were stained for DNA fragmentation and P-Erk1/2. DNA fragmentation was evident only in granule cells and CA3 pyramidal cells of the stimulated side 48 h after 24 h of PPS. PPS resulted in robust phosphorylation of Erk1/2 that displayed a stereotyped timecourse, appearing first in hilar neurons on the ipsilateral side and later in hilar neurons, granule cells, hippocampal pyramidal and non-neuronal cell populations on both the stimulated and contralateral sides. Both Erk1/2 phosphorylation and DNA fragmentation show definite and reproducible staining patterns after PPS that vary based on duration of stimulation. Populations displaying Erk1/2 activation appeared to differ from those showing DNA fragmentation and neuronal injury.

MR spectroscopic changes in the rat hippocampus following proton radiosurgery

Purpose: To identify MR spectroscopic changes in the rat hippocampus following proton radiosurgery. Methods and Materials: A group of 12 rats were treated with Bragg peak proton beam irradiation involving the right hippocampus. Single doses of 30 CGE, 50 CGE, 70 CGE, 90 CGE were delivered to groups of 3 animals using single fraction technique. Animals were imaged using a standard 3 T GE Signa MRI at 4 months following treatment. An untreated animal was also studied. A 3″ surface coil was employed to obtain T1 weighted coronal pre- and post-gadolinium images (TR 600 and TE 30) and dual echo T2 weighted coronal images (TR 3000, TE 30/90). Volumetric analysis with custom software was done to evaluate areas of increased signal on T2 weighted images and the development of hydrocephalus was examined. Animals were sacrificed and specimens of the treated hippocampus were harvested for High Resolution Magic Angle Spinning MR Spectroscopy (HRMAS) followed by histopathology of the tissue samples. Peak values of choline, creatine, N-acetyl aspartate and lipids were evaluated and compared. Results: Peak tissue injury occurred in the surviving 90 CGE animal by both T2 weighted and post-gadolinium imaging. Gadolinium enhancement was seen in decreasing volumes of tissue at dosage levels from 90 to 50 CGE. Hydrocephalus was seen on the untreated side in the 90 CGE animal likely because of mass effect, while it was seen in small degrees in the side of treatment in the 70 and 50 CGE animals. Histopathology showed changes at 90 and 70 CGE, but not at 50 or 30 CGE at this time point using H and E stains. HRMAS showed spectroscopic changes in the surviving 90 and 70 CGE animals but not in the 50 and 30 CGE animals. Statistical significance was not reached because of the small sample size. Conclusions: Following single dose proton radiosurgery of rat hippocampus, HRMAS is able to identify metabolic changes induced by radiation. Studies built on these principles may help develop non-invasive MR spectroscopic methods to distinguish radiation changes from tumor recurrence.

MRI changes in the rat hippocampus following proton radiosurgery.

Purpose: To define radiographic dose-response relationships for proton radiosurgery using a rat brain model. Methods and Materials: A group of 23 rats was treated with Bragg peak proton beam irradiation involving the right hippocampus. Single doses of 5, 12, 20, 30, 60, 90 and 130 cobalt gray equivalents (CGE) were delivered to groups of 3 animals using single fraction technique. One extra animal was included at the 130- and 30-CGE doses. Animals were imaged using a standard 1.5-tesla GE Signa MRI. A 3-inch surface coil was employed to obtain T1-weighted sagittal images (TR 600 and TE 30) and dual echo T2-weighted coronal images (TR 3,000 and TE 30/90). Animals were imaged at 1.5, 3, 4.5, 6 and 9 months. Volumetric analysis with custom software was done to evaluate areas of increased signal on T2-weighted images, and signal change versus time curves were generated. Gadolinium-enhanced T1-weighted imaging was also done at the 9-month time point to further evaluate tissue injury. The development of hydrocephalus was also examined. Results: Peak tissue injury was greater and occurred earlier with higher versus lower doses of radiation. Statistically significant differences were seen between the 130- and 90-CGE animals and between the 90- and 60-CGE animals (p < 0.0016) using ANOVA. Signal changes can be seen in at least 1 of the animals at 20 CGE. The largest volume of tissue enhancement at 9 months was seen in animals at 60 CGE, which may represent an intermediate zone of tissue injury and gliosis compared with greater tissue loss at higher doses and less injury at lower doses. Hydrocephalus developed first in the untreated hemisphere in 130- and 90-CGE animals as a result of mass effect while it occurred at a later time in the treated hemisphere in lower-dose animals. Conclusions: Following single-dose proton radiosurgery of rat hippocampus, serial MRIs show T2 signal changes in animals ranging from 130 down to 20 CGE as well as the development of hydrocephalus. Dose-effect relationships using proton radiosurgery in rats will be a helpful step in guiding further studies on radiation injury to brain tissue.

ANEURYSMAL RUPTURE DURING COILING

OBJECTIVE To study the incidence and clinical outcomes of intraoperative aneurysm rupture (IOR) during endovascular coil embolization at a single large volume center and to review the literature on this subject to determine whether IOR rupture rate and mortality correlate with volume of aneurysms treated at a given center and years since the institution of Guglielmi detachable coils as a treatment modality. METHODS We reviewed the aneurysm database at the Center for Endovascular Surgery since its inception (1997-2003) and reviewed 600 consecutively treated intracranial aneurysms in which coiling was attempted. All patients who sustained an IOR were studied. Procedural and follow-up angiograms as well as clinical outcomes were retrospectively reviewed. A literature review was conducted. RESULTS Six patients (1.0%) experienced IOR (1.4% in acutely ruptured lesions, 0% in unruptured). All six had presented with diffuse subarachnoid hemorrhage (Fisher Grade 3) and in good clinical grade (Hunt & Hess Grades 1-3). One patient was rendered permanently disabled secondary to delay in controlling the IOR. All others were neurologically unchanged. A review of the literature revealed a trend in correlation between volume of aneurysms treated and IOR rate; no statistically significant correlation was found between volume of aneurysms treated or years since the introduction of GDC technology and IOR rates or mortality. CONCLUSION IOR remains a serious risk of endosaccular coiling of intracranial aneurysms, with aneurysms presenting with subarachnoid hemorrhage at greater risk for this complication. This risk can be minimized with very low associated morbidity and mortality (incidence 1%, 17% morbidity, 0% mortality at our institution).