Abstract: Radiation-Induced Tissue Damage
Radiation-induced tissue damage in the central nervous system is a wellknown complication of interstitial brachytherapy for brain tumors, yet imaging correlates have historically been based upon computed tomographic scans. We now present magnetic resonance imaging (MRI) to correlate radiation changes after interstitial brachytherapy with the histopathology.
August 1992, Volume 31, Number 2
336 Radionecrosis Secondary to Interstitial
Brachytherapy: Correlation of Magnetic Resonance
Imaging and Histopathology
AUTHOR(S): Oppenheimer, Jeffrey H., M.D.;
Levy, Michael L., M.D.; Sinha, Uttam, M.D.; El-
Kadi, Hikmat, M.D.; Apuzzo, Michael L. J., M.D.;
Luxton, Gary, Ph.D.; Petrovich, Zbigniew, M.D.;
Zee, Chi-Shing, M.D.; Miller, Carol A., M.D.
Departments of Neurological Surgery (JHO, MLL,
HEK, MLJA), Pathology (US, CAM), Radiation
Oncology (GL, ZP), and Radiology, Division of
Neuroradiology (CSZ), University of Southern
California Medical Center and the LAC-USC Medical
Center, Los Angeles, California
Neurosurgery 31; 336-343, 1992
ABSTRACT: RADIATION-INDUCED TISSUE DAMAGE
in the central nervous system is a wellknown complication of interstitial brachytherapy for brain tumors, yet imaging correlates have historically been based upon computed tomographic scans. We now present magnetic resonance imaging (MRI) to correlate radiation changes after interstitial brachytherapy with the histopathology. The central nervous system of a 38-year-old man with a left frontal cerebral glioma diagnosed by stereotactic biopsy was treated with interstitial brachytherapy (iridium-192, 47 Gy), followed by limited-field irradiation (45 Gy). With progressive deterioration, a second biopsy demonstrated radiation changes. Despite aggressive medical management, the patient died 9 months after completion of radiation therapy.
Postmortem evaluation compared MRI scans of the intact, fixed brain with the histopathology. Axial sections (10 mm) obtained by MRI scan and autopsy brain slices were cut in the identical plane. Neuroimaging and histopathological findings of the lesion correlated within 2 to 3 mm resolution. In the peripheral white matter, MRI scan did not indicate the extent of radiation effect histologically. We suggest that MRI has limited utility in assessing the extent of impact of radiation on surrounding brain.
Radiation-induced tissue damage in the central nervous system is a well-known complication of interstitial brachytherapy for brain tumors. Histological changes induced by tissue analysis were first described by Fischer and Holfelder in 1930 (15). More recent reports have demonstrated effects on the vasculature of the neural parenchyma adjacent to a tumor, as well as edema, necrosis, and reactive gliosis (4,22,28). Imaging correlates of radiation changes have been more difficult to obtain. Most of the earlier data was based on computed tomographic (CT) scans on patients suspected of having radiation necrosis (4-6,8,9, 32). With magnetic resonance imaging (MRI), better comprehension of the extent of radiation-induced tissue damage is evolving (7,9,39). The purpose of this report is to correlate radiation changes on MR imaging with the histopathology of an intrinsic glial neoplasm treated with curative intent with a combination of teletherapy and brachytherapy.
MATERIALS AND METHODS
A 38-year-old man initially experienced a generalized seizure and, 4 years later, a second event involving focal seizures of the right arm. The patient was subsequently hospitalized. Upon admission, the neurological examination was normal. CT scan of the head demonstrated a 3-cm, low-density lesion with minimal contrast enhancement in the left frontal centrum semiovale and subcortical white matter (Fig. 1, A and B). The patient was placed on glucocorticoids and anticonvulsants. CT scan-guided stereotactic biopsy revealed anaplastic astrocytoma.
Figure 1. Preoperative CT scan demonstrating lowdensity lesion in the left frontal centrum semiovale and subcortical white matter (open arrows). There is minimal contrast enhancement. A, without contrast; B, with contrast.
The treatment plan consisted of interstitial brachytherapy to be followed by external-beam irradiation with a field to extend beyond the tumor by a 2-cm margin. Surgical resection was excluded because of possible unacceptable morbidity based on the proximity of the lesion to the motor cortex. Interstitial brachytherapy was carried out 1 week after the biopsy. Two Silastic catheters were each afterloaded with seven 3-mm length iridium-192
seeds, spaced 4 mm along their length center-tocenter, for an overall active length of 2.7 cm. The two catheters were stereotactically implanted into the anterior and posterior aspects of the lesion 2.25 cm apart (Fig. 2). The choice of separation distance was based on the goal of encompassing the dimensions of the tumor, while minimizing the possibility of infection by limiting the implant to two catheters. To minimize the possibility of necrosis caused by high individual seed activity, seed strength was limited by treatment protocol to a maximum of 1 mg radium equivalent. Pretreatment calculations showed that with 1 mg radium equivalent seeds, a target dose rate of 50 cGy/hr would extend at least 0.5 cm beyond the ends of the perpendicular line joining the two catheters over their full active length. The anteroposterior lesion length of 3 cm thus required a separation of approximately 2 cm. The actual 2.25-cm distance between catheters reflected a suspicion that the anteroposterior dimension of the lesion might in fact be larger.
Figure 2. CT scan performed after stereotactic placement of two silastic catheters 2.25 cm apart in the lesion. They were afterloaded with multiple iridium-192 sources. Isodose curves for brachytherapy treatment are superimposed. The 40 cGy/hr line was used, and the patient was treated for a total of 47.1 hours for a total dose of 47.1 Gy. Increments of 1 cm are delineated on the x and y axes.
Seed activity at the time of the implant was 1.70 mCi, corresponding to 0.97 mg radium equivalent. The bulk of the tumor was encompassed by the 40 cGy/hr dose rate surface, and the 40 cGy/hr dose rate was chosen to designate the treatment. A total dose of 47 Gy was administered to a volume of 21 cm3 during a 4.9-day implant. As a measure of the nonhomogeneity of the brachytherapy treatment, we note that a total dose of 70 Gy (150% of treatment dose) was administered to a volume of 11 cm3 (52% of treatment volume). The 47-Gy treatment volume was quasi-rectangular in shape, of approximate maximum dimensions 2.2 cm wide ´ 4.2 cm anteroposteriorly ´ 3.3 cm vertically. The maximum dose rate to the inner table of the calvarium was less than 20 cGy/hr for a total dose of less than 23.5 Gy over an area of approximately 2 ´ 2.5 cm.
External beam radiotherapy was begun on the 16th day after removal of the implant using an x-ray beam from a 4 MV linear accelerator. A total of 45 Gy was delivered through two beam portals: a rotated left lateral and a left superior-anterior oblique 60-degree wedged pair of custom-blocked fields with outer dimensions of 7.5 ´ 7.5 cm. The beam shaping allowed for a 2-cm margin beyond the region of enhancement on the axial CT scan that demonstrated
the lesion in its largest diameter. The external-beam irradiation was accomplished in 1.8-Gy daily fractions, 5 days per week over an interval of 35 days.
The patient experienced a focal seizure during initial catheter implantation and was treated with Dilantin and phenobarbital. There were no other acute
toxicities during this phase of therapy. During the subsequent 5 months after completion of therapy, the patient functioned well at home without deficits, except for some mild hemiparesis. The patient then experienced progressive right hemisensory symptoms, weakness, and a pronounced expressive
aphasia. CT scan at this time demonstrated a ringenhancement of the lesion now 3.4 cm in size and an 0.8-cm shift of the midline structures (Fig. 3A).
Repeat biopsy of the lesion demonstrated radiationinduced changes but no evidence of tumor. The patient was treated with high-dose glucocorticoids with some improvement, but his mental status deteriorated over the following 2 months. Interval CT examination demonstrated an enlargement of the diameter of the ring enhancement to 4.5 cm and an increase in the shift of the midline structures to 1.2 cm (Fig. 3B). The patient refused surgical decompression and vigorous medical management was undertaken using glucocorticoids, mannitol, furosemide, and hyperventilation. The patient had a limited response to this treatment and died 9 months after completion of the radiation therapy.
Figure 3. Interval contrast-enhanced CT scans performed after completion of interstitial brachytherapy and external-beam radiation treatment demonstrating progressive radiation changes. A, 7 months; B, 9 months.
The brain was removed at autopsy and fixed in 10% phosphate-buffered formalin. Detailed imaging was obtained on the formalin-fixed brain in three planes using a Vista American Health 0.5 Tesla MRI scanner. T1 (TR, 2000 ms; TE, 20 ms) and T2 (TR, 2000 ms; TE, 100 ms) weighted images were taken.
Serial horizontal brain slices were cut 10 mm in thickness and correlated with the axial MRI images of the same size. These were arbitrarily numbered 1 to 8 from the vertex to the base of the brain. Levels 1 through 3 included the vertex. Level 4 was approximately 15 mm superior to the surface of the
corpus callosum; level 5 cut through the bodies of the lateral ventricles, approximately 5 cm superior to the corpus callosum; level 6 cut horizontally through the internal capsule, splenium, and genu of the corpus callosum; and level 7 included the basal ganglia, thalamus, and insular cortex and was inferior to the body of the corpus callosum. Level 8 included the basilar portion of the brain, Circle of Willis, frontoorbital cortex, inferior temporal lobe and their pial
surfaces (Table 1).
Three levels, 4, 5, and 6, that encompassed the lesion were correlated with their matching axial MRI scan cuts. The images of the tumor margin were then outlined using camera lucida tracings (Fig. 4, A and B). Camera lucida tracings were made by placing a clear plastic sheet over the MRI or CT scan of interest. The images of the tumor margin were then outlined using a permanent alcohol/waterproof fine tip marker. Superimposed on the entire affected hemisphere, and including each outlined section, a graph composed of 10-mm grids designated the tissue blocks to be cut. Blocks were selected for histological examination that formed an intersect over the lesion and samples extended to medial, lateral, anterior, and posterior pial surfaces. Each block was marked on all six surfaces with color dyes (Davidson Dyes, San Diego, CA) to insure precise and consistent anatomic orientation. After paraffin embedding, 8-m sections were cut from the superior surface of each block and stained with hematoxylin and eosin. Additional blocks were selected randomly from both cerebral hemispheres, mesencephalon, cerebellum, pons, medulla, and cervical, thoracic, and lumbar spinal cord. A total of over 80 sections were microscopically examined.
Figure 4. Reproductions of camera lucida tracings from an axial MRI scan taken through level 5 at the center of the lesion. Tissue blocks are on 1 ´ 1 cm grids. Hematoxylin and eosin stained sections demonstrated necrosis of the lesion (A) and radiation necrosis (B) as delineated by the darkened borders. The 1 ´ 1 cm sections are designated initially by level 5, then by position: anterior (A), posterior (P), lateral (L), and medial (M), and finally by distance from the center of the lesion. Note that necrosis was evident from 5A7 to 5P4 and from 5M2 to 5L3. Note that the lesion was evident from 5A2 to 5P1 and from 5M2 to 5L2.
MRI scan of the intact brain in formalin revealed a focal area of increased signal surrounded by a circumscribed rim of relatively less signal on T1- and T2-weighted images (Fig. 5, A and B). Brain sections correlated with MRI scan axial slices within approximately 3-mm resolution.
Figure 5. MRI scan of the postmortem brain after fixation. Axial section taken through level 5 at the center of the lesion. A, T1-weighted image (TR, 200 ms; TE, 20 ms). B, T2-weighted image (TR, 200 ms; TE, 100 ms).
Postmortem examination of the dura revealed a circumscribed defect 2 to 3 mm in diameter in the left frontoparietal region, surrounded by a hemorrhagic
ring 10 mm in diameter. This lesion corresponded approximately to the site of stereotactic biopsies and brachytherapy implantation. The leptomeningeal
vessels over the cerebral convexities were moderately distended. The left cerebral hemisphere was swollen and tense, especially adjacent to a circumscribed, hemorrhagic, softened area, 10 mm in diameter, underlying the dural defect. The left cingulate gyrus was softened because of subfalcial herniation.
Gross examination of the horizontally oriented brain sections revealed a circumscribed, cavitary lesion measuring 4 cm in diameter in the left frontoparietal white matter. The diffusely marginated lesion was variegated with several small focal hemorrhages and a central, necrotic, gray area
surrounded by a brownish rim of tissue. The surrounding tissues were softened and edematous. The lesion extended in the dorsoventrad orientation
from the level of the superior surface of the corpus callosum to the midthalamic region. Superiorly, it extended 15 mm above the corpus callosum at the
level of the bodies of the lateral ventricles and posteriorly through the internal capsule (levels 4, 5, Redistribution of this article permitted only in accordance with the publisher’s copyright provisions. and 6). The corpus callosum was normal in width and color, but slightly softened in onsistency. The contralateral cerebral hemisphere, cerebellum, and brain stem were normal. No tracts from biopsy cannulas or brachytherapy catheters were identified.
Microscopic examination revealed necrotic tumor cells in the cerebral white matter superior to the level of the corpus callosum and extending through the internal capsule (levels 5 and 6). The lesion consisted of infiltrating nests of moderate sized, pleomorphic “ghost-like” cells containing dense, pyknotic nuclei (Fig. 6A). No section contained tumor cells with nuclei of distinguishable internal architecture, such as distinct nucleoli and clustered chromatin, as seen in the original (before radiation) surgical specimen. These changes, located in the site of the original tumor focus, suggest that the lesion is not viable. At all levels, the tumor cells were confined to the white matter, and there was no extension to the cortical
ribbon, corpus callosum, contralateral hemisphere, or brain stem.
Radiation effects observed as far as 5 cm peripheral to the margin of the mass included fibrinoid necrosis, acute necrotizing vasculitis of arterioles, and focal hemorrhage (Fig. 6B). Coagulation necrosis and edema of the adjacent neuropil extended to the cortical surface anteriorly and laterally and 25 mm posteriorly from the center of the lesion. These changes extended to the level of the bodies of the lateral ventricles 5 mm above the corpus callosum (level 5).
Histopathological changes also included radiation vasculitis, with blood vessels showing hyaline thickening and mononuclear cell infiltrates. Telangiectatic vessels, particularly in the subcortical and cortical regions, were also present (Fig. 6B). Gemistocytic astrocytes were scattered throughout the affected white matter, particularly associated with the acute vasculitis. Surprisingly, there was a paucity of macrophages throughout the brain, including areas adjacent to necrotic tissue. A few bizarre cells with large pleomorphic nuclei were scattered around some necrotic vessels, suggesting radiation effects within the endothelia. These findings were more abundant adjacent to the margin of the mass. A dense fibrillary meshwork of glial processes was seen in all of the lateral sections including those immediately adjacent to the mass. Gliosis was also present in the most cephalad brain slice (level 1).
Figure 6. Photomicrographs of representative sections (´1600). A, cellular detail is blurred because of widespread necrosis of the tumor cells. Note dark, homogeneous nuclei with indistinct chromatin. B, radiation vasculitis. This transversely sectioned arteriole is surrounded by a mononuclear infiltrate. The vessel wall shows focal hyalinization. C, radiation effect; vascular telangiectasia. This longitudinally sectioned vessel has a wide lumen and is lined by a single layer of endothelium.
Correlation of imaging with histopathology
The tumor mass noted histologically corresponded to the central hyperintense lesion surrounded by a ring of lower intensity signal on both T1- and T2-
weighted images. The image correlated within a 0.5- cm resolution with the tissue cuts. The image of the lesion itself had a maximum diameter of 4 cm at a level of the bodies of the lateral ventricles and 5 mm above the corpus callosum (level 5). At all levels, the necrotic mass was confined to the white matter of the centrum semiovale with no extension to the cortices, corpus callosum, contralateral hemisphere, or brain stem. In the circumferential margin of the lesion represented by a relatively lower intensity, there was diffuse vasculitis consisting of neutrophils, a minority of mononuclear cells, and rare macrophage infiltrates (Fig. 7).
Figure 7. Axial gross necropsy section correlating to MRI scan plane, level 5. Focal area of irradiated mass is shown in the left frontoparietal centrum semiovale and subcortical regions. Teletherapy isodose curves are superimposed to scale.
Combined effects of radiation vasculitis and necrosis were represented on MRI scan by a heterogenous high signal of T1- and T2-weighted images at the section taken through the center of the lesion (slice 5). When comparing tissue block plots with axial MRI scans, the MRI scan apparently underestimated the histopathological findings by as much as 46% in the most cephalad sections (Table 2). Histologically, radiation changes became less severe on sections taken from tissue blocks further away from the center of the mass. These peripheral changes did not appear on MRI scans in vivo, and, for this reason, the extent of radiation injury was substantially underestimated.
Relationship of histopathological findings to planned radiation treatment fields
As expected, sections of the contralateral cerebral hemisphere, brain stem, and cerebellum were normal, as these structures extended well outside of the planned treatment fields. Radiation changes were found as far as 3 cm outside of the 45-Gy isodose line for the external-beam treatment field.
Microscopic analysis of the brain section that contained the mass in its widest diameter revealed radiation necrosis and vasculitis extending beyond
both the interstitial brachytherapy and external-beam treatment fields. No radiation histopathology was demonstrated beyond the midline boundary of the
dosimetry profiles for either radiotherapeutic modality. Changes were seen 5 cm beyond the 40- cGy/hr isodose line for the interstitial brachytherapy
in the anterior direction. Posteriorly, medially, and laterally, these changes extended 2, 1.25, and 2.75 cm beyond the planned isodose curve, respectively. Both vasculitis and necrosis extended 3.5 and 0.5 cm, respectively, beyond the anterior and posterior margins of the planned teletherapy field. In the medial and lateral dimensions, these changes were each within 0.5 cm of the field boundary in the MRI scan sections containing the lesion with maximal extent.
MRI evaluation on an intact, fixed brain with subsequent neuropathological analysis provides a correlation of three modalities: dosimetry, MRI signal intensity, and histopathology. We have presented a patient with anaplastic astrocytoma treated with interstitial brachytherapy and beam radiation therapy. Serial imaging in this patient demonstrated that the radiation changes may have contributed to a progression in mass effect, shift of
the midline structures, and herniation. Clinical signs and symptoms of mass effect secondary to biopsyproven radiation changes were evident 7 months after cessation of therapy and led to his death 2 months later. The patient failed to respond to glucocorticoids, although other reports favorably advocate this therapy (30,42). Anticoagulation had also been reported as the treatment of choice for radiation toxicity (34).
Several reports suggest that surgical cytoreduction of these lesions also has a role in their management (1,14,16,18,21,29,33,35).
Evaluation of the combined effects of interstitial brachytherapy and external-beam therapy indicated that MRI scan underestimated the extent of the
radiation effect. The brain surrounding the tumor had confirmed radiation histopathology and, on MRI scan, a less intense signal than the tumor, but
hyperintense relative to normal white matter.
The radionecrosis extended to the cortical ribbon, and MRI scan significantly underestimated the extent of these changes, confirming a previous report (23). Histologically, radiation changes included an area 46% greater than the regions on MRI scan represented by relative signal increase on T1- and T2- weighted images. The most peripheral portions of white matter, near the cerebral vertices and extending to the gray-white interfaces, had neuropathological markers of radiation response including hyaline neurons of vessels and edema. These changes represented on MRI scan by a signal of lesser intensity were progressively sparser than those seen near the mass and, for this reason, may have been more difficult to image. Other histological hallmarks of radiation necrosis found within the white matter included vascular proliferation, coagulation and fibrinoid necrosis, and telangiectatic proliferation of blood vessels. No radiation effect was found in the contralateral hemisphere, brain stem or cerebellum.
Differences between the MRI scan of the living patient and the postmortem, formalin-fixed brain have also been reported (2,3,24). The effect of formalin in altering brain water content may have contributed to decreased MRI scan sensitivity in demonstrating these peripheral radiation changes (40).
Late radiation injury has been extensively investigated (6,10,19,21,25,26,28,29,36,37). Radiation necrosis of the brain has been reported in a wide dose range from 20 to 69 Gy. Clinical manifestations have occurred from 9 to 28 months from the time of completion of radiotherapy (31). Symptoms of
radionecrosis in this report were seen 7 months after the cessation of both teletherapy (45 Gy) and interstitial brachytherapy (47 Gy) treatments. The
rapidity of onset of symptoms in our patient might be explained by the combined effects on white matter of teletherapy and brachytherapy.
In treating the patient with interstitial brachytherapy, we realized that there was an obvious trade-off between the considerable nonhomogeneity
that would result from limiting the implant to two catheters and the potential to minimize the possibility of infection. Such nonhomogeneity results from any limitation of the maximal number of seeds, whether clinically or otherwise restricted, and represents a generic problem with interstitial implants. Given any potential limitations to the treatment with the maximal number of seeds, certain principles are followed. The choice of separation distance will be based on the goal of encompassing the dimensions of the tumor, while attempting to minimize the possibility of necrosis caused by high individual seed activity.
Tumor was represented by the greatest increase in signal intensity on MRI scan and correlated closely with the dimensions of the tumor mass, with
localization within 0.5 cm resolution. Tumor cells were seen only in the central region and histologically were necrotic in all samples, suggesting successful eradication of the tumor. Our T1 findings contrast with others that have shown MRI scan to be inadequate in differentiating radiation
changes from tumor recurrence, as both appear as an increase signal on T1 and T2 sequences. Dooms et al. (12) used normal white matter as a reference and found a general increase in signal on T2-weighted images in periventricular regions in patients with histologically confirmed radiation effect. Focal tumor recurrences and regions of radiation effect were both demonstrated on MRI scan as a low signal on T1- and a high signal on T2-weighted images. However, many of these patients were stereotactically assayed, lacking autopsy confirmation (12). Grossman et al. (17) used an experimental model of radiation injury and their findings correlated well with those presented in our report. They found that the gadolinium diethylene-triamine-pentaacetic acidenhanced proton image was the most sensitive for the detection of radiation injury. As our study was conducted on postmortem tissues, we were unable to investigate the role of gadolinium diethylenetriamine-pentaacetic acid.
Most reports on MRI scan of radiation necrosis of the brain have demonstrated T1 and T2 relaxation times that are relatively higher than that of normal
white matter (7,20). The contrast-enhanced CT scan is a poor alternative to MRI scan as it is inadequate in differentiating tumor recurrence from post-radiation changes or necrosis, each of which appear as contrastenhancing entities (4,8,27). Positron emission tomography has been found to be the most specific in differentiating radiation changes from tumor recurrence (11,13,41).
The correlation between tissue radiation dosimetry and pathological findings is understood less than the relationship between the MRI scan and pathological findings. A dose of 59 Gy delivered to a tumor volume of 21 cm3 caused extensive radiation changes well outside of the planned field of radiation. Histological changes most distal from the lesion in the frontal subcortical region were clearly outside the radiation portals and the field created by the iridium- 192 implants. The iridium-192 isodose line at which the patient received a total dose of 23.5 Gy is 5 cm away from the most distal histopathological radiation changes seen in the anterior frontal region. This area received less than a 2.2-Gy dose of teletherapy radiation. The presence of these histological changes indicative of post-radiation injury may be a result of secondary effects of radiation vasculitis on end
vessels. This pathophysiological mechanism might resemble that of lacunar infarction (22,28). Other investigators have suggested autoimmune mechanisms causing demyelination. Radiation vasculopathy has been studied in patients that have had radiosurgery for arteriovenous malformations. One suggested mechanism for the ensuing thrombosis Redistribution of this article permitted only in accordance with the publisher’s copyright provisions. is sclerosis of the large vessels from occlusion of the vasa vasorum (38). The interplay between interstitial brachytherapy and external-beam therapy in the brain warrants further clinical, radiological, and histopathological study.
This report provides clinicopathological analysis of the effects of combined modalities of external-beam radiation and interstitial brachytherapy on focal anaplastic astrocytoma and the surrounding brain. MRI scan of the fixed brain correlated closely with the subsequent detailed neuropathological
examination. Toxic radiation effects yielded increased signal intensity on T1- and T2-weighted images. Histopathologically, there was widespread
necrosis of the tumor, suggesting that the lesion had been eradicated by the combined modality radiotherapy. Combined radiation effects extended
beyond the original tumor mass well beyond planned external beam and interstitial brachytherapy treatment fields. The MRI scan provides a useful tool
in providing anatomical definition of tumor mass but may have limited utility in assessing the impact of radiation on the surrounding brain.
The authors are grateful for the expert clerical
assistance of Phillip Mora, Jeanette Espinosa, and
Karen Levy, R.N.
This work was supported in part by National
Institute of Aging Grant 5-P50-AG05142 (C.A.M.)
and Training Grant AG0-0093 (U.S.).
Received, April 16, 1991.
Accepted, February 13, 1992.
Reprint requests: Carol A. Miller, M.D.,
Department of Pathology, USC School of Medicine,
McKibben Annex, Room 345, 2011 Zonal Avenue,
Los Angeles, CA 90033.
1. Bernstein M, Laperriere N, Leung P, McKenzie S: Interstitial brachytherapy for malignant brain tumors: Preliminary results. Neurosurgery 26:371-380, 1990.
2. Braffman BH, Zimmerman RA, Trojanowski JQ, Gonatas NK, Hickey WF, Schlaepfer WW: Brain MR: Pathologic correlation with gross and histopathology. 1. Lacunar infarction and Virchow-Robin spaces. AJNR 9:621-628, 1988.
3. Braffman BH, Zimmerman RA, Trojanowski JQ, Gonatas NK, Hickey WF, Schlaepfer WW: Brain MR: Pathologic correlation with gross and histopathology. 2. Hyperintense white-matter foci in the elderly. AJNR 9:629-636, 1988.
4. Burger PC: Pathologic anatomy and CT correlations in the glioblastoma multiforme. Appl Neurophysiol 46:180-187, 1983.
5. Burger PC, Duboid PJ, Schold SC, Smith KR, Odom GL, Crafts DC, Giangasperio F: Computerized tomographic and pathologic studies of the untreated, quiescent, and recurrent glioblastoma. J Neurosurg 58:159- 169, 1983.
6. Burger PC, Mahaley Jr SM, Dudka L, Vogel FS: The morphologic effects of radiation administered therapeutically for intracranial gliomas. A postmortem study of 25 cases. Cancer 44:1256-1272, 1979.
7. Curnes JT, Laster DW, Ball MR, Moody DM, Witcofski RL: Magnetic resonance imaging of radiation injury to the brain. AJNR 7:389-394, 1986.
8. Davis PC, Hoffman JC Jr, Pearl GS, Braun IF: CT evaluation of efforts of cranial radiation therapy in children. AJNR 7:639- 644, 1986.
9. Davis RL, Burger GR, Gutin PH, Phillips TL: Response of human malignant gliomas and CNS tissue to I-125 brachytherapy: A study of seven autopsy cases. Acta Neurochir (Wien) [Suppl] 33:301- 305, 1984.
10. De Reuck J, Eecken HV: The anatomy of late radiation encephalopathy. Eur Neurol 13:481- 484, 1975.
11. Di Chiro G, Oldfield E, Wright DC, De Michelle D, Katz DA, Patronas NJ, Doppman JL, Larson SM, Ito M, Kufta CV: Cerebral necrosis after radiotherapy and/or intraarterial chemotherapy for brain tumors: PET and neuropathologic studies. AJNR 150:189-197, 1988.
12. Dooms GC, Hecht S, Brant-Zawadzki M, Berthiaume Y, Norman D, Newton TH: Brain radiation lesions: MR imaging. Radiology 158:149-155, 1986.
13. Doyle WK, Budinger TF, Valk PE, Levin VA, Gutin PH: Differentiation of cerebral radiation necrosis from tumor recurrence by [18F]FDG and 82Rb positron emission tomography. J Comput Assist Tomogr 11:563-570, 1987.
14. Edwards MS, Wilson CB: Treatment of radiation necrosis, in Gilbert HA, Kagan AR (eds): Radiation Damage to the Nervous System. New York, Raven Press, 1980, pp 129-143.
15. Fischer AW, Holfelder H: Lokales Amyloid in Gehirn. Ein Spatfolge von Rontegenbestrahlungen. Z Chir 227:475-483, 1930.
16. Glass PJ, Hwang TL, Leavens ME, Libshitz HI: Cerebral radiation necrosis following treatment of extracranial malignancies. Cancer 54:1966-1972, 1984.
17. Grossman RI, Hecht-Leavitt CM, Evans SM, Lenkinski RE, Holland GA, Van Winkle TJ, McGrath JT, Curran WF, Shetty A, Joseph PM: Experimental radiation injury: Combined MR imaging and spectroscopy. Radiology 169:305-309, 1988.
18. Gutin PH, Leibel SA, Wara WM, Choucair A, Levin VA, Phillips TL, Silver P, Da Silva V, Edwards MSB, Davis RL, Weaver KA, Lamb S: Recurrent malignant gliomas: Survival following interstitial brachytherapy with high Redistribution of this article permitted only in accordance with the publisher’s copyright provisions. activity iodine- 125 sources. J Neurosurg 67:864-873, 1987.
19. Haymaker W, Ibrahim MZM, Miquel J, Call N: Delayed radiation effects in the brains of monkeys exposed to X- and gamma-rays. J Neuropathol Exp Neurol 27:50-79, 1968.
20. Hect-Leavitt C, Grossman RI, Curran Jr WJ, McGrath JT, Biery DN, Joseph PM, Nelson DF: MR of brain radiation injury: Experimental studies in cats. AJNR 8:427-430, 1987.
21. Hohwieler ML, Lo TCM, Silverman ML, Freidberg SR: Brain necrosis after radiotherapy for primary intracerebral tumor. Neurosurgery 18:67-74, 1986.
22. Hopewell JW, Young CMA: Changes in the microcirculation of normal tissues after irradiation. Int J Radiat Oncol Biol Phys 4:53-58, 1978.
23. Johnson PC, Hunt SJ, Drayer BP: Human cerebral gliomas: Correlation of postmortem MR imaging and neuropathologic findings. Radiology 170:211-217, 1989.
24. Kamman Rl, Go KG, Stomp GP, Hulstaert CE, Berendsen HJC: Changes in relaxation times T1 and T2 after biopsy and fixation. Magn Reson Imaging 3:245-250, 1985.
25. Kramer S, Lee KF: Complications of radiation therapy: The central nervous system. Semin Roentgenol 9:75-83, 1974.
26. Kramer S: The hazards of therapeutic irradiation of the central nervous system. Clin Neurosurg 15:301-318, 1968.
27. Kumar PP, Good RR, Jones EO, Skultety FM, Leibrock LG, McComb RD: Contrastenhancing computed tomography ring in glioblastoma multiforme after intraoperative endocurietherapy. Cancer 61:1759- 1765, 1988.
28. Liena JF, Cespedes G, Hirano A, Simmerman HM, Feiring EH, Fine D: Vascular alteration in delayed radiation necrosis of the brain. Arch Pathol Lab Med 100:531- 534, 1976.
29. Lorenza ND, Nolletti A, Palma L: Late cerebral radionecrosis. Surg Neurol 10:281- 290, 1978.
30. Martins AN, Severance RE, Henry JM, Doyle TF: Experimental delayed radiation necrosis of the brain: Part 1. Effect of early dexamethasone treatment. J Neurosurg 51:587-596, 1979.
31. Martins AN, Johnston JS, Henry JM, Stoffel TJ, Di Chiro G: Delayed radiation necrosis of the brain. J Neurosurg 47:336- 345, 1977.
32. Mikhael MA: Radiation necrosis of the brain: Correlation between patterns of computed tomography and dose of radiation. J Comput Assist Tomogr 3:241-249, 1979.
33. Prados M, Leibel S, Barnett CM, Gutin P: Interstitial brachytherapy for metastatic brain tumors. Cancer 63:657-660, 1989.
34. Rizzoli HV, Pagnonelli DM: Treatment of delayed radiation necrosis of the brain: A clinical observation. J Neurosurg 60:589- 594, 1984.
35. Rottenberg DA, Chernik NL, Beck MD, Ellis F, Posner JB: Cerebral necrosis following radiotherapy of extracranial neoplasms. Ann Neurol 1:339-357, 1977.
36. Rubinstein LJ: Atlas of Tumor Pathology, second series, fascicle 6: Tumors of the Central Nervous System. Washington, DC, Armed Forces Institute of Pathology, 1972, pp 349-360.
37. Sheline GE, Wara WM, Smith V: Therapeutic irradiation and brain injury. Int J Radiat Oncology Biol Phys 6:1215-1228, 1980. 38. Steiner L: Radiosurgery in cerebral arteriovenous malformations, in Fein JM, Flamm ES (eds): Cerebrovascular Surgery. New York, Springer Verlag, 1984, vol 4, pp
39. Tsuruda JS, Kortman KE, Bradley WG, Wheller DC, Van Dalsem W, Bradley TP: Radiation effects on cerebral white matter: MR evaluation. AJNR 8:431-437, 1987.
40. Unger EC, Gado MH, Fulling KF, Littlefield JF: Acute cerebral infarction in monkeys: An experimental study using MR imaging. Radiology 162:789-795, 1987.
41. Valk PE, Budinger TF, Levin VA, Silver P, Gutin PH, Doyle WK: PET of malignant cerebral tumors after interstitial brachytherapy. J Neurosurg 69:830-838, 1988.
42. Woo E, Lam K, Yu Yl, Lee PWH, Huang CY: Cerebral radionecrosis: Is surgery necessary? J Neurol Neurosurg Psychiatry 50:1407-1414, 1987.
The survival rate of patients with malignant gliomas remains dismal, with median, 2-year, and 4- year survivals of 9 months, 10 to 15%, and 5%, respectively, reported in a recent large clinical trial (4). The vast majority of patients who die do so from local tumor recurrence at the primary tumor site despite external beam radiation therapy with or without BCNU chemotherapy. Interstitial brachytherapy has been reported as a treatment modality for patients with recurrent malignant gliomas (2) and for those with newly diagnosed malignant gliomas (in conjunction with externalbeam
radiation therapy) (3). Median total radiation doses, including the interstitial brachytherapy and the external-beam radiation therapy, are in the range of 11,000 to 13,000 cGy (rad). Such patients are at high risk for the development of radiation necrosis, manifest with progressive neurological symptoms and signs, have a contrast-enhancing lesion with surrounding edema and mass effect on imaging studies of the brain, and possible steroid dependency. The overall clinical picture cannot be distinguished from persistent or progressive tumor. A second operation has been required in up to 49% of patients receiving interstitial brachytherapy for symptomatic necrosis (1). This case report by Oppenheimer and co- Redistribution of this article permitted only in accordance with the publisher’s copyright provisions. workers provides an example of the outcome of an untreated case of radiation necrosis, i.e., death, and correlative imaging and histopathological studies. The neuro-oncology team dealing with malignant glioma patients, including the neurosurgeon, neurologist, radiation oncologist, medical oncologist, diagnostic radiologist, and pathologist all must be aware of the clinical-radiological- pathological spectrum presented by this new treatment modality. A similar spectrum is anticipated for patients with malignant gliomas treated by gamma knife or linear accelerator radiosurgery.
Edward G. Shaw
1. Gutin P, Wara W, Larson D, Leibel S, Prados M, Levin V, Sneed P, Weaver K, Silver P, Lamborn K, Phillips T: Stereotaxic interstitial iodine-125 “boost” in the initial management of malignant gliomas. Int J Radiat Oncol Biol Phys 19[Suppl 1]:204, 1990 (abstr).
2. Gutin PH, Leibel SA, Wara WM, Choucair A, Levin VA, Philips TL, Silver P, Da Silva V, Edwards MSB, Davis RL, Weaver KA, Lamb S: Recurrent malignant gliomas: Survival following interstitial brachytherapy with highactivity iodine-125 sources. J Neurosurg 67:864-873, 1987.
3. Leibel SA, Gutin PH, Wara WM, Silver PS, Larson DA, Edwards MSB, Lamb SA, Ham B, Weaver KA, Barnett C, Phillips TL: Survival and quality of life after interstitial implantation of removable high- activity iodine-125 sources for the treatment of patients with recurrent malignant gliomas. Int
J Radiat Oncol Biol Phys 17:1129-1139, 1989.
4. Nelson DF, Diener-West M, Horton J, Chang CH, Schoenfeld D, Nelson JS: Combined modality approach to treatment of malignant gliomas: Re-evaluation of TROG 7401/ECOG 1374 with long- term follow-up: A joint study of the Radiation Therapy Oncology Group and the Eastern Cooperative Oncology Group.
NCI Monogr 6:279-284, 1988.