Incidence of cavernoma development in children after radiotherapy for brain tumors

Transcription

1 J Neurosurg (5 Suppl Pediatrics) 106: , 2007 Incidence of cavernoma development in children after radiotherapy for brain tumors SASHA BURN, F.R.C.S.(SN), 1 ROXANA GUNNY, F.R.C.R., 2 KIM PHIPPS, B.SC. (HONS), 1 MARK GAZE, M.D., 3,4 AND RICHARD HAYWARD, F.R.C.S. 1 1 Departments of Neurosurgery, 2 Neuroradiology, and 3 Oncology, Great Ormond Street Hospital for Children; and 4 Department of Oncology, University of London Hospitals NHS Foundation Trust, London, United Kingdom Object. Cavernous hemangiomas (cavernomas) are benign vascular malformations that may cause seizures and/or hemorrhage when they develop in the brain. The incidence of cavernoma development after brain radiotherapy is unknown. The aim of this study was to assess the prevalence of cavernoma formation in patients who had previously received radiotherapy for brain tumors during childhood. Methods. All patients were identified who were younger than 16 years of age and had undergone radiotherapy for brain tumors within a 16-year period (January 1, 1988 December 31, 2003). The patient data that were ascertained included age at diagnosis, sex, histopathology results, initial preoperative magnetic resonance (MR) imaging results, date of radiotherapy, and date of detection of cavernoma. Children who were followed up for less than 1 month after radiotherapy or who died during treatment were excluded, as were children with brainstem tumors. All patients had undergone diagnostic MR imaging before receiving radiotherapy, and no vascular malformations were revealed. Of the 379 patients identified, 297 satisfied the inclusion criteria. Ten patients (3.4%) developed cavernomas after radiation therapy. The ages of these patients ranged from 2 to 11 years at the time of radiotherapy (median 7 years), and the latency interval between radiotherapy and cavernoma development was 3 to 102 months (median 37 months). Conclusions. The prevalence of cavernomas in the present study was more than six times greater than the prevalence rate cited in the literature for this population. The authors conclude that there is an increased risk of developing an intracranial cavernoma after radiotherapy for brain tumors. The possibility of this complication arising should be mentioned when informed consent is sought before treatment using radiotherapy. KEY WORDS brain tumor cavernoma vascular malformation radiotherapy children pediatric neurosurgery R ADIATION-INDUCED vascular malformations are unusual but well-recognized sequelae of radiotherapy to the brain; other sequelae include diffuse white matter necrosis, demyelination, reactive astrocytosis, brain atrophy, and dystrophic mineralization, any of which may become long-term complications. 3,6,20 Cavernous hemangiomas (cavernomas) are composed histopathologically of closely apposed, dilated vascular channels with little or no intervening brain parenchyma. The blood vessel walls are usually thin and consist of endothelium and collagenous adventitia but may show areas of calcification and even ossification. 7 There is no feeding artery or early-filling draining vein to a cavernoma and thus it is an angiographically occult lesion. On computed tomographic images, cavernomas appear as hyperdense calcified lesions. Typical MR imaging features of cavernomas include a characteristic mixed-signal, central popcorn appearance, which includes a shortening of foci as seen on T 1 -weighted images (representing blood products Abbreviations used in this paper: DVA = developmental venous anomaly; MR = magnetic resonance. of different ages including methemoglobin) surrounded by a rim of signal hypointensity as noted on T 1 - and T 2 -weighted images. This rim of signal hypointensity corresponds to the peripheral ring of hemosiderin seen in histopathological specimens. Results from a previous retrospective study 5 showed that five out of 189 children or adults who had received previous brain radiotherapy received a diagnosis of cavernoma using MR imaging, and suggested that children ( 10 years of age) might be at a greater risk for developing these tumors than adults. In their study the authors reported that three (2%) of 143 children who had received radiotherapy developed cavernomas, whereas only two (0.05%) of 4065 patients older than 10 years of age developed a cavernoma. Yet the risk of developing a cavernoma following radiotherapy to the brain in children up to 16 years of age is unknown. The aim of this study was to determine the incidence of brain cavernoma formation in a large cohort of children who had received radiotherapy for a primary brain tumor outside the pons. Determining any increased risk of developing a cavernoma following brain radiotherapy has important im- 379

2 S. Burn et al. plications for both advising children and their families and obtaining consent when these children are offered radiotherapy as a treatment option. Clinical Material and Methods All patients younger than 16 years of age and who had undergone radiotherapy for brain tumors within a 16-year period (January 1, 1988 December 31, 2003) were identified using the Great Ormond Street Hospital pediatric neurooncology database. The following patient data were ascertained from a neurosurgical patient database, a computerized radiology information system, and patient clinical records: age at diagnosis, sex, histopathology results, initial preoperative MR imaging results, date of radiotherapy, and date of detection of cavernoma. Children who were followed up for less than 1 month after radiotherapy or who died during treatment were excluded, as were children with brainstem tumors as they were believed to be unlikely to survive long enough to develop brain cavernomas. The available radiology reports of all patients were retrieved, and a pediatric neuroradiologist reviewed all MR imaging results of reported cavernomas. Imaging that was not performed or reported by pediatric neuroradiologists at Great Ormond Street Hospital was also reviewed. All patients had diagnostic MR imaging performed before receiving radiotherapy, and no vascular malformations were revealed. Patients were treated according to standard neuroradiology protocols. The preoperative brain MR imaging protocol included obtaining an axial T 2 -weighted image, a coronal precontrast T 1 -weighted image, postcontrast T 1 -weighted fast spin echo images in three planes, and coronal fluidattenuated inversion-recovery images by using a 1.5 tesla MR imaging unit (Siemens Magnetom, Vision and Symphony, Siemens). Imaging was performed beginning 3 to 6 months postoperatively and was repeated every 3 to 6 months for the first year, and then yearly or every other year thereafter for at least 5 years. 4,17,18 The diagnosis of cavernoma was confirmed if the lesion had a complete ring of hypointensity as seen on T 2 -weighted images and a heterogeneous central popcorn appearance, including regions of T 1 shortening. The size, location, signal characteristics, presence of perilesional edema, and temporal evolution of the cavernoma were noted (Fig. 1). Results Of the 379 total patients identified, 297 patients satisfied the inclusion criteria. Ten patients (five male and five female) were identified as having developed cavernomas following radiotherapy. Six patients had high-grade tumors and four had low-grade tumors. One patient presented with seizures and one patient presented with hemiparesis secondary to a hemorrhage at the site of the cavernoma. The cavernomas in the other eight patients that were detected using surveillance imaging have been asymptomatic to date. The age range of the patients who developed cavernomas was 2 to 11 years (median 7 years) at the time of radiotherapy. The latency interval between radiation therapy and the development of the cavernoma was 3 to 102 months (median 37 months). A summary of patient diagnoses, ages at radiotherapy, FIG. 1. Axial T 2 -weighted (A C) and T 1 -weighted (D) MR images obtained in a child who developed a cavernoma after receiving radiotherapy for a pineal region germinoma. A and B: Images showing multiple cavernomas within the radiotherapy field (arrows). C and D: Images obtained 2 years and 9 months after radiotherapy showing signal hypointensity within the inferior left frontal lobe with some central areas of hyperintensity, characteristic of a cavernoma within the radiotherapy field. The postcontrast image (D) shows some linear regions of enhancement presumed to represent anomalous veins (arrows) converging on a central region of T 1 shortening that was hyperintense before contrast enhancement (arrow in C). These regions of T 1 signal shortening and linear contrast enhancement were not present on earlier postoperative MR images. and intervals until cavernoma development following radiotherapy is given in Table 1. This information was then compared with the age at radiotherapy of all patients in the study (Fig. 2). The majority of patients in our study received radiotherapy between the ages of 3 and 10 years (211 patients). There was no obvious statistically significant clustering of patients with cavernomas in any particular age group. Table 2 summarizes the radiotherapy and chemotherapy details in the 10 patients with cavernomas. The radiotherapy dose was compared with the latency period for cavernoma development and no statistically significant relationship was found. There was also no statistically significant relationship between patients who received (112) or did not receive (175) chemotherapy and the development of cavernomas (five patients in each group, p = 0.52, chi-square test). The clinical and radiological course of the cavernomas is detailed in Table 3. Discussion The strength of this study is that it is the only systematic review of a defined group of children who had received brain radiotherapy that focuses on the consequent development of a cavernoma. The importance of studying cavernoma formation lies in the recognition that cavernomas may become symptomatic. We need to be aware that cavernomas and other occult vascular malformations 14,15 may form months after radiotherapy. Cavernoma formation is therefore an issue that may need to be discussed with patients and their families when recommending radiotherapy. The prevalence of cavernoma in this study was 3.4%, 380

3 Cavernoma after radiotherapy for brain tumors in children TABLE 1 Summary of characteristics of patients with postirradiation cavernomas* Case Duration Until No. Age at Cavernoma De- Presenting (sex) Diagnosis Rx (yrs) velopment (mos) Symptoms 1 (M) medulloblastoma seizures 2 (M) low-grade astrocytoma asymptomatic 3 (M) medulloblastoma hemorrhage 4 (M) suprasellar low-grade asymptomatic astrocytoma 5 (F) medulloblastoma asymptomatic 6 (F) basal ganglia PNET asymptomatic 7 (F) DNT asymptomatic 8 (F) cervicomedullary low asymptomatic grade astrocytoma 9 (M) pineal region germino asymptomatic ma 10 (F) suprasellar low-grade asymptomatic astrocytoma * DNT = dysembryonic neuroepithelial tumor; PNET = primary neuroectodermal tumor; Rx = radiotherapy. FIG. 2. Bar graph showing the number of patients with and without cavernomas at each age from 1 to 16 years old. more than six times the prevalence cited in the general population (as high as 0.53%), 5 thus demonstrating that there is an increased risk of developing cavernomas following radiotherapy in children than previously reported. It has been reported that cavernoma development following radiotherapy is related to both the dose of radiotherapy and the age of the patient at the time of treatment. 5,6,9,14 Although we noted a similar trend, the number of patients in our study was too small to be of statistical significance. The increased incidence of occult vascular malformation in children after radiotherapy may be linked to the high levels of angiogenic factors found in children compared with adults. 6 Cavernomas were detected in our study using standard tumor surveillance protocols that included T 1 - and T 2 -weighted MR imaging. Gradient echo T 2 *-weighted imaging, which is particularly sensitive for detecting susceptibility artifacts from blood products (specifically deoxyhemoglobin and hemosiderin), is a standard MR imaging technique that has an increased sensitivity for detecting cavernomas and was not used in these patients. Therefore we may be underestimating the true number of radiotherapy-induced cavernomas in this population. There is a wide variation in the latency interval between radiotherapy and the development of cavernomas, and therefore there is a need to survey patients over a long period of time. Our study revealed a latency interval of 3 months to 8.5 years. There have been reports of cavernoma formation as long as 41 years after radiotherapy. 6 It is possible that more patients in our study may eventually develop cavernomas as our follow-up period was only 14 years. The findings of our study compare favorably with those of similar studies 1,3,5,6,10,15 in that we also found a high incidence of cavernoma formation in children who have had radiotherapy for brain tumors. We discussed similar potential risk factors such as age at treatment with radiotherapy and the possible added risk of chemotherapy, although none of these factors were statistically significant in our study. In other similar studies, patients with cavernomas were identified following clinical events secondary to cavernoma development. Most of the cavernomas in our study were detected as asymptomatic lesions on surveillance imaging. Many of the lesions did not fulfill all the diagnostic criteria for cavernomas on the initial images but more complete features were revealed after further imaging. This delay in identification should be noted when abnormal blood products, distant from the resection zone, are seen on surveillance images. Historically there has been a debate regarding the etiology of cavernomas. Many argue that cavernomas are congenital abnormalities, 11,19 and cavernoma formation following radiotherapy has been reported, but true spontaneous development is rarely reported. All of the patients in the present study had undergone preradiotherapy imaging that did not reveal any evidence of a cavernoma. A cavernoma outside the field of radiotherapy developed in one patient, and this should be regarded as the spontaneous development of a cavernoma, therefore this patient was not included in our results. The pathogenesis of cavernous hemangiomas following radiotherapy is believed to result from injury to vascular structures because the radiotherapy stimulates the upregulation of angiogenesis factors such as vascular endothelial growth factor and basic fibroblast growth factor. 8,12 This upregulation results in mineralizing microangiopathy, which leads to luminal narrowing in the smaller blood vessels in the brain. Stimulation of such growth factors, which are found in cavernous hemangiomas, may explain increased cavernoma formation following radiotherapy. Of note, 24 of the patients in our study received radiotherapy before 3 years of age; most of these patients were diagnosed in the 1980s and followed treatment plans that are no longer in practice. Despite the young age of the patients in this group at the time of radiotherapy, only one of these children developed a cavernoma. The follow-up period for this subgroup was very similar to the total group, thus ruling out a short follow-up period as an explanation for the low number of cavernomas that developed in this subgroup. The patient in Case 5 was noted to have a prominent DVA before receiving radiotherapy in the region where the cavernoma subsequently developed, which was within the radiotherapy field. The link between DVAs and cavernoma development has often been noted, especially since the advent of computed tomography and MR imaging, 2,15,16 and this observation may suggest there was already a predisposition to cavernoma development in this child. Most of the patients in our study developed a single cavernoma, although one patient who developed multiple cavernomas also had multiple regions of linear enhancement 381

4 S. Burn et al. TABLE 2 Summary of radiotherapy and chemotherapy details in 10 patients with cavernomas* Case Age at Cranial No. Rx (yrs) Diagnosis Rx (Gy) Rx Boost Dose (Gy) Chemotherapy Treatment medulloblastoma CHP low-grade astrocytoma low-grade glioma protocol (vincristine & carboplatin) medulloblastoma NA low-grade astrocytoma (suprasellar) 50 NA NA medulloblastoma infant PNET protocol (cyclophosphamide, carboplatin, & vincristine) PNET supratentorial boost CHP rt parietal DNT 50 NA NA low-grade astrocytoma (cervicomedullary NA tumor, leptomeningeal spread) germinoma NA low-grade astrocytoma (suprasellar) 50 NA CHP 455 * CHP = Children s Hospital of Philadelphia Protocol no. 455; NA = not applicable. on the MR image that were believed to represent abnormal enhancing venous structures. It is possible that all of these changes may be due to a generalized field change following radiotherapy. At 9 years of age this child was one of the older children at the time of radiotherapy. He received a moderate dose of radiotherapy (a cranial dose of 24 Gy), and the first cavernoma developed 33 months after radiotherapy. In contrast, the mean duration until the development of a cavernoma in this study was 41 months. Once the cavernoma has been identified, can we predict which ones are more likely to become symptomatic? The incidence of hemorrhage has been reported to be between and 22.9% 13 per year for incidental cavernomas. The incidence of hemorrhage is recognized to be higher in cavernomas that have bled previously. It has been reported that cavernomas in deep anatomical locations such as the thalamus and basal ganglia are more likely to hemorrhage. 13 This may simply be due to their location within or close to eloquent structures resulting in early detection of hemorrhage, or it may have a more structural explanation due to the difference in neuronal structure within white matter compared with gray matter. Only one of the cavernomas in our study was located deep within the brain (within the thalamus). The patient with this cavernoma was excluded from the study because the cavernoma occurred outside the field of radiotherapy. This cavernoma was asymptomatic initially but then hemorrhaged 41 months after it was diagnosed in the patient. There did not seem to be any relationship between the size of the cavernoma and subsequent course of the cavernoma in our study, or in previous studies. We recognize that we have relied upon the accuracy of reports from our own radiology department and external radiology departments for many of the images that were classified as negative for cavernoma and that more sensitive T 2 *-weighted imaging was not routinely performed. It is therefore possible that the number of cavernomas detected in our study is underestimated. All images reported as positive for cavernomas were reviewed by a neuroradiologist. TABLE 3 Summary of the course of cavernomas in 10 patients Duration Until Case Cavernoma De- No. velopment (mos) Site of Cavernoma Course of Cavernoma(s) 1 14 lt parietal lobe patient presented w/ seizures & died prior to further imaging rt frontal & rt parietal lobes early features of cavernoma present 48 mos postradiotherapy; full features of cavernoma developed over following 54 mos; cavernomas then reduced in size over next 6 mos 3 53 rt frontal lobe precentral gyrus cavernoma & edema on initial image reduced over 3 mos 4 41 lt temporal lobe subcortical white matter abnormality began as small region of T 2 hypointensity (2 3 mm) & then developed all the features of a cavernoma over 17 mos 5 16 lt frontal corona radiata extending to prominent DVA on a preradiotherapy image; first signs of a develventricular margin oping cavernoma at same site as DVA detected at 16 mos following Rx; cavernoma increased in size over next 7 mos 6 3 rt posterior limb of internal capsule, increase in cavernoma size over 40 mos thalamus, caudate, corona radiata 7 60 central midbrain no change over 38 mos cerebellar vermis no change over 17 mos 9 33 multiple cavernomas: largest in rt occip- lt insula lesion increased, rt mesial occipital lesion reduced over 13 ital lobe; others in lt frontal, lt mesial mos frontal, lt insula, & lt temporal lobes rt putamen, insular cortex & external/ increase in cavernoma size over the initial 30 mos; then decrease extreme capsules in size over the subsequent 51 mos 382

5 Cavernoma after radiotherapy for brain tumors in children We also appreciate that there were a considerable number of patients lost to follow up in our study. Yet the strength of our study is that it is the only systematic review of a defined group of children who had received brain radiotherapy that focuses on the later development of a cavernoma. Most investigators have examined groups of patients that have presented with symptoms secondary to a cavernoma. Conclusions There is an increased incidence of intracranial cavernoma development following radiotherapy for brain tumors in childhood. This conclusion could be further strengthened by comparing the incidence of cavernoma formation in children with brain tumors who had not received radiotherapy with those who had received radiotherapy. The diagnosis of cavernoma needs to be considered when a child previously treated using radiotherapy presents with either seizures or cerebral hemorrhage in the absence of obvious tumor recurrence. It is therefore important to regularly use MR imaging surveillance in patient follow up to facilitate early detection of a cavernoma and to consider using T 2 *-weighted imaging as part of this routine tumor surveillance. If a cavernoma is detected, appropriate treatment strategies could then be discussed with the patient and family. References 1. Baumgartner JE, Ater JL, Ha CS, Kuttesch JF, Leeds NE, Fuller GN, et al: Pathologically proven cavernous angiomas of the brain following radiation therapy for pediatric brain tumors. Pediatr Neurosurg 39: , Campeau NG, Lane JI: De novo development of a lesion with the appearance of a cavernous malformation adjacent to an existing developmental venous anomaly. AJNR Am J Neuroradiol 26: , Gaensler EH, Dillon WP, Edwards MS, Larson DA, Rosenau W, Wilson CB: Radiation-induced telangiectasia in the brain simulates cryptic vascular malformations at MR imaging. Radiology 193: , Good CD, Wade AM, Hayward RD, Phipps KP, Michalski AJ, Harkness WFJ, et al: Surveillance neuroimaging in childhood intracranial ependymoma: how effective, how often, and for how long? J Neurosurg 94:27 32, Heckl S, Aschoff A, Kunze S: Radiation-induced cavernous hemangiomas of the brain: a late effect predominantly in children. Cancer 94: , Jain R, Robertson PL, Gandhi D, Gujar SK, Muraszko KM, Gebarski S: Radiation-induced cavernomas of the brain. AJNR Am J Neuroradiol 26: , Kalimo H, Kaste M, Haltia M: Vascular diseases, in Graham DI, Lantos PL (eds): Greenfield s Neuropathology, ed 6. London: Arnold, 1997, Vol 1, pp Keene DL, Johnston DL, Grimard L, Michaud J, Vassilyadi M, Ventureyra E: Vascular complications of cranial radiation. Childs Nerv Syst 22: , Koike S, Aida N, Hata M, Fujita K, Ozawa Y, Inoue T: Asymptomatic radiation-induced telangiectasia in children after cranial irradiation: frequency, latency, and dose relation. Radiology 230:93 99, Lew SM, Morgan JN, Psaty E, Lefton DR, Allen JC, Abbott R: Cumulative incidence of radiation-induced cavernomas in longterm survivors of medulloblastoma. J Neurosurg 104 (2 Suppl Pediatrics): , Moran NF, Fish DR, Kitchen N, Shorvan S, Kendall BE, Stevens JM: Supratentorial cavernous haemangiomas and epilepsy: a review of the literature and case series. J Neurol Neurosurg Psychiatry 66: , Novelli PM, Reigel DH, Langham Gleason P, Yunis E: Multiple cavernous angiomas after high-dose whole-brain radiation therapy. Pediatr Neurosurg 26: , Porter PJ, Willinsky RA, Harper W, Wallace MC: Cerebral cavernous malformations: natural history and prognosis after clinical deterioration with or without hemorrhage. J Neurosurg 87: , Pozzati E, Acciari N, Tognetti F, Marliani F, Giangaspero F: Growth, subsequent bleeding, and de novo appearance of cerebral cavernous angiomas. Neurosurgery 38: , Pozzati E, Giangaspero F, Marliani F, Acciarri N: Occult cerebrovascular malformations after irradiation. Neurosurgery 39: , Rigamonti D, Spetzler RF: The association of venous and cavernous malformations. Report of four cases and discussion of the pathophysical, diagnostic, and therapeutic implications. Acta Neurochir (Wien) 92: , Saunders DE, Hayward RD, Phipps KP, Chong WK, Wade AM: Surveillance neuroimaging of intracranial medulloblastoma in children: how effective, how often, and for how long? J Neurosurg 99: , Saunders DE, Phipps KP, Wade AM, Hayward RD: Surveillance imaging strategies following surgery and/or radiotherapy for childhood cerebellar low grade astrocytoma. J Neurosurg 102 (2 Suppl Pediatrics): , Stefan H, Hammen T: Cavernous haemangiomas, epilepsy and treatment strategies. Acta Neurol Scand 110: , Valk PE, Dillon WP: Radiation injury of the brain. AJNR Am J Neuroradiol 12:45 62, 1991 Manuscript submitted July 29, Accepted January 18, Address reprint requests to: Sasha Burn, F.R.C.S.(SN), 6 Appleby Road, London E8 3ET, United Kingdom. sashaburn@ hotmail.com. 383