There are well-established differences in cancer epidemiology

Transcription

1 J Neurosurg Pediatrics 14: , 2014 AANS, 2014 Brain metastases in patients diagnosed with a solid primary cancer during childhood: experience from a single referral cancer center Clinical article Dima Suki, Ph.D., 1 Rami Khoury Abdulla, M.D., 1 Minming Ding, M.S., 1 Soumen Khatua, M.D., 2 and Raymond Sawaya, M.D. 1 Departments of 1 Neurosurgery and 2 Pediatrics, The University of Texas M.D. Anderson Cancer Center, Houston, Texas Object. Metastasis to the brain is frequent in adult cancer patients but rare among children. Advances in primary tumor treatment and the associated prolonged survival are said to have increased the frequency of brain metastasis in children. The authors present a series of cases of brain metastases in children diagnosed with a solid primary cancer, evaluate brain metastasis trends, and describe tumor type, patterns of occurrence, and prognosis. Methods. Patients with brain metastases whose primary cancer was diagnosed during childhood were identified in the Tumor Registry at The University of Texas M.D. Anderson Cancer Center. A review of their hospital records provided demographic data, history, and clinical data, including primary cancer sites, number and location of brain metastases, sites of extracranial metastases, treatments, and outcomes. Results. Fifty-four pediatric patients (1.4%) had a brain metastasis from a solid primary tumor. Sarcomas were the most common (54%), followed by melanoma (15%). The patients median ages at diagnosis of the primary cancer and the brain metastasis were years and years, respectively. The primary cancer was localized at diagnosis in 48% of patients and disseminated regionally in only 14%. The primary tumor and brain metastasis presented synchronously in 15% of patients, and other extracranial metastases were present when the primary cancer was diagnosed. The remaining patients were diagnosed with brain metastasis after initiation of primary cancer treatment, with a median presentation interval of 17 months after primary cancer diagnosis (range 2 77 months). At the time of diagnosis, the brain metastasis was the first site of systemic metastasis in only 4 (8%) of the 51 patients for whom data were available. Up to 70% of patients had lung metastases when brain metastases were found. Symptoms led to the brain metastasis diagnosis in 65% of cases. Brain metastases were single in 60% of cases and multiple in 35%; 6% had only leptomeningeal disease. The median Kaplan-Meier estimates of survival after diagnoses of primary cancer and brain metastasis were 29 months (95% CI months) and 9 months (95% CI 6 11 months), respectively. Untreated patients survived for a median of 0.9 months after brain metastasis diagnosis (95% CI months). Those receiving treatment survived for a median of 8 months after initiation of therapy (95% CI 6 11 months). Conclusions. The results of this study challenge the current notion of an increased incidence of brain metastases among children with a solid primary cancer. The earlier diagnosis of the primary cancer, prior to its dissemination to distant sites (especially the brain), and initiation of presumably more effective treatments may support such an observation. However, although the actual number of cases may not be increasing, the prognosis after the diagnosis of a brain metastasis remains poor regardless of the management strategy. ( Key Words childhood brain metastasis solid primary tumor leptomeningeal disease prognosis oncology There are well-established differences in cancer epidemiology between the adult and pediatric populations. Of particular interest relative to this paper is the differential incidence of brain metastasis between the Abbreviations used in this paper: CNS = central nervous system; LMD = leptomeningeal disease; M.D. Anderson = The University of Texas M.D. Anderson Cancer Center; SRS = stereotactic radiosurgery; WBRT = whole-brain radiotherapy. two groups. Whereas solid brain metastases are known to occur frequently in adult patients with a solid cancer primary tumor and are the most common intracranial tumors in this group, they are uncommon in the pediatric population. Estimates from published clinical reports put the frequency of brain metastases in adults at 20% 40% of all patients with a solid primary cancer versus 1% 10% in the corresponding pediatric population. The type of primary cancer and the presence of pul- 372 J Neurosurg: Pediatrics / Volume 14 / October 2014

2 Trends in childhood brain metastasis from solid primary cancers monary metastases have been most frequently noted as determinants of brain metastases in children with solid primary cancers in published case reports and small series. More effective treatment for the primary cancer and prolonged survival over the recent years 12 may have resulted in an increased frequency of brain metastases in the pediatric population. Improvement in neuroimaging techniques may also have led to improved detection and earlier diagnosis of brain metastasis, with greater certainty in both groups. We present a relatively large series of cases of brain metastasis in patients from a single center who were diagnosed with a solid primary cancer and brain metastasis during childhood, evaluate the brain metastasis trends, and describe multiple variables, including tumor type, patterns of occurrence, and prognosis. Methods The study was conducted under the auspices of a protocol approved by institutional review board of The University of Texas M.D. Anderson Cancer Center (M.D. Anderson). Patients with a solid brain metastasis or leptomeningeal disease (LMD) (referred to collectively as brain metastases) whose primary cancer was diagnosed during childhood were identified from a review of the Tumor Registry records at M.D. Anderson. A retrospective review of the patients hospital records was conducted. Included in our study were patients with parenchymal, epidural, and subdural metastases. Also included were patients with calvarial and supragaleal metastases invading the epidural space or beyond and patients with leptomeningeal involvement. Excluded were patients with metastases to the calvaria only and patients with the direct extension of an extracranial primary tumor into the brain. Autopsy reports were not reviewed as part of this study. Data obtained included patient sex and date of birth; details on the primary cancer, including site, histology, date of diagnosis and treatment; site and timing of extracranial metastasis, if any; presence of symptoms leading to the diagnosis of the brain metastasis; details on the brain metastases, including number, size, location, date of diagnosis, and treatment; and survival information. The date of death (when available) was also confirmed with the Social Security Death Index. Frequencies and descriptive statistics of the demographic and clinical variables were obtained. The data were analyzed using IBM SPSS v.19. The Kaplan-Meier method was used to estimate overall survival proportions, measured from the dates of diagnosis of the primary cancer and brain metastasis and from the date of brain metastasis treatment. All tests were 2-tailed. A p value 0.05 was considered statistically significant. J Neurosurg: Pediatrics / Volume 14 / October 2014 Results A review of the Tumor Registry records identified 7625 pediatric patients with cancer registering at the institution between January 1, 1990 and September 30, Of those, 3950 had a solid primary cancer outside the central nervous system (CNS). During that same period, 54 patients (1.4%) were found to have a solid brain metastasis and/or leptomeningeal disease. These patients constitute our study population. Of the 54 patients with brain metastases, 44% were female and 56% were male. The median ages at diagnosis of the primary cancer and the brain metastasis were years (range birth years) and years (range 3 months years), respectively. The family history was positive for cancer in 63% of patients. Thirty of the brain metastasis histologies (56%) were confirmed by pathology. The most frequent primary cancer was sarcoma, seen in 29 patients (54%) (Table 1). We did not identify any case of Wilms tumor with brain metastasis (0 of 159 patients). Table 2 shows the incidence of brain metastasis among different types of primary cancer based on the records of the Tumor Registry at M.D. Anderson. The imaging modality used for initial brain metastasis diagnosis was MRI in 53% of the 47 patients for whom data were available, CT scans in 43%, and PET-CT and octreotide scan in 1 patient each (4%). The primary cancer was localized at the time of its diagnosis in 48% of the cases. In an additional 14%, the patients had only regional dissemination (to lymph nodes linked to the primary site of involvement). The presentation of the primary tumor and brain metastasis was synchronous in 8 patients (15%). This synchronicity was observed in 7% of the sarcoma cases and in 24% of those with other primary cancers (p = 0.12). In all of these cases, the patient had other extracranial metastatic sites at the time of diagnosis of the primary malignancy. The profile of extracranial metastatic sites included lung metastases in 4 patients, 3 each with bone and lymph node involvement, and 2 with spinal metastasis. In 3 of the 8 patients in whom the primary tumor and brain metastasis were synchronous (1 patient with testicular choriocarcinoma, 1 with unclassified sarcoma, and 1 with high-grade neuroendocrine carcinoma from an unknown primary, all presenting at years of age), CNS symptoms were the initial manifestation of brain metastasis rather than any direct symptoms from the primary malignancy. In 4 of the 8 cases, brain metastasis was identified on routine surveillance scans while the patients remained asymptomatic. One patient s medical records were insufficient for analysis. The rest of the patients were diagnosed with brain metastases after initiation of treatment for the primary cancer, with a median presentation interval of 17 months after diagnosis of the primary cancer (range 2 77 months). The primary cancer was progressing in only 12% of patients at the time of diagnosis of the brain metastasis, but 86% had an extracranial metastasis outside the primary site. Three additional patients had a history of eradication of an extracranial metastasis (bringing the total to 92% with an extracranial metastasis at the time of brain metastasis diagnosis or conversely 8% in whom brain metastasis was the first site of systemic metastasis). As many as 70% of the patients had lung metastasis by the time they presented with brain metastasis, with no statistically significant difference according to tumor histology. None of the 3 patients with neuroblastoma had a lung metastasis or lung invasion at the time of diagnosis 373

3 D. Suki et al. TABLE 1: Characteristics of 54 pediatric patients with brain metastases* Characteristic No. of Pts w/ Data Available Value age at primary cancer Dx 53 median (yrs) range (yrs) age at brain metastasis Dx 53 median (yrs) range (yrs) sex 54 female 24 (44) male 30 (56) family history of cancer 41 yes 26 (63) no 15 (37) family history of CNS malignancy 41 yes 3 (7) no 38 (93) Sx leading to brain metastasis Dx (65) headache 16 (33) nausea/vomiting 13 (27) seizure 11 (23) hemiparesis 6 (12) motor disturbance 3 (6) drowsiness 3 (6) speech difficulty 3 (6) sensory disturbance 2 (4) confusion 2 (4) cranial nerve deficit 2 (4) visual problem 1 (2) irritability 1 (2) dizziness 1 (2) time from Sx to brain metastasis Dx 16 median (days) 3 range (days) 0 23 brain metastasis diagnosed by imaging (100) type of scan 47 MRI 25 (53) CT 20 (43) octreotide nuclear scan 1 (2) PET-CT 1 (2) no. of brain metastases at Dx 52 single 31 (60) multiple 18 (35) LMD only throughout 3 (6) brain metastasis type 50 parenchymal 36 (72) dura-based 10 (20) (continued) TABLE 1: Characteristics of 54 pediatric patients with brain metastases* (continued) Characteristic No. of Pts w/ Data Available Value brain metastasis type (continued) 50 LMD only 3 (6) calvarial & epidural 1 (2) location of solid brain metastasis 45 infratentorial 4 (9) supratentorial 40 (89) both 1 (2) largest diameter of largest solid brain 33 metastasis median (cm) 3.0 range (cm) solid brain metastasis hemorrhagic on 30 imaging yes 18 (60) no 12 (40) solid brain metastasis calcified on 22 imaging yes 5 (23) no 17 (77) solid brain metastasis cystic on imaging 26 yes 9 (35) no 17 (65) path confirmation of brain metastasis (56) resection 26 CSF analysis 2 biopsy 2 LMD status at presentation of brain 51 metastasis excl cases w/ LMD only LMD at presentation before Tx 3 (6) LMD after initiation of Tx of brain 5 (10) metastasis no LMD 43 (84) LMD Dx 11 imaging only 8 (73) CSF only 0 (0) both 3 (27) site of LMD 10 spinal cord 2 (20) brain 4 (40) both 4 (40) primary cancer types excl hematol 54 cancers sarcoma 29 (54) osteosarcoma 11 sarcoma, unclassified 6 Ewing sarcoma 3 (continued) 374 J Neurosurg: Pediatrics / Volume 14 / October 2014

4 Trends in childhood brain metastasis from solid primary cancers TABLE 1: Characteristics of 54 pediatric patients with brain metastases* (continued) Characteristic No. of Pts w/ Data Available Value primary cancer types excl hematol 54 cancers (continued) sarcoma (continued) 29 (54) clear cell sarcoma of kidney 3 rhabdomyosarcoma, alveolar 3 rhabdomyosarcoma, embryonal 1 angiosarcoma 1 pleuropulmonary blastoma 1 malignant melanoma 8 (15) testicular 3 (6) choriocarcinoma 2 endodermal sinus tumor 1 thyroid 3 (6) papillary carcinoma 3 adrenal 3 (6) neuroblastoma 3 lung 1 (2) carcinoid 1 unknown primary 2 (4) choriocarcinoma 1 neuroendocrine carcinoma, high 1 grade gastric 1 (2) signet ring cell carcinoma 1 paraspinal 1 (2) ppnet 1 eye 1 (2) retinoblastoma 1 mediastinum 1 (2) malignant epithelioid neoplasia 1 consistent w/ YST thigh (soft tissue) 1 (2) round cell malignancy w/ mela- 1 nocytic differentiation stage of primary cancer at the time of 50 its Dx localized 24 (48) regional 7 (14) distant 19 (38) primary cancer diagnosed before brain 54 metastasis yes 46 (85) no 8 (15) time from primary cancer Dx to brain 52 metastasis median (mos) 15 range (mos) 0 77 J Neurosurg: Pediatrics / Volume 14 / October 2014 (continued) TABLE 1: Characteristics of 54 pediatric patients with brain metastases* (continued) Characteristic No. of Pts w/ Data Available Value time from primary cancer Dx to brain 45 metastasis among pts whose brain metastasis was diagnosed after primary median (mos) 17 range (mos) 2 77 status of primary cancer at time of brain 32 metastasis Dx stable/responding 5 (16) progressing 4 (12) no evidence of disease 23 (72) extracranial metastasis at time of brain (86) metastasis Dx lung (64) lymph nodes (26) bone, extraspinal 54 8 (15) spine 54 6 (11) liver 53 4 (8) soft tissue 54 2 (4) other** 54 5 (9) extracranial metastasis by time of brain (92) metastasis Dx lung (70) lymph nodes (26) bone, extraspinal 54 8 (15) spine 54 6 (11) liver 53 4 (8) soft tissue 54 2 (4) other** 54 5 (9) * Values represent numbers of patients (%) unless otherwise specified. Percentage totals may not equal 100% due to rounding. Dx = diagnosis; excl = excluding; hematol = hematological; LMD = leptomeningeal disease; path = pathological; ppnet = peripheral primitive neuroectodermal tumor; pt = patient; Sx = symptom(s); Tx = treatment; YST = yolk sac tumor. A value of 0 indicates diagnosis at birth. One additional patient whose brain metastasis was initially diagnosed through screening did subsequently develop headache, hemiparesis, and sensory disturbance prior to treatment of the brain metastasis. Among patients with symptoms and available data. Some of the non-cns metastases (3 lung metastases) had been treated and were no longer present at the time of diagnosis of the brain metastasis, hence the difference in the data between by the time of brain metastasis and at the time of brain metastasis. ** Paraspinal, ovarian, peritoneal, renal, and pancreatic (one each). of the brain metastasis. Two of these patients had calvarial tumors with intracranial extensions, and the third had a dura-based metastasis. In regard to clinical presentation of brain metastases, 65% of the patients were symptomatic at presentation. 375

5 TABLE 2: Frequency of brain metastasis relative to tumor histology in pediatric patients at The University of Texas M.D. Anderson Cancer Center, * Histology Pts w/ Brain Metastasis Denominator From Tumor Registry Over Same Interval Frequency of Brain Metastasis (%) choriocarcinoma clear cell sarcoma of the kidney signet ring cell carcinoma angiosarcoma neuroendocrine tumors endodermal sinus tumor & YST unclassified sarcoma malignant melanoma carcinoid ppnet osteosarcoma pleuropulmonary blastoma rhabdomyosarcoma Ewing sarcoma neuroblastoma papillary carcinoma retinoblastoma <1 round cell malignancy w/ melanocytic differentiation 1 Wilms tumor * Data from The University of Texas M.D. Anderson Cancer Center Tumor Registry records. Brain and other CNS primary tumors were excluded from the total. Values represent numbers of patients unless otherwise specified. Total represents all thyroid tumors. D. Suki et al. Headache (33%), nausea/vomiting (27%), and seizures (23%) were the most common symptoms, with no difference in patients with sarcoma versus those with other primary cancers. Thirty-five percent of brain metastases were discovered during staging or screening. Of the 52 patients for whom data were available on the number of brain metastases at initial brain metastasis presentation, 31 (60%) had a single tumor, 18 (35%) had multiple tumors (6 had 2 brain metastases, 9 had 3 or more brain metastases, and 3 had an unknown number), and 3 (6%) had only LMD. The tumors were parenchymal in 36 cases (72%) and dura-based in 10 (20%; Table 1). One patient had a calvarial/epidural tumor. Regarding location of the solid brain metastases, 89% of patients for whom data were available had supratentorial tumors, 9% had infratentorial tumors, and 2% had tumors in both locations. The numbers of patients were too small to show statistically significant differences in number of tumors or tumor location among sarcoma, melanoma, and other primaries. Seven melanoma patients had supratentorial tumors; one had both supratentorial and infratentorial tumors. The median largest tumor diameter was 3.0 cm (range cm); 60% of tumors for which data were available appeared hemorrhagic on imaging scans, and 35% were cystic. All 3 patients with LMD as the sole manifestation of brain metastasis developed the LMD after treatment for the primary cancer. Two patients had LMD in conjunction with the solid brain metastases at initial discovery of the brain metastasis (one with retinoblastoma in the spinal cord and one with unclassified sarcoma in both brain and spinal cord); one developed brain metastasis 3 months later, prior to any brain treatment (melanoma in the brain). Five patients (one with neuroblastoma in the brain, one with alveolar rhabdomyosarcoma in both the brain and the spinal cord, one with papillary thyroid carcinoma in the spinal cord, one with malignant melanoma from an unknown site, and one with round cell malignancy with melanocytic differentiation in the brain) developed LMD later in the course of the disease. Most patients (91%) received treatment for brain metastases as shown in Table 3 (treatment data were not available for 1 patient). Of the 3 patients with LMD as the sole manifestation of brain metastasis mentioned above, one received no treatment (peripheral primitive neuroectodermal tumor); one received WBRT (alveolar rhabdomyosarcoma); and one received chemotherapy followed by WBRT (signet ring cell gastric carcinoma). All died within 2 months of LMD diagnosis. The other 3 patients with a solid brain metastasis and an LMD diagnosis prior to initiation of brain treatment received stereotactic radiosurgery (SRS, administered to the metastasis, an unclassified sarcoma), whole-brain radiotherapy (WBRT, for melanoma), or chemotherapy followed by WBRT + intensitymodulated radiation therapy (for retinoblastoma). Of the remaining 48 patients, 45% received combination therapy as initial brain metastasis treatment (28% of those had resection with WBRT and/or chemotherapy, and 17% received chemotherapy followed by WBRT); 28% un- 376 J Neurosurg: Pediatrics / Volume 14 / October 2014

6 Trends in childhood brain metastasis from solid primary cancers TABLE 3: Treatments and outcomes in pediatric patients with brain metastases* Variable No. of Pts w/ Data Available Value brain metastasis treated during lifetime (91) if treated, initial Tx was the only Tx 45 yes 29 (64) no 16 (36) reason initial Tx was the only Tx 26 no brain metastasis recurrence 22 (85) no Tx despite recurrence 2 (8) death soon after brain metastasis Dx 1 (4) went to hospice 1 (4) initial Tx type 52 none 5 (10) resection 13 (25) WBRT 5 (10) SRS 2 (4) chemo 4 (8) resection & WBRT 4 (8) resection & chemo 3 (6) chemo & WBRT 6 (12) chemo & SRS 3 (6) chemo & proton therapy 1 (2) resection, chemo, & WBRT 3 (6) resection, chemo, & SRS 2 (4) resection, chemo, & radiation (unspecified) 1 (2) vital status at last follow-up 54 alive 12 (22) dead 42 (78) overall survival after primary cancer Dx (mos) median 29 95% CI overall survival after brain metastasis Dx (mos) median 9 95% CI 6 11 overall survival after brain metastasis Tx (mos) median 8 95% CI 6 11 brain metastasis among the patients remaining alive at the end of the study was 24 months (range months). All but 2 patients had durations of follow-up or survival of 3.5 years or less. The patient with the longest follow-up (160.5 months) was a 17-year-old male initially diagnosed with a Clark s Level 2 melanoma on his midback. The patient had a wide local resection without residual tumor. Five years later, he developed right facial palsy that led to the diagnosis of a hemorrhagic lesion in the left frontal lobe. The tumor was resected and WBRT was administered. Further evaluation during the same interval showed evidence of systemic metastatic disease in the liver, lungs, and soft tissue, as well as lymphadenopathy, for which the patient received biochemotherapy. Two years later, at last study follow-up, there was continued response and no evidence of brain metastasis recurrence. The patient with the second longest follow-up (141.7 months) was a male child diagnosed with a localized lung carcinoid at age 11 and with a liver metastasis and a supratentorial brain metastasis at around age 16. The brain metastasis was asymptomatic and was diagnosed through staging. It was completely resected, and there has been no recurrence as of the last follow-up visit. The median Kaplan-Meier estimate of survival after primary tumor diagnosis was 29 months (95% CI months) (Fig. 1). The median Kaplan-Meier estimate of survival after brain metastasis diagnosis was 9 months (95% CI 6 11 months; Fig. 2). The 6-month and 1-year survival rates were 66% and 38%, respectively. After excluding the 6 patients with an LMD diagnosis prior to initial treatment, the median survival time after brain metastasis diagnosis was 9 months (95% CI 6 13 months). Patients not receiving treatment survived for a median of 0.9 months after brain metastasis diagnosis (95% CI months); those receiving treatment survived for a median of 8 months after initiation of therapy (95% CI 6 11 months). Corresponding figures, after excluding patients with LMD at presentation, are 0.9 months (95% CI months) and 8 months (95% CI 6 11 months), respectively. * Values represent numbers of patients (%) unless otherwise specified. Percentage totals may not equal 100% due to rounding. chemo = chemotherapy; SRS = stereotactic radiosurgery. derwent resection alone; 9% received WBRT alone; 9% received chemotherapy alone; and 9% received no treatment. The numbers were too small to permit detection of statistically significant differences in type of therapy administered to patients or tumor characteristics such as status of the primary or systemic noncerebral cancers at the time of initial brain metastasis treatment, number or location of brain metastases, or tumor histology. The median duration of follow-up after diagnosis of J Neurosurg: Pediatrics / Volume 14 / October 2014 Fig. 1. Kaplan-Meier plot of overall survival after primary cancer diagnosis in a series of 54 pediatric patients with brain metastasis from a solid primary cancer. 377

7 D. Suki et al. Fig. 2. Kaplan-Meier plot of overall survival after diagnosis of brain metastasis in a series of 54 pediatric patients with brain metastasis from a solid primary cancer. Discussion Metastasis to the CNS is the most common type of brain tumor in the adult population, contributing almost one-third of all cases with such a diagnosis. It is estimated that 20% 40% of all adults with cancer will develop brain metastasis. Brain metastases originate mainly from lung, breast, melanoma, kidney, and gastrointestinal primary cancers. Tumors from unknown primary cancers are not uncommon (15% of brain metastasis cases). In the pediatric population, brain seeding is common in patients with leukemia. On the other hand, brain metastasis from a solid extracranial primary tumor is uncommon, with reported frequencies of 1.5% 10% 1,8,14,22,27 from various clinical sources and 6% 13% from autopsy series. 7,24 Differences in frequency and patterns of occurrence between adults and children have been loosely linked to a number of factors, including but not limited to the different spectrum of primary cancer histologies seen between the 2 populations; physiological/biological differences between children and adults, such as a less mature blood-brain barrier in the former (enabling penetration of the drugs into the cerebral parenchyma); and the different treatment methodologies used. 20 The 3 factors of histology, physiology, and treatment are not independent of each other, as the type of tumor and its sensitivity to various treatments also dictate the treatment methodology chosen, which in turn may impact physiology (particularly the integrity of the blood-brain barrier and its capacity to prevent neoplastic cells from invading the brain). 21 These differences limit attempts to extrapolate data from the adult literature to treat the pediatric population having brain metastases. Many recent publications on pediatric brain metastasis suggest 1) that more effective treatment of the primary cancer and an increase in survival in recent years may have resulted in an increased incidence of brain metastases in the pediatric population and 2) that improved neuroimaging techniques may have in turn led to improved detection, with brain metastases being diagnosed at an earlier stage and with greater certainty. Owing to the rarity of the condition in children, the earlier published reports were only anecdotal. In more recent decades, there has been a surge in published small series, bolstering support for the notion that the incidence of the condition and/ or its rate of detection could be on the increase (Tables 4 and 5). These series do not include more than 30 patients, hindering the extraction of solid conclusions as to epidemiology, natural history, uniformity, and efficacy of various treatment approaches and detailed patient outcomes. Our series estimate of a 1.4% frequency of brain metastases in pediatric cancer patients registered at M.D. Anderson and the trend over time (Fig. 3) may or may not be construed as evidence against an increase in the incidence of brain metastasis. These data were obtained from the hospital s Tumor Registry. At the Tumor Registry, abstracting of patient records takes place either 6 months after initial registration, upon determining that a patient has developed a second primary, or at death. Ascertainment of metastases is based on information in the patient s record at the time of each abstracting. Consequently, metastases in patients not known to be dead, who have not developed a second primary, or who have not returned to M.D. Anderson may not be reflected in the Tumor Registry data, making our 1.4% figure an underestimate of the true frequency of brain metastases. On the other hand, the 1.4% figure could be a true estimate of the incidence of brain metastases in our patient population, reflecting an earlier diagnosis of the primary cancer prior to its dissemination to a distant site, and particularly, prior to seeding of the brain or the use of more effective treatments. The latter hypothesis is supported by the finding that the primary cancer was localized at the time of its diagnosis in 48% of our cases and that there was only regional dissemination in another 14%. The current notion that there is an increased incidence of brain metastasis among children with a primary solid cancer is therefore challenged in our series. The solid primary cancer histologies most commonly reported in the pediatric population are sarcomas, neuroblastomas, nephroblastomas, and germ cell tumors. 6 The tendency of these non-cns primary cancers to metastasize varies among different reports. An earlier study described Wilms tumor as a common precursor to brain tumors. 24 No case of brain metastasis in a patient with Wilms tumor was found in our series, which is consistent with the findings of others, 1,7 a probable reflection of the effectiveness of chemotherapy against this particular histological type. Graus et al. 7 reported a high incidence of brain metastasis for germ cell tumors. We found a brain metastasis incidence of 43% among the patients with choriocarcinoma. Within the sarcoma category, different subhistologies are said to have different predilections for metastasis to the brain. The incidence of brain metastases among various sarcoma subtypes in our series was equally low. In the adult population, melanoma has the highest propensity to metastasize to the brain. In children, melanoma is cited among the primary tumors associated with a relatively higher incidence of metastasis (8% to 18%). 3,16,17 In our series, melanoma accounted for 15% of the cases of metastasis, but the overall prevalence of melanoma in our pediatric patient population was only 3%. Brain metastasis was diagnosed at the same time as the primary cancer in 8 patients (15%), another unusual 378 J Neurosurg: Pediatrics / Volume 14 / October 2014

8 Trends in childhood brain metastasis from solid primary cancers TABLE 4: Summary of the published literature on pediatric cases of brain metastasis from a solid primary cancer: Part 1 Authors & Year Study Years No. of Pts w/ Brain Metastases Primary Cancer Histology Distribution of Primary Cancer Histology Cases w/ Only LMD Included? Cases w/ Intracranial Extension From Skull: Dura Included? Autopsy Results Included? Pt Age at Dx* Time From Primary Dx to Brain Metastasis Dx Kramer et al., 2001 Kebudi et al., 2005 Bouffet et al., 1997 Rodriguez- Galindo et al., 1997 Graus et al., 1983 Stefanowicz et al., 2011 Tasdemiroglu & Patchell, neuroblastoma (Stage IV) all extracranial solid tumors including lymphoma all solid primaries excl lymphoma & primary brain tumors neuroblastoma, 11 (100) yes no no primary: median 37 mos (range mos) sarcoma, 12 (75 ); Wilms tumor, 2 (12.5); germ cell tumors, 2 (12.5) Ewing sarcoma, 3 (25); neuroblastoma, 3 (25); Wilms tumor, 1 (8.3); softtissue sarcoma, 1 (8.3); osteosarcoma, 3 (25); retinoblastoma, 1 (8.3) no no no unspecified: 10.5 yrs (range 1 16 yrs) no no no unspecified: median 9 yrs (range 3 17 yrs) melanoma melanoma, 8 (100) NR NR no primary: median all solid extracranial tumors excl lymphoma all solid extracranial malignancies (total w/ parenchymal, calvarial, & dural mets; the 7 cases of parenchymal were described further) all non-cns primaries excl lymphoma & leukemia for the 18 cases diagnosed postmortem: osteosarcoma, 5 (27.8); rhabdomyosarcoma, 3 (16.7); Ewing sarcoma, 2 (11.1); germ cell tumor, 4 (22.2); melanoma, 1 (5.6); angiosarcoma, 1 (5.6); malignant schwannoma, 2 (11.1) soft-tissue sarcoma, 3 (30); germ cell tumors, 3 (30); neuroblastoma, 1 (10); clear cell sarcoma of the kidney, 1 (10); osteosarcoma, 1 (10); pleuropulmonary blastoma, 1 (10) Wilms tumor, 2 (28.6); neuroblastoma, 2 (28.6); melanoma, 1 (14.3); hepatocellular carcinoma, 1 (14.3); angiosarcoma, 1 (14.3) 15 yrs (range yrs) median 12.2 mos (range 5 32 mos) median 16 mos (range 1 70 mos) median 15 mos (range 9 24 mos) median 20 mos (range 0 50 mos) NR no yes NR osteosarcoma, median 22 mos (range mos); germ cell tumor, median 13 mos (range 0 39 mos); rhabdomyosarcoma, median 8.5 mos (range mos); other tumors, median 20 mos (range mos) NR no no metastasis: median 13.8 yrs (range yrs) NR no no unspecified: median 6 yrs (range 1 18 yrs) median 8 mos (range 1 32 mos) mean 327 days (range days) (continued) J Neurosurg: Pediatrics / Volume 14 / October

9 D. Suki et al. TABLE 4: Summary of the published literature on pediatric cases of brain metastasis from a solid primary cancer: Part 1 (continued) Authors & Year Study Years No. of Pts w/ Brain Metastases Primary Cancer Histology Distribution of Primary Cancer Histology Cases w/ Only LMD Included? Cases w/ Intracranial Extension From Skull: Dura Included? Autopsy Results Included? Pt Age at Dx* Time From Primary Dx to Brain Metastasis Dx Parasuraman et al., 1999 Weyl-Ben Arush et al., 1995 Vannuci & Baten, 1974 Marina et al., 1993 Spunt et al., 2004 Postovsky et al., 2003 Paulino et al., rhabdomyosarcoma & Ewing sarcoma all solid primaries excl brain tumors & lymphoproliferative disorders all extracranial solid malignancies osteosarcoma except head & neck primaries Ewing sarcoma, 11 (52.4); rhabdomyosarcomas, 10 (47.6) neuroblastoma, 3 (50); osteosarcoma, 1 (16.7); rhabdomyosarcoma, 1 (16.7); hepatocarcinoma, 1 (16.7) neuroblastoma, 2 (15.4); embryonal rhabdomyosarcoma, 3 (23); Wilms tumor, 4 (30.7); osteogenic sarcoma, 1 (7.7); hepatocellular carcinoma, 1 (7.7); melanoma, 1 (7.7); ovarian carcinoma, 1 (7.7) no yes yes primary: median 10.4 yrs (range yrs) NR NR no NR median 13 mos (range 1 36 mos) NR no yes primary: median 58.5 mos (range 1 mo 13 yrs) osteosarcoma, 13 (100) NR NR yes primary: range 4 25 yrs germ cell tumors germ cell tumors, 16 (100) NR NR yes primary: median 6.5 yrs (range yrs) sarcoma sarcoma, 18 (100) no no no primary: mean 15.4 yrs (range 4 24 yrs); metastasis: mean 17.4 yrs (range yrs) sarcoma, neuroblastoma, & Wilms tumor sarcoma, 20 (66.7); neuroblastoma, 9 (30); Wilms tumor, 1 (3.4) no no yes metastasis: median 14 yrs (range 8 mos 20 yrs) NR median 23 mos (range 5 48 mos) median 3 mos (range 15 days 33 mos) median 7 mos (range 0 37 mos) for the 10 pts w/ localized disease at Dx, mean 34.3 mos (range mos); for the 8 pts w/ metastatic disease at time of Dx, mean 11.1 mos (range, 0 24 mos) 5 mos (range 1 43 mos) (continued) 380 J Neurosurg: Pediatrics / Volume 14 / October 2014

10 Trends in childhood brain metastasis from solid primary cancers TABLE 4: Summary of the published literature on pediatric cases of brain metastasis from a solid primary cancer: Part 1 (continued) Time From Primary Dx to Brain Metastasis Dx Autopsy Results Included? Pt Age at Dx* Cases w/ Intracranial Extension From Skull: Dura Included? Cases w/ Only LMD Included? Distribution of Primary Cancer Histology Primary Cancer Histology No. of Pts w/ Brain Metastases Study Years Authors & Year 16 mos (range 2 77 mos) yes yes no primary: median 11 yrs (range 0 18 yrs); metastasis: 13 yrs (range 4 mos 20 yrs) sarcoma, 24 (54.5); melanoma, 5 (11.4); choriocarcinoma, 3 (6.8); endodermal sinus tumor, 2 (4.5); neuroblastoma, 3 (6.8); thyroid, 2 (4.5); signet cell gastric carcinoma, 1 (2.3); carcinoid, 1 (2.3); ppnet, 1 (2.3); neuroendocrine tumor, 1 (2.3); retinoblastoma, 1 (2.3) present case solid tumors outside the CNS * Values are given for age at diagnosis of primary tumor or brain metastasis as indicated. For values marked as unspecified, the reference point was unclear. mets = metastases; NR = not reported. finding in this study. This frequency is higher than is usually reported in children (5% 10% 1 ) and more similar to that reported in adults. To our knowledge, brain metastasis has rarely been reported as the site of initial presentation of a solid primary cancer in children. In 3 of the 8 cases, the brain symptom was the initial manifestation of the cancer and led to the diagnosis of the brain metastasis and the subsequent prompt establishment of the primary cancer diagnosis. The rest of the patients in our series developed brain metastasis after initiation of treatment for the primary tumor, within a median of 17 months after the primary cancer diagnosis (range 2 77 months). This figure falls within the range of medians from other series (13 22 months). 1,4,5,8,27 In 10% of adult patients with a solid primary cancer, the brain is the primary site of systemic metastasis after the diagnosis of the primary cancer. 20 In the other 90%, a brain metastasis is preceded by a systemic noncerebral metastasis. Similarly, in the pediatric population, brain metastasis generally follows extensive systemic metastasis and disease progression. 1,6 8,14,20 Our results are consistent with these previous findings, but we did find 4 cases (8%) in which the brain was the first site of metastasis after the diagnosis of the primary tumor. Bouffet et al. 1 reported 2 patients (17%) in their series of 12 pediatric patients with metastatic brain tumors, in whom the brain was the first and only site of metastasis. It is possible that although systemic chemotherapy effectively suppresses systemic metastases, it is kept out of the CNS by the blood-brain barrier and hence is ineffective in preventing brain metastasis. Lung cancer, both primary and secondary, is associated with a high incidence of brain metastasis in the adult population, supporting the hypothesis that brain metastasis results from hematogenous spread. Primary lung tumors are rare in pediatric patients, a likely explanation for the lower incidence of brain metastases observed in pediatric patients. However, secondary lung tumors do occur, and these have been reported to be often associated with a brain metastasis. 1,4,7,24,25 In our series, as many as 70% of patients had a lung metastasis at the time brain metastasis was found, suggesting a route of transmission similar to that in adults. There was no statistically significant difference relative to tumor histology, possibly owing to the small number of cases in the various histological groups. Although a majority of patients had lung metastasis, one-third or so did not. In their 1983 series on brain metastases in children, Graus et al. 7 reported an absence of neuroblastoma cases and attributed this finding to the rarity of pulmonary metastasis from that disease; so did Shaw and Eden 18 in their 1992 paper and Westphal et al. in their 2003 paper. 26 In our series, among the patients with no pulmonary metastasis, we found 3 instances of brain metastasis from a neuroblastoma primary. Owing to the absence of pulmonary metastasis in that disease, brain metastases from neuroblastoma are rarely parenchymal. 8,22 These lesions generally arise from the adjacent bone. Likewise, none of the 3 tumors from a neuroblastoma primary tumor in our series was parenchymal. Regarding brain metastasis characteristics, the tumors were symptomatic in 65% of patients, and headache, J Neurosurg: Pediatrics / Volume 14 / October

11 D. Suki et al. TABLE 5: Summary of the published literature on pediatric cases of brain metastasis from a solid primary cancer: Part 2* Authors & Year Brain Metastasis Incidence Brain Metastasis Dx (screening vs Sx vs postmortem) Brain Metastasis Location No. of Brain Mets CNS Mets at Time of Primary Dx Cases w/ Other Organ Involvement at Dx of Primary Non-CNS Mets at Dx of Brain Metastasis Duration of Survival After Primary Dx Survival After Brain Metastasis Dx Kramer et al., 2001 Kebudi et al., 2005 Bouffet et al., 1997 Rodriguez- Galindo et al., 1997 Graus et al., 1983 Stefanowicz et al., 2011 Tasdemiroglu & Patchell, 1997 Parasuraman et al., 1999 Weyl-Ben Arush et al., 1995 Vannuci & Baten, 1974 Marina et al., /251 (4.3) NR NR NR 0 (0) 11 (100) 3 (27) NR median 6.7 mos (no range given) 16/1100 (1.4) NR supratentorial in 11 pts; infratentorial in 3; both in 2 12/486 (2.4) Sx in 10 cases, screening in 2 8/44 (18), only 6 described in study Sx in 4 cases, screening in 2 31/139 (22.3) Sx in 25 cases; postmortem in 6 10/511 (2) Sx in 8 cases, screening in 2 12/154 (8); 7 (4.5) parenchymal 11/335 (3.3) for Ewing sarcoma; 10/419 (2.4) for rhabdomyosarcoma supratentorial in 8 pts; infratentorial in 1; both in 3 supratentorial in 4 pts; infratentorial in 1; both in 1 single in 8 pts; multiple in 8 single in 5 pts; multiple in 7 single in 3 pts; multiple in 3 NR single in 7 pts; multiple in 24 supratentorial in 10 pts single in 5 pts; multiple in 5 NR NR single in 4 pts; multiple in 3 NR NR as a summary, but as case by case w/ incomplete information single in 15 pts; multiple in 16 4 (25) 12 (75) 16 (100) NR for the pts who died (n = 15), median 2 mos (range 2 days 6 mos); 1 patient was still alive at 20 mos 0 (0) 9 (75) 10 (83.3) NR for the pts who died (n = 11) median 3 mos (range 1 day 38 mos); 1 pt was still alive at 6 yrs 1 (16.7%) 4 (67.7) 5 (83.4) NR for the pts who died (n = 5), median 5 mos (range 2 10 mos); 1 pt was still alive at 34 mos NR NR 30 (96.7) NR not enough data for a summary 2 (20) 7 (70) NR NR not enough data for a summary NR NR NR NR median 30 days (range days) 1 (4.7) 12 (57) NR 76.2% ± 9% at 1 yr; 28.6% ± 9.1% at 3 yrs median 2.7 mos; 1-yr survival 23.8% ± 8.5% 6/61 (10) NR NR NR NR NR NR NR 9.8 mos (range 2 54 mos) 13/217 (6) Sx in 10 cases, postmortem in 3 supratentorial in 13 pts single in 4 pts; multiple in 9 NR NR NR NR information available in 9 cases; median 31.5 days (range days) 13/254 (5%) Sx in all 13 cases NR NR 0 (0) 6 (46.1) NR NR median 16 mos (range mos) (continued) 382 J Neurosurg: Pediatrics / Volume 14 / October 2014

12 Trends in childhood brain metastasis from solid primary cancers TABLE 5: Summary of the published literature on pediatric cases of brain metastasis from a solid primary cancer: Part 2* (continued) Authors & Year Brain Metastasis Incidence Brain Metastasis Dx (screening vs Sx vs postmortem) Brain Metastasis Location No. of Brain Mets CNS Mets at Time of Primary Dx Cases w/ Other Organ Involvement at Dx of Primary Non-CNS Mets at Dx of Brain Metastasis Duration of Survival After Primary Dx Survival After Brain Metastasis Dx Spunt et al., 2004 Postovsky et al., 2003 Paulino et al., /206 (8) Sx in 12 cases, screening in 2, postmortem in 2 supratentorial in 9 pts; infratentorial in 3; both in 4 18/411 (4) Sx in 18 cases supratentorial in 15 pts; infratentorial in 2; both in 1 30/611 (5) Sx in 25 cases, screening in 1 (remainder unavailable) current series 44/3613 (1.2) Sx in 26 cases, screening in 10, findings on other images in 4, no data in 4 NR 2 (12.5) NR 15 (87.5) NR information available in 11 cases; for those who died (n = 9), median survival 4 mos (range 0 19); 2 pts still alive (at 4 yrs & 18 yrs) single in 7 pts; multiple in 11 not enough to classify single in 18 pts; multiple in 12 supratentorial in 30 pts; infratentorial in 4; both in 1 for solid brain mets: single in 23 pts; multiple in 16 (where data were available) 0 (0) 8 (44.4) 13 (72.2) NR mean 5.03 ± 5.8 mos 5 (16.7) NR 29 (96.7) NR median 4 mos (range 1 wk 16 mos); 6-mo survival rate 27%; 1- yr survival rate 11.5% 8 (18) 22 (55) 39 (95) 28 mos (95% CI mos) median 8 mos (95% CI 5 10 mos) * Values represent numbers of patients (%) unless otherwise specified. Means are presented with SDs. J Neurosurg: Pediatrics / Volume 14 / October

13 D. Suki et al. Fig. 3. Yearly trends of brain metastasis diagnosis. Data are from The University of Texas M.D. Anderson Cancer Center Tumor Registry, nausea/vomiting, and seizures were the most common symptoms. These symptoms occurred among patients with various tumor histological types and ages at the time of diagnosis of brain metastasis. Our findings with regard to symptoms are similar to those previously published. 1,7,8,13,14 About one-third of the instances of brain metastasis were discovered during staging or screening. A few years ago, the rarity of the condition in pediatric patients did not warrant routine surveillance imaging in cases of a solid primary tumor. 6 However, the alleged change in the epidemiology toward an increasing trend may have somewhat altered clinical practice. In our series, only 1 of 17 cases of metastasis (in a patient with testicular choriocarcinoma) discovered during screening or staging was from the 1990s. The absence of denominator data, however, does not allow a firm conclusion of a true change in practice in favor of screening for brain metastases. Overall, 3 patients had LMD as the initial presentation of the brain metastasis. In their series from the University of Iowa Hospitals and Clinics, Paulino et al. 14 used lumbar puncture at initial diagnosis to report an LMD incidence of 28.5% in 7 patients with brain metastasis. There was a high prevalence of LMD at autopsy (found in 6 of 7 children). The authors recommended performing a lumbar puncture as well as imaging of the spine in pediatric patients with a diagnosis of a brain metastasis. The overall prognosis of pediatric patients with brain metastasis is poor because the disease is already very advanced when diagnosed. Treatment only improved survival marginally, with only 2 children in this series surviving beyond 10 years after diagnosis of brain metastasis. The heterogeneity of the patient population and the small number of patients prevent the formulation of definitive recommendations regarding the optimal treatment of brain metastasis in children. Like all other published series, this series is based on a retrospective review of available patient records and suffers from many of the biases inherent in such reviews, including information and selection biases. It is nevertheless one of the largest, if not the largest, published series on pediatric patients with brain metastases. Another caveat in our series concerns the fact that the most common solid tumor was osteosarcoma. We see a larger number of patients at our institution with pediatric osteosarcoma because of the referral pattern and the institutional expertise with this tumor, thus contributing to its higher incidence. However, in other major published series, 1,7,14,24 Ewing sarcoma and soft tissue sarcoma, followed closely by osteosarcoma, are the main primary tumors that metastasize to the brain. The poor prognosis and possible change in natural history of the condition should be a concern to pediatric oncologists. Some authors have argued in favor of CNS prophylaxis in pediatric patients with solid cancers, especially Ewing sarcoma. 11 Their argument is predominantly based on the premise that systemic chemotherapy prevents systemic metastases but is prevented by the bloodbrain barrier from reaching the CNS and protecting it against brain metastasis. Trigg et al. 23 reviewed data from National Cancer Institute protocols through July 1980 and compared the incidence of CNS involvement in Ewing sarcoma between 92 patients receiving CNS prophylaxis (WBRT of 2000 rads) and a single dose of intrathecal methotrexate (12 mg/m 2 ) and 62 patients not receiving any CNS prophylaxis. They found no difference between the 2 groups in the incidence of CNS metastasis. More studies to identify new agents that are likely to cross the blood-brain barrier are warranted. Some authors have also argued in favor of routine surveillance in all pediatric patients with a solid primary cancer 10 or visceral metastasis. 17 Those same authors are skeptical as to whether early detection is likely to impact outcome with the currently available therapies. 10 According to Buzaid et al., 2 routine surveillance for brain metastasis may not be warranted in all patients. The authors referred to the rate of false positives obtained with the use of CT as the main reason for not following this practice. Other reasons include the 384 J Neurosurg: Pediatrics / Volume 14 / October 2014

14 Trends in childhood brain metastasis from solid primary cancers advanced disease status of most patients at brain metastasis diagnosis and the facts that many brain metastases are usually symptomatic and associated with poor outcome regardless of the therapy used. 13 In the absence of routine surveillance, changes in neurological status should alert these oncologists to the possible presence of a brain metastasis. Phase II studies of promising new agents in the pediatric population, particularly certain subsets of it, may be warranted. Conclusions From the findings of this study of brain metastases in children, we conclude that heightened awareness and prompt diagnosis of the primary cancer increase the chance of its identification before it can disseminate and may result in improved chances of survival by preventing complications such as brain metastasis. 3 Acknowledgments We thank David M. Wildrick, Ph.D., for editorial assistance and Stephanie Jenkins for assistance with manuscript preparation. Disclosure This work was supported by financial assistance from the Apache Corporation Brain Tumor Data Management Fund, a donor who had no direct involvement in this study. Author contributions to the study and manuscript preparation include the following. Conception and design: Suki, Khatua, Sawaya. Acquisition of data: Suki, Khoury Abdulla, Ding. Analysis and interpretation of data: all authors. Drafting the article: Suki. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Suki. Statistical analysis: Suki. Administrative/technical/material support: Sawaya. References 1. Bouffet E, Doumi N, Thiesse P, Mottolese C, Jouvet A, Lacroze M, et al: Brain metastases in children with solid tumors. Cancer 79: , Buzaid AC, Tinoco L, Ross MI, Legha SS, Benjamin RS: Role of computed tomography in the staging of patients with localregional metastases of melanoma. J Clin Oncol 13: , Davidoff AM, Cirrincione C, Seigler HF: Malignant melanoma in children. Ann Surg Oncol 1: , Deutsch M, Albo V, Wollman MR: Radiotherapy for cerebral metastases in children. Int J Radiat Oncol Biol Phys 8: , Deutsch M, Orlando S, Wollman M: Radiotherapy for metastases to the brain in children. Med Pediatr Oncol 39:60 62, Goldman S, Echevarría ME, Fangusaro J: Pediatric brain metastasis from extraneural malignancies: a review. Cancer Treat Res 136: , Graus F, Walker RW, Allen JC: Brain metastases in children. J Pediatr 103: , Kebudi R, Ayan I, Görgün O, Ağaoğlu FY, Vural S, Darendeliler E: Brain metastasis in pediatric extracranial solid tumors: survey and literature review. J Neurooncol 71:43 48, Kramer K, Kushner B, Heller G, Cheung NK: Neuroblastoma metastatic to the central nervous system. Cancer 91: , Marina NM, Pratt CB, Shema SJ, Brooks T, Rao B, Meyer WH: Brain metastases in osteosarcoma. Report of a long-term survivor and review of the St. Jude Children s Research Hospital experience. Cancer 71: , Marsa GW, Johnson RE: Altered pattern of metastasis following treatment of Ewing s sarcoma with radiotherapy and adjuvant chemotherapy. Cancer 27: , Miller RW, McKay FW: Decline in US childhood cancer mortality through JAMA 251: , Parasuraman S, Langston J, Rao BN, Poquette CA, Jenkins JJ, Merchant T, et al: Brain metastases in pediatric Ewing sarcoma and rhabdomyosarcoma: the St. Jude Children s Research Hospital experience. J Pediatr Hematol Oncol 21: , Paulino AC, Nguyen TX, Barker JL Jr: Brain metastasis in children with sarcoma, neuroblastoma, and Wilms tumor. Int J Radiat Oncol Biol Phys 57: , Postovsky S, Ash S, Ramu IN, Yaniv Y, Zaizov R, Futerman B, et al: Central nervous system involvement in children with sarcoma. Oncology 65: , Pratt CB, Palmer MK, Thatcher N, Crowther D: Malignant melanoma in children and adolescents. Cancer 47: , Rodriguez-Galindo C, Pappo AS, Kaste SC, Rao BN, Cain A, Jenkins JJ, et al: Brain metastases in children with melanoma. Cancer 79: , Shaw PJ, Eden T: Neuroblastoma with intracranial involvement: an ENSG Study. Med Pediatr Oncol 20: , Spunt SL, Walsh MF, Krasin MJ, Helton KJ, Billups CA, Cain AM, et al: Brain metastases of malignant germ cell tumors in children and adolescents. Cancer 101: , Stefanowicz J, Iżycka-Świeszewska E, Szurowska E, Bień E, Szarszewski A, Liberek A, et al: Brain metastases in paediatric patients: characteristics of a patient series and review of the literature. Folia Neuropathol 49: , Subramanian A, Harris A, Piggott K, Shieff C, Bradford R: Metastasis to and from the central nervous system the relatively protected site. Lancet Oncol 3: , Tasdemiroglu E, Patchell RA: Cerebral metastases in childhood malignancies. Acta Neurochir (Wien) 139: , Trigg ME, Makuch R, Glaubiger D: Actuarial risk of isolated CNS involvement in Ewing s sarcoma following prophylactic cranial irradiation and intrathecal methotrexate. Int J Radiat Oncol Biol Phys 11: , Vannucci RC, Baten M: Cerebral metastatic disease in childhood. Neurology 24: , Wagner AS, Fleitz JM, Kleinschmidt-Demasters BK: Pediatric adrenal cortical carcinoma: brain metastases and relationship to NF-1, case reports and review of the literature. J Neurooncol 75: , Westphal M, Heese O, de Wit M: Intracranial metastases: therapeutic options. Ann Oncol 14 Suppl 3:iii4 iii10, Weyl-Ben Arush M, Stein M, Perez-Nachum M, Dale J, Babilsky H, Zelnik N, et al: Neurologic complications in pediatric solid tumors. Oncology 52:89 92, 1995 Manuscript submitted June 21, Accepted July 10, Please include this information when citing this paper: published online August 15, 2014; DOI: / PEDS Address correspondence to: Dima Suki, Ph.D., Department of Neurosurgery, The University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd., Unit 442, Houston, TX dsuki@mdanderson.org. J Neurosurg: Pediatrics / Volume 14 / October