J Neurooncol (2014) 120:371–379 DOI 10.1007/s11060-014-1561-8

CLINICAL STUDY

Clinical outcomes of children and adults with central nervous system primitive neuroectodermal tumor Rachael A. Lester • Lindsay C. Brown • Laurence J. Eckel • Robert T. Foote • Amulya A. NageswaraRao • Jan C. Buckner • Ian F. Parney • Nicholas M. Wetjen Nadia N. Laack

•

Received: 28 January 2014 / Accepted: 21 July 2014 / Published online: 13 August 2014 Ó Springer Science+Business Media New York 2014

Abstract Central nervous system primitive neuroectodermal tumors (CNS PNETs) predominantly occur in children and rarely in adults. Because of the rarity of this tumor, its outcomes and prognostic variables are not well characterized. The purpose of this study was to evaluate clinical outcomes and prognostic factors for children and adults with CNS PNET. The records of 26 patients (11 children and 15 adults) with CNS PNET from 1991 to 2011 were reviewed retrospectively. Disease-free survival (DFS) and overall survival (OS) were estimated with the Kaplan– Meier method, and relevant prognostic factors were ana-

Presented at the 2013 Meeting of Pediatric Radiation Oncology Society, Louisville, Kentucky, May 7–11, 2013. R. A. Lester Mayo Medical School, Mayo Clinic College of Medicine, Rochester, MN, USA L. C. Brown
R. T. Foote
N. N. Laack (&) Department of Radiation Oncology, Mayo Clinic, 200 First St SW, Rochester, MN 55905, USA e-mail: [email protected] L. J. Eckel Department of Radiology, Mayo Clinic, Rochester, MN, USA A. A. NageswaraRao Division of Pediatric Hematology and Oncology, Mayo Clinic, Rochester, MN, USA J. C. Buckner Division of Medical Oncology, Mayo Clinic, Rochester, MN, USA I. F. Parney
N. M. Wetjen Department of Neurologic Surgery, Mayo Clinic, Rochester, MN, USA

lyzed. For the cohort, both the 5-year DFS and the OS were 46 %. For pediatric patients, the 5-year DFS was 78 %; for adult patients, it was 22 % (P = 0.004). Five-year OS for the pediatric and adult patients was 67 and 33 %, respectively (P = 0.07). With bivariate analysis including chemotherapy regimen (high dose vs. standard vs. nonstandard) or risk stratification (standard vs. high) and age, the increased risk of disease recurrence in adults persisted. A nonsignificant tendency toward poorer OS in adult patients relative to pediatric patients also persisted. High-dose chemotherapy with stem cell rescue was associated with a statistically significant improvement in OS and a tendency toward improved DFS, although the findings were mitigated when the effect of age was considered. Local recurrence was the primary pattern of treatment failure in both adults and children. Our results suggest that adult patients with CNS PNETs have inferior outcomes relative to the pediatric cohort. Further research is needed to improve outcomes for CNS PNET in populations of all ages. Keywords Central nervous system
Pediatric
Primitive neuroectodermal tumor
Risk factors
Survival Abbreviations CD Cluster of differentiation CCG Children’s Cancer Group CNS Central nervous system DFS Disease-free survival GTR Gross total resection HDCT High-dose chemotherapy HR Hazard ratio OS Overall survival PNET Primitive neuroectodermal tumor NF Neurofilament

123

372

Introduction

J Neurooncol (2014) 120:371–379 Table 1 Patient and treatment characteristics Characteristic

Primitive neuroectodermal tumors (PNETs) comprise a heterogeneous group of embryonic malignant tumors that affect both the central and peripheral nervous systems. Central nervous system (CNS) PNETs can arise in the cerebral hemispheres, pineal gland, brain stem, and spinal cord [1]. Although medulloblastomas were traditionally categorized as CNS PNETs, recent data from cytogenetic and molecular genetic analyses suggest that CNS PNETs are genetically distinct from medulloblastoma [2–5]. For example, the loss of CDKN2A/CDKN2B and gains of 19p are associated with CNS PNET only [4], whereas isochromosome 17q and monosomy of chromosome 6 have been reported in a subset of medulloblastomas but not in CNS PNETs [6]. Patients with CNS PNETs have a worse prognosis than those with diagnosed medulloblastomas; the difference in treatment response and outcome may be due to the molecular and genetic alterations that define these two distinct groups [6– 10]. Despite the increasingly evident differences between CNS PNET and medulloblastoma, patients with CNS PNETs are often treated with protocols designed for high-risk medulloblastoma patients. Furthermore, adult and pediatric CNS PNET patients are treated similarly, despite the paucity of knowledge regarding potential differences in tumor behavior and treatment outcomes between these two age groups. A recent study proposes that the clinical and radiological features of supratentorial PNET in adults are similar to those in children, suggesting that supratentorial PNET in adults might be a variant form of the pediatric tumor with delayed onset [11]. Little else has been described regarding CNS PNET similarities and differences among adults and children. The 3-to-5-year overall survival (OS) in pediatric patients with CNS PNETs is reported to be 40 to 50 %. Factors associated with poor prognosis in pediatric patients include young age, disseminated disease, and nonpineal location [9, 12–15]. Aside from the aforementioned study of supratentorial PNET [11], there is a dearth of information about survival and prognostic factors for adults with CNS PNET. Consequently, the overall prognostic implication of age (i.e. adult vs pediatric) in CNS PNET has yet to be determined. Further elucidation of the effect of age on outcomes is essential, both to allow more accurate definition of prognosis for adult and pediatric patients and to aid in determination of optimal treatment strategies.

Methods Patients With permission from the Mayo Clinic Institutional Review Board, the records of patients with histologically

123

Sex Male Female Primary tumor site Brain

Patients, no. (%) All (adult and pediatric)

Pediatrica

Adultb

26 (100)

11 (42)

15 (58)

13 (50)

7 (64)

6 (40)

13 (50)

4 (36)

9 (60)

24 (92)

10 (91)

14 (93)

Spinal cord

1 (4)

0 (0)

1 (7)

Unknown

1 (4)

1 (9)

0 (0)

M stage M0

21 (81)

8 (73)

13 (87)

M?

5 (19)

3 (27)

2 (13)

Risk stratification Standard risk

12 (46)

6 (55)

6 (40)

High risk

14 (54)

5 (45)

9 (60)

Chemotherapy Nonstandard

5 (19)

1 (9)

4 (27)

Standard

8 (31)

1 (9)

7 (47)

13 (50)

9 (82)

4 (27)

14 (54)

5 (45)

9 (60)

12 (46)

6 (55)

6 (40)

Biopsy

7 (27)

3 (27)

4 (27)

STR

7 (27)

2 (18)

5 (33)

GTR

12 (46)

6 (55)

6 (40)

No

12 (46)

9 (82)

3 (19)

Yes

14 (54)

2 (18)

12 (80)

10 (71)

High dose Radiation therapy Nonstandard Standard Surgical resection

Recurrence

Recurrence location Tumor bed

2 (100)

8 (67)

Brain, non-tumor bed

3 (21)

0 (0)

3 (25)

Spine, non-tumor bed

1 (7)

0 (0)

1 (8)

14 (54)

3 (27)

11 (73)

12 (46)

8 (73)

4 (27)

Death Dead Alive Treatment period (years) 1991–1999

6 (23)

2 (18)

4 (27)

2000–2011

20 (77)

9 (82)

11 (73)

GTR gross total resection, STR subtotal resection a

Age at diagnosis, \18 years

b

Age at diagnosis, C18 years

confirmed CNS PNETs seen at Mayo Clinic in Rochester, Minnesota, from 1991 to 2011 were retrospectively reviewed. A search of the tumor registry of those who authorized participation in research studies yielded 68 cases of PNETs. Forty-two were excluded on the basis of infratentorial location, pathologic characteristics of

J Neurooncol (2014) 120:371–379

glioblastomas, or incomplete information. No patient with infratentorial disease was included, to ensure that misdiagnosed medulloblastoma was not introduced into our population. A total of 26 patients were included in the study. Demographic data are reported in Table 1. Standard treatment of adult and pediatric CNS PNETs generally follows the high-risk medulloblastoma multimodality protocol, which includes maximal surgical resection, radiotherapy, and chemotherapy. M staging was graded with the following criteria: M0, no evidence of spread outside the local site; M1, positive cerebrospinal fluid cytologic testing; M2, infratentorial or supratentorial seeding; M3, spinal seeding; and M4, seeding outside the neuraxis [16]. In patients who did not undergo lumbar puncture to evaluate cerebrospinal fluid, whole-spine magnetic resonance imaging was used for staging. Patients were dichotomized into M0 and M? (i.e., M1, M2, M3, and M4) for prognostic analysis. Patients were also dichotomized into high- and standard-risk groups. Patients were considered to be at high risk if either their CNS PNET was M?, they were \3 years of age, or they had an incompletely resected primary tumor, with residual disease measuring [1.5 cm2. All other cases were considered to be standard risk. Imaging At diagnosis, brain and spine magnetic resonance imaging was performed, as well as a lumbar puncture in cases where the result would influence the treatment plan. Generally, preoperative magnetic resonance imaging showed characteristics previously described in PNETs—namely, a variety of hemispheric and midline, relatively large, well-circumscribed, heterogeneous but often partially cystic, enhancing tumors without substantial peritumoral edema [17, 18]. More specifically, on diffusion-weighted imaging, some degree of restricted diffusion was often present, a feature that helps to distinguish these highly cellular tumors from other brain tumor types [17] (Fig. 1). Midline tumors often showed hydrocephalus. Diffuse leptomeningeal enhancement was observed in 2 patients, compatible with subarachnoid tumor spread. Pathologic diagnosis Histopathologic classification of CNS PNET continues to be a challenge. Pathologic diagnosis of our samples was made on the basis of morphologic characteristics, with the aid of immunohistochemical staining and genetic profiling (Table 2). All CNS PNETs were diagnosed on the basis of the World Health Organization classification. Its current criteria identify supratentorial PNET as a Grade IV ‘‘embryonal tumour in the cerebrum or suprasellar region

373

composed of undifferentiated or poorly differentiated neuroepithelial cells which have the capacity for or display divergent differentiation along neuronal, astrocytic, ependymal, muscular or melanocytic lines’’ [19]. Each sample was analyzed by at least 2 expert neuropathologists at our institution, including the specimens of patients who had their initial diagnosis elsewhere. In addition to the stains listed in Table 2, immunohistochemistry profiling of C2 samples included stains for vimentin, keratin, neurofilament protein, desmin, cluster of differentiation (CD) 3, CD20, CD30, CD45, CD56, CD57, CD79a, CD99, terminal deoxynucleotide transferase, epithelial membrane antigen, neuron-specific enolase, neuronal nuclei, thyroid transcription factor-1, INI-1 (BAF47), a-fetoprotein, chromogranin, reticulin, and melan A. Cytogenetic testing of one or more samples included detection of p53, Ckit, 1p, and 16q. Eight patients’ samples were analyzed with electron microscopy. Treatment Surgery Maximum safe surgical resection of the primary tumor was performed. The extent of surgical resection was determined by postoperative magnetic resonance imaging and defined as gross total resection (GTR) (no visible residual tumor), subtotal resection (visible residual tumor remaining), or biopsy only. The product of the greatest perpendicular dimensions was used to quantify residual tumor for risk stratification. Radiation therapy Radiation therapy was individualized. Craniospinal irradiation was recommended for most patients, in accordance with the standard high-risk medulloblastoma protocol, followed by a radiotherapy boost to the tumor bed. The median radiation dose was 3,600 cGy (range 2,340–3,960 cGy) to the craniospinal axis in 180 cGy fractions 5 days a week, with boost to achieve a total of median dose of 5,580 cGy (range 3,800–6,000 cGy) to the tumor bed. Three adults received cranial irradiation only. One pediatric patient younger than 3 years did not receive radiotherapy because it was believed that potential toxicity of craniospinal radiation at such a young age outweighed its potential benefits. The other 22 patients received craniospinal irradiation. Chemotherapy Chemotherapy regimens were individualized, with selection based on M stage, age, and the standard treatment at

123

374

J Neurooncol (2014) 120:371–379

Fig. 1 Upper panels are representative magnetic resonance imaging of primitive neuroectodermal tumor of a 6-year-old boy evaluated for headache and hand tremor. Axial T2-weighted (a), fluid-attenuated inversion recovery (b), diffusion-weighted imaging (c), and postcontrast, fat-saturated T1 (d) images show a well-circumscribed mass in the midline posterior third ventricle, with restricted diffusion (asterisk in c) and heterogeneous enhancement (arrow in d) but no peritumoral edema. Lower panels are representative magnetic resonance imaging of primitive neuroectodermal tumor in a 64-year-old Table 2 Pathologic staining

Patients

woman with recent progressive headache, presenting with first-time seizure. Axial T2-weighted (e), fluid-attenuated inversion recovery (f), diffusion-weighted imaging (g), and post-contrast-medium, fatsaturated T1 (h) images show a well-circumscribed cystic and solid intra-axial mass in the left frontal lobe, with partial restricted diffusion (asterisk in g) in the region of mild enhancement (arrow in h) and minimal peritumoral edema (curved arrow in f)

Stains, ratio (%) Synaptophysin

GFAP

S-100

MiB-1

NF

CAM5.2

6/11 High (55)

2/11 (18)

0/11 (0)

Pediatric Positive

11/11 (100)

6/11 (55)

2/11 (18)

1/11 Moderate (9) Negative

0/11 (0)

4/11 (36)

2/11 (18)

0/11 (0)

4/11 (36)

3/11 (27)

Not reported

0/11 (0)

1/11 (9)

7/11 (64)

4/11 (36)

5/11 (45)

8/11 (73)

10/15 (67)

4/15 (27)

5/15 (33)

8/15 High (53)

1/15 (7)

0/15 (0)

Adults Positive

2/15 Moderate (13) 1/15 Low (7) GFAP glial fibrillary acidic protein, NF neurofilament

Negative

5/15 (33)

10/15 (67)

7/15 (47)

0/15 (0)

9/15 (60)

7/15 (47)

Not reported

0/15 (0)

1/15 (7)

3/15 (20)

4/15 (27)

5/15 (33)

8/15 (53)

time of diagnosis. Chemotherapy regimens were most commonly the standard medulloblastoma treatment combination of vincristine, cisplatin/carboplatin, and cyclophosphamide, or high-dose regimens followed by stem cell rescue per protocol ACNS0331, St. Jude Medulloblastoma (SJMB)-96 [20], Children’s Cancer Group [CCG]-99702, or CCG-99703. Both CCG protocols are high-dose regimens containing carboplatin and thiotepa. Nonstandard regimens included Ewing sarcoma-type regimens (e.g. adriamycin, ifosfamide) and temozolomide. All patients except one pediatric patient and 1 adult patient received

123

chemotherapy. These two patients were determined to be poor candidates because of rapidly progressing disease. Follow-up After completion of treatment, patients underwent clinical evaluation and magnetic resonance imaging of the brain and whole spine every 3–6 months for the first 2 years and annually thereafter. Toxicity was scored retrospectively in accordance with the Common Terminology Criteria for Adverse Events [21].

J Neurooncol (2014) 120:371–379

375

Statistical methods The Kaplan–Meier method was used to obtain 5-year estimates of disease-free survival (DFS) and OS with 95 % confidence interval. Univariate associations of variables with both DFS and OS were assessed using Cox proportional hazards regression model, with the results reported as hazard ratio (HR) with 95 % confidence interval. The data of patients dying without recurrence were censored for DFS at the time of death. A recurrence was defined as a distant relapse or local relapse, whichever came first. OS was calculated from the date of diagnosis to death, or last patient follow-up. The a level was set at 0.05 for statistical significance. Results Patient characteristics are listed in Table 2. The median age at diagnosis for the 26 patients was 20.4 years (range 2.8–64 years). Eleven patients were \18 years old and 15 were C18 years old. Median follow-up time for surviving patients and all patients was 8.7 years (range 1.6–21.7 years) and 2.5 years (range 0.5–21.7 years), respectively. For the entire patient population, 5-year DFS and OS were 46 % (95 % CI 28–72 %) and 46 % (95 % CI 29–72 %), respectively (Fig. 2). The 5-year DFS for pediatric and adult patients was 78 and 22 % (P = 0.004) (Fig. 3); the 5-year OS was 67 and 33 % (P = 0.07). These findings persisted when the population was dichotomized by \30 and C30 years of age, with older patients having poorer outcomes. When separated into three groups—age 0–17, 18–29, and C30 years—patients in the adolescent (17–29 years) and adult ( C30 years) groups had poorer outcomes relative to the pediatric (0–17 years) cohort. There were no statistical differences in outcome when the adolescent group and the adult group were compared. Thus, all subsequent analyses are presented for pediatric (\18 years) versus adult (C18 years) cohorts. All patients except 1 individual (aged 2.8 years) received radiotherapy, as previously described. Statistical relationships remained unchanged when this patient was excluded. Three adults received cranial radiation therapy only. The median radiation dose for the patients receiving craniospinal irradiation was 3,600 cGy (range 2,340–3,960 cGy) to the craniospinal axis and 5,580 cGy (range 3,800–6,000 cGy) to the tumor bed. All patients except 1 pediatric patient and 1 adult patient received chemotherapy, as previously described. The types of chemotherapy administered (standard, high dose, or nonstandard) differed significantly for pediatric, compared with adult, patients (P = 0.03), with more pediatric patients receiving high-dose chemotherapy (HDCT)

Fig. 2 Survival of all patients with central nervous system primitive neuroectodermal tumors. a Kaplan–Meier method estimates for disease-free survival (DFS). b Kaplan–Meier method estimates for overall survival (OS)

followed by stem cell rescue. The 5-year DFS with HDCT, standard chemotherapy, and nonstandard chemotherapy were 55, 25, and 60 %, respectively. Relative to treatment with standard chemotherapy, treatment with HDCT was associated with a nonsignificant tendency toward improved DFS (HR, 0.4; 95 % CI 0.1–1.1; P = 0.07). The 5-year OS with high-dose, standard, and nonstandard chemotherapy was 64, 25, and 40 %, respectively. Relative to treatment with standard chemotherapy, HDCT was associated with improved OS (HR, 0.3; 95 % CI 0.1–1.0; P = 0.05). Both adult and pediatric patients who received radiotherapy had late adverse effects attributed to radiotherapy. One pediatric patient had grade 3 retinopathy and grade 3 esophageal fibrosis and 1 adult patient had grade 3 radiation vasculopathy. Two pediatric patients had grade 2 growth retardation with radiation-induced scoliosis. One adult patient and one pediatric patient had grade 2 cataracts. Four pediatric patients and one adult patient had grade 2 hypothyroidism, and two pediatric patients had grade 2 panhypopituitarism. Two pediatric and three adult patients reported a decline in cognitive processing or memory; no formal neurocognitive testing was available.

123

376

Fig. 3 Survival of pediatric and adult patients with central nervous system primitive neuroectodermal tumors. a Kaplan–Meier method estimates for disease-free survival (DFS). b Kaplan–Meier method estimates for overall survival (OS)

Chemotherapy-related toxicity was most common in adult patients who received HDCT. Four adult patients received the high-dose treatment. One patient had grade 5 renal toxicity and a second had grade 5 pulmonary toxicity. A third patient had grade 4 venoocclusive disease, which necessitated chemotherapy cessation. Among the nine pediatric patients who received HDCT, one patient had grade 3 reversible renal toxicity and two patients had grade 4 bilateral ototoxicity. We analyzed previously reported prognostic factors and their association with the population as a whole and with each subgroup. Extent of resection showed a significant association with 5-year OS for the population as a whole but a nonsignificant association with 5-year DFS, with a 5-year OS for GTR vs non-GTR of 64 and 32 % (P = 0.05) and a 5-year DFS of 56 and 40 % (P = 0.18), respectively. Extent of surgical resection was not predictive of DFS or OS in the adult or pediatric subgroups. Risk stratification did not have a statistically significant association with DFS or OS. For the population as a whole, median DFS was 4.2 years for patients with standard risk and 1.2 years for patients with high risk, and median OS was 5.9 years for patients with standard risk and 1.8 years

123

J Neurooncol (2014) 120:371–379

for patients with high risk. M stage (M0 vs. M?), radiotherapy doses to the tumor bed (standard vs nonstandard), primary location (spine vs. brain or pineal vs. nonpineal sites), and year of treatment (before vs. after the year 2000) did not show a statistically significant correlation with DFS or OS for the population as a whole or either subgroup. Because of the small sample size, multivariate analysis was not possible. However, bivariate analyses were performed, including age and other potential confounding factors. The poorer DFS in adult patients persisted when additional variables were included in the model. When risk group (high vs standard) and extent of surgical resection (GTR vs non-GTR) were analyzed along with age, the DFS HR for an adult vs pediatric patient continued to be significant at 9.8 (95 % CI 2.5–44.8) and 9.2 (95 % CI 2.1–41.8), respectively. With the inclusion of the chemotherapy type (high dose vs. nonstandard vs standard), the DFS HR for adult vs pediatric patients was 8.2 (95 % CI 1.7–39.1). Similarly, the tendency for poorer OS in adults relative to pediatric patients persisted with bivariate modeling. Including risk status, chemotherapy regimen, and extent of surgical resection plus age, the HR for an adult patient was 3.1 (95 % CI 0.86–11.2), 2.4 (95 % CI 0.6–9.5), and 2.9 (95 % CI 0.80–10.4), respectively. Overall, 14 recurrences occurred, with 12 in adults and two in pediatric patients. Eight adult recurrences (67 %) developed in the tumor bed site, 3 (25 %) in the brain in a non–tumor bed location, and 1 (8 %) in the spine in a nontumor bed location. All recurrences in pediatric patients were in the tumor bed site. One adult patient and 1 pediatric patient are still alive after recurrence. At recurrence, patients were treated with various combinations of chemotherapy, surgery, and radiotherapy. Because local recurrence was the primary pattern of treatment failure, further analysis was undertaken to evaluate factors that may be associated with local failure. Of the 10 patients with local recurrence, 5 (50 %) had GTR and 5 (50 %) had non-GTR. One patient (10 %) received nonstandard chemotherapy; 5 (50 %), standard chemotherapy; and 4 (40 %), HDCT. Two recurrences (20 %) occurred with a radiation dose \5,580 cGy; 7 (70 %), a dose of 5,580 cGy; and 1 (10 %), a dose [5,580 cGy.

Discussion There is a paucity of data describing outcomes in adult patients versus pediatric patients diagnosed with CNS PNET. In our study, DFS was significantly poorer in the adult population than in the pediatric population. There also was a nonsignificant tendency toward poorer 5-year OS in the adult population. These differences persist when

J Neurooncol (2014) 120:371–379

risk stratification, chemotherapy type, and extent of surgical resection are taken into account. Therefore, this study suggests that adults diagnosed with CNS PNET have poorer outcomes than pediatric patients. Because of these different outcomes, it may be reasonable to consider treatment intensification for adults in order to improve outcomes. In addition, we reviewed other potential prognostic factors for the pediatric and adult patient subgroups. Current literature suggests that poor prognostic factors for pediatric patients with CNS PNETs include evidence of dissemination at diagnosis (M?), young age (\3 years), and non-pineal location [9, 12–15]. In the present study, M? status in a pediatric patient did not predict worse DFS or OS than pediatric patients with M0 stage. However, relatively few pediatric patients (three patients) of our population had M? disease, and therefore the power to detect differences was limited. Only one patient was younger than 3 years and only one had a pineal tumor, precluding meaningful analysis of the prognostic significance of age or pineal location. Because of the rarity of CNS PNETs, particularly in adults, there is a lack of data regarding outcomes and prognostic indicators for adults. We were not able to identify prognostic factors in the adult population because of the small number of patients in our analysis. Patients receiving HDCT showed a statistically significant improvement in OS and a tendency toward improved DFS. In bivariate analysis, however, this difference was mitigated by inclusion of age. Thus, the optimal chemotherapy regimen is still unclear, particularly in the adult population, in whom HDCT was poorly tolerated. Local recurrence was the predominant pattern of failure in our series, in keeping with other reports [13, 22–24]. It appears that our current CNS PNET regimen, in which nearly all patients receive trimodality therapy, is often inadequate in achieving primary tumor control. Possible strategies for improving each modality and its resulting contribution to local control must be entertained. Many studies suggest that complete resection is a good prognostic factor in children with CNS PNETs [9, 12, 13, 25–28]. We found a significant tendency toward improved OS with GTR and a nonsignificant tendency toward improved DFS in our population. It is widely recognized that this association may be due in part to confounding factors, such as smaller tumor size. However, it is accepted that when feasible and reasonably safe, total resection of tumor is ideal. Continued improvements in neurosurgical practice, such as increasingly sensitive imaging and stereotactic surgical techniques, may allow for further improvements in the percentage of patients who undergo GTR. Even with GTR, adjuvant therapy is paramount in improving local control, and potential for optimization of

377

radiotherapy and chemotherapy regimens must also be examined. Strategies that may be investigated to improve efficacy of radiotherapy include dose escalation and advances in concurrent synergistic systemic therapy. Improved radiotherapy efficacy must always be balanced with potential for toxicity, however. Particularly when considering radiation to the pediatric brain, neurotoxicity must be strongly considered, as radiation to the developing nervous system can have profound effects on future cognition and function [29, 30]. Considering variability in tumor behavior and response to treatment for pediatric vs adult populations, the concept of age-stratified protocols becomes increasingly attractive. For example, adult patients unable to tolerate HDCT regimens and less sensitive to the neurotoxic effects of radiation may be best treated with dose escalation to 6,000 cGy or greater to the primary site, with less intense chemotherapy. Pediatric patients, conversely, are better able to tolerate intensive chemotherapy regimens and more likely to have radiation-related neurotoxicity and thus may be better suited for increasingly intensive HDCT regimens with maintenance of the current standard radiation doses of 3,600 cGy to the craniospinal axis followed by a boost to the tumor bed to 5,580 cGy. Important limitations of our study include its retrospective nature and the lack of molecular characterization of included tumors. Recent studies suggest diagnostic and prognostic molecular markers for CNS PNET, including LIN28 and OLIG2, that we were unable to assess. Furthermore, CNS PNET is a poorly defined entity with considerable heterogeneity, and even after careful histopathologic reanalysis, it is difficult to define a typical profile for these tumors [31]. They pose a particular diagnostic challenge with respect to differentiating CNS PNETs from high-grade ‘‘small cell’’ glioblastoma variants, which may exhibit a similar phenotype [32]. Given the dismal prognosis of small cell glioblastoma, our pathologists make a concerted effort to define these cases appropriately, and thus we have a high degree of confidence that few or no such patients were included in our cohort [33]. Similarly, genome-wide analysis suggests significant genetic heterogeneity among CNS PNETs, although the significance of the impact of this genetic variability on clinical phenotype is unknown. However, given the ongoing evolution of CNS PNET classification, this study contributes valuable information regarding age-dependent prognosis in CNS PNET. We provided exhaustive detail about the diagnostic process of our patients so that clinicians can determine the relevance of these data to their own patients now and in the future as the diagnostic criteria continue to evolve. In conclusion, our study confirms the universally poor prognosis of patients with CNS PNET. We demonstrate that adult patients with CNS PNETs have inferior

123

378

J Neurooncol (2014) 120:371–379

outcomes relative to the pediatric cohort. The poor survival with this disease suggests that treatment intensification should be considered, particularly in adults. Radiation dose escalation to the primary tumor site could be considered in adult patients, and chemotherapy intensification in the form of HDCT with stem cell rescue could be further studied in children and younger adults who have favorable performance status and organ function. Current advances in molecular biology also may help to better stratify these tumors and improve current treatment approaches. Because of the rarity of these tumors in pediatric and, even more so, adult populations, further progress in characterizing and improving outcomes may be achieved only through large collaborative trials.

Conflict of interest

None.

References 1. Berger MS, Prados MD, Al-Mefty O (2005) Textbook of neurooncology. Elsevier Saunders, Philadelphia 2. Inda MM, Perot C, Guillaud-Bataille M, Danglot G, Rey JA, Bello MJ et al (2005) Genetic heterogeneity in supratentorial and infratentorial primitive neuroectodermal tumours of the central nervous system. Histopathology 47(6):631–637 3. McCabe MG, Ichimura K, Liu L, Plant K, Backlund LM, Pearson DM et al (2006) High-resolution array-based comparative genomic hybridization of medulloblastomas and supratentorial primitive neuroectodermal tumors. J Neuropathol Exp Neurol 65(6):549–561 4. Pfister S, Remke M, Toedt G, Werft W, Benner A, Mendrzyk F et al (2007) Supratentorial primitive neuroectodermal tumors of the central nervous system frequently harbor deletions of the CDKN2A locus and other genomic aberrations distinct from medulloblastomas. Genes Chromosomes Cancer 46(9):839–851 5. Li M, Lee KF, Lu Y, Clarke I, Shih D, Eberhart C et al (2009) Frequent amplification of a chr19q13.41 microRNA polycistron in aggressive primitive neuroectodermal brain tumors. Cancer Cell 16(6):533–546 Erratum in: Cancer Cell. 2010 Apr 13;17(4):413 6. Miller S, Rogers HA, Lyon P, Rand V, Adamowicz-Brice M, Clifford SC et al (2011) Genome-wide molecular characterization of central nervous system primitive neuroectodermal tumor and pineoblastoma. Neuro Oncol 13(8):866–879 7. Pizer BL, Weston CL, Robinson KJ, Ellison DW, Ironside J, Saran F et al (2006) Analysis of patients with supratentorial primitive neuro-ectodermal tumours entered into the SIOP/ UKCCSG PNET 3 study. Eur J Cancer 42(8):1120–1128 Epub 2006 Apr 24 8. Sung KW, Yoo KH, Cho EJ, Koo HH, do Koo H, Shin HJ et al (2007) High-dose chemotherapy and autologous stem cell rescue in children with newly diagnosed high-risk or relapsed medulloblastoma or supratentorial primitive neuroectodermal tumor. Pediatr Blood Cancer 48(4):408–415 9. Cohen BH, Zeltzer PM, Boyett JM, Geyer JR, Allen JC, Finlay JL et al (1995) Prognostic factors and treatment results for supratentorial primitive neuroectodermal tumors in children using radiation and chemotherapy: a Childrens Cancer Group Randomized Trial. J Clin Oncol 13(7):1687–1696

123

10. Russo C, Pellarin M, Tingby O, Bollen AW, Lamborn KR, Mohapatra G et al (1999) Comparative genomic hybridization in patients with supratentorial and infratentorial primitive neuroectodermal tumors. Cancer 86(2):331–339 11. Kim DG, Lee DY, Paek SH, Chi JG, Choe G, Jung HW (2002) Supratentorial primitive neuroectodermal tumors in adults. J Neurooncol 60(1):43–52 12. Reddy AT, Janss AJ, Phillips PC, Weiss HL, Packer RJ (2000) Outcome for children with supratentorial primitive neuroectodermal tumors treated with surgery, radiation, and chemotherapy. Cancer 88(9):2189–2193 13. Timmermann B, Kortmann RD, Kuhl J, Rutkowski S, Meisner C, Pietsch T et al (2006) Role of radiotherapy in supratentorial primitive neuroectodermal tumor in young children: results of the German HIT-SKK87 and HIT-SKK92 trials. J Clin Oncol 24(10):1554–1560 14. Chintagumpala M, Hassall T, Palmer S, Ashley D, Wallace D, Kasow K et al (2009) A pilot study of risk-adapted radiotherapy and chemotherapy in patients with supratentorial PNET. Neuro Oncol. 11(1):33–40 Epub 2008 Sep 16 15. Johnston DL, Keene DL, Lafay-Cousin L, Steinbok P, Sung L, Carret AS et al (2008) Supratentorial primitive neuroectodermal tumors: a Canadian pediatric brain tumor consortium report. J Neurooncol 86(1):101–108 Epub 2007 Jul 10 16. Chang CH, Housepian EM, Herbert C Jr (1969) An operative staging system and a megavoltage radiotherapeutic technic for cerebellar medulloblastomas. Radiology 93(6):1351–1359 17. Chawla A, Emmanuel JV, Seow WT, Lou J, Teo HE, Lim CC (2007) Paediatric PNET: pre-surgical MRI features. Clin Radiol 62(1):43–52 18. Dai AI, Backstrom JW, Burger PC, Duffner PK (2003) Supratentorial primitive neuroectodermal tumors of infancy: clinical and radiologic findings. Pediatr Neurol 29(5):430–434 19. Rorke LB, Hart MN, McLendon RE (2000) Supratentorial primitive neuroectodermal tumour (PNET). In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumours of the nervous system. International Agency for Research on Cancer, Lyon, p 141 20. Gajjar A, Chintagumpala M, Ashley D, Kellie S, Kun LE, Merchant TE et al (2006) Risk-adapted craniospinal radiotherapy followed by high-dose chemotherapy and stem-cell rescue in children with newly diagnosed medulloblastoma (St Jude Medulloblastoma-96): long-term results from a prospective, multicentre trial. Lancet Oncol 7(10):813–820. Erratum in: Lancet Oncol. 2006 Oct;7(10):797 21. National Institutes of Health, National Cancer Institute. Common Terminology Criteria for Adverse Events (CTCAE). [Internet] Version 4.0. NIH Publication No. 09-5410. Bethesda (MD): U. S. Department of Health and Human Services; 2009 May 28 [updated 2010 Jun 14; cited 2014 May 16]. Available from: http://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-14_ QuickReference_5x7.pdf 22. Friedrich C, von Bueren AO, von Hoff K, Gerber NU, Ottensmeier H, Deinlein F et al (2013) Treatment of young children with CNS-primitive neuroectodermal tumors/pineoblastomas in the prospective multicenter trial HIT 2000 using different chemotherapy regimens and radiotherapy. Neuro Oncol 15(2):224–234 Epub 2012 Dec 7 23. McBride SM, Daganzo SM, Banerjee A, Gupta N, Lamborn KR, Prados MD et al (2008) Radiation is an important component of multimodality therapy for pediatric non-pineal supratentorial primitive neuroectodermal tumors. Int J Radiat Oncol Biol Phys 72(5):1319–1323 Epub 2008 May 15 24. Fangusaro J, Finlay J, Sposto R, Ji L, Saly M, Zacharoulis S et al (2008) Intensive chemotherapy followed by consolidative myeloablative chemotherapy with autologous hematopoietic cell

J Neurooncol (2014) 120:371–379

25.

26.

27.

28.

29.

rescue (AuHCR) in young children with newly diagnosed supratentorial primitive neuroectodermal tumors (sPNETs): report of the Head Start I and II experience. Pediatr Blood Cancer 50(2):312–318 Albright AL, Wisoff JH, Zeltzer P, Boyett J, Rorke LB, Stanley P et al (1995) Prognostic factors in children with supratentorial (nonpineal) primitive neuroectodermal tumors: a neurosurgical perspective from the Children’s Cancer Group. Pediatr Neurosurg 22(1):1–7 Dirks PB, Harris L, Hoffman HJ, Humphreys RP, Drake JM, Rutka JT (1996) Supratentorial primitive neuroectodermal tumors in children. J Neurooncol 29(1):75–84 Yang HJ, Nam DH, Wang KC, Kim YM, Chi JG, Cho BK (1999) Supratentorial primitive neuroectodermal tumor in children: clinical features, treatment outcome and prognostic factors. Childs Nerv Syst 15(8):377–383 Mikaeloff Y, Raquin MA, Lellouch-Tubiana A, Terrier-Lacombe MJ, Zerah M, Bulteau C et al (1998) Primitive cerebral neuroectodermal tumors excluding medulloblastomas: a retrospective study of 30 cases. Pediatr Neurosurg 29(4):170–177 Merchant TE, Kiehna EN, Li C, Shukla H, Sengupta S, Xiong X et al (2006) Modeling radiation dosimetry to predict cognitive

379

30.

31.

32.

33.

outcomes in pediatric patients with CNS embryonal tumors including medulloblastoma. Int J Radiat Oncol Biol Phys 65(1):210–221 Epub 2006 Feb 10 Mabbott DJ, Barnes M, Laperriere N, Landry SH, Bouffet E (2007) Neurocognitive function in same-sex twins following focal radiation for medulloblastoma. Neuro Oncol. 9(4):460–464 Epub 2007 Aug 17 Picard D, Miller S, Hawkins CE, Bouffet E, Rogers HA, Chan TS et al (2012) Markers of survival and metastatic potential in childhood CNS primitive neuro-ectodermal brain tumours: an integrative genomic analysis. Lancet Oncol 13(8):838–848 Epub 2012 Jun 11 Li MH, Bouffet E, Hawkins CE, Squire JA, Huang A (2005) Molecular genetics of supratentorial primitive neuroectodermal tumors and pineoblastoma. Neurosurg Focus 19(5):E3 Perry A, Miller CR, Gujrati M, Scheithauer BW, Zambrano SC, Jost SC et al (2009) Malignant gliomas with primitive neuroectodermal tumor-like components: a clinicopathologic and genetic study of 53 cases. Brain Pathol 19(1):81–90 Epub 2008 Apr 29

123