SEARCH

SEARCH BY CITATION

Keywords:

  • supratentorial primitive neuroectodermal tumor;
  • primitive neuroectodermal tumors arising supratentorially (SPNET);
  • pineoblastoma;
  • chemotherapy;
  • radiation

Abstract

  1. Top of page
  2. Abstract
  3. CLINICAL MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

The outcome of a child with a primitive neuroectodermal tumors arising supratentorially (SPNET) is not well characterized and may differ from the outcome of a patient with a histologically similar cerebellar tumor (medulloblastoma [MB]). Recently, 5-year progression free survival rates as high as 80% have been reported for children with MB treated with craniospinal radiation (CRT) and chemotherapy including cisplatin, lomustine (CCNU), and vincristine (VCR).

METHODS

The authors reviewed the outcome of 22 consecutive patients age 3 years and older (mean age, 10 years; range, 3-18 years) with SPNET who were treated at the study institutions between 1981 and 1996. Tumor location included was 13 pineal, 6 cortical, and 3 thalamic or suprasellar. Five patients had disease dissemination at diagnosis. All patients underwent surgery and staging, followed by CRT and chemotherapy with cisplatin, CCNU, and VCR.

RESULTS

Of the 22 patients, 13 had developed disease progression and 10 had died at the time of last follow-up. Overall progression free survival (PFS) was 47% ± 11% at 3 years and 37% ± 11% at 5 years. There was a significant difference in PFS between patients with localized disease versus those with disseminated disease (P = 0.04). There was no statistical association between tumor location and survival. Although not significant (P = 0.21), there was a trend toward better survival of those patients with complete or near-complete resection compared with those with partial resection or biopsy.

CONCLUSIONS

The results of the current study demonstrate that the outcome for children with SPNET treated with radiation and chemotherapy appears worse than for children with MB treated with identical therapy. This suggests that there may be biologic differences between supratentorial and infratentorial primitive neuroectodermal tumors, thus requiring refinements in treatment. Cancer 2000;88:2189–93. © 2000 American Cancer Society.

Primitive neuroectodermal tumor (PNET) is the most common malignant brain tumor for children. This class of tumor is characterized by small, round, poorly differentiated blue cells. Primitive neuroectodermal tumor can occur in any location in the central nervous system but most frequently arises in the cerebellum in which it is commonly referred to as medulloblastoma. Medulloblastoma constitutes approximately 20% of brain tumors in children, whereas supratentorial primitive neuroectodermal tumors (SPNETs) account for 2.5% of pediatric brain tumors.1, 2 The more common supratentorial locations include the pineal gland and cerebral cortex. Tumors in these locations may be referred to as pineoblastoma (pineal gland) and cerebral medulloblastoma. Primitive neuroectodermal tumor also may arise in the basal ganglia, thalamus, or diencephalon. Because these tumors appear histologically identical, the treatment of SPNET has been extrapolated from the data on medulloblastoma. The outcome for children with SPNET generally has been reported to be poor and worse than medulloblastoma.3 There are, however, few reports utilizing systematic studies on the outcome for children with SPNET. It is therefore unclear if there are differences in outcome for patients with poor risk medulloblastoma and SPNET.

Currently, patients with medulloblastoma are broadly divided into two groups, average risk and poor risk, based on evidence of dissemination at diagnosis and degree of resection. Although survival rates vary from one series to another, even patients with poor risk disease have been reported to have good outcomes, with 5-year disease free survival reported to be 85% ± 6% in one study.4 These patients were treated with surgical resection, followed by craniospinal radiation with local boost to the tumor bed and adjuvant chemotherapy that included lomustine, cisplatin, and vincristine. We reviewed our experience with patients age 3 years and older with SPNET treated with surgery, radiation, and chemotherapy. The treatment regimen was identical to what was used to treat children with high risk medulloblastoma at our institutions.

CLINICAL MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. CLINICAL MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

The medical records for all patients with SPNET in the tumor registries at Children's National Medical Center and Children's Hospital of Philadelphia were reviewed. All patients age 3–18 years, whose tumors were diagnosed between 1981 and 1996, who met eligibility criteria were included in the study. There were 22 consecutive patients with SPNET treated at these institutions utilizing the same protocol that was used to treat poor risk medulloblastoma. These in-house protocols were identical and subsequently became a component of Children's Cancer Group medulloblastoma studies. Informed consent was obtained from each patient's parent or legal guardian before entry onto the study.

All patients underwent surgery for histologic diagnosis and removal of as much tumor as was possible. Extent of surgical resection was classified by the operating neurosurgeon as total, near total (greater than 90% but less than total), partial (50–90%), or biopsy. All patients underwent staging evaluation that included pre- and postoperative magnetic resonance imaging (MRI) of the brain, MRI of the spine, and lumbar puncture for cerebrospinal fluid cytology. Patients were considered to have disseminated disease at diagnosis if there was imaging evidence of disease anywhere in the neuroaxis other than the primary site and/or positive cerebrospinal fluid (CSF) cytology. Pathologic diagnosis of PNET was made by the pediatric pathologist at the institution of diagnosis.

Treatment Protocol

All patients underwent radiation therapy after initial recovery from surgery. Therapy was begun within 28 days of surgery. Twenty-one patients received 3400–4000 centigray (cGy) to the craniospinal axis. One patient received only 2300 cGy to the craniospinal axis because of young age at the time of diagnosis. After craniospinal radiation, all patients received a 1080- to 2000-cGy boost of radiation to the local tumor site (up to a total dose of 4680–6000 cGy). Radiation boost also was given to site of metastasis when indicated. Daily fractions of 180 cGy were used.

The chemotherapy consisted of three agents: vincristine, lomustine (CCNU), and cisplatin. Patients were given vincristine, at a dose of 1.5 mg/sq m (up to a dose of 2 mg), weekly during radiation treatment. After 6 weeks of rest, patients were given a regimen of CCNU at 75 milligrams per metered square (mg/sq m), cisplatin at 68 mg/sq m every 6 weeks, and vincristine at 1.5 mg/sq m (up to 2 mg) weekly for 3 consecutive weeks. Eight 6-week cycles were planned.

A contrast enhanced computed tomography or MRI study of the brain was performed every 3 months (after every 2 cycles of therapy) while the patient was receiving treatment and at 6-month intervals thereafter for the first 3 years after treatment. In patients with metastatic disease at diagnosis, cerebrospinal fluid cytology and spinal neuroimaging were performed 6 weeks after finishing radiation treatment and every 3 months until there was complete disappearance of disseminated disease. In patients without disseminated disease at diagnosis, cerebrospinal fluid examination and spinal neuroimaging were performed if there was clinical suspicion of leptomeningeal tumor recurrence.

Patient Population

Between 1981 and 1996, 22 consecutive patients diagnosed with SPNET were treated with the above regimen. Patients were between age 3 and 18 years with a mean of 10 years, at the time of diagnosis. Tumor location included 13 pineal, 6 cortical, 2 thalamic, and 1 suprasellar. Five of 22 patients had tumor dissemination at the time of diagnosis. Ten patients underwent complete or near complete resection. No patient in this group had dissemination at diagnosis. Nine patients underwent partial resection, and 3 underwent biopsy only.

Statistical Analysis

Patients were examined for time to tumor progression and overall survival. Progression free survival rate was measured from the date of initial diagnosis to the date of progressive disease or last contact. Overall survival rate was measured from the time of diagnosis to death or last contact. Distributions of progression free survival and overall survival were estimated using the technique of Kaplan and Meier product limit method and compared using the log rank test.

RESULTS

  1. Top of page
  2. Abstract
  3. CLINICAL MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

There are 9 of 22 (40.9%) patients presently alive and free of progressive disease. All of these patients had a complete response to treatment. Kaplan–Meier estimation of progression free survival rate at 3 years was 47% ± 11% and at 5 years was 37% ± 11% (Fig. 1). The overall survival rate at 3 years is 59% ± 11% and at 5 years is 53% ± 11% (Fig. 2).

thumbnail image

Figure 1. Kaplan–Meier curve for progression free survival.

Download figure to PowerPoint

thumbnail image

Figure 2. Kaplan–Meier curve for overall survival.

Download figure to PowerPoint

There are 13 (59.1%) patients who developed progressive disease during the study period. Ten (45.5%) of these patients died of progressive disease. All five patients who presented with tumor dissemination have died of their disease. The remaining three patients are alive despite disease progression and have been treated with other regimens.

Progression free survival and overall survival were not adversely affected by tumor location. The 5-year progression free survival rate for patients with pineal tumors was 38% compared with 33% for those with cortical or other supratentorial tumors (P = 0.95) (Fig. 3). The 5-year progression free survival for patients who underwent complete or near complete resections was 53% compared with 25% for those who underwent partial resection or biopsy (P = 0.22) (Fig. 4). Patients with disseminated disease at the time of diagnosis had a 5-year progression free survival rate of 0% compared with 49% for those patients with localized disease at diagnosis (P = 0.04) (Fig. 5).

thumbnail image

Figure 3. Kaplan–Meier curve by location of disease.

Download figure to PowerPoint

thumbnail image

Figure 4. Kaplan–Meier curves by degree of resection.

Download figure to PowerPoint

thumbnail image

Figure 5. Kaplan–Meier survival curves by dissemination of disease.

Download figure to PowerPoint

One patient who had a complete response to treatment and is alive developed pre B cell acute lymphocytic leukemia (pre-B ALL) in 1995 that was treated with bone marrow transplant (BMT). He subsequently developed colonic/rectal carcinoma in 1997 that was treated with surgery and radiation. Right hemispheric glioblastoma multiforme was diagnosed in 1998 and was treated with surgery and chemotherapy.

DISCUSSION

  1. Top of page
  2. Abstract
  3. CLINICAL MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Survival rates for children with supratentorial PNET seem to be worse than for children with medulloblastoma. Comparing these two groups is problematic because of the lack of an accepted, easily applicable staging system for SPNET compared with the TNM staging system routinely used for medulloblastoma and the relative infrequency of SPNET as compared with the incidence of medulloblastoma. Furthermore, the definition of high risk medulloblastoma has varied some between series as has the treatment strategy used. The 5-year progression free survival rate of 37% and overall survival rate of 53% for the 22 patients in this series seems worse than the progression free survival of 85% reported by Packer et al. for children with poor risk medulloblastoma treated with identical therapy.4 In the later series, even 15 patients with metastatic disease had a 5-year progression free survival of 67% ± 15%.

In one of the only other published series as far as we know of patients with SPNET, Cohen et al. reported a 3-year progression free survival rate of 45% for children age 18 months to 21 years, treated with surgery, radiation, and chemotherapy.5 These results are similar to the 3-year progression free survival rate of 47% in our series. In a recently published series by Yang, 27 patients treated with SPNET treated with surgery, chemotherapy, and/or radiation had a 3-year progression free survival of 60%, but outcome at 5 years fell to 38%.6 In Cohen et al.'s series, there was no difference in outcome for children treated with one of two chemotherapy regimens that included either CCNU, vincristine, and prednisone or eight-in-one chemotherapy. Their series did report a significantly better outcome for children with pineal PNET. Neither our group nor Dirks et al. found improved outcome for patients with pineal PNET3. In fact, in the study reported here, there was no difference in outcome based on tumor location. These results may be limited because there were relatively few patients in the study groups. There was also no statistical difference in outcome based on extent of surgery. There appears, however, to be a trend toward better outcome (P = 0.22) for children with completely resected or near completely resected tumor, and our analysis may be limited by statistical power. Dissemination at diagnosis was clearly a significant indicator of poor outcome.

It is not clear why children with SPNET seem to do worse than those with medulloblastoma. It would make some intuitive sense that SPNET occurs in locations that often preclude safe complete surgical resection. But even when the tumor is completely or near completely resected, the 5-year progression free survival in our series was 53%. All patients in this group had localized disease at diagnosis. It is also not clear why survival rates for patients who present with disseminated disease are so poor. Although it is possible that these patients have a more aggressive or advanced form of the disease, no consistent histologic or genetic differences have been found between disseminated and local disease in SPNET.

Although medulloblastoma and SPNET appear identical histologically, there may be biologic differences between them. Analyses of medulloblastoma have shown loss of chromosome arm 17p as the most frequent genetic abnormality. This abnormality was not found in a series of eight SPNET.7 In a different series, investigators found a tendency toward a poor outcome (P = 0.1125) for patients with medulloblastoma or SPNET who were positive for N-myc expression.8 These investigators did not separate out those patients with SPNET so it is unclear whether their tumors are more likely to express N-myc than medulloblastoma. This series looked at a total of 19 tumors. A larger series will need to be examined to confirm that N-myc expression is a poor prognostic factor in SPNET.

More recent work by Russo et al. indicates that different genetic aberrations do occur in SPNET as compared with medulloblastoma9 Using comparative genomic hybridization, 53 PNETs were analyzed to determine copy number aberrations. There were 10 SPNETs analyzed, and all had copy number aberrations. Loss of 14q was detected in 4 of 10 SPNETs but not in any of the medulloblastoma cases. Loss of 19q was detected in 4 of 10 SPNETs but only in 1 of the medulloblastoma cases. The most common copy number aberration in medulloblastoma (16 of 43 cases) was again gain of chromosome 17q, which was not found in any SPNET case.

These biologic results and the clinical difference found in our review indicate that SPNET and medulloblastoma appear to be distinct entities despite their similar histologic appearance. A continued understanding of the biologic underpinnings of SPNET, and how they differ from medulloblastoma, may not only explain their more aggressive nature, but also lead to refinements in treatment that are clearly needed.

REFERENCES

  1. Top of page
  2. Abstract
  3. CLINICAL MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  • 1
    Gaffney CC, Sloan, JP, Bradley NJ, Bloom HJ. Primitive neuroectodermal tumours of the cerebrum. Pathology and treatment. J Neurooncol 1985; 3: 2333.
  • 2
    Bruno RL, Norris DG. Primitive neruoectodermal tumors of infancy and childhood. In: HumphreyGB, DehnerLP, GrindeyGB, editors. Pediatric oncology. vol. 1. Boston: Nijhoff, 1981: 2657.
  • 3
    Dirks PB, Harris L, Hoffman HJ, Humphreys RP, Drake JM, Rutka JT. Supratentorial primitive neuroectodermal tumors in children. J Neurooncol 1996; 29: 7584.
  • 4
    Packer RJ, Sutton LN, Elterman R, Lange B, Goldwein J, Nicholson HS, et al. Outcome for children with medulloblastoma treated with radiation and cisplatin, CCNU, and vincristine chemotherapy. J Neurosurg 1994; 81: 6908.
  • 5
    Cohen BH, Zeltzer PM, Boyett JM, Geyer JR, Allen JC, Finlay, et al. Prognostic factors and treatment results for supratentorial primitive neuroectodermal tumors in children using radiation and chemotherapy: a Children's Cancer Group randomized trial. J Clin Oncol 1995; 13: 168796.
  • 6
    Yang HJ, Nam DH, Wang KC, Kim YM, Chi JG, Cho BK. Supratentorial primitive neuroectodermal tumor in children: clinical features, treatment outcome and prognostic factors. Childs Nerv Syst 1999; 15: 37783.
  • 7
    Burnett ME, White EC, Sih S, von Haken MS, Cogen PH. Chromosome arm 17p deletion analysis reveals molecular genetic heterogeneity in supratentorial and infratentorial primitive neuroectodermal tumors of the central nervous system. Cancer Genet Cytogenet 1997; 97: 2531.
  • 8
    Moriuchi S, Shimizu K, Miyao Y, Hayakawa T. An immunohistochemical analysis of medulloblastoma and PNET with emphasis on N-myc protein expression. Anticancer Res 1996; 16: 268792.
  • 9
    Russo C, Pellarin M, Tingby O, Bollen AW, Lamborn KR, Mohapatra G, et al. Comparative genomic hybridization in patients with supratentorial and infratentorial primitive neuroectodermal tumors. Cancer 1999; 86: 3319.