Impact of H3.3 K27M Mutation on Prognosis and Survival of Grade IV Spinal Cord Glioma on the Basis of New 2016 World Health Organization Classification of the Central Nervous System

Impact of H3.3 K27M Mutation on Prognosis and Survival of Grade IV Spinal Cord Glioma on the... Abstract BACKGROUND Spinal cord glioma grade IV is a rare, diffuse midline glioma. H3 K27M-mutant was classified in a different entity in the 2016 World Health Organization (WHO) classification recently. No reports about prognosis of spinal cord glioma grade IV are available yet. OBJECTIVE To analyze the prognostic factors for spinal cord glioma grade IV. METHODS Twenty-five patients with spinal cord glioma of grade IV who underwent surgery in a single institute were selected. All grade IV spinal cord glioma histologically confirmed as glioblastoma or “diffuse midline glioma with H3 K27M-mutant” by the 2016 WHO classification of the central nervous system were included. Basic demographics, treatment modalities, and pathological tumor molecular profiles were investigated for prognosis. RESULTS Mean age was 39.1 yr; male to female ratio was 18 : 7. Tumor was located in thoracic cord (53.3%), cervical cord (40%), and lumbar area (6.7%). Median overall survival was 37.1 mo; median disease-free survival was 18.5 mo. Treatment modality showed no statistical difference. Only K27M profile showed significant prognostic value, 20 patients (80%) showed K27M mutation positive, K27M mutation patients showed longer overall survival (40.07 mo) than K27M negative patients (11.63 mo, P < .0001), and disease-free survival (20.85 vs 8.72 mo, P = .0241). CONCLUSION This study is the first and largest report of the prognosis of primary spinal cord grade IV glioma using the new WHO classification. This study reported survival analysis and prognostic factors, and revealed that H3.3 K27M mutation is not a major poor prognostic factor. Further studies to explore K27M mutations needed for risk stratification and therapy optimization. H3 K27M, Grade IV, Prognosis, Spinal cord, Glioma, Survival analysis, WHO classification ABBREVIATIONS ABBREVIATIONS BSA Body Surface Area CCRT concurrent chemoradiation therapy CNS central nervous system DFS disease-free survival FISH fluorescent in situ hybridization GBM glioblastoma multiforme IHC immunohistochemistry OS overall survival TMZ temozolomide WHO World Health Organization Intramedullary astrocytoma constitutes 6% to 8% of all primary spinal cord tumors.1 Primary spinal glioblastomas compose 7.5% of all intramedullary gliomas.2,3 Fewer than 200 cases of spinal glioblastoma were ever reported in the literature. A few studies enrolling fewer than 10 patients have been published because of the rarity of cases of spinal glioblastoma.4,5 The small number of cases in the literature about spinal cord glioblastoma and its survival analysis, the prognostic factors on disease-free survival (DFS) and overall survival (OS), made spinal cord glioma to be understood poorly. The optimal surgical management of high-grade intramedullary spinal cord astrocytoma continues to be controversial. The effect of the extent of surgery of low-grade spinal cord astrocytoma on the prognosis is inconsistent. 6-9 For malignant astrocytoma, the survival analysis with the extent of surgery remains largely unstudied.7 TABLE 1. Summary of Recent Literature on Primary Spinal Cord Glioblastoma Study design  Author  Source  Year  Study design and period  No of GBM  OS (mo)  Good Px  Poor Px  No effect  Single institute  Yana- madala39  J Clin Neurosci  2016  Single institute (MGH, USA, 1990-2015)  GBM 6  No report    Ryu40  Eur Spine J  2015  Single institute (Gangnam Severance, Korea, 1983-2014)  GBM 6  12  Histologic lower grade        Babu41  Spine  2014  Single institute (Duke Univ, USA, 1992-2012)  GBM 9  12.1    High Gr, dissemination Span of level  Resection    McGirt31  Neuro- surgery  2008  Single institute (New York Univ, USA, 1990-2002)  GBM 8  9  Radical resection      Meta-analysis  Beyer42  PLOS ONE  2016  Meta-analysis (Web of Science & Pubmed 1938-2015)  GBM 155  9  Age < 60, Tx in 1980-2015    GTR, gender, location    Konar43  World Neurosurg  2016  Meta-analysis (Pubmed 1938-2015)  GBM 128  11  Age 18-65 surgery + RTx&CTx  Radical surgery + RTx disseminate      Hernández- Durán44  J Clin Neurosci  2015  Meta-analysis (MeSH & pubmed 1970-2014)  GBM 64  16 or 10      TMZ use    Adams10  Spine  2012  SEER Population-based analysis 1973-2007  GBM 59  10  AA, adult, male, radical resection        Santi11  Cancer  2003  AFIP database 1962-2000  GBM 23  10    Age > 40  Gender, gr,Tx option, resection  Study design  Author  Source  Year  Study design and period  No of GBM  OS (mo)  Good Px  Poor Px  No effect  Single institute  Yana- madala39  J Clin Neurosci  2016  Single institute (MGH, USA, 1990-2015)  GBM 6  No report    Ryu40  Eur Spine J  2015  Single institute (Gangnam Severance, Korea, 1983-2014)  GBM 6  12  Histologic lower grade        Babu41  Spine  2014  Single institute (Duke Univ, USA, 1992-2012)  GBM 9  12.1    High Gr, dissemination Span of level  Resection    McGirt31  Neuro- surgery  2008  Single institute (New York Univ, USA, 1990-2002)  GBM 8  9  Radical resection      Meta-analysis  Beyer42  PLOS ONE  2016  Meta-analysis (Web of Science & Pubmed 1938-2015)  GBM 155  9  Age < 60, Tx in 1980-2015    GTR, gender, location    Konar43  World Neurosurg  2016  Meta-analysis (Pubmed 1938-2015)  GBM 128  11  Age 18-65 surgery + RTx&CTx  Radical surgery + RTx disseminate      Hernández- Durán44  J Clin Neurosci  2015  Meta-analysis (MeSH & pubmed 1970-2014)  GBM 64  16 or 10      TMZ use    Adams10  Spine  2012  SEER Population-based analysis 1973-2007  GBM 59  10  AA, adult, male, radical resection        Santi11  Cancer  2003  AFIP database 1962-2000  GBM 23  10    Age > 40  Gender, gr,Tx option, resection  MGH, Massachusetts General Hospital; SEER, Surveillance, Epidemiology and End Results; AFIP, Armed Forces Institute of Pathology database; OS. overall median survival. View Large TABLE 1. Summary of Recent Literature on Primary Spinal Cord Glioblastoma Study design  Author  Source  Year  Study design and period  No of GBM  OS (mo)  Good Px  Poor Px  No effect  Single institute  Yana- madala39  J Clin Neurosci  2016  Single institute (MGH, USA, 1990-2015)  GBM 6  No report    Ryu40  Eur Spine J  2015  Single institute (Gangnam Severance, Korea, 1983-2014)  GBM 6  12  Histologic lower grade        Babu41  Spine  2014  Single institute (Duke Univ, USA, 1992-2012)  GBM 9  12.1    High Gr, dissemination Span of level  Resection    McGirt31  Neuro- surgery  2008  Single institute (New York Univ, USA, 1990-2002)  GBM 8  9  Radical resection      Meta-analysis  Beyer42  PLOS ONE  2016  Meta-analysis (Web of Science & Pubmed 1938-2015)  GBM 155  9  Age < 60, Tx in 1980-2015    GTR, gender, location    Konar43  World Neurosurg  2016  Meta-analysis (Pubmed 1938-2015)  GBM 128  11  Age 18-65 surgery + RTx&CTx  Radical surgery + RTx disseminate      Hernández- Durán44  J Clin Neurosci  2015  Meta-analysis (MeSH & pubmed 1970-2014)  GBM 64  16 or 10      TMZ use    Adams10  Spine  2012  SEER Population-based analysis 1973-2007  GBM 59  10  AA, adult, male, radical resection        Santi11  Cancer  2003  AFIP database 1962-2000  GBM 23  10    Age > 40  Gender, gr,Tx option, resection  Study design  Author  Source  Year  Study design and period  No of GBM  OS (mo)  Good Px  Poor Px  No effect  Single institute  Yana- madala39  J Clin Neurosci  2016  Single institute (MGH, USA, 1990-2015)  GBM 6  No report    Ryu40  Eur Spine J  2015  Single institute (Gangnam Severance, Korea, 1983-2014)  GBM 6  12  Histologic lower grade        Babu41  Spine  2014  Single institute (Duke Univ, USA, 1992-2012)  GBM 9  12.1    High Gr, dissemination Span of level  Resection    McGirt31  Neuro- surgery  2008  Single institute (New York Univ, USA, 1990-2002)  GBM 8  9  Radical resection      Meta-analysis  Beyer42  PLOS ONE  2016  Meta-analysis (Web of Science & Pubmed 1938-2015)  GBM 155  9  Age < 60, Tx in 1980-2015    GTR, gender, location    Konar43  World Neurosurg  2016  Meta-analysis (Pubmed 1938-2015)  GBM 128  11  Age 18-65 surgery + RTx&CTx  Radical surgery + RTx disseminate      Hernández- Durán44  J Clin Neurosci  2015  Meta-analysis (MeSH & pubmed 1970-2014)  GBM 64  16 or 10      TMZ use    Adams10  Spine  2012  SEER Population-based analysis 1973-2007  GBM 59  10  AA, adult, male, radical resection        Santi11  Cancer  2003  AFIP database 1962-2000  GBM 23  10    Age > 40  Gender, gr,Tx option, resection  MGH, Massachusetts General Hospital; SEER, Surveillance, Epidemiology and End Results; AFIP, Armed Forces Institute of Pathology database; OS. overall median survival. View Large Previously reported spinal cord glioma case series have reported that the OS of patients with spinal cord glioblastoma is about 10 mo (Table 1), compared with the better survival of 14 mo of brain astrocytoma.10,11 Recent developments in genomic studies have enabled researchers to determine the molecular profile and characteristics of glioma to figure out biological behavior and categorize them into similar groups from an etiological perspective. Researchers suggested that some of the newly identified mutations are associated with tumor subtype and location of tumor and patient age. Sturm et al12 reported that “the H3F3A K27M mutation was predominantly identified in malignant astrocytoma arising in the midline structure such as the thalamus, brain stem, and spinal cord and is mainly prevalent in pediatric and young adult patients.”12 As the results of molecular analysis in the midline structure, the diffuse midline glioma, H3 K27M-mutant, was newly categorized as a separate pathological entity in the 2016 WHO classification.13 Several articles have reported diffuse midline glioma, H3 K27M-mutant in the midline structure of brain including the thalamus and brain stem in adults; however, diffuse midline glioma, H3 K27M-mutant cases in spinal cord are rarely reported.14 In addition, the genetic and molecular profiles of diffuse midline glioma, H3 K27M-mutant in spinal cord remain unclear due to a lack of relevant studies. The purpose of this study was to perform survival analysis of spinal cord grade IV glioma on the basis of new 2016 WHO classification of central nervous system (CNS) tumor by elucidating the association of the molecular profiles with the prognosis and to determine the role of H3 K27M mutant in prediction of the OS. This study is the first report of the prognosis of spinal cord grade IV glioma on the basis of the new 2016 WHO classification of central nervous system tumors and includes the largest numbers of patients from a single institute over 18 yr. METHODS Patient Population Twenty-five patients who were surgically treated for spinal cord glioma of grade IV, from 1998 to 2016 in a single institute, were selected for study, and general demographics are shown in Table 2. TABLE 2. Demographics of Patients Variable    n  %  Mean age (years ± SD)    39.1 ± 16.2    Gender  Total  25      Male  7  28    Female  18  72  Extent of surgery          Biopsy  3  12    PR  4  16    STR  9  36    GTR  9  36  Treatment          Surgery  3  12    Surgery + RTx  6  24    Surgery + CTx  1  4    Surgery + RTx + CTx  15  60  Variable    n  %  Mean age (years ± SD)    39.1 ± 16.2    Gender  Total  25      Male  7  28    Female  18  72  Extent of surgery          Biopsy  3  12    PR  4  16    STR  9  36    GTR  9  36  Treatment          Surgery  3  12    Surgery + RTx  6  24    Surgery + CTx  1  4    Surgery + RTx + CTx  15  60  PR, partial resection; STR, subtotal resection; GTR, gross total resection; RTx, radiotherapy; CTx, chemotherapy. View Large TABLE 2. Demographics of Patients Variable    n  %  Mean age (years ± SD)    39.1 ± 16.2    Gender  Total  25      Male  7  28    Female  18  72  Extent of surgery          Biopsy  3  12    PR  4  16    STR  9  36    GTR  9  36  Treatment          Surgery  3  12    Surgery + RTx  6  24    Surgery + CTx  1  4    Surgery + RTx + CTx  15  60  Variable    n  %  Mean age (years ± SD)    39.1 ± 16.2    Gender  Total  25      Male  7  28    Female  18  72  Extent of surgery          Biopsy  3  12    PR  4  16    STR  9  36    GTR  9  36  Treatment          Surgery  3  12    Surgery + RTx  6  24    Surgery + CTx  1  4    Surgery + RTx + CTx  15  60  PR, partial resection; STR, subtotal resection; GTR, gross total resection; RTx, radiotherapy; CTx, chemotherapy. View Large All grade IV spinal cord gliomas that were histologically confirmed as glioblastoma as well as diffuse midline glioma with H3 K27M-mutant by the 2016 WHO classification of the central nervous system13 were included in this study. Patient demographics—radiological, surgical, and pathological—and clinical follow-up outcomes were analyzed. Demographic factors including age and gender were identified. The patient's symptoms, signs, and the neurological examination findings were gathered as preoperative records of patients. The tumor factors including size, location, and number of spanning spinal cord levels were identified. The surgical factors included the extent of resection and adjuvant chemotherapy and radiation therapy. The postoperative factors included neurological status, recurrence pattern, localized or distant metastasis, complication, time to recurrence, and survival. The dates of death were determined from patient's medical records. Moreover, the characteristics of grade IV spinal cord glioma were compared with brain glioblastoma multiforme (GBM) on the basis of the same, single institutional brain tumor database. This study was approved by institutional review board in our institute, waiving patient written informed consent as the retrospective study (4-2017-1251). Treatment Modality Extent of surgery was defined as gross total resection (≥90%), subtotal resection (≥50% and <90%), partial resection (<50%), or open biopsy. Patients’ chemotherapy regimen history was identified as well from medical records. Postoperative adjuvant radiation therapy involving local area or whole neuro-axis was identified for individual patients. The chemotherapy regimen and radiation therapy followed a standard protocol of glioma treatment in brain. Most patients started concurrent chemoradiation therapy (CCRT) with using Stupp regimen, daily 75 mg/Body Surface Area (BSA) of temozolomide (TMZ) plus 45 to 54 Gy of radiotherapy at a dose of 1.8 Gy per fraction given once a day, 5 d per week over 5 to 6 wk, within 3 wk of surgery.15 Twenty-eight days following the completion of CCRT, TMZ was started; 6 cycles on every 4 wk were administered unless adverse events or progression was detected. The first cycle regimen is 150 mg/BSA(m2) for 5 d; the second to sixth cycle regimen is 200 mg/BSA for 5 d in each cycle. Radiation volume included tumor bed, based on preoperative and postoperative magnetic resonance imaging, and 2 to 5 cm longitudinal margins. In cases of leptomeningeal seeding, the entire spinal cord or craniospinal irradiation was chosen. In cases of recurrence beyond 6 mo after initial treatment, reoperation was considered a priority when the tumor was resectable. Maximal safe resection was attempted using the method used for initial surgery. Repeat radiotherapy was again recommended after surgery or unresectable recurrent tumor. Adjuvant TMZ administration, using the same regimen (with initial treatment), was also considered following repeat radiotherapy. If the tumor recurred within 6 mo after initial treatment or re-recurred, bevacizumab/iriotecan was considered as salvage treatment. Prognostic Factors for Survival Basic demographics including gender, age, tumor location, and tumor spanning length were investigated. Extent of surgery, radiotherapy, and chemotherapy were investigated for treatment modality. Pathological tumor molecular profiles including MGMT methylation, IDH1 mutation, K27M, 1p/19q del, and EGFR amplification were identified and analyzed. Pathological Diagnoses All cases were reviewed (by 2 neuropathologists, SHK and JC). Specifically, pilocytic astrocytoma and ependymoma were excluded. All cases were graded according to the 2016 World Health Organization (WHO) classification, as WHO grade II, III, or IV.16 H3.3 K27M mutation was examined by immunohistochemistry (IHC) (Figures 1 and 2) and pyrosequeuncing. The pyrosequencing for identification of H3.3 gene (H3F3A K27M and G34V) was preformed according to previously described methods.17 IHC with a Ventana BenchMark XT autostainer (Ventana Medical System Inc, Tucson, Arizona) according to the protocol was performed. The antibodies used were Anti-Histone H3 Antibody, K27M mutant (ABE 419, 1 : 300 dilution, Millipore, Burlington, Massachusetts), and Anti-Histone H3 Antibody, K27me3 (07-449, 1 : 300 dilution, Millipore). When the distinct nuclear staining of K27M mutant and loss of K27me3 were identified in glioma cells, these cells were regarded as “mutant”/“positive.” From pyrosequencing results, there was no G34V mutation. All cases showing K27M mutation in pyrosequencing were positive in IHC. K27M mutant was regarded as diffuse midline glioma, K27M mutant (the 2016 WHO grade IV), regardless of histological grade according to new WHO classification. FIGURE 1. View largeDownload slide Diffuse astrocytoma and K27M. Two diffuse astrocytomas (A and B, H-E ×200) show the same histological grade but different H3.3 K27M immunoreactivity patterns (C and D, K27M × 200). The K27M mutant protein strongly expresses in the tumor cell nuclei in C. FIGURE 1. View largeDownload slide Diffuse astrocytoma and K27M. Two diffuse astrocytomas (A and B, H-E ×200) show the same histological grade but different H3.3 K27M immunoreactivity patterns (C and D, K27M × 200). The K27M mutant protein strongly expresses in the tumor cell nuclei in C. FIGURE 2. View largeDownload slide Glioblastoma and K27M. Two glioblastomas show the same histological patterns (A and B, H-E ×200) and H3.3 K27M immunoreactivity patterns (C and D K27M ×200). FIGURE 2. View largeDownload slide Glioblastoma and K27M. Two glioblastomas show the same histological patterns (A and B, H-E ×200) and H3.3 K27M immunoreactivity patterns (C and D K27M ×200). Additional molecular profiles, such as 1p/19q co-deletion, IDH mutation, MGMT hypermethylation, EGFR amplification, and H3.3 K27M mutation, were examined. The 1p/19q co-deletion was examined by Fluorescent in situ hybridization (FISH) (1p/19q fluorescent probe kit, Vysis, Abbott Molecular, Chicago, Illinois), as previously described.18 The assessment and interpretation of FISH results were performed according to the guidelines defined by the SIOP Europe Neuroblastoma Pathology and Biology and Bone Marrow Group.19 More than 100 nonoverlapping nuclei were enumerated per hybridization for each probe. Tumor cells that showed more than 30% of nuclei with DNA loss were identified as chromosomal loss. IDH1 mutation was examined by IHC. With a representative section, we performed IHC with a Ventana BenchMark XT autostainer (Ventana Medical System Inc) according to the protocol. The antibodies were anti-human IDH1 R132H mouse monoclonal (Clone H09L, 1 : 80 dilution, Dianova, Hamburg, Germany). When the cytoplasmic expression of the IDH1 R132H was identified in glioma cells, we regarded those cases as “mutant”/“positive.” The DNA methylation status of CpG islands at the MGMT promoter was determined by methylation-specific polymerase chain reaction as previously described, with some modifications.20,21 EGFR amplification was examined by FISH (EGFR/CEP7 dual color probe kit, Vysis, Abbott Molecular), as described previously.22 Specimens were considered to be amplified when more than 10% of tumor cells exhibited either an EGFR/CEP7 ratio of >2 or innumerable tight clusters of signals of the locus probe. Statistical Analysis Student t-tests and chi-square tests were used for individual variables. Kaplan-Meier curves estimated time to OS, DFS, and recurrence, and were compared using the log-rank test. We conducted only univariable log-rank test, because the number of samples was not enough. Two-sided P values smaller than .05 were considered to be significant. Statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, North Carolina). RESULTS Patients Demographics General demographics of 25 patients who were surgically treated for primary spinal cord glioma of grade IV from 1998 to 2016 in a single institute are shown in Table 2. Mean age was 39.1 yr (39.1 ± 16.2), and mean follow-up duration was 26.4 mo. Male to female was 18:7, male was predominant. The number of expired patients during follow-up and investigation was 13 (52%). Long-term survivor (survival longer than 3 yr) numbers were 6 patients, and 5 patients among them expired during follow-up. The longest survivor was a 52-yr-old male patient, histological subtype was anaplastic astrocytoma with K27M (+), and he has survived 94.3 mo until now without recurrence. Tumor Characteristics The total number of segments involved by tumor was 75 segments. The most frequent location is location of tumor C5, 6, 7, and T12 (Figure 3). Tumor was located in thoracic cord (53.3%), cervical cord (40%), and lumbar area (6.7%) in order of frequency. Average length of span by tumor was 3 segments. FIGURE 3. View largeDownload slide Location and involved segments of grade IV spinal cord glioma. Tumor was located in the thoracic cord (53.3%), cervical cord (40%), and lumbar area (6.7%) in order of frequency. FIGURE 3. View largeDownload slide Location and involved segments of grade IV spinal cord glioma. Tumor was located in the thoracic cord (53.3%), cervical cord (40%), and lumbar area (6.7%) in order of frequency. Treatment Strategy Extent of surgery was identified as the following (Table 2): gross total resection in 9 patients (36%), subtotal resection in 9 patients (36%), partial resection in 4 patients (16%), and open biopsy in 3 patients (12%). Extent of surgery did not show statistical differences in survival (Table 3). Adjuvant treatment was identified (Table 2): radiation therapy with chemotherapy, 15 patients (60%); radiation therapy only, 6 patients (24%); chemotherapy only, 1 patient (4%); surgery only, 3 patients (12%). Adjuvant radiotherapy or chemotherapy showed no statistical significance in improving survival (Table 3). The numbers of chemotherapy cycles and radiation therapy were diverse depending on the patient's survival, metastasis, recurrence pattern, and general condition, providing a heterogeneous treatment protocol in our series despite the small number. TABLE 3. Statistical Analysis for Prognostic Factors. Factors    OS (mo)  P  DFS (mo)  P  Gender      .7727    .3441    Male  37.05    19.23      Female  51.34    10.62    Extent of surgery      .6719    .0025    GTR  29.74    14.82      STR  48.87    40.07      PR  37.05    23.84      Bx  51.34    18.49    RTx      .9801    .9388    Yes  37.05    19.97      No  16.13    11.04    CTx      .1016    .9831    Yes  29.74    16.95      No  Not reached    19.34    K27M mutation      <.0001    .0241    Positive  40.07    20.85      Negative  11.63    8.72    Factors    OS (mo)  P  DFS (mo)  P  Gender      .7727    .3441    Male  37.05    19.23      Female  51.34    10.62    Extent of surgery      .6719    .0025    GTR  29.74    14.82      STR  48.87    40.07      PR  37.05    23.84      Bx  51.34    18.49    RTx      .9801    .9388    Yes  37.05    19.97      No  16.13    11.04    CTx      .1016    .9831    Yes  29.74    16.95      No  Not reached    19.34    K27M mutation      <.0001    .0241    Positive  40.07    20.85      Negative  11.63    8.72    OS, overall median survival; DFS, disease-free survival; PR, partial resection; STR, subtotal resection; GTR, gross total resection. View Large TABLE 3. Statistical Analysis for Prognostic Factors. Factors    OS (mo)  P  DFS (mo)  P  Gender      .7727    .3441    Male  37.05    19.23      Female  51.34    10.62    Extent of surgery      .6719    .0025    GTR  29.74    14.82      STR  48.87    40.07      PR  37.05    23.84      Bx  51.34    18.49    RTx      .9801    .9388    Yes  37.05    19.97      No  16.13    11.04    CTx      .1016    .9831    Yes  29.74    16.95      No  Not reached    19.34    K27M mutation      <.0001    .0241    Positive  40.07    20.85      Negative  11.63    8.72    Factors    OS (mo)  P  DFS (mo)  P  Gender      .7727    .3441    Male  37.05    19.23      Female  51.34    10.62    Extent of surgery      .6719    .0025    GTR  29.74    14.82      STR  48.87    40.07      PR  37.05    23.84      Bx  51.34    18.49    RTx      .9801    .9388    Yes  37.05    19.97      No  16.13    11.04    CTx      .1016    .9831    Yes  29.74    16.95      No  Not reached    19.34    K27M mutation      <.0001    .0241    Positive  40.07    20.85      Negative  11.63    8.72    OS, overall median survival; DFS, disease-free survival; PR, partial resection; STR, subtotal resection; GTR, gross total resection. View Large Survival Analysis Overall Survival and Disease-Free Survival Median OS was 37.1 mo. Survival at 1 and 5 yr was 82.3% and 18.6%, respectively (Figure 4). Median DFS was 18.5 mo; survival at 1 and 5 yr was 64% and 4%, respectively (Figure 4). FIGURE 4. View largeDownload slide Survival analysis. Median OS was 37.1 mo, survival at 1 and 5 yr was 82.3% and 18.6%, respectively. Median DFS was 18.5 mo, survival at 1 and 5 yr was 64% and 4%, respectively. FIGURE 4. View largeDownload slide Survival analysis. Median OS was 37.1 mo, survival at 1 and 5 yr was 82.3% and 18.6%, respectively. Median DFS was 18.5 mo, survival at 1 and 5 yr was 64% and 4%, respectively. Prognostic Factors Male gender showed a longer survival than female gender in OS and DFS, but had no statistical significance. Extent of surgery and extent of tumor involvement did not reveal any prognostic importance. Radiotherapy and chemotherapy showed heterogeneous results without statistical significance, and did not show any effect on prognosis and OS. Patients with positive tumor molecular profile of K27M showed significant longer survival than negative K27M patients (Figure 5, Table 3). FIGURE 5. View largeDownload slide OS analysis by K27M status. Positive K27M mutation patients showed longer OS (40.07 mo) than K27M negative patients (11.63 mo, P < .0001) FIGURE 5. View largeDownload slide OS analysis by K27M status. Positive K27M mutation patients showed longer OS (40.07 mo) than K27M negative patients (11.63 mo, P < .0001) Tumor Molecular Profile Molecular profiles, such as 1p/19q co-deletion, IDH mutation, MGMT hypermethylation, EGFR amplification, and H3.3 K27M mutation, were examined (Table 4). Among the molecular profiles, only the K27M profile showed prognostic value. In 25 patients with grade IV spinal cord glioma, 20 patients (80%) showed K27M mutation positive and 5 patients (20%) showed negative. Positive K27M mutation patients showed longer OS (40.07 mo) (11.63 mo, P < .0001) and DFS (20.85 vs 8.72 mo, P = .0241) than negative K27M mutation patients (Figure 5) in grade IV category. TABLE 4. Characteristics of Molecular Profile. Molecular profile  n    n  %  H3.3 K27M mutation  25  Positive  20  80      Negative  5  20  MGMT hypermethylation  25  Methylation  1  4      Unmethylation  24  96  IDH1 mutation  25  Wild  25  100      Mutation  0  0  1p/19q co-deletion  25  Yes  0  0      No  25  100  EGFR amplification  25  Yes  1  4      no  24  96  Molecular profile  n    n  %  H3.3 K27M mutation  25  Positive  20  80      Negative  5  20  MGMT hypermethylation  25  Methylation  1  4      Unmethylation  24  96  IDH1 mutation  25  Wild  25  100      Mutation  0  0  1p/19q co-deletion  25  Yes  0  0      No  25  100  EGFR amplification  25  Yes  1  4      no  24  96  View Large TABLE 4. Characteristics of Molecular Profile. Molecular profile  n    n  %  H3.3 K27M mutation  25  Positive  20  80      Negative  5  20  MGMT hypermethylation  25  Methylation  1  4      Unmethylation  24  96  IDH1 mutation  25  Wild  25  100      Mutation  0  0  1p/19q co-deletion  25  Yes  0  0      No  25  100  EGFR amplification  25  Yes  1  4      no  24  96  Molecular profile  n    n  %  H3.3 K27M mutation  25  Positive  20  80      Negative  5  20  MGMT hypermethylation  25  Methylation  1  4      Unmethylation  24  96  IDH1 mutation  25  Wild  25  100      Mutation  0  0  1p/19q co-deletion  25  Yes  0  0      No  25  100  EGFR amplification  25  Yes  1  4      no  24  96  View Large When we analyzed the impact of K27M mutation within the histological grade 4 (glioblastoma), similar results were obtained. K27M mutant with histological grade 4 (glioblastoma) showed longer OS (20.85 mo) than K27M negative patients (11.64 mo, P = .0276) and DFS (21.18 vs 8.72 mo, P = .0758) than negative K27M mutation patients (Figure 6). The OS of K27M positive in GBM was 87.5% (1 yr) and 15% (5 yr). K27M negative in GBM showed survival as 40% (1 yr) and 0% (5 yr). FIGURE 6. View largeDownload slide OS analysis by K27M status in glioblastoma (histological grade IV). K27M mutant with histological grade 4 (glioblastoma) showed longer OS (20.85 mo) than K27M negative patients (11.64 mo, P = .0276). FIGURE 6. View largeDownload slide OS analysis by K27M status in glioblastoma (histological grade IV). K27M mutant with histological grade 4 (glioblastoma) showed longer OS (20.85 mo) than K27M negative patients (11.64 mo, P = .0276). Available profile of 1p/19q co-deletion, MGMT methylation, and IDH1 mutation were shown in Table 4 and Figure 7. All of the molecular profiles were analyzed. Characteristics of grade IV glioma molecular profile from available data were uniformity in IDH 1 negative, no 1p/19q co deletion. Most patients had MGMT unmethylation except one who showed hypermethylation (Figure 7). FIGURE 7. View largeDownload slide A map illustrating patient demographics and molecular profiles. (Location: C, cervical; CT, cervicothoracic; T, thoracic; TL, thoracolumbar; L, lumbar; Extent of surgery 1, open biopsy; 2, partial resection; 3, subtotal resection; 4, gross total resection.) FIGURE 7. View largeDownload slide A map illustrating patient demographics and molecular profiles. (Location: C, cervical; CT, cervicothoracic; T, thoracic; TL, thoracolumbar; L, lumbar; Extent of surgery 1, open biopsy; 2, partial resection; 3, subtotal resection; 4, gross total resection.) DISCUSSION Intramedullary spinal cord tumors are infrequent tumors, only accounting for 2% to 4% of all central nervous system origin tumors.1,11 Helseth et al23 reported that “astrocytoma and ependymoma comprise the majority of intramedullary spinal cord tumors, they primarily occur in different patient populations.”23 Among the intramedullary spinal cord tumor, astrocytoma is the most common tumor found in young age, comprising 59% to 90%,1,12 whereas ependymoma is most often found in adults.1,24 High-grade or malignant tumor composed 7% to 25% of all spinal cord astrocytoma in young age and 10% to 30% of spinal cord astrocytoma in adults.25-31 Spinal cord astrocytoma most frequently involves the cervical spinal cord, mainly the lower cervical segment, followed by the thoracic spine.29,32,33 The consensus on the treatment for spinal cord glioblastoma is not clearly determined because of the rarity of cases and limited research. The standard treatment of spinal cord glioblastoma includes gross total resection or biopsy followed by radiotherapy and chemotherapy.31,34 New 2016 WHO classification made a change in standard treatment for glioma. We have 6 patients of radiation only group. Among them, 4 patients had diffuse astrocytoma (H3 K27M-mutant) and 1 patient had anaplastic astrocytoma (H3 K27M-mutant). These patients followed standard protocol (surgery and radiation therapy only) for grade II and III glioma, before the 2016 WHO classification grouped them to grade IV. The remaining 1 patient with glioblastoma (H3 K27M mutant) suffered from upper gastrointestinal perforation after initial surgery and could not receive chemotherapy. The major concern after establishing new WHO CNS tumor classification is the validation of new the classification supported by clinical results and prognosis. The major changes in the 2016 WHO classification of CNS tumors include “major restructuring of diffuse gliomas with incorporation of genetically defined entities, addition of newly recognized entities with variants and patterns (IDH-wild-type and IDH-mutant glioblastoma, Diffuse midline glioma, H3 K27M-mutant).”13 Solomon et al14 recently reported 47 patients (50%) with diffuse midline gliomas, H3 K27M mutation among a total of 94 CNS tumor patients, including 9 patients (53%) with spinal cord glioma with H3 K27M mutation among a total of 17 spinal cord glioma tumor patients. Our study revealed that 51.3% of all spinal cord gliomas showed K27M mutation; additionally, more frequent K27M mutation was found in higher histological grade (Table 5). K27M mutation status of all 39 cases according to the 2007 WHO classification was shown in Table 5. Shadowed cells are grade IV tumor according to the 2016 WHO classification. Ten cases (light gray cells) that had diffuse and anaplastic astrocytoma histology were reclassified as grade IV in the 2016 WHO classification. TABLE 5. K27M Mutation Status of All 39 Cases According to the 2007 WHO Classification.   K27M mutation status    WHO grade by histology  K27M+  K27M−  Total  % K27M+  Diffuse astrocytoma  4a  8  12  33.3  Anaplastic astrocytoma  6a  6  12  50  Glioblastoma  10b  5b  15  66.7  Total  20  19  39  51.3    K27M mutation status    WHO grade by histology  K27M+  K27M−  Total  % K27M+  Diffuse astrocytoma  4a  8  12  33.3  Anaplastic astrocytoma  6a  6  12  50  Glioblastoma  10b  5b  15  66.7  Total  20  19  39  51.3  Shadowed cells are grade IV tumor according to the 2016 WHO classification. aTen cases that had diffuse and anaplastic astrocytoma histology were reclassified as grade IV in 2016 WHO classification. bFifteen cases of glioblastoma are also grade IV under 2007 and 2016 WHO classification. View Large TABLE 5. K27M Mutation Status of All 39 Cases According to the 2007 WHO Classification.   K27M mutation status    WHO grade by histology  K27M+  K27M−  Total  % K27M+  Diffuse astrocytoma  4a  8  12  33.3  Anaplastic astrocytoma  6a  6  12  50  Glioblastoma  10b  5b  15  66.7  Total  20  19  39  51.3    K27M mutation status    WHO grade by histology  K27M+  K27M−  Total  % K27M+  Diffuse astrocytoma  4a  8  12  33.3  Anaplastic astrocytoma  6a  6  12  50  Glioblastoma  10b  5b  15  66.7  Total  20  19  39  51.3  Shadowed cells are grade IV tumor according to the 2016 WHO classification. aTen cases that had diffuse and anaplastic astrocytoma histology were reclassified as grade IV in 2016 WHO classification. bFifteen cases of glioblastoma are also grade IV under 2007 and 2016 WHO classification. View Large Xiao et al35 reported that the age at surgery was the only significant prognostic factor of spinal cord astrocytoma in a 13-patient series,35 though they analyzed molecular prognostic factors but they did not include K27M mutation in the study. Our results showed advanced results with no significant molecular prognostic factors except K27M mutation (Table 4). Although the K27M mutation was frequently observed in adult brainstem and thalamic gliomas, this mutation was associated with a poorer prognosis in brainstem gliomas but not in thalamic gliomas. These differences in the adult brainstem and thalamic gliomas may result from the differences in genetic aberrations, and may provide location-specific molecular targets for treatment strategy.36 It is unclear whether spinal cord astrocytoma could share genetic alterations with supratentorial glioma. Because the spinal cord has to be considered a midline structure, a high incidence of H3 K27M mutations in spinal cord glioma could be hypothesized. The spinal high-grade glioma in children and adults frequently showed H3 K27M mutations. Though the homogeneity of the molecular profile was revealed in our series such as all IDH 1 negative, no 1p/19q codeletion, only 1 case of methylation, Cheng et al37 recently reported methylation in 8 of 14 patients, IDH 1 mutation in 1/14, EGFR overexpression in 11/14 in their glioblastoma study. A large grade IV glioma cohort is required to conclude on molecular profile characteristics.37 Limitations The present study has several limitations, including its retrospective design, which may introduce selection bias. Though this study is a large cohort of spinal cord gliomas, there are limitations to the number available, because of the rarity of the disease, and power to assess enough variables. The number of chemotherapy cycles and radiation therapy were diverse depending on the patient's survival, metastasis, recurrence pattern, and general condition, to present a heterogeneous treatment protocol in our series despite the small number. Additionally, comparison between 20 patients with H3 K27M mutations and 5 patients with no mutations does suggest importance but is limited by only 5 patients in the poorly performing cohort. Further studies should be focused on investigation of the possible association of H3F3AK27M mutations with patient survival and used for therapy optimization and the prediction of outcomes.38 In the brain, there are many known prognostic markers such as IDH, MGMT, and 1p/19q-codeletion. However, those molecular changes are known to be very rare in spinal cord. That turned out to be true to a certain degree in our study. It may reflect that the biology of glioma differs independent of “location.” Further studies with more cases are needed to answer these questions. CONCLUSION This study is the first report of the prognosis of primary spinal cord grade IV glioma on the basis of the new WHO classification, and includes the largest numbers of patients in a single institute over 18 yr. This study performed integrative survival analysis, elucidating the association of the various prognostic factors, and revealed that H3.3 K27M mutation is not a major poor prognostic factor, unlike in brain glioma. Further studies are needed to explore K27M mutations’ use for prediction of outcome and treatment strategy. Disclosures This study was supported by a faculty research grant of Yonsei University College of Medicine for 2017 (6-2017-0067), the National Research Foundation of Korea (NRF) grant funded by the Korea government (NRF-2014M3A7B4051596) and the Technology Innovation Program funded by the Ministry of Trade, Industry, & Energy, Korea (No. 10062712, Development of spinal fusion implant and its manufacturing system; the functionality optimized, patient-customized in terms of bioactive materials to meet the clinical needs). The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. REFERENCES 1. Reimer R, Onofrio BM. Astrocytomas of the spinal cord in children and adolescents. J Neurosurg . 1985; 63( 5): 669- 675. Google Scholar CrossRef Search ADS PubMed  2. Raco A, Esposito V, Lenzi J, Piccirilli M, Delfini R, Cantore G. Long-term follow-up of intramedullary spinal cord tumors: a series of 202 cases. Neurosurgery . 2005; 56( 5): 972- 981; discussion 972-981. Google Scholar PubMed  3. Miller DC. Surgical pathology of intramedullary spinal cord neoplasms. J Neurooncol . 2000; 47( 3): 189- 194. Google Scholar CrossRef Search ADS PubMed  4. Gee TS, Ghani AR, Idris B, Awang MS. Case report: a rare case of pediatric conus medularis glioblastoma multiforme. Med J Malaysia . 2012; 67( 4): 438- 441. Google Scholar PubMed  5. Morais N, Mascarenhas L, Soares-Fernandes JP, Silva A, Magalhaes Z, Costa JA. Primary spinal glioblastoma: a case report and review of the literature. Oncol Lett . 2013; 5( 3): 992- 996. Google Scholar CrossRef Search ADS PubMed  6. Hardison HH, Packer RJ, Rorke LB, Schut L, Sutton LN, Bruce DA. Outcome of children with primary intramedullary spinal cord tumors. Child's Nerv Syst . 1987; 3( 2): 89- 92. Google Scholar CrossRef Search ADS   7. Minehan KJ, Shaw EG, Scheithauer BW, Davis DL, Onofrio BM. Spinal cord astrocytoma: pathological and treatment considerations. J Neurosurg . 1995; 83( 4): 590- 595. Google Scholar CrossRef Search ADS PubMed  8. Samii M, Klekamp J. Surgical results of 100 intramedullary tumors in relation to accompanying syringomyelia. Neurosurgery . 1994; 35( 5): 865- 873; discussion 873. Google Scholar CrossRef Search ADS PubMed  9. Sandler HM, Papadopoulos SM, Thornton AF Jr, Ross DA. Spinal cord astrocytomas: results of therapy. Neurosurgery . 1992; 30( 4): 490- 493. Google Scholar PubMed  10. Adams H, Avendano J, Raza SM, Gokaslan ZL, Jallo GI, Quinones-Hinojosa A. Prognostic factors and survival in primary malignant astrocytomas of the spinal cord: a population-based analysis from 1973 to 2007. Spine . 2012; 37( 12): E727- E735. Google Scholar CrossRef Search ADS PubMed  11. Santi M, Mena H, Wong K, Koeller K, Olsen C, Rushing EJ. Spinal cord malignant astrocytomas. Clinicopathologic features in 36 cases. Cancer . 2003; 98( 3): 554- 561. Google Scholar CrossRef Search ADS PubMed  12. Sturm D, Witt H, Hovestadt V et al.   Hotspot mutations in H3F3A and IDH1 define distinct epigenetic and biological subgroups of glioblastoma. Cancer Cell . 2012; 22( 4): 425- 437. Google Scholar CrossRef Search ADS PubMed  13. Louis DN, Perry A, Reifenberger G et al.   The 2016 World Health Organization classification of tumors of the central nervous system: a summary. Acta Neuropathol . 2016; 131( 6): 803- 820. Google Scholar CrossRef Search ADS PubMed  14. Solomon DA, Wood MD, Tihan T et al.   Diffuse midline gliomas with histone H3-K27M mutation: a series of 47 cases assessing the spectrum of morphologic variation and associated genetic alterations. Brain Pathol . 2016; 26( 5): 569- 580. Google Scholar CrossRef Search ADS PubMed  15. Stupp R, Mason WP, van den Bent MJ et al.   Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med . 2005; 352( 10): 987- 996. Google Scholar CrossRef Search ADS PubMed  16. Komori T. The 2016 WHO classification of tumours of the central nervous system: the major points of revision. Neurol Med Chir (Tokyo) . 2017; 57( 7): 301- 311. Google Scholar CrossRef Search ADS PubMed  17. Gielen GH, Gessi M, Hammes J, Kramm CM, Waha A, Pietsch T. H3F3A K27M mutation in pediatric CNS tumors: a marker for diffuse high-grade astrocytomas. Am J Clin Pathol . 2013; 139( 3): 345- 349. Google Scholar CrossRef Search ADS PubMed  18. Woehrer A, Sander P, Haberler C et al.   FISH-based detection of 1p 19q codeletion in oligodendroglial tumors: procedures and protocols for neuropathological practice - a publication under the auspices of the Research Committee of the European Confederation of Neuropathological Societies (Euro-CNS). Clin Neuropathol . 2011; 30( 3): 47- 55. Google Scholar CrossRef Search ADS PubMed  19. Ambros PF, Ambros IM, Siop Europe Neuroblastoma Pathology B, Bone Marrow G. Pathology and biology guidelines for resectable and unresectable neuroblastic tumors and bone marrow examination guidelines. Med Pediatr Oncol . 2001; 37( 6): 492- 504. Google Scholar CrossRef Search ADS PubMed  20. Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia. Cancer Res . 1999; 59( 4): 793- 797. Google Scholar PubMed  21. Esteller M, Toyota M, Sanchez-Cespedes M et al.   Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is associated with G to A mutations in K-ras in colorectal tumorigenesis. Cancer Res . 2000; 60( 9): 2368- 2371. Google Scholar PubMed  22. Zhao LL, Xu KL, Wang SW, Hu BL, Chen LR. Pathological significance of epidermal growth factor receptor expression and amplification in human gliomas. Histopathology . 2012; 61( 4): 726- 736. Google Scholar CrossRef Search ADS PubMed  23. Helseth A, Mork SJ. Primary intraspinal neoplasms in Norway, 1955 to 1986. A population-based survey of 467 patients. J Neurosurg . 1989; 71( 6): 842- 845. Google Scholar CrossRef Search ADS PubMed  24. Milano MT, Johnson MD, Sul J et al.   Primary spinal cord glioma: a Surveillance, Epidemiology, and End Results database study. J Neurooncol . 2010; 98( 1): 83- 92. Google Scholar CrossRef Search ADS PubMed  25. Bouffet E, Pierre-Kahn A, Marchal JC et al.   Prognostic factors in pediatric spinal cord astrocytoma. Cancer . 1998; 83( 11): 2391- 2399. Google Scholar CrossRef Search ADS PubMed  26. Cooper PR. Outcome after operative treatment of intramedullary spinal cord tumors in adults: intermediate and long-term results in 51 patients. Neurosurgery . 1989; 25( 6): 855- 859. Google Scholar CrossRef Search ADS PubMed  27. Cristante L, Herrmann HD. Surgical management of intramedullary spinal cord tumors: functional outcome and sources of morbidity. Neurosurgery . 1994; 35( 1): 69- 76; discussion 74-66. Google Scholar CrossRef Search ADS PubMed  28. Epstein FJ, Farmer JP. Pediatric spinal cord tumor surgery. Neurosurg Clin N Am . 1990; 1( 3): 569- 590. Google Scholar PubMed  29. Epstein FJ, Farmer JP, Freed D. Adult intramedullary astrocytomas of the spinal cord. J Neurosurg . 1992; 77( 3): 355- 359. Google Scholar CrossRef Search ADS PubMed  30. Goh KY, Velasquez L, Epstein FJ. Pediatric intrameduuary spinal cord tumors: is surgery alone enough? Pediatr Neurosurg . 1997; 27( 1): 34- 39. Google Scholar CrossRef Search ADS PubMed  31. McGirt MJ, Goldstein IM, Chaichana KL, Tobias ME, Kothbauer KF, Jallo GI. Extent of surgical resection of malignant astrocytomas of the spinal cord: outcome analysis of 35 patients. Neurosurgery . 2008; 63( 1): 55- 61; discussion 60-51. Google Scholar CrossRef Search ADS PubMed  32. Guidetti B, Mercuri S, Vagnozzi R. Long-term results of the surgical treatment of 129 intramedullary spinal gliomas. J Neurosurg . 1981; 54( 3): 323- 330. Google Scholar CrossRef Search ADS PubMed  33. Pollack IF. Intramedullary spinal cord astrocytomas in children. Pediatr Blood Cancer . 2004; 43( 6): 617- 618. Google Scholar CrossRef Search ADS PubMed  34. Sposto R, Ertel IJ, Jenkin RD et al.   The effectiveness of chemotherapy for treatment of high grade astrocytoma in children: results of a randomized trial. A report from the Childrens Cancer Study Group. J Neuro-Oncol . 1989; 7( 2): 165- 177. Google Scholar CrossRef Search ADS   35. Xiao R, Abdullah KG, Miller JA et al.   Molecular and clinical prognostic factors for favorable outcome following surgical resection of adult intramedullary spinal cord astrocytomas. Clin Neurol Neurosurg . 2016; 144: 82- 87. Google Scholar CrossRef Search ADS PubMed  36. Feng J, Hao S, Pan C et al.   The H3.3 K27M mutation results in a poorer prognosis in brainstem gliomas than thalamic gliomas in adults. Hum Pathol . 2015; 46( 11): 1626- 1632. Google Scholar CrossRef Search ADS PubMed  37. Cheng X, Lou S, Huang S, Chen H, Liu J. Primary spinal cord glioblastoma multiforme: a retrospective study of patients at a single institution. World Neurosurg . 2017; 106: 113- 119. Google Scholar CrossRef Search ADS PubMed  38. Gessi M, Gielen GH, Dreschmann V, Waha A, Pietsch T. High frequency of H3F3A K27M mutations characterizes pediatric and adult high-grade gliomas of the spinal cord. Acta Neuropathol . 2015; 130( 3): 435- 437. Google Scholar CrossRef Search ADS PubMed  39. Yanamadala V, Koffie RM, Shankar GM et al.   Spinal cord glioblastoma: 25years of experience from a single institution. J Clin Neurosci . 2016; 27: 138- 141. Google Scholar CrossRef Search ADS PubMed  40. Ryu SJ, Kim JY, Kim KH et al.   A retrospective observational study on the treatment outcomes of 26 patients with spinal cord astrocytoma including two cases of malignant transformation. Eur Spine J . 2016; 25( 12): 4067- 4079. Google Scholar CrossRef Search ADS PubMed  41. Babu R, Karikari IO, Owens TR, Bagley CA. Spinal cord astrocytomas: a modern 20-year experience at a single institution. Spine . 2014; 39( 7): 533- 540. Google Scholar CrossRef Search ADS PubMed  42. Beyer S, von Bueren AO, Klautke G et al.   A systematic review on the characteristics, treatments and outcomes of the patients with primary spinal glioblastomas or gliosarcomas reported in literature until March 2015. PLoS One . 2016; 11( 2): e0148312. Google Scholar CrossRef Search ADS PubMed  43 Konar SK, Maiti TK, Bir SC, Kalakoti P, Bollam P, Nanda A. Predictive factors determining the overall outcome of primary spinal glioblastoma multiforme: an integrative survival analysis. World Neurosurg . 2016; 86: 341- 348.e3. Google Scholar CrossRef Search ADS PubMed  44. Hernández-Durán S, Bregy A, Shah AH, Hanft S, Komotar RJ, Manzano GR. Primary spinal cord glioblastoma multiforme treated with temozolomide. J Clin Neurosci . 2015; 22( 12): 1877- 1882. Google Scholar CrossRef Search ADS PubMed  Acknowledgment The authors would like to thank Dong-Su Jang, MFA (Medical Illustrator) for his help with the illustrations. Copyright © 2018 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Neurosurgery Oxford University Press

Impact of H3.3 K27M Mutation on Prognosis and Survival of Grade IV Spinal Cord Glioma on the Basis of New 2016 World Health Organization Classification of the Central Nervous System

Loading next page...
 
/lp/ou_press/impact-of-h3-3-k27m-mutation-on-prognosis-and-survival-of-grade-iv-t31C1CbzSs
Publisher
Congress of Neurological Surgeons
Copyright
Copyright © 2018 by the Congress of Neurological Surgeons
ISSN
0148-396X
eISSN
1524-4040
D.O.I.
10.1093/neuros/nyy150
Publisher site
See Article on Publisher Site

Abstract

Abstract BACKGROUND Spinal cord glioma grade IV is a rare, diffuse midline glioma. H3 K27M-mutant was classified in a different entity in the 2016 World Health Organization (WHO) classification recently. No reports about prognosis of spinal cord glioma grade IV are available yet. OBJECTIVE To analyze the prognostic factors for spinal cord glioma grade IV. METHODS Twenty-five patients with spinal cord glioma of grade IV who underwent surgery in a single institute were selected. All grade IV spinal cord glioma histologically confirmed as glioblastoma or “diffuse midline glioma with H3 K27M-mutant” by the 2016 WHO classification of the central nervous system were included. Basic demographics, treatment modalities, and pathological tumor molecular profiles were investigated for prognosis. RESULTS Mean age was 39.1 yr; male to female ratio was 18 : 7. Tumor was located in thoracic cord (53.3%), cervical cord (40%), and lumbar area (6.7%). Median overall survival was 37.1 mo; median disease-free survival was 18.5 mo. Treatment modality showed no statistical difference. Only K27M profile showed significant prognostic value, 20 patients (80%) showed K27M mutation positive, K27M mutation patients showed longer overall survival (40.07 mo) than K27M negative patients (11.63 mo, P < .0001), and disease-free survival (20.85 vs 8.72 mo, P = .0241). CONCLUSION This study is the first and largest report of the prognosis of primary spinal cord grade IV glioma using the new WHO classification. This study reported survival analysis and prognostic factors, and revealed that H3.3 K27M mutation is not a major poor prognostic factor. Further studies to explore K27M mutations needed for risk stratification and therapy optimization. H3 K27M, Grade IV, Prognosis, Spinal cord, Glioma, Survival analysis, WHO classification ABBREVIATIONS ABBREVIATIONS BSA Body Surface Area CCRT concurrent chemoradiation therapy CNS central nervous system DFS disease-free survival FISH fluorescent in situ hybridization GBM glioblastoma multiforme IHC immunohistochemistry OS overall survival TMZ temozolomide WHO World Health Organization Intramedullary astrocytoma constitutes 6% to 8% of all primary spinal cord tumors.1 Primary spinal glioblastomas compose 7.5% of all intramedullary gliomas.2,3 Fewer than 200 cases of spinal glioblastoma were ever reported in the literature. A few studies enrolling fewer than 10 patients have been published because of the rarity of cases of spinal glioblastoma.4,5 The small number of cases in the literature about spinal cord glioblastoma and its survival analysis, the prognostic factors on disease-free survival (DFS) and overall survival (OS), made spinal cord glioma to be understood poorly. The optimal surgical management of high-grade intramedullary spinal cord astrocytoma continues to be controversial. The effect of the extent of surgery of low-grade spinal cord astrocytoma on the prognosis is inconsistent. 6-9 For malignant astrocytoma, the survival analysis with the extent of surgery remains largely unstudied.7 TABLE 1. Summary of Recent Literature on Primary Spinal Cord Glioblastoma Study design  Author  Source  Year  Study design and period  No of GBM  OS (mo)  Good Px  Poor Px  No effect  Single institute  Yana- madala39  J Clin Neurosci  2016  Single institute (MGH, USA, 1990-2015)  GBM 6  No report    Ryu40  Eur Spine J  2015  Single institute (Gangnam Severance, Korea, 1983-2014)  GBM 6  12  Histologic lower grade        Babu41  Spine  2014  Single institute (Duke Univ, USA, 1992-2012)  GBM 9  12.1    High Gr, dissemination Span of level  Resection    McGirt31  Neuro- surgery  2008  Single institute (New York Univ, USA, 1990-2002)  GBM 8  9  Radical resection      Meta-analysis  Beyer42  PLOS ONE  2016  Meta-analysis (Web of Science & Pubmed 1938-2015)  GBM 155  9  Age < 60, Tx in 1980-2015    GTR, gender, location    Konar43  World Neurosurg  2016  Meta-analysis (Pubmed 1938-2015)  GBM 128  11  Age 18-65 surgery + RTx&CTx  Radical surgery + RTx disseminate      Hernández- Durán44  J Clin Neurosci  2015  Meta-analysis (MeSH & pubmed 1970-2014)  GBM 64  16 or 10      TMZ use    Adams10  Spine  2012  SEER Population-based analysis 1973-2007  GBM 59  10  AA, adult, male, radical resection        Santi11  Cancer  2003  AFIP database 1962-2000  GBM 23  10    Age > 40  Gender, gr,Tx option, resection  Study design  Author  Source  Year  Study design and period  No of GBM  OS (mo)  Good Px  Poor Px  No effect  Single institute  Yana- madala39  J Clin Neurosci  2016  Single institute (MGH, USA, 1990-2015)  GBM 6  No report    Ryu40  Eur Spine J  2015  Single institute (Gangnam Severance, Korea, 1983-2014)  GBM 6  12  Histologic lower grade        Babu41  Spine  2014  Single institute (Duke Univ, USA, 1992-2012)  GBM 9  12.1    High Gr, dissemination Span of level  Resection    McGirt31  Neuro- surgery  2008  Single institute (New York Univ, USA, 1990-2002)  GBM 8  9  Radical resection      Meta-analysis  Beyer42  PLOS ONE  2016  Meta-analysis (Web of Science & Pubmed 1938-2015)  GBM 155  9  Age < 60, Tx in 1980-2015    GTR, gender, location    Konar43  World Neurosurg  2016  Meta-analysis (Pubmed 1938-2015)  GBM 128  11  Age 18-65 surgery + RTx&CTx  Radical surgery + RTx disseminate      Hernández- Durán44  J Clin Neurosci  2015  Meta-analysis (MeSH & pubmed 1970-2014)  GBM 64  16 or 10      TMZ use    Adams10  Spine  2012  SEER Population-based analysis 1973-2007  GBM 59  10  AA, adult, male, radical resection        Santi11  Cancer  2003  AFIP database 1962-2000  GBM 23  10    Age > 40  Gender, gr,Tx option, resection  MGH, Massachusetts General Hospital; SEER, Surveillance, Epidemiology and End Results; AFIP, Armed Forces Institute of Pathology database; OS. overall median survival. View Large TABLE 1. Summary of Recent Literature on Primary Spinal Cord Glioblastoma Study design  Author  Source  Year  Study design and period  No of GBM  OS (mo)  Good Px  Poor Px  No effect  Single institute  Yana- madala39  J Clin Neurosci  2016  Single institute (MGH, USA, 1990-2015)  GBM 6  No report    Ryu40  Eur Spine J  2015  Single institute (Gangnam Severance, Korea, 1983-2014)  GBM 6  12  Histologic lower grade        Babu41  Spine  2014  Single institute (Duke Univ, USA, 1992-2012)  GBM 9  12.1    High Gr, dissemination Span of level  Resection    McGirt31  Neuro- surgery  2008  Single institute (New York Univ, USA, 1990-2002)  GBM 8  9  Radical resection      Meta-analysis  Beyer42  PLOS ONE  2016  Meta-analysis (Web of Science & Pubmed 1938-2015)  GBM 155  9  Age < 60, Tx in 1980-2015    GTR, gender, location    Konar43  World Neurosurg  2016  Meta-analysis (Pubmed 1938-2015)  GBM 128  11  Age 18-65 surgery + RTx&CTx  Radical surgery + RTx disseminate      Hernández- Durán44  J Clin Neurosci  2015  Meta-analysis (MeSH & pubmed 1970-2014)  GBM 64  16 or 10      TMZ use    Adams10  Spine  2012  SEER Population-based analysis 1973-2007  GBM 59  10  AA, adult, male, radical resection        Santi11  Cancer  2003  AFIP database 1962-2000  GBM 23  10    Age > 40  Gender, gr,Tx option, resection  Study design  Author  Source  Year  Study design and period  No of GBM  OS (mo)  Good Px  Poor Px  No effect  Single institute  Yana- madala39  J Clin Neurosci  2016  Single institute (MGH, USA, 1990-2015)  GBM 6  No report    Ryu40  Eur Spine J  2015  Single institute (Gangnam Severance, Korea, 1983-2014)  GBM 6  12  Histologic lower grade        Babu41  Spine  2014  Single institute (Duke Univ, USA, 1992-2012)  GBM 9  12.1    High Gr, dissemination Span of level  Resection    McGirt31  Neuro- surgery  2008  Single institute (New York Univ, USA, 1990-2002)  GBM 8  9  Radical resection      Meta-analysis  Beyer42  PLOS ONE  2016  Meta-analysis (Web of Science & Pubmed 1938-2015)  GBM 155  9  Age < 60, Tx in 1980-2015    GTR, gender, location    Konar43  World Neurosurg  2016  Meta-analysis (Pubmed 1938-2015)  GBM 128  11  Age 18-65 surgery + RTx&CTx  Radical surgery + RTx disseminate      Hernández- Durán44  J Clin Neurosci  2015  Meta-analysis (MeSH & pubmed 1970-2014)  GBM 64  16 or 10      TMZ use    Adams10  Spine  2012  SEER Population-based analysis 1973-2007  GBM 59  10  AA, adult, male, radical resection        Santi11  Cancer  2003  AFIP database 1962-2000  GBM 23  10    Age > 40  Gender, gr,Tx option, resection  MGH, Massachusetts General Hospital; SEER, Surveillance, Epidemiology and End Results; AFIP, Armed Forces Institute of Pathology database; OS. overall median survival. View Large Previously reported spinal cord glioma case series have reported that the OS of patients with spinal cord glioblastoma is about 10 mo (Table 1), compared with the better survival of 14 mo of brain astrocytoma.10,11 Recent developments in genomic studies have enabled researchers to determine the molecular profile and characteristics of glioma to figure out biological behavior and categorize them into similar groups from an etiological perspective. Researchers suggested that some of the newly identified mutations are associated with tumor subtype and location of tumor and patient age. Sturm et al12 reported that “the H3F3A K27M mutation was predominantly identified in malignant astrocytoma arising in the midline structure such as the thalamus, brain stem, and spinal cord and is mainly prevalent in pediatric and young adult patients.”12 As the results of molecular analysis in the midline structure, the diffuse midline glioma, H3 K27M-mutant, was newly categorized as a separate pathological entity in the 2016 WHO classification.13 Several articles have reported diffuse midline glioma, H3 K27M-mutant in the midline structure of brain including the thalamus and brain stem in adults; however, diffuse midline glioma, H3 K27M-mutant cases in spinal cord are rarely reported.14 In addition, the genetic and molecular profiles of diffuse midline glioma, H3 K27M-mutant in spinal cord remain unclear due to a lack of relevant studies. The purpose of this study was to perform survival analysis of spinal cord grade IV glioma on the basis of new 2016 WHO classification of central nervous system (CNS) tumor by elucidating the association of the molecular profiles with the prognosis and to determine the role of H3 K27M mutant in prediction of the OS. This study is the first report of the prognosis of spinal cord grade IV glioma on the basis of the new 2016 WHO classification of central nervous system tumors and includes the largest numbers of patients from a single institute over 18 yr. METHODS Patient Population Twenty-five patients who were surgically treated for spinal cord glioma of grade IV, from 1998 to 2016 in a single institute, were selected for study, and general demographics are shown in Table 2. TABLE 2. Demographics of Patients Variable    n  %  Mean age (years ± SD)    39.1 ± 16.2    Gender  Total  25      Male  7  28    Female  18  72  Extent of surgery          Biopsy  3  12    PR  4  16    STR  9  36    GTR  9  36  Treatment          Surgery  3  12    Surgery + RTx  6  24    Surgery + CTx  1  4    Surgery + RTx + CTx  15  60  Variable    n  %  Mean age (years ± SD)    39.1 ± 16.2    Gender  Total  25      Male  7  28    Female  18  72  Extent of surgery          Biopsy  3  12    PR  4  16    STR  9  36    GTR  9  36  Treatment          Surgery  3  12    Surgery + RTx  6  24    Surgery + CTx  1  4    Surgery + RTx + CTx  15  60  PR, partial resection; STR, subtotal resection; GTR, gross total resection; RTx, radiotherapy; CTx, chemotherapy. View Large TABLE 2. Demographics of Patients Variable    n  %  Mean age (years ± SD)    39.1 ± 16.2    Gender  Total  25      Male  7  28    Female  18  72  Extent of surgery          Biopsy  3  12    PR  4  16    STR  9  36    GTR  9  36  Treatment          Surgery  3  12    Surgery + RTx  6  24    Surgery + CTx  1  4    Surgery + RTx + CTx  15  60  Variable    n  %  Mean age (years ± SD)    39.1 ± 16.2    Gender  Total  25      Male  7  28    Female  18  72  Extent of surgery          Biopsy  3  12    PR  4  16    STR  9  36    GTR  9  36  Treatment          Surgery  3  12    Surgery + RTx  6  24    Surgery + CTx  1  4    Surgery + RTx + CTx  15  60  PR, partial resection; STR, subtotal resection; GTR, gross total resection; RTx, radiotherapy; CTx, chemotherapy. View Large All grade IV spinal cord gliomas that were histologically confirmed as glioblastoma as well as diffuse midline glioma with H3 K27M-mutant by the 2016 WHO classification of the central nervous system13 were included in this study. Patient demographics—radiological, surgical, and pathological—and clinical follow-up outcomes were analyzed. Demographic factors including age and gender were identified. The patient's symptoms, signs, and the neurological examination findings were gathered as preoperative records of patients. The tumor factors including size, location, and number of spanning spinal cord levels were identified. The surgical factors included the extent of resection and adjuvant chemotherapy and radiation therapy. The postoperative factors included neurological status, recurrence pattern, localized or distant metastasis, complication, time to recurrence, and survival. The dates of death were determined from patient's medical records. Moreover, the characteristics of grade IV spinal cord glioma were compared with brain glioblastoma multiforme (GBM) on the basis of the same, single institutional brain tumor database. This study was approved by institutional review board in our institute, waiving patient written informed consent as the retrospective study (4-2017-1251). Treatment Modality Extent of surgery was defined as gross total resection (≥90%), subtotal resection (≥50% and <90%), partial resection (<50%), or open biopsy. Patients’ chemotherapy regimen history was identified as well from medical records. Postoperative adjuvant radiation therapy involving local area or whole neuro-axis was identified for individual patients. The chemotherapy regimen and radiation therapy followed a standard protocol of glioma treatment in brain. Most patients started concurrent chemoradiation therapy (CCRT) with using Stupp regimen, daily 75 mg/Body Surface Area (BSA) of temozolomide (TMZ) plus 45 to 54 Gy of radiotherapy at a dose of 1.8 Gy per fraction given once a day, 5 d per week over 5 to 6 wk, within 3 wk of surgery.15 Twenty-eight days following the completion of CCRT, TMZ was started; 6 cycles on every 4 wk were administered unless adverse events or progression was detected. The first cycle regimen is 150 mg/BSA(m2) for 5 d; the second to sixth cycle regimen is 200 mg/BSA for 5 d in each cycle. Radiation volume included tumor bed, based on preoperative and postoperative magnetic resonance imaging, and 2 to 5 cm longitudinal margins. In cases of leptomeningeal seeding, the entire spinal cord or craniospinal irradiation was chosen. In cases of recurrence beyond 6 mo after initial treatment, reoperation was considered a priority when the tumor was resectable. Maximal safe resection was attempted using the method used for initial surgery. Repeat radiotherapy was again recommended after surgery or unresectable recurrent tumor. Adjuvant TMZ administration, using the same regimen (with initial treatment), was also considered following repeat radiotherapy. If the tumor recurred within 6 mo after initial treatment or re-recurred, bevacizumab/iriotecan was considered as salvage treatment. Prognostic Factors for Survival Basic demographics including gender, age, tumor location, and tumor spanning length were investigated. Extent of surgery, radiotherapy, and chemotherapy were investigated for treatment modality. Pathological tumor molecular profiles including MGMT methylation, IDH1 mutation, K27M, 1p/19q del, and EGFR amplification were identified and analyzed. Pathological Diagnoses All cases were reviewed (by 2 neuropathologists, SHK and JC). Specifically, pilocytic astrocytoma and ependymoma were excluded. All cases were graded according to the 2016 World Health Organization (WHO) classification, as WHO grade II, III, or IV.16 H3.3 K27M mutation was examined by immunohistochemistry (IHC) (Figures 1 and 2) and pyrosequeuncing. The pyrosequencing for identification of H3.3 gene (H3F3A K27M and G34V) was preformed according to previously described methods.17 IHC with a Ventana BenchMark XT autostainer (Ventana Medical System Inc, Tucson, Arizona) according to the protocol was performed. The antibodies used were Anti-Histone H3 Antibody, K27M mutant (ABE 419, 1 : 300 dilution, Millipore, Burlington, Massachusetts), and Anti-Histone H3 Antibody, K27me3 (07-449, 1 : 300 dilution, Millipore). When the distinct nuclear staining of K27M mutant and loss of K27me3 were identified in glioma cells, these cells were regarded as “mutant”/“positive.” From pyrosequencing results, there was no G34V mutation. All cases showing K27M mutation in pyrosequencing were positive in IHC. K27M mutant was regarded as diffuse midline glioma, K27M mutant (the 2016 WHO grade IV), regardless of histological grade according to new WHO classification. FIGURE 1. View largeDownload slide Diffuse astrocytoma and K27M. Two diffuse astrocytomas (A and B, H-E ×200) show the same histological grade but different H3.3 K27M immunoreactivity patterns (C and D, K27M × 200). The K27M mutant protein strongly expresses in the tumor cell nuclei in C. FIGURE 1. View largeDownload slide Diffuse astrocytoma and K27M. Two diffuse astrocytomas (A and B, H-E ×200) show the same histological grade but different H3.3 K27M immunoreactivity patterns (C and D, K27M × 200). The K27M mutant protein strongly expresses in the tumor cell nuclei in C. FIGURE 2. View largeDownload slide Glioblastoma and K27M. Two glioblastomas show the same histological patterns (A and B, H-E ×200) and H3.3 K27M immunoreactivity patterns (C and D K27M ×200). FIGURE 2. View largeDownload slide Glioblastoma and K27M. Two glioblastomas show the same histological patterns (A and B, H-E ×200) and H3.3 K27M immunoreactivity patterns (C and D K27M ×200). Additional molecular profiles, such as 1p/19q co-deletion, IDH mutation, MGMT hypermethylation, EGFR amplification, and H3.3 K27M mutation, were examined. The 1p/19q co-deletion was examined by Fluorescent in situ hybridization (FISH) (1p/19q fluorescent probe kit, Vysis, Abbott Molecular, Chicago, Illinois), as previously described.18 The assessment and interpretation of FISH results were performed according to the guidelines defined by the SIOP Europe Neuroblastoma Pathology and Biology and Bone Marrow Group.19 More than 100 nonoverlapping nuclei were enumerated per hybridization for each probe. Tumor cells that showed more than 30% of nuclei with DNA loss were identified as chromosomal loss. IDH1 mutation was examined by IHC. With a representative section, we performed IHC with a Ventana BenchMark XT autostainer (Ventana Medical System Inc) according to the protocol. The antibodies were anti-human IDH1 R132H mouse monoclonal (Clone H09L, 1 : 80 dilution, Dianova, Hamburg, Germany). When the cytoplasmic expression of the IDH1 R132H was identified in glioma cells, we regarded those cases as “mutant”/“positive.” The DNA methylation status of CpG islands at the MGMT promoter was determined by methylation-specific polymerase chain reaction as previously described, with some modifications.20,21 EGFR amplification was examined by FISH (EGFR/CEP7 dual color probe kit, Vysis, Abbott Molecular), as described previously.22 Specimens were considered to be amplified when more than 10% of tumor cells exhibited either an EGFR/CEP7 ratio of >2 or innumerable tight clusters of signals of the locus probe. Statistical Analysis Student t-tests and chi-square tests were used for individual variables. Kaplan-Meier curves estimated time to OS, DFS, and recurrence, and were compared using the log-rank test. We conducted only univariable log-rank test, because the number of samples was not enough. Two-sided P values smaller than .05 were considered to be significant. Statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, North Carolina). RESULTS Patients Demographics General demographics of 25 patients who were surgically treated for primary spinal cord glioma of grade IV from 1998 to 2016 in a single institute are shown in Table 2. Mean age was 39.1 yr (39.1 ± 16.2), and mean follow-up duration was 26.4 mo. Male to female was 18:7, male was predominant. The number of expired patients during follow-up and investigation was 13 (52%). Long-term survivor (survival longer than 3 yr) numbers were 6 patients, and 5 patients among them expired during follow-up. The longest survivor was a 52-yr-old male patient, histological subtype was anaplastic astrocytoma with K27M (+), and he has survived 94.3 mo until now without recurrence. Tumor Characteristics The total number of segments involved by tumor was 75 segments. The most frequent location is location of tumor C5, 6, 7, and T12 (Figure 3). Tumor was located in thoracic cord (53.3%), cervical cord (40%), and lumbar area (6.7%) in order of frequency. Average length of span by tumor was 3 segments. FIGURE 3. View largeDownload slide Location and involved segments of grade IV spinal cord glioma. Tumor was located in the thoracic cord (53.3%), cervical cord (40%), and lumbar area (6.7%) in order of frequency. FIGURE 3. View largeDownload slide Location and involved segments of grade IV spinal cord glioma. Tumor was located in the thoracic cord (53.3%), cervical cord (40%), and lumbar area (6.7%) in order of frequency. Treatment Strategy Extent of surgery was identified as the following (Table 2): gross total resection in 9 patients (36%), subtotal resection in 9 patients (36%), partial resection in 4 patients (16%), and open biopsy in 3 patients (12%). Extent of surgery did not show statistical differences in survival (Table 3). Adjuvant treatment was identified (Table 2): radiation therapy with chemotherapy, 15 patients (60%); radiation therapy only, 6 patients (24%); chemotherapy only, 1 patient (4%); surgery only, 3 patients (12%). Adjuvant radiotherapy or chemotherapy showed no statistical significance in improving survival (Table 3). The numbers of chemotherapy cycles and radiation therapy were diverse depending on the patient's survival, metastasis, recurrence pattern, and general condition, providing a heterogeneous treatment protocol in our series despite the small number. TABLE 3. Statistical Analysis for Prognostic Factors. Factors    OS (mo)  P  DFS (mo)  P  Gender      .7727    .3441    Male  37.05    19.23      Female  51.34    10.62    Extent of surgery      .6719    .0025    GTR  29.74    14.82      STR  48.87    40.07      PR  37.05    23.84      Bx  51.34    18.49    RTx      .9801    .9388    Yes  37.05    19.97      No  16.13    11.04    CTx      .1016    .9831    Yes  29.74    16.95      No  Not reached    19.34    K27M mutation      <.0001    .0241    Positive  40.07    20.85      Negative  11.63    8.72    Factors    OS (mo)  P  DFS (mo)  P  Gender      .7727    .3441    Male  37.05    19.23      Female  51.34    10.62    Extent of surgery      .6719    .0025    GTR  29.74    14.82      STR  48.87    40.07      PR  37.05    23.84      Bx  51.34    18.49    RTx      .9801    .9388    Yes  37.05    19.97      No  16.13    11.04    CTx      .1016    .9831    Yes  29.74    16.95      No  Not reached    19.34    K27M mutation      <.0001    .0241    Positive  40.07    20.85      Negative  11.63    8.72    OS, overall median survival; DFS, disease-free survival; PR, partial resection; STR, subtotal resection; GTR, gross total resection. View Large TABLE 3. Statistical Analysis for Prognostic Factors. Factors    OS (mo)  P  DFS (mo)  P  Gender      .7727    .3441    Male  37.05    19.23      Female  51.34    10.62    Extent of surgery      .6719    .0025    GTR  29.74    14.82      STR  48.87    40.07      PR  37.05    23.84      Bx  51.34    18.49    RTx      .9801    .9388    Yes  37.05    19.97      No  16.13    11.04    CTx      .1016    .9831    Yes  29.74    16.95      No  Not reached    19.34    K27M mutation      <.0001    .0241    Positive  40.07    20.85      Negative  11.63    8.72    Factors    OS (mo)  P  DFS (mo)  P  Gender      .7727    .3441    Male  37.05    19.23      Female  51.34    10.62    Extent of surgery      .6719    .0025    GTR  29.74    14.82      STR  48.87    40.07      PR  37.05    23.84      Bx  51.34    18.49    RTx      .9801    .9388    Yes  37.05    19.97      No  16.13    11.04    CTx      .1016    .9831    Yes  29.74    16.95      No  Not reached    19.34    K27M mutation      <.0001    .0241    Positive  40.07    20.85      Negative  11.63    8.72    OS, overall median survival; DFS, disease-free survival; PR, partial resection; STR, subtotal resection; GTR, gross total resection. View Large Survival Analysis Overall Survival and Disease-Free Survival Median OS was 37.1 mo. Survival at 1 and 5 yr was 82.3% and 18.6%, respectively (Figure 4). Median DFS was 18.5 mo; survival at 1 and 5 yr was 64% and 4%, respectively (Figure 4). FIGURE 4. View largeDownload slide Survival analysis. Median OS was 37.1 mo, survival at 1 and 5 yr was 82.3% and 18.6%, respectively. Median DFS was 18.5 mo, survival at 1 and 5 yr was 64% and 4%, respectively. FIGURE 4. View largeDownload slide Survival analysis. Median OS was 37.1 mo, survival at 1 and 5 yr was 82.3% and 18.6%, respectively. Median DFS was 18.5 mo, survival at 1 and 5 yr was 64% and 4%, respectively. Prognostic Factors Male gender showed a longer survival than female gender in OS and DFS, but had no statistical significance. Extent of surgery and extent of tumor involvement did not reveal any prognostic importance. Radiotherapy and chemotherapy showed heterogeneous results without statistical significance, and did not show any effect on prognosis and OS. Patients with positive tumor molecular profile of K27M showed significant longer survival than negative K27M patients (Figure 5, Table 3). FIGURE 5. View largeDownload slide OS analysis by K27M status. Positive K27M mutation patients showed longer OS (40.07 mo) than K27M negative patients (11.63 mo, P < .0001) FIGURE 5. View largeDownload slide OS analysis by K27M status. Positive K27M mutation patients showed longer OS (40.07 mo) than K27M negative patients (11.63 mo, P < .0001) Tumor Molecular Profile Molecular profiles, such as 1p/19q co-deletion, IDH mutation, MGMT hypermethylation, EGFR amplification, and H3.3 K27M mutation, were examined (Table 4). Among the molecular profiles, only the K27M profile showed prognostic value. In 25 patients with grade IV spinal cord glioma, 20 patients (80%) showed K27M mutation positive and 5 patients (20%) showed negative. Positive K27M mutation patients showed longer OS (40.07 mo) (11.63 mo, P < .0001) and DFS (20.85 vs 8.72 mo, P = .0241) than negative K27M mutation patients (Figure 5) in grade IV category. TABLE 4. Characteristics of Molecular Profile. Molecular profile  n    n  %  H3.3 K27M mutation  25  Positive  20  80      Negative  5  20  MGMT hypermethylation  25  Methylation  1  4      Unmethylation  24  96  IDH1 mutation  25  Wild  25  100      Mutation  0  0  1p/19q co-deletion  25  Yes  0  0      No  25  100  EGFR amplification  25  Yes  1  4      no  24  96  Molecular profile  n    n  %  H3.3 K27M mutation  25  Positive  20  80      Negative  5  20  MGMT hypermethylation  25  Methylation  1  4      Unmethylation  24  96  IDH1 mutation  25  Wild  25  100      Mutation  0  0  1p/19q co-deletion  25  Yes  0  0      No  25  100  EGFR amplification  25  Yes  1  4      no  24  96  View Large TABLE 4. Characteristics of Molecular Profile. Molecular profile  n    n  %  H3.3 K27M mutation  25  Positive  20  80      Negative  5  20  MGMT hypermethylation  25  Methylation  1  4      Unmethylation  24  96  IDH1 mutation  25  Wild  25  100      Mutation  0  0  1p/19q co-deletion  25  Yes  0  0      No  25  100  EGFR amplification  25  Yes  1  4      no  24  96  Molecular profile  n    n  %  H3.3 K27M mutation  25  Positive  20  80      Negative  5  20  MGMT hypermethylation  25  Methylation  1  4      Unmethylation  24  96  IDH1 mutation  25  Wild  25  100      Mutation  0  0  1p/19q co-deletion  25  Yes  0  0      No  25  100  EGFR amplification  25  Yes  1  4      no  24  96  View Large When we analyzed the impact of K27M mutation within the histological grade 4 (glioblastoma), similar results were obtained. K27M mutant with histological grade 4 (glioblastoma) showed longer OS (20.85 mo) than K27M negative patients (11.64 mo, P = .0276) and DFS (21.18 vs 8.72 mo, P = .0758) than negative K27M mutation patients (Figure 6). The OS of K27M positive in GBM was 87.5% (1 yr) and 15% (5 yr). K27M negative in GBM showed survival as 40% (1 yr) and 0% (5 yr). FIGURE 6. View largeDownload slide OS analysis by K27M status in glioblastoma (histological grade IV). K27M mutant with histological grade 4 (glioblastoma) showed longer OS (20.85 mo) than K27M negative patients (11.64 mo, P = .0276). FIGURE 6. View largeDownload slide OS analysis by K27M status in glioblastoma (histological grade IV). K27M mutant with histological grade 4 (glioblastoma) showed longer OS (20.85 mo) than K27M negative patients (11.64 mo, P = .0276). Available profile of 1p/19q co-deletion, MGMT methylation, and IDH1 mutation were shown in Table 4 and Figure 7. All of the molecular profiles were analyzed. Characteristics of grade IV glioma molecular profile from available data were uniformity in IDH 1 negative, no 1p/19q co deletion. Most patients had MGMT unmethylation except one who showed hypermethylation (Figure 7). FIGURE 7. View largeDownload slide A map illustrating patient demographics and molecular profiles. (Location: C, cervical; CT, cervicothoracic; T, thoracic; TL, thoracolumbar; L, lumbar; Extent of surgery 1, open biopsy; 2, partial resection; 3, subtotal resection; 4, gross total resection.) FIGURE 7. View largeDownload slide A map illustrating patient demographics and molecular profiles. (Location: C, cervical; CT, cervicothoracic; T, thoracic; TL, thoracolumbar; L, lumbar; Extent of surgery 1, open biopsy; 2, partial resection; 3, subtotal resection; 4, gross total resection.) DISCUSSION Intramedullary spinal cord tumors are infrequent tumors, only accounting for 2% to 4% of all central nervous system origin tumors.1,11 Helseth et al23 reported that “astrocytoma and ependymoma comprise the majority of intramedullary spinal cord tumors, they primarily occur in different patient populations.”23 Among the intramedullary spinal cord tumor, astrocytoma is the most common tumor found in young age, comprising 59% to 90%,1,12 whereas ependymoma is most often found in adults.1,24 High-grade or malignant tumor composed 7% to 25% of all spinal cord astrocytoma in young age and 10% to 30% of spinal cord astrocytoma in adults.25-31 Spinal cord astrocytoma most frequently involves the cervical spinal cord, mainly the lower cervical segment, followed by the thoracic spine.29,32,33 The consensus on the treatment for spinal cord glioblastoma is not clearly determined because of the rarity of cases and limited research. The standard treatment of spinal cord glioblastoma includes gross total resection or biopsy followed by radiotherapy and chemotherapy.31,34 New 2016 WHO classification made a change in standard treatment for glioma. We have 6 patients of radiation only group. Among them, 4 patients had diffuse astrocytoma (H3 K27M-mutant) and 1 patient had anaplastic astrocytoma (H3 K27M-mutant). These patients followed standard protocol (surgery and radiation therapy only) for grade II and III glioma, before the 2016 WHO classification grouped them to grade IV. The remaining 1 patient with glioblastoma (H3 K27M mutant) suffered from upper gastrointestinal perforation after initial surgery and could not receive chemotherapy. The major concern after establishing new WHO CNS tumor classification is the validation of new the classification supported by clinical results and prognosis. The major changes in the 2016 WHO classification of CNS tumors include “major restructuring of diffuse gliomas with incorporation of genetically defined entities, addition of newly recognized entities with variants and patterns (IDH-wild-type and IDH-mutant glioblastoma, Diffuse midline glioma, H3 K27M-mutant).”13 Solomon et al14 recently reported 47 patients (50%) with diffuse midline gliomas, H3 K27M mutation among a total of 94 CNS tumor patients, including 9 patients (53%) with spinal cord glioma with H3 K27M mutation among a total of 17 spinal cord glioma tumor patients. Our study revealed that 51.3% of all spinal cord gliomas showed K27M mutation; additionally, more frequent K27M mutation was found in higher histological grade (Table 5). K27M mutation status of all 39 cases according to the 2007 WHO classification was shown in Table 5. Shadowed cells are grade IV tumor according to the 2016 WHO classification. Ten cases (light gray cells) that had diffuse and anaplastic astrocytoma histology were reclassified as grade IV in the 2016 WHO classification. TABLE 5. K27M Mutation Status of All 39 Cases According to the 2007 WHO Classification.   K27M mutation status    WHO grade by histology  K27M+  K27M−  Total  % K27M+  Diffuse astrocytoma  4a  8  12  33.3  Anaplastic astrocytoma  6a  6  12  50  Glioblastoma  10b  5b  15  66.7  Total  20  19  39  51.3    K27M mutation status    WHO grade by histology  K27M+  K27M−  Total  % K27M+  Diffuse astrocytoma  4a  8  12  33.3  Anaplastic astrocytoma  6a  6  12  50  Glioblastoma  10b  5b  15  66.7  Total  20  19  39  51.3  Shadowed cells are grade IV tumor according to the 2016 WHO classification. aTen cases that had diffuse and anaplastic astrocytoma histology were reclassified as grade IV in 2016 WHO classification. bFifteen cases of glioblastoma are also grade IV under 2007 and 2016 WHO classification. View Large TABLE 5. K27M Mutation Status of All 39 Cases According to the 2007 WHO Classification.   K27M mutation status    WHO grade by histology  K27M+  K27M−  Total  % K27M+  Diffuse astrocytoma  4a  8  12  33.3  Anaplastic astrocytoma  6a  6  12  50  Glioblastoma  10b  5b  15  66.7  Total  20  19  39  51.3    K27M mutation status    WHO grade by histology  K27M+  K27M−  Total  % K27M+  Diffuse astrocytoma  4a  8  12  33.3  Anaplastic astrocytoma  6a  6  12  50  Glioblastoma  10b  5b  15  66.7  Total  20  19  39  51.3  Shadowed cells are grade IV tumor according to the 2016 WHO classification. aTen cases that had diffuse and anaplastic astrocytoma histology were reclassified as grade IV in 2016 WHO classification. bFifteen cases of glioblastoma are also grade IV under 2007 and 2016 WHO classification. View Large Xiao et al35 reported that the age at surgery was the only significant prognostic factor of spinal cord astrocytoma in a 13-patient series,35 though they analyzed molecular prognostic factors but they did not include K27M mutation in the study. Our results showed advanced results with no significant molecular prognostic factors except K27M mutation (Table 4). Although the K27M mutation was frequently observed in adult brainstem and thalamic gliomas, this mutation was associated with a poorer prognosis in brainstem gliomas but not in thalamic gliomas. These differences in the adult brainstem and thalamic gliomas may result from the differences in genetic aberrations, and may provide location-specific molecular targets for treatment strategy.36 It is unclear whether spinal cord astrocytoma could share genetic alterations with supratentorial glioma. Because the spinal cord has to be considered a midline structure, a high incidence of H3 K27M mutations in spinal cord glioma could be hypothesized. The spinal high-grade glioma in children and adults frequently showed H3 K27M mutations. Though the homogeneity of the molecular profile was revealed in our series such as all IDH 1 negative, no 1p/19q codeletion, only 1 case of methylation, Cheng et al37 recently reported methylation in 8 of 14 patients, IDH 1 mutation in 1/14, EGFR overexpression in 11/14 in their glioblastoma study. A large grade IV glioma cohort is required to conclude on molecular profile characteristics.37 Limitations The present study has several limitations, including its retrospective design, which may introduce selection bias. Though this study is a large cohort of spinal cord gliomas, there are limitations to the number available, because of the rarity of the disease, and power to assess enough variables. The number of chemotherapy cycles and radiation therapy were diverse depending on the patient's survival, metastasis, recurrence pattern, and general condition, to present a heterogeneous treatment protocol in our series despite the small number. Additionally, comparison between 20 patients with H3 K27M mutations and 5 patients with no mutations does suggest importance but is limited by only 5 patients in the poorly performing cohort. Further studies should be focused on investigation of the possible association of H3F3AK27M mutations with patient survival and used for therapy optimization and the prediction of outcomes.38 In the brain, there are many known prognostic markers such as IDH, MGMT, and 1p/19q-codeletion. However, those molecular changes are known to be very rare in spinal cord. That turned out to be true to a certain degree in our study. It may reflect that the biology of glioma differs independent of “location.” Further studies with more cases are needed to answer these questions. CONCLUSION This study is the first report of the prognosis of primary spinal cord grade IV glioma on the basis of the new WHO classification, and includes the largest numbers of patients in a single institute over 18 yr. This study performed integrative survival analysis, elucidating the association of the various prognostic factors, and revealed that H3.3 K27M mutation is not a major poor prognostic factor, unlike in brain glioma. Further studies are needed to explore K27M mutations’ use for prediction of outcome and treatment strategy. Disclosures This study was supported by a faculty research grant of Yonsei University College of Medicine for 2017 (6-2017-0067), the National Research Foundation of Korea (NRF) grant funded by the Korea government (NRF-2014M3A7B4051596) and the Technology Innovation Program funded by the Ministry of Trade, Industry, & Energy, Korea (No. 10062712, Development of spinal fusion implant and its manufacturing system; the functionality optimized, patient-customized in terms of bioactive materials to meet the clinical needs). The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. REFERENCES 1. Reimer R, Onofrio BM. Astrocytomas of the spinal cord in children and adolescents. J Neurosurg . 1985; 63( 5): 669- 675. Google Scholar CrossRef Search ADS PubMed  2. Raco A, Esposito V, Lenzi J, Piccirilli M, Delfini R, Cantore G. Long-term follow-up of intramedullary spinal cord tumors: a series of 202 cases. Neurosurgery . 2005; 56( 5): 972- 981; discussion 972-981. Google Scholar PubMed  3. Miller DC. Surgical pathology of intramedullary spinal cord neoplasms. J Neurooncol . 2000; 47( 3): 189- 194. Google Scholar CrossRef Search ADS PubMed  4. Gee TS, Ghani AR, Idris B, Awang MS. Case report: a rare case of pediatric conus medularis glioblastoma multiforme. Med J Malaysia . 2012; 67( 4): 438- 441. Google Scholar PubMed  5. Morais N, Mascarenhas L, Soares-Fernandes JP, Silva A, Magalhaes Z, Costa JA. Primary spinal glioblastoma: a case report and review of the literature. Oncol Lett . 2013; 5( 3): 992- 996. Google Scholar CrossRef Search ADS PubMed  6. Hardison HH, Packer RJ, Rorke LB, Schut L, Sutton LN, Bruce DA. Outcome of children with primary intramedullary spinal cord tumors. Child's Nerv Syst . 1987; 3( 2): 89- 92. Google Scholar CrossRef Search ADS   7. Minehan KJ, Shaw EG, Scheithauer BW, Davis DL, Onofrio BM. Spinal cord astrocytoma: pathological and treatment considerations. J Neurosurg . 1995; 83( 4): 590- 595. Google Scholar CrossRef Search ADS PubMed  8. Samii M, Klekamp J. Surgical results of 100 intramedullary tumors in relation to accompanying syringomyelia. Neurosurgery . 1994; 35( 5): 865- 873; discussion 873. Google Scholar CrossRef Search ADS PubMed  9. Sandler HM, Papadopoulos SM, Thornton AF Jr, Ross DA. Spinal cord astrocytomas: results of therapy. Neurosurgery . 1992; 30( 4): 490- 493. Google Scholar PubMed  10. Adams H, Avendano J, Raza SM, Gokaslan ZL, Jallo GI, Quinones-Hinojosa A. Prognostic factors and survival in primary malignant astrocytomas of the spinal cord: a population-based analysis from 1973 to 2007. Spine . 2012; 37( 12): E727- E735. Google Scholar CrossRef Search ADS PubMed  11. Santi M, Mena H, Wong K, Koeller K, Olsen C, Rushing EJ. Spinal cord malignant astrocytomas. Clinicopathologic features in 36 cases. Cancer . 2003; 98( 3): 554- 561. Google Scholar CrossRef Search ADS PubMed  12. Sturm D, Witt H, Hovestadt V et al.   Hotspot mutations in H3F3A and IDH1 define distinct epigenetic and biological subgroups of glioblastoma. Cancer Cell . 2012; 22( 4): 425- 437. Google Scholar CrossRef Search ADS PubMed  13. Louis DN, Perry A, Reifenberger G et al.   The 2016 World Health Organization classification of tumors of the central nervous system: a summary. Acta Neuropathol . 2016; 131( 6): 803- 820. Google Scholar CrossRef Search ADS PubMed  14. Solomon DA, Wood MD, Tihan T et al.   Diffuse midline gliomas with histone H3-K27M mutation: a series of 47 cases assessing the spectrum of morphologic variation and associated genetic alterations. Brain Pathol . 2016; 26( 5): 569- 580. Google Scholar CrossRef Search ADS PubMed  15. Stupp R, Mason WP, van den Bent MJ et al.   Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med . 2005; 352( 10): 987- 996. Google Scholar CrossRef Search ADS PubMed  16. Komori T. The 2016 WHO classification of tumours of the central nervous system: the major points of revision. Neurol Med Chir (Tokyo) . 2017; 57( 7): 301- 311. Google Scholar CrossRef Search ADS PubMed  17. Gielen GH, Gessi M, Hammes J, Kramm CM, Waha A, Pietsch T. H3F3A K27M mutation in pediatric CNS tumors: a marker for diffuse high-grade astrocytomas. Am J Clin Pathol . 2013; 139( 3): 345- 349. Google Scholar CrossRef Search ADS PubMed  18. Woehrer A, Sander P, Haberler C et al.   FISH-based detection of 1p 19q codeletion in oligodendroglial tumors: procedures and protocols for neuropathological practice - a publication under the auspices of the Research Committee of the European Confederation of Neuropathological Societies (Euro-CNS). Clin Neuropathol . 2011; 30( 3): 47- 55. Google Scholar CrossRef Search ADS PubMed  19. Ambros PF, Ambros IM, Siop Europe Neuroblastoma Pathology B, Bone Marrow G. Pathology and biology guidelines for resectable and unresectable neuroblastic tumors and bone marrow examination guidelines. Med Pediatr Oncol . 2001; 37( 6): 492- 504. Google Scholar CrossRef Search ADS PubMed  20. Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia. Cancer Res . 1999; 59( 4): 793- 797. Google Scholar PubMed  21. Esteller M, Toyota M, Sanchez-Cespedes M et al.   Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is associated with G to A mutations in K-ras in colorectal tumorigenesis. Cancer Res . 2000; 60( 9): 2368- 2371. Google Scholar PubMed  22. Zhao LL, Xu KL, Wang SW, Hu BL, Chen LR. Pathological significance of epidermal growth factor receptor expression and amplification in human gliomas. Histopathology . 2012; 61( 4): 726- 736. Google Scholar CrossRef Search ADS PubMed  23. Helseth A, Mork SJ. Primary intraspinal neoplasms in Norway, 1955 to 1986. A population-based survey of 467 patients. J Neurosurg . 1989; 71( 6): 842- 845. Google Scholar CrossRef Search ADS PubMed  24. Milano MT, Johnson MD, Sul J et al.   Primary spinal cord glioma: a Surveillance, Epidemiology, and End Results database study. J Neurooncol . 2010; 98( 1): 83- 92. Google Scholar CrossRef Search ADS PubMed  25. Bouffet E, Pierre-Kahn A, Marchal JC et al.   Prognostic factors in pediatric spinal cord astrocytoma. Cancer . 1998; 83( 11): 2391- 2399. Google Scholar CrossRef Search ADS PubMed  26. Cooper PR. Outcome after operative treatment of intramedullary spinal cord tumors in adults: intermediate and long-term results in 51 patients. Neurosurgery . 1989; 25( 6): 855- 859. Google Scholar CrossRef Search ADS PubMed  27. Cristante L, Herrmann HD. Surgical management of intramedullary spinal cord tumors: functional outcome and sources of morbidity. Neurosurgery . 1994; 35( 1): 69- 76; discussion 74-66. Google Scholar CrossRef Search ADS PubMed  28. Epstein FJ, Farmer JP. Pediatric spinal cord tumor surgery. Neurosurg Clin N Am . 1990; 1( 3): 569- 590. Google Scholar PubMed  29. Epstein FJ, Farmer JP, Freed D. Adult intramedullary astrocytomas of the spinal cord. J Neurosurg . 1992; 77( 3): 355- 359. Google Scholar CrossRef Search ADS PubMed  30. Goh KY, Velasquez L, Epstein FJ. Pediatric intrameduuary spinal cord tumors: is surgery alone enough? Pediatr Neurosurg . 1997; 27( 1): 34- 39. Google Scholar CrossRef Search ADS PubMed  31. McGirt MJ, Goldstein IM, Chaichana KL, Tobias ME, Kothbauer KF, Jallo GI. Extent of surgical resection of malignant astrocytomas of the spinal cord: outcome analysis of 35 patients. Neurosurgery . 2008; 63( 1): 55- 61; discussion 60-51. Google Scholar CrossRef Search ADS PubMed  32. Guidetti B, Mercuri S, Vagnozzi R. Long-term results of the surgical treatment of 129 intramedullary spinal gliomas. J Neurosurg . 1981; 54( 3): 323- 330. Google Scholar CrossRef Search ADS PubMed  33. Pollack IF. Intramedullary spinal cord astrocytomas in children. Pediatr Blood Cancer . 2004; 43( 6): 617- 618. Google Scholar CrossRef Search ADS PubMed  34. Sposto R, Ertel IJ, Jenkin RD et al.   The effectiveness of chemotherapy for treatment of high grade astrocytoma in children: results of a randomized trial. A report from the Childrens Cancer Study Group. J Neuro-Oncol . 1989; 7( 2): 165- 177. Google Scholar CrossRef Search ADS   35. Xiao R, Abdullah KG, Miller JA et al.   Molecular and clinical prognostic factors for favorable outcome following surgical resection of adult intramedullary spinal cord astrocytomas. Clin Neurol Neurosurg . 2016; 144: 82- 87. Google Scholar CrossRef Search ADS PubMed  36. Feng J, Hao S, Pan C et al.   The H3.3 K27M mutation results in a poorer prognosis in brainstem gliomas than thalamic gliomas in adults. Hum Pathol . 2015; 46( 11): 1626- 1632. Google Scholar CrossRef Search ADS PubMed  37. Cheng X, Lou S, Huang S, Chen H, Liu J. Primary spinal cord glioblastoma multiforme: a retrospective study of patients at a single institution. World Neurosurg . 2017; 106: 113- 119. Google Scholar CrossRef Search ADS PubMed  38. Gessi M, Gielen GH, Dreschmann V, Waha A, Pietsch T. High frequency of H3F3A K27M mutations characterizes pediatric and adult high-grade gliomas of the spinal cord. Acta Neuropathol . 2015; 130( 3): 435- 437. Google Scholar CrossRef Search ADS PubMed  39. Yanamadala V, Koffie RM, Shankar GM et al.   Spinal cord glioblastoma: 25years of experience from a single institution. J Clin Neurosci . 2016; 27: 138- 141. Google Scholar CrossRef Search ADS PubMed  40. Ryu SJ, Kim JY, Kim KH et al.   A retrospective observational study on the treatment outcomes of 26 patients with spinal cord astrocytoma including two cases of malignant transformation. Eur Spine J . 2016; 25( 12): 4067- 4079. Google Scholar CrossRef Search ADS PubMed  41. Babu R, Karikari IO, Owens TR, Bagley CA. Spinal cord astrocytomas: a modern 20-year experience at a single institution. Spine . 2014; 39( 7): 533- 540. Google Scholar CrossRef Search ADS PubMed  42. Beyer S, von Bueren AO, Klautke G et al.   A systematic review on the characteristics, treatments and outcomes of the patients with primary spinal glioblastomas or gliosarcomas reported in literature until March 2015. PLoS One . 2016; 11( 2): e0148312. Google Scholar CrossRef Search ADS PubMed  43 Konar SK, Maiti TK, Bir SC, Kalakoti P, Bollam P, Nanda A. Predictive factors determining the overall outcome of primary spinal glioblastoma multiforme: an integrative survival analysis. World Neurosurg . 2016; 86: 341- 348.e3. Google Scholar CrossRef Search ADS PubMed  44. Hernández-Durán S, Bregy A, Shah AH, Hanft S, Komotar RJ, Manzano GR. Primary spinal cord glioblastoma multiforme treated with temozolomide. J Clin Neurosci . 2015; 22( 12): 1877- 1882. Google Scholar CrossRef Search ADS PubMed  Acknowledgment The authors would like to thank Dong-Su Jang, MFA (Medical Illustrator) for his help with the illustrations. Copyright © 2018 by the Congress of Neurological Surgeons This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

NeurosurgeryOxford University Press

Published: May 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off