Recent reports have shown that many oligodendroglial

Transcription

1 Allelic Losses in Oligodendroglial and Oligodendroglioma-like Neoplasms Analysis Using Microsatellite Repeats and Polymerase Chain Reaction Mahlon D. Johnson, MD, PhD; Cindy L. Vnencak-Jones, PhD; Steven A. Toms, MD; Paul M. Moots, MD; Robert Weil, MD Context. Oligodendroglial tumors are heterogenous neoplasms with histologic features shared with other central nervous system tumors, such as dysembryoplastic neuroepithelial tumors. Objective. We examined a series of tumors, identified as possessing oligodendroglial components at the time of intraoperative examination, to see if molecular subsets based on the oligodendroglial component could be recognized. Design. DNA was extracted from fresh brain tumor tissue and corresponding peripheral blood or normal tissues. Genotypes for multiple loci were determined by polymerase chain reaction amplification using fluorescent-labeled primers for markers on chromosomes 1p, 17p, and 19q. Results. Of the 12 oligodendrogliomas, 6 (60%) of 10 informative cases for 1p exhibited loss of heterozygosity (LOH). Six (50%) of 12 informative cases for 19q exhibited LOH. Each case also showed LOH at 1p. Three (25%) of 12 informative cases exhibited LOH at 17p for the dinucleotide repeat within the TP53 gene. In oligoastrocytomas, none of 4 informative cases showed LOH at 1p, 1 (25%) showed LOH at 19q, and 2 (50%) at 17p. One case also displayed microsatellite instability at 3 of 8 markers. In the 3 anaplastic oligodendrogliomas, 1 was not informative for 1p and none of the informative tumors exhibited LOH at 1p or 17p; 1 case (33%) exhibited LOH at 19q. Of the 14 informative anaplastic oligoastrocytomas, LOH was seen in 5 (36%) at both 1p and 19q and in 2 (14%) at 17p. Those with allelic loss at TP53 were astrocytoma predominant. No dysembryoplastic neuroepithelial tumors exhibited LOH at any marker on 1p, 17p, or 19q. Conclusions. These findings suggest that routine screening for allelic losses, in samples intraoperatively determined to have an oligodendroglial component, will reveal prognostically or therapeutically relevant information in the majority of cases. (Arch Pathol Lab Med. 2003;127: ) Recent reports have shown that many oligodendroglial tumors have a better prognosis and respond to specific chemotherapy better than astrocytic neoplasms of comparable grade. 1 5 Recently, this difference has been correlated with allelic loss on chromosomes 1p and/or 19q in World Health Organization (WHO) grade II and III oligodendrogliomas and oligoastrocytomas Loss of heterozygosity (LOH) on chromosome 1p appears to correlate with response to procarbazine, N-(2-chloroethyl)-N cyclohexyl-n-nitrosourea (CCNU)/lomustine, and vincristine (PCV) therapy. 8,10 12 Moreover, in anaplastic oligodendrogliomas, LOH of 1p and/or 19q correlates with distinctly different survival rates. 11 Nonetheless, a consensus has not been reached on the prevalence of these Accepted for publication July 29, From the Departments of Pathology (Drs Johnson and Vnencak- Jones), Neurosurgery (Drs Toms and Weil), and Neurology (Neuro-oncology) (Dr Moots), Vanderbilt Medical School, Nashville, Tenn; and Tennessee Valley Health Care System, Nashville, Tenn (Dr Johnson). Dr Toms is now with the Brain Tumor Institute, Cleveland Clinic Foundation, Cleveland, Ohio; and Dr Weil is with the Surgical Neurology Branch, National Institute of Neurological Disorders and Stroke, Bethesda, Md. Reprints: Mahlon Johnson, MD, PhD, Department of Pathology, T3218 MCN, Vanderbilt Medical School, 21st Avenue S, Nashville, TN ( mahlon.johnson@vanderbilt.edu). allelic losses in oligodendroglial neoplasms or the value of routinely applying LOH analysis to brain biopsies. Equally uncertain is whether other central nervous system tumors with oligodendroglial elements exhibit similar allelic losses and might confound interpretation of LOH studies. Other tumors possess an oligodendroglial component that, particularly in a small biopsy, may be mistaken for an oligodendroglioma and trigger LOH analysis. Most notable are simple and complex dysembryoplastic neuroepithelial tumors (DNTs) Recognition and differentiation of DNTs from oligodendrogliomas is particularly important since they are usually lower grade and can be treated successfully by surgery alone In typical cases, clinical features may distinguish DNTs from oligodendroglial tumors, as the former are usually encountered in patients with long-standing partial complex seizures and often present at an earlier age (mean, 9 years). Moreover, grossly DNTs involve and often expand the cortex. 13,14 However, as with oligodendrogliomas, calcification occurs frequently, 13,14 and DNTs may exhibit histologic features of an oligodendroglioma between the nodular glioneuronal elements Thus, in fragmented biopsies, in which the characteristic nodular architecture has been distorted, DNTs may resemble oligodendrogliomas or oligoastrocytomas In addition, some DNTs may present in some- Arch Pathol Lab Med Vol 127, December 2003 Allelic Losses in Oligodendroglial Neoplasms Johnson et al 1573

2 Tumor Age, Mean y Table 1. Patient Characteristics* Sex, M/F Frontal Parietal Temporal Occipital Dysembryoplastic neuroepithelial tumor, WHO grade I 26 3/ Oligodendroglioma, WHO grade II 40 6/ Oligoastrocytoma, WHO grade II 40 2/ Anaplastic oligodendroglioma, WHO grade III 36 1/ Anaplastic oligoastrocytoma, WHO grade III 52 5/ * WHO indicates World Health Organization. Predominant site. what older individuals as frontal lobe masses with limited histories, and their oligodendroglial component may predominate, which can cause significant diagnostic dilemmas in small or piecemeal tissue samples. 14,15 To our knowledge, only limited fluorescence in situ hybridization (FISH) and polymerase chain reaction (PCR) analyses of 3 DNTs have been reported to date, and thus the genetic profiles and utility with multiple microsatellite markers on these tumors has not been established In the past several years, a growing body of evidence has suggested that, beyond the utility of LOH at 1p and 19q to predict survival in patients with oligodendrogliomas, molecular pathologic investigation may provide additional insight into the pathogenesis and treatment of these tumors. For example, anaplastic oligodendrogliomas with LOH at 1p and 19q are more likely to respond to PCV chemotherapy than those without such alterations, and there is evidence that even a partial oligodendroglial component in a mixed oligoastrocytoma improves response to therapy, as well as survival, compared with pure astrocytomas. 4 Similarly, recent studies by Zlatescu et al 19 and Mueller et al 20 have demonstrated a molecular genetic dichotomy in the distribution of oligodendroglial tumors of the cerebral hemispheres. Extratemporal oligodendrogliomas and oligoastrocytomas are more likely to have LOH at 1p and 19q, and a normal TP53, than are tumors found in the temporal lobes. 19,20 Given the slowly growing literature supporting the potential importance of molecular classifications for the diagnosis, treatment, and survival of patients with oligodendrogliomas, as well as the potential for regional heterogeneity of molecular parameters in these tumors, we sought to analyze a consecutive series of oligodendroglial and oligodendroglial-like tumors (DNTs) using a series of PCR-based microsatellite markers to compare and contrast the distribution of molecular subsets of these tumors. MATERIALS AND METHODS Tissue Samples Tumor tissue and blood were collected from a series of patients after intraoperative analysis identified an oligodendroglial or DNT-like component. Fresh tumor adjacent to tumor identified at frozen section analysis was submitted for PCR analysis. Only those specimens subsequently found on permanent sections to meet WHO criteria for oligodendroglial tumors or DNT were included in this study. 21 Grading was according to WHO criteria. 21 Patient characteristics are summarized in Table 1. Nearly all cases were reevaluated and diagnoses were confirmed by review at Vanderbilt University Medical School (Nashville, Tenn) or the Radiation Treatment Oncology Group. For most oligodendroglial neoplasms, fresh tumor and blood were collected prospectively at Vanderbilt University Medical Center at the time of initial surgery under Institutional Review Board approved protocols. For 1 anaplastic oligodendroglioma, 1 anaplastic oligoastrocytoma, and all DNTs, formalin-fixed, paraffin-embedded tumor was analyzed with uninvolved brain used as a control. DNA Extraction DNA was extracted from fresh brain tissue and corresponding peripheral blood using the PUREGENE kit (Gentra Systems, Minneapolis, Minn) according to the manufacturer s protocol. DNA from paraffin-embedded brain tissue was extracted using the PUREGENE kit with minor modifications, as follows. Ten 10- m-thick sections were collected, deparaffinized with xylene, washed with ethanol, and digested with cell lysis buffer and proteinase K (200 g/ml) for 16 to 60 hours. Treatment of the specimen with RNase was not performed. Proteins were precipitated and the DNA extracted with isopropanol. Glycogen was added as a carrier molecule as needed, and purified DNA was resuspended in varying amounts of PUREGENE hydration solution. Genotypes for multiple loci were determined by PCR amplification using fluorescent primers tagged with HEX, FAM, or TET (Applied Biosystems, Foster City, Calif) for markers on chromosome 1p (D1S80, FGR, D1S253, and D1S482); chromosome 17p (TP53 and D17S520) and chromosome 19q 13 (D19S178, APOC2, D19S601, and D19S180). Nucleotide sequences and mapping information were retrieved from the Human Genome Database. All amplifications were performed in 15-mL reaction volumes and included an initial denaturation at 94 C of6 minutes and a final extension at 72 C for 10 minutes. Polymerase chain reaction conditions for markers D1S253, D1S482, D19S178, APOC2, D19S601, D19S180, and D17S520 included 30 cycles with denaturation at 94 C for 45 seconds, annealing at 57 C for 45 seconds, and a 1-minute extension at 72 C. Polymerase chain reaction conditions for markers TP53 and FGR included 30 cycles of 94 C for 1 minute, 59 C for 1 minute, and 72 C for 1 minute. Polymerase chain reaction conditions for VNTR marker D1S80 were 35 cycles of 94 C for 1 minute, 65 C for 1 minute, and 72 C for 1.5 minutes. Amplicons were subjected to 5% polyacrylamide denaturing gel electrophoresis on an ABI-377 DNA sequencer and analyzed using GeneScan software. In DNA samples demonstrating heterozygosity for a particular marker, LOH was determined by measuring the peak height from each of the alleles produced from both the tumor and corresponding normal DNA. The formula (T1/T2)/(N1/N2) was applied, where T1 and N1 are the peak heights generated from PCR products from the smaller allele from tumor and normal tissue, respectively, and T2 and N2 are the peak heights generated from the PCR products from the larger allele of tumor and normal tissue, respectively. 22 Diagnostic criteria for LOH required the calculated ratios to be less than In cases in which the allele ratio was greater than 1.00, the ratio was converted to 1/[(T1/T2)/(N1/N2)] Arch Pathol Lab Med Vol 127, December 2003 Allelic Losses in Oligodendroglial Neoplasms Johnson et al

3 Table 2. Loss of Heterozygosity (LOH)* Tumor LOH at 1p (Inconclusive) LOH at 19q (Inconclusive) LOH at 17p (Inconclusive) Dysembroplastic neuroepithelial tumor, WHO grade I 0/4 (0/4) 0/4 (0/4) 0/4 (0/4) Oligodendroglioma, WHO grade II 6/12 (2/12) 6/12 (0/12) 3/12 (0/12) Oligoastrocytoma, WHO grade II 0/4 (0/4) 1/4 (0/4) 2/4 (0/4) Anaplastic oligodendroglioma, WHO grade III 0/3 (1/3 NI) 1/3 (0/3) 0/3 (0/3) Anaplastic oligoastrocytoma, WHO grade III 5/14 (0/14) 5/14 (0/14) 2/14 (0/14) * WHO indicates World Health Organization; NI, noninformative. Table 3. Heterozygous Not heterozygous Not determined Frequency of Heterozygosity for 1p, 19q, and 17p Markers in Loss of Heterozygosity Studies for Gliomas 1p, No. (%) D1S80* FGR* D1S253 D1S (70) 7 (21) 3 (9) 15 (45) 18 (55) 19q, No. (%) D19S178 APOC2 D19S601 D19S180* 26 (74) 9 (26) 30 (86) 5 (14) 27 (77) 7 (20) 1 (3) 24 (73) 9 (27) 17p, No. (%) TP53 33 (94) 2 (6) D17S (86) 5 (14) Total * These markers were not included in the initial analysis of 2 cases. Amplifications of DNA extracted from some paraffin-embedded tissue specimens were unsuccessful for markers with amplicons greater than 200 bp in length. RESULTS Thirty-three patients with an oligodendroglial tumor and 4 patients with DNT were examined (Table 1). Corresponding blood or normal brain tissues were analyzed in concert with the tumor for LOH at chromosomes 1p, 17p, and 19q. Findings are summarized in Tables 2 and 3. In WHO grade II oligodendrogliomas, 6 (60%) of 10 informative cases exhibited LOH at 1p, and 6 (50%) of 12 informative cases showed LOH at 19q (Figure 2); each of these cases also showed LOH at 1p. Three (25%) of 12 cases exhibited LOH at 17p for the dinucleotide repeat within the TP53 gene. In WHO grade II oligoastrocytomas, none of 4 informative cases showed LOH at 1p, 1 (25%) of 4 showed LOH at 19q, and 2 (50%) of 4 exhibited LOH at 17p, encompassing the TP53 gene. One case (25%) also displayed microsatellite instability at 3 of 8 loci (Figure 1). In the 3 anaplastic oligodendrogliomas, 1 case was noninformative, and neither of 2 informative cases showed LOH at 1p. All 3 cases were informative for the other markers, but none showed LOH for 17p; and 1 (33%) of 3 exhibited LOH at 19q. Of the 14 anaplastic oligoastrocytomas, 6 were designated astrocytoma predominant and 2 were oligodendroglioma predominant. Loss of heterozygosity was seen in 5 (36%) of 14 cases at 1p; 5 (36%) of 14 showed LOH at 19q, and 2 (14%) of 14 demonstrated LOH at 17p, encompassing the TP53 gene. Those with allelic loss of TP53 were astrocytoma predominant. Of the 4 DNTs examined, all were informative but none of the tumors exhibited LOH at 1p, 17p, or 19q. We found no microsatellite instability. The utility of the markers used in these studies is outlined in Table 3. Six (18%) of 33 patients were not heterozygous and thus not informative for markers D1S80 or FGR on 1p, and in an additional 2 patients, these markers were not used in the analysis. Thus for 8 patients, markers D1S253 and D1S484 were evaluated, and 4 patients were heterozygous for one or both markers. The utility of the markers used in this study is best measured by the frequency of heterozygosity for each marker. For 1p, although 70% of patients were heterozygous for marker D1S80, the amplicons derived from this variable number tandem repeat marker are large and are not successfully amplified in DNA extracted from paraffin-embedded tissue. Also, because of the large variation in size between alleles for this marker, interpretation of LOH must be made with caution, since allelic drop-out of the larger allele can happen during amplification and can lead to a false-positive result. For 19q, markers APOC2 and D19S601 had a high frequency of heterozygosity and were most useful. Marker APOC2 could provide a slightly better advantage than marker D19S601 for paraffin-embedded tissues, since most amplicons derived from APOC2 range from 125 bp to 170 bp, compared to those generated from D19S601, in which most range from 200 bp to 230 bp. For 17p, both TP53 and D17S520 were quite useful markers, with 94% and 86% of patients informative for each, respectively. In addition, both markers yield amplicons smaller than 150 bp, which is easily amplified from DNA extracted from paraffin-embedded tissues. Combining the data obtained for all markers at each chromosomal region, the evaluation of LOH in tumor tissue was successful in 32 (91%) of 35 patients for 1p; 34 (97%) of 35 patients for 17p; and 35 (100%) of 35 patients for 19q. COMMENT In this study, 6 (60%) of 10 informative oligodendrogliomas exhibited LOH at 1p, similar to the range of 33% to 73% described by others. 7,10 Six (50%) of 12 showed LOH at 1p and 19q, consistent with the previously reported range of 40% to 70%. 6,7,9,10 Only 1 of 3 temporal oligodendrogliomas exhibited LOH at 1p and 19q, while 5 of 9 oligodendrogliomas at other sites exhibited these changes. These findings are similar to those of Mueller et al, 20 who found LOH 1p and Arch Pathol Lab Med Vol 127, December 2003 Allelic Losses in Oligodendroglial Neoplasms Johnson et al 1575

4 Figure 1. Gel electropherograms generated following amplification of DNA extracted from blood (A) and tumor tissue (B) from a patient with an oligoastrocytoma. Microsatellite instability was observed in DNA from tumor tissue for 3 of 8 markers. DNA from peripheral blood produced FGR amplicons at 139-bp fragments (green peak), D19S180 amplicons at 165 and 169 bp (blue peaks), and D19S601 amplicons at 208 and 220 bp (green peaks). DNA extracted from the tumor tissue produced unique amplicons of 141 bp at the FGR locus, 171 bp at the D19S180 locus, and 216 bp at the D19S601 locus. The red line represents baseline, and the small peaks represent known DNA size standards. C, Oligodendroglioma histology with uniformly round nuclei. D, Zone of astrocytoma with elongated irregular nuclei (hematoxylin-eosin, original magnification 400). 19q less often in temporal oligodendroglial neoplasms. Although each of our cases had well-defined oligodendroglial areas, none of our oligoastrocytomas exhibited LOH at 1p, in contrast with a previous report of 50% to 52%, 7,22 and only 1 of our cases (25%) exhibited LOH at 19q, which is less than the 70% to 75% found by some. 7,22 Half of our oligoastrocytomas were astrocytoma predominant and exhibited LOH on chromosome 17 in the region of the TP53 gene, consistent with work by Maintz et al, 22 which demonstrated that at least 30% of oligoastrocytomas may show 1576 Arch Pathol Lab Med Vol 127, December 2003 Allelic Losses in Oligodendroglial Neoplasms Johnson et al

5 Figure 2. Gel electropherograms generated following amplification of DNA extracted from blood (A) and tumor tissue (B) from a patient with an oligodendroglioma. TP53 amplicons in similar amounts were generated from normal and tumor DNA (black peaks at 100 and 124 bp). In contrast, FGR amplicons generated from 1p (green peaks at 137 and 139 bp) and D19S601 amplicons generated for 19q (green peaks at 208 and 216 bp) show significantly reduced amounts of polymerase chain reaction products generated from FGR allele 139 and D19S601 allele 216 from tumor DNA. These findings are consistent with loss of heterozygosity at 1p and 19q with (T1/T2)/(N1/N2) ratios less than C and D, Oligodendroglioma. Tumor cells exhibit round nuclei with halos (hematoxylin-eosin, original magnifications 250 [C] and 400 [D]). genetic alterations similar to astrocytomas, that is, LOH at 17p but not at 1p or 19q. Although 1 case was noninformative, none of 3 anaplastic oligodendrogliomas exhibited LOH at 1p, and only 33% showed LOH at 19q. In contrast, previous studies have reported LOH at 1p in 58% and LOH at 19q in 66% of cases. 11 These differences likely reflect our smaller sample size and/or differences in inclusion criteria of the cases. 16 In these cases, the T1/T2 to N1/N2 ratios were essentially the same, and thus we do not feel that the lack of detectable LOH reflects contami- Arch Pathol Lab Med Vol 127, December 2003 Allelic Losses in Oligodendroglial Neoplasms Johnson et al 1577

6 nation of tumor sample with normal brain. Five (36%) of 14 anaplastic oligoastrocytomas exhibited LOH at both 1p and 19q. This number is somewhat lower than previous reports of LOH at 1p in 46% of cases and at 19q in 75%. 7 Collectively, our findings confirm the prevalence of LOH at 1p and 19q in oligodendroglial tumors, but they also suggest that in unselected cases defined by current criteria, LOH at 1p and 19q may be somewhat less frequent in oligoastrocytomas than previously reported. The reported prevalence of microsatellite instability in adult gliomas has varied from 1.9% 23 to 20% 45% of malignant astrocytic tumors. 24,25 The prevalence in oligodendrogliomas is as high as 17% 26 and has been noted in oligoastrocytomas. 27 In the present study, 1 oligoastrocytoma exhibited microsatellite instability, but this was not detected in oligodendrogliomas. This finding suggests that microsatellite instability does not contribute to the pathogenesis of oligodendrogliomas. Microsatellite instability was also absent in anaplastic oligodendroglial tumors, suggesting that it is not associated with progression. Similarly, in astrocytic tumors, microsatellite instability is not associated with progression. 28 It is widely recognized that simple DNTs contain neoplastic oligodendrocytes and that complex DNTs often have histologic oligodendroglioma in zones between glioneuronal elements. Nonetheless, these histologic features do not appear to significantly alter the favorable long-term prognosis of these WHO grade I tumors, which suggests that the biology of these tumor cells is different than that of oligodendroglioma cells. Findings presented here indicate that these tumors lack the allelic losses on chromosomes 1p, 17p, and 19q identified in oligodendrogliomas. These findings are identical to the recently reported FISH analysis by Perry et al. 16 Furthermore, just prior to submission of these findings, Prayson et al 17 reported that FISH analysis revealed an intact chromosome 1p in DNTs, and Fujisawa et al 18 reported negative PCR findings. Nonetheless, FISH cannot precisely determine LOH. The present study, which uses specific PCR analysis of alleles on key chromosomes implicated in the pathogenesis of most oligodendrogliomas, confirms and extends these findings. Our work and that of others suggests that the pathogenesis of DNTs is different than that of oligodendroglial neoplasms. Currently, recognition of DNTs and differentiation from oligodendrogliomas or oligoastrocytomas are largely contingent on clinical features and identification of nodular architecture associated with glioneuronal elements However, in small fragmented biopsies in which architecture is hard to assess, distinguishing between these entities may be difficult yet imperative owing to differences in prognosis and treatment. Our evaluation of several relatively small samples suggests that analysis of LOH may be helpful in distinguishing between DNTs and oligodendroglial tumors, in this setting, in at least some cases. Although the absence of LOH would not solve the diagnostic dilemma, identification of LOH at 1p, 17p, or 19q would suggest the presence of an oligodendroglial or oligoastrocytic neoplasm. Additional studies may clarify the utility of PCR analysis for LOH as a tool for differentiating these tumors. Although FISH is used in some centers, the present study illustrates the clinical utility of microsatellite markers and PCR for the detection of LOH in gliomas. In addition, PCR can detect small deletions and can specifically localize the regions contained in the deletion. This information could be critical in defining the genes involved in the pathogenesis of these tumors. For example, PCR, in contrast to FISH, has the potential to detect small deletions in the cyclin-dependent inhibitor gene CDKN2C at 1p32, which occurs in a limited number of oligodendrogliomas. 29 Identification of TP53 mutations, critical to the analysis of oligoastrocytomas, can only be achieved by PCR. Most mutations in TP53 are due to loss of 1 wildtype allele and mutations in the other, which is duplicated during mitotic recombination. Thus, PCR analysis would detect LOH, while FISH would detect 2 alleles and fail to detect an abnormality. 30 In contrast to FISH, PCR (1) allows for easy analysis of numerous allelic alterations, (2) allows for detection of inactivating mutations for TP53, (3) provides a permanent record of results, and (4) is free of in situ hybridization artifacts, 30 in contrast to FISH, in which the signal degrades with time depending on the light exposure. 30 Polymerase chain reaction analysis, however, appears less reliable in analysis of samples containing significant amounts of normal or reactive tissue. In this setting, PCR signals and determination of LOH by ratios may be confounded by unwitting amplification of normal alleles obscuring LOH. While this can be avoided by careful selection of tumor tissue (guided by smears), frozen sections, or microscopic selection of tissue to be removed from formalin-fixed, paraffin-embedded tissues, it remains a limitation in the analysis of biopsies from the edge of tissue and represents a scenario in which FISH analysis may be more useful. 30 In summary, the present study confirms previous observations of LOH at 1p and 19q in approximately half and two thirds of oligodendrogliomas, respectively. In addition, we found no LOH at 1p, 17p, or 19q in DNTs, raising the possibility that PCR evaluation using microsatellite primers for these markers, if positive, may facilitate differentiation of an oligodendroglioma from a DNT in a fragmented biopsy. This work, considered with prior studies, demonstrates the value and feasibility of routine LOH analysis in a clinical practice setting. References 1. Mork SJ, Halvorsen TB, Lindegaard KF, Fida GF. Oligodendroglioma: histologic evaluation and prognosis. J Neuropathol Exp Neurol. 1986;45: Glass J, Hochberg FH, Gruber ML, Louis DN, Smith D, Rattner B. The treatment of oligodendrogliomas and mixed oligoastrocytomas with PCV chemotherapy. J Neurosurg. 1992;76: Cairncross G, Macdonald D, Ludwin S, et al. Chemotherapy for anaplastic oligodendroglioma. J Clin Oncol. 1994;12: Kim L, Hochberg FH, Thornton AF, et al. Procarbazine, lomustine, and vincristine (PCV) chemotherapy for grade III and IV oligoastrocytomas. J Neurosurg. 1996;85: Mason W, Louis DN, Cairncross JG. Chemosensitive gliomas in adults: which ones and why? J Clin Oncol. 1997;15: Reifenberger J, Reifenberger G, Liu L, James CD, Wechsler W, Collins VP. Molecular genetic analysis of oligodendroglial tumors shows preferential allelic deletions on 19q and 1p. Am J Pathol. 1994;145: Kraus JA, Koopmann J, Kaskel P, et al. Shared allelic losses on chromosomes 1p and 19q suggest a common origin of oligodendroglioma and oligoastrocytoma. J Neuropathol Exp Neurol. 1995;54: Cairncross JG, Ueki K, Zlatescu MC, et al. Specific genetic predictors of chemotherapeutic response and survival in patients with anaplastic oligodendrogliomas. J Natl Cancer Inst (Bethesda). 1998;90: Smith JS, Alderete B, Minn Y, et al. Localization of common deletion regions on 1p and 19q in human gliomas and their association with histological subtype. Oncogene. 1999;18: Smith JS, Perry A, Borrell TJ, et al. Alterations of chromosome arms 1p and 19q as predictors of survival in oligodendrogliomas, astrocytomas and oligoastrocytomas. J Clin Oncol. 2000;18: Ino Y, Betensky RA, Zlatescu MC, et al. Molecular subtypes of anaplastic 1578 Arch Pathol Lab Med Vol 127, December 2003 Allelic Losses in Oligodendroglial Neoplasms Johnson et al

7 oligodendroglioma: implications for patient management at diagnosis. Clin Cancer Res. 2001;7: Louis DN, Holland EC, Cairncross JG. Glioma classification: a molecular reappraisal. Am J Pathol. 2001;159: Daumas-Duport C, Scheithauer B, Chodkiewicz J-P, Laws ER Jr, Vedrenne C. Dysembryoplastic neuroepithelial tumor: a surgically curable tumor of young patients with intractable seizures. Neurosurgery. 1988;23: Daumas-Duport C. Dysembryoplastic neuroepithelial tumours. Brain Pathol. 1993;3: Honavar M, Janota I, Polkey CE. Histological heterogeneity of dysembryoplastic neuroepithelial tumor: identification and differential diagnosis in a series of 74 cases. Histopathology. 1999;34: Perry A, Fuller CE, Banerjee R, Brat DJ, Scheithauer BW. Ancillary FISH analysis for 1p and 19q status: preliminary observations in 287 gliomas and oligodendroglioma mimics. Front Biosci. 2003;8:A1 A Prayson RA, Castilla EA, Hartke M, Pettay J, Tubbs RR, Barnett GH. Chromosome 1p allelic loss by fluorescence in situ hybridization is not observed in dysembryoplastic neuroepithelial tumors. Am J Clin Pathol. 2002;118: Fujisawa H, Marukawa K, Hasegawa M, et al. Genetic differences between neurocytoma and dysembryoplastic neuroepithelial tumor and oligodendroglial tumors. J Neurosurg. 2002;97: Zlatescu MC, TehraniYazdi A, Sasaki H, et al. Tumor location and growth pattern correlate with genetic signature in oligodendroglial neoplasms. Cancer Res. 2001;61: Mueller W, Hartmann C, Hoffman A, et al. Genetic signature of oligoastrocytomas correlates with tumor location and denotes distinct molecular subsets. Am J Pathol. 2002;161: Reifenberger G, Kros JM, Burger PC, Louis DN, Collins VP. Oligodendroglioma. In: Kleihues P, Cavenee WK, eds. World Health Organization Classification of Tumours of the Nervous System. Lyon, France: IARC; Maintz D, Fiedler K, Koopman J, et al. Molecular genetic subtypes for oligoastrocytomas. J Neuropathol Exp Neurol. 1997;56: Wooster R, Cleton-Jansen AM, Collins N, et al. Instability of short tandem repeats (microsatellites) in human cancers. Nat Genet. 1994;6: Izumoto S, Arita N, Ohnishi T, et al. Microsatellite instability and mutated type II transforming growth factor-b receptor gene in gliomas. Cancer Lett. 1997; 112: Dams E, van de Kelft EJZ, Martin JJ, Verlooy J, Willems PJ. Instability of microsatellites in human gliomas. Cancer Res. 1995;55: Zhu JJ, Santarius T, Wu X, et al. Screening for loss of heterozygosity and microsatellite instability in oligodendrogliomas. Genes Chromosom Cancer. 1998;21: Leung SY, Chan TL, Chung LP, et al. Microsatellite instability and mutation of DNA mismatch repair genes in gliomas. Am J Pathol. 1998;153: Gomori E, Fulop Z, Meszaros I, Doczi T, Matolcsy A. Microsatellite analysis of primary and recurrent glial tumors suggests different modalities of clonal evolution of tumor cells. J Neuropathol Exp Neurol. 2002;61: Husemann K, Wolter M, Buschges R, Bostrom J, Sabel M, Reifenberger G. Identification of two distinct deleted regions on the short arm of chromosome 1 and rare mutation of the CDKN2C gene from 1p32 in oligodendroglial tumors. J Neuropathol Exp Neurol. 1999;58: Fuller CE, Perry A. Fluorescence in situ hybridization (FISH) in diagnostic and investigative neuropathology. Brain Pathol. 2002;12: Arch Pathol Lab Med Vol 127, December 2003 Allelic Losses in Oligodendroglial Neoplasms Johnson et al 1579