| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
Clatterbridge Cancer Research Trust, J. K. Douglas Laboratories, Clatterbridge Hospital, Bebington, Wirral, CH63 4JY [C. W., K. A. J., Y. M., J. T-H.]; Walton Centre for Neurology and Neurosurgery, Liverpool, L97LJ [D. G. d. P., J. C. B., P. C. W.]; and Department of Neurological Science, University of Liverpool, Liverpool, L9 7LJ [D. G. d. P., S. A. A. H., P. C. W.], United Kingdom
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Design: We have used laser capture microdissection to sample the various histological phenotypes present in 42 tumors from 25 glioma cases with either inter- or intratumoral histological heterogeneity, and multiple simultaneous PCR amplification of microsatellite markers and capillary electrophoresis to determine allelic imbalance in chromosomes 1p, 19q, 17p, 10p, and 10q.
Results: Loss of 1p36 and 19q13 was seen only in oligodendroglial histology in 7 of 13 oligodendrogliomas. 17p13 loss was found in 14 of 41 tumors in astrocytic, oligoastrocytic, oligodendroglial, and glioblastomatous histologies. Chromosome 10 loss was seen in all of the high-grade histologies in 7 of 7 glioblastomas with an oligodendroglial component and in 1 of 5 low-grade oligodendroglial regions present within high-grade tumors. Seven tumors from 5 cases had no detectable losses of any markers investigated. In 13 tumors with intratumoral heterogeneity, identical genetic losses were present in all areas of histological differentiation. Additional losses were seen in some but not all of the histologies within 2 tumors and were associated with progression in 3 cases.
Conclusions: The gliomas in this study were more homogeneous in their genotype than their histological phenotype with regions of differing histological subtype indistinguishable by the genetic markers investigated, supporting a monoclonal origin of these tumors.
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Histopathological diagnosis of gliomas, based largely on morphological criteria as described in the WHO classification system (5) , is frequently challenging, particularly in those with an oligodendroglial component (6, 7, 8) . Gliomas frequently show considerable histological heterogeneity. Anaplastic changes in gliomas can be diffuse, or it may be possible to distinguish areas with low- and high-grade histological features that may be indicative of progression in some parts of the tumor to higher grade (9) . In addition, many gliomas show mixed histology with features of more than one histological subtype present within the same tumor sample or in samples obtained at different times (9, 10, 11) . These difficulties in the accurate histopathological classification of gliomas are compounded when the biopsy size is small or has been taken from the edge of a tumor diffusely infiltrating the surrounding normal brain tissue.
Recent research suggests that molecular classification may assist histopathological diagnosis and that gliomas with oligodendroglial or astrocytic differentiation arise via different histogenetic routes (6 , 12, 13, 14) . Astrocytomas often have loss of chromosome 17p and mutations in the p53 gene, whereas oligodendrogliomas have frequent losses of chromosomes 1p and 19q, and a low incidence of p53 mutation (12 , 13 , 15 , 16) . Oligoastrocytomas, which display features of both astrocytic and oligodendroglial differentiation, commonly show either p53 mutation or allelic losses of chromosomes 1p and 19q, indicating genetic subsets of oligoastrocytoma (17, 18, 19) . Evidence is now accumulating that some genetic alterations may have prognostic or predictive value. Allelic loss of chromosome 1p is emerging as a marker of chemotherapeutic response, and prolonged survival in oligodendrogliomas (20, 21, 22, 23) and may be associated with a more favorable clinical course in other high-grade gliomas (24) . Combined losses of 1p and 19q have been shown to predict prolonged survival independent of tumor grade in oligodendrogliomas (25) and to indicate durable responses to chemotherapy in anaplastic oligodendrogliomas (26) . Loss of chromosome 10 is common in high-grade astrocytomas and has been associated with poor prognosis in a number of studies (23 , 27, 28, 29) .
Identification of tumors with genetic losses in loci of potential prognostic or predictive value is likely to play a significant role in the future clinical management of glioma patients, but the relationship between these genetic losses and histological phenotypic heterogeneity is not yet fully understood. Whereas the available evidence suggests that gliomas are monoclonal in their origin (30, 31, 32) , there is also evidence for genetic heterogeneity within gliomas (30 , 33, 34, 35) . With the trend toward minimally invasive surgery and stereotactic biopsy, it is particularly important to determine whether genetic losses in prognostic or predictive markers are present throughout the tumor tissue and detectable in any sample, or whether only some regions of specific histological phenotype bear the aberration (36) .
In this study, we have investigated the relationship between histological heterogeneity and genetic alterations in a series of histologically heterogeneous gliomas using multiple simultaneous PCR amplification of microsatellite markers to determine allelic imbalance in loci on chromosomes 1p36, 19q13, 17p13, 10q22–26, and 10p11–15, and samples obtained by laser capture microdissection.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Regions of histological heterogeneity either within a single tumor or between samples obtained at different surgical procedures were noted. Between two and six regions with discrete histological appearance were identified for laser capture microdissection per patient; in total 83 regions of tumor tissue were investigated. The histopathological phenotype of these regions was classified according to the same WHO criteria, but categorized as if viewed in isolation from the tumor as a whole. Tumors with an overall diagnosis of oligoastrocytoma contained at least 25% oligodendroglial elements, whether distinct or admixed.
|
Microsatellite Analysis.
Investigation of allelic imbalance was carried out using paired normal and tumor samples as described previously (38)
, but using a two-round PCR procedure, followed by capillary electrophoresis (39)
. The microsatellite panel included the polymorphic markers: chromosome 1- D1S2667, D1S508, D1S214 (1p36.3); chromosome 19-D19S412, D19S112, D19S596 (19q13); chromosome 17-D17S796, D17S1176, D17S1353 (17p13); chromosome 10p-D10S89, D10S189, D10S179 (10p11–10p15); and chromosome 10q- D10S537, D10S1687, AFMa086wg9, D10s2491 (10q22–10q23) and D10S215, D10S583, D10S587, D10S1723, D10S212 (10q24–10q26). Twenty-µl first round PCR reactions contained 0.1 µM unlabelled primers for all of the microsatellite markers, 0.2 mM dNTPs, 1.5 mM MgCl2, 1x Hotstar PCR buffer, 0.5 units Hotstar DNA Polymerase (Qiagen Inc., Valencia, CA) and 10-µl tissue sample or 15-ng lymphocyte DNA purified from blood (QIAamp DNA minikit; Qiagen Inc.). The hot start enzyme was activated by incubation at 96°C for 13 min, followed by 20 cycles of 96°C for 30 s, 55°C for 30 s, 72°C for 30 s, and a 10-min extension at 72°C. Five µl of first round products diluted with 1:16 with H2O were used as target in the second PCR round using primers for a single microsatellite marker. Reverse primers were labeled at their 5' ends with fluorescent 6-carboxyfluorescein or 6-carboxy-2',7'-dimethoxy-4',5'-dichlorofluorescein. PCR conditions were identical to the first round except that 40 cycles were used. PCR was carried out in triplicate for each sample in the first round and in duplicate for each first-round PCR product in the second round.
DNA from HN5 and A431 cell lines (40) , purified by standard phenol chloroform extraction techniques, was mixed in known proportions ranging from 100% HN5:0% A431 to 0% HN5:100% A431 before assay and 0.3–15 ng used as target in PCR reactions using either a single PCR round with primers for individual microsatellite markers (38) or the two-round procedure that included all 21 of the primers in the first round and individual primers in the second round.
PCR products were mixed into panels according to expected amplicon size range and fluorescent label to permit discrimination of individual markers after electrophoretic separation using a 96-lane capillary electrophoresis system (MegaBACE; Molecular Dynamics). Six panels were used for the electrophoresis of the 21-microsatellite markers investigated. Sixteen microsatellite markers and four panels were used for some samples. A 96-well PCR plate format and transfers using eight channel pipettes were used throughout. Individual PCR products (2.5 µl) were mixed (2–5 markers/panel) and 2.5 µl of the mixture added to 2.5 µl H2O, 0.5 µl MegaBACE ET 400-R size standard (Amersham Pharmacia Biotech Inc.), and 4.5 µl MegaBACE loading solution (Amersham Pharmacia Biotech Inc.). Samples were incubated at 90°C for 70 s and cooled rapidly on ice before capillary electrophoresis through MegaBACE long-read matrix.
After electrophoresis, data analysis was carried out using Genetic Profiler (Molecular Dynamics) and Excel (Microsoft) software to calculate peak height ratios. Where allelic imbalance was observed rather than complete loss, the ratio of the mean peak heights of the two alleles was calculated for normal (N1/N2) and tumor (T1/T2). Allelic imbalance was calculated from the ratio of the tumor signal to that of the normal signal T1/T2 over N1/N2. Complete loss or ratios of <0.33 indicated loss of heterozygosity at that locus. Partial loss was indicated if significant skewing occurred (ratio >0.33–<0.75) and P < 0.05 when tested by the Mann Whitney test (38) .
p53 Mutation Detection.
For p53 mutation detection a novel methodology developed for use with small amounts of microdissected tissue was used as reported previously (38
, 40)
.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. These results suggest that the assay could be used to detect allelic imbalance when the tumor component is 20% or more of the tissue sample. Allelic imbalance detected previously in tissues using single microsatellite markers and a one-round PCR procedure (38)
was found using the multiple simultaneous amplification and two-round procedure.
|
. Seven of 8 cases that presented as low-grade gliomas had oligodendroglial or oligoastrocytic features and 5 had progressed to higher grade. All of the cases that presented as grade III tumors and 7 of the glioblastomas had an oligodendroglial component. Three cases had tumors classified as gliosarcomas. In total 83 regions of discrete histological phenotype were identified for microdissection and genetic analysis. Twenty-two of 25 of the cases studied had regions of histology with an oligodendroglial or an oligoastrocytic phenotype represented in 44 of these regions. Three tumors presented in childhood, and 1 had a clinical history suggesting development during childhood.
Analysis of Allelic Imbalance.
Allelic imbalance on chromosome 1p36 was investigated in 74 loci of which 82% were informative. Sixty-nine polymorphic loci were examined on 19q13, of which 71% were informative. Of the microsatellite markers, 79%, 69%, and 80% were informative in the 73, 201, and 66 loci investigated for 17p13, 10q22–26, and 10p11–15, respectively. The marker panel chosen yielded at least one informative marker for all of the chromosomal regions for all but 1 case, for which analysis of 19q13 was not possible.
Genetic Alterations and Histopathological Diagnosis.
The relationship between the overall histopathological diagnosis and genetic alterations is given in Table 2
and Fig. 2
. Loss of both 1p36 and 19q13 was found only in gliomas with "pure oligodendroglial features" and was seen in the presenting tumor in 4 cases, but only after progression to an anaplastic tumor in case 3. No tumor showed loss of 19q13 only, but loss of chromosome 1p36 markers without 19q13 loss was seen in a glioblastoma and its gliosarcomatous recurrence (case 18). Chromosome 17p13 loss was detected in 8 cases and in tumors of all histologies. In 5 of these, 17p13 loss was evident in the presenting tumor. In contrast, a glioblastoma recurrence showed complete loss of two informative chromosome 17 markers, whereas only partial loss of one marker was detected in the presenting anaplastic oligodendroglioma (case 11). 17p13 loss was accompanied by loss of 19q13 in an astrocytoma (case 1), and by loss of 1p36 and 19q13 in a grade III oligodendroglioma (case 14). Chromosome 10 loss was only found in high-grade gliomas. Ten cases showed loss of both arms of chromosome 10, whereas an anaplastic oligodendroglioma showed loss of 10p11–15 associated with progression to a glioblastoma but retained 10q22–26 (case 10). An oligoastrocytoma grade III (case 16) and a glioblastoma (case 21) had loss only of the q arm. Five cases did not show detectable loss of any of the chromosomal loci investigated in the presenting glioma. Four of these had further surgery, two showed no losses of the investigated loci in the second surgical sample, and two cases (3 and 10) had losses associated with progression. Case 15 showed partial loss of two markers, but this finding could not be substantiated in the postmortem tumor because of failed PCR amplification.
|
|
|
, and the results of genetic analyses are illustrated in Fig. 2, A–C
. Genetic losses of chromosomes 1p36, 19q13, and 17p13 if present were found in all regions of differing histology within a single tumor and were retained in all histologies of the recurrences, with two exceptions (cases 1 and 25). In these cases, additional genetic losses were seen in some but not all regions of similar histological grade within a single tumor. Additional loss of a chromosome 17 marker was seen in the glioblastoma and undifferentiated components but not in the gemistocytic component of the fourth surgical sample for case 1. Loss of chromosome 19q13 markers was seen only in the undifferentiated component of a gliosarcoma (case 25).
|
Of the 25 tumors that showed more than one histological phenotype within a single tumor, 16 tumors from 15 cases showed regions of oligodendroglial histology as well as regions typical of high-grade astrocytomas. Of these, 4 cases had no detectable losses of the markers investigated, whereas 3 cases had losses of chromosome 17p13, and 1 had loss of 17p13, 1p36, and 19q13 present in oligodendroglial and more astrocytic components. The 7 glioblastomas with oligodendroglial features showed only loss of chromosome 10 in all of the high-grade histologies irrespective of histological differentiation. None of the cases with a histopathology diagnosis of "oligodendroglioma" and 1p36/19q13 loss had regions of more astrocytic histology.
Identical genetic losses were seen in both the gliomatous and sarcomatous components of 2 gliosarcomas (cases 18 and 20) and in undifferentiated tumor tissue or regions of bone, epitheliod, or rhabdoid differentiation in case 25.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
At the time of starting this study, the impact of morphological heterogeneity on histopathological diagnosis of gliomas and the relationship with molecular genetic alterations was uncertain. Few previous studies had used laser capture microdissection to address issues of the regional heterogeneity of genotype and phenotype. In selection of cases for this study, we concentrated on tumors with distinct regions of histology that if sampled in the absence of other tissue would be given a histopathological classification that differed from the overall pathology classification. These regions might otherwise be ignored in the diagnosis of large tumor specimens but might lead to erroneous diagnosis if only small biopsy samples were available. We wished to determine whether these tumors were as heterogeneous in their genotype as in their phenotype.
We have used an approach based on laser capture microdissection and multiple simultaneous PCR amplification of microsatellite markers. The method has high sensitivity, being capable of detection of tumor cells present in tissues in minor proportions, whereas capillary electrophoresis enables relatively high throughput. Allelic imbalance reflects the average of the cells within the tissue samples compared, but cytogenetic heterogeneity such as polyploidy or aneuploidy is not detected. Similar multiplex approaches have been used successfully in the genetic analysis of breast and other cancers (39 , 43) without the need for whole genome amplification (44) . The use of a single microdissected sample for analysis of multiple loci enables PCR from higher initial target amounts, minimizing the possibility of allele-specific PCR, which may result from low concentrations of target DNA, particularly when formalin-fixed, paraffin-embedded tissue is used. In addition, replication of PCR reactions allows assessment of reproducibility of the analysis. Microdissected "uninvolved brain tissue," such as cortex, white matter, thrombosed blood vessels, and dura, may be used as a source of constitutional DNA and, in the absence of a blood sample, enables analysis of cases with a paucity of "normal" tissue. This approach is applicable to glioma samples from the range of surgical procedures and standard fixation techniques, and could be adapted for diagnostic use.
Because of the selection of cases based on histological heterogeneity rather than overall histopathological diagnosis, the numbers in each diagnostic group are small, preventing statistical comparisons between groups or with other studies. As in other studies, the molecular classification did not show a high concordance with the histopathological diagnosis. However, combined loss of 1p36 and 19q13 was found only in gliomas with a histopathology diagnosis of "pure" oligodendroglioma, in 54% of these tumors. Oligodendrogliomas that lacked 1p36/19q13 loss, or regions of oligodendroglial histology in oligoastrocytomas or glioblastomas, showed either loss of 17p13 or no detectable losses of the 1p, 19q, or 17p markers investigated. Similarly, oligoastrocytomas or regions of oligoastrocytic or astrocytic histology showed either loss of 17p13 or no detectable losses of the 1p, 19q, or 17p markers investigated. Some genetic alterations occur early in tumorigenesis and would be expected in virtually any areas of tumor, whereas others occur later in progression and may show greater regional variability (45 , 46) . Losses of 17p13 and 1p36/19q13 are considered early genetic events in gliomagenesis (47 , 48) . In this study, these genetic alterations, if present, were detected in all histologies of the earliest clinical sample and were retained in all of the histologies of subsequent samples, except for 2 cases, where progression to higher grade was associated with loss of 1p36/19q13 or 17p13. A number of morphologically heterogeneous oligodendrogliomas or oligoastrocytomas of grades II and III in temporal or nontemporal locations failed to show loss of any of the markers investigated, despite the use of a sensitive assay and microdissection. Similarly, absence of either 1p/19q loss or p53 mutation has been reported in astrocytomas (42) and in oligodendrogliomas/oligoastrocytomas predominantly with a temporal location (18) , suggesting additional genetic pathways in their histogenesis. Alternatively, other means of inactivating the p53 or 1p/19q pathways, or alterations such as P16/CDNK2A deletions or EGFR2 amplification, as in de novo glioblastomas (49) or anaplastic oligodendrogliomas (23 , 26) , may be important in genesis of these tumors.
Other previous studies have investigated genetic alterations in regions of oligodendroglial or astrocytic histology separated by microdissection, but thus far only in a few cases. Identical losses of 1p and 19q were found in both histologies in three oligoastrocytomas investigated by Kraus et al. (17) , and more recently in four biphasic oligoastrocytomas analyzed by Dong et al. (30) , indicating a monoclonal origin. Similarly, Watanabe et al. (42) showed loss of 1p and 19q in both histologies in one low-grade diffuse astrocytoma with a small oligodendroglial component, and the same p53 mutation in both histologies in an another. Dong et al. (30) found additional genetic losses in either or both components of five oligoastrocytomas that showed 1p and 19q loss in both oligodendroglial and astrocytic histology. They propose that these tumors are monoclonal in origin with heterogeneous genetic changes arising during clonal expansion. In contrast, two oligoastrocytomas had divergent allelic loss patterns in the two histological components, suggesting that these cases may have had a biclonal origin (30) . In the present study, the tumors investigated were more homogeneous in their genotype than in their histological phenotype. Regions of differing histology showed identical genetic analyses, and regions of differing histological subtype could not be distinguished by the genetic markers investigated, supporting a monoclonal origin.
As in other studies, identical genetic alterations were observed in regions of differing histological differentiation within gliosarcomas (50, 51, 52) , including a case with regions of uncommon sarcomatous phenotype. In addition to chromosome 10 loss, one of the gliosarcomas had loss of 1p36, which has been associated in glioblastomas with prolonged survival (24) ; however, this case survived only 8.5 months after diagnosis of the presenting glioblastoma.
Chromosome 10 loss, if present, was found in of the high-grade histologies irrespective of histological subtype within a tumor specimen and was detected in some but not all of the low-grade oligodendroglial histologies present within these high-grade gliomas. These results are in keeping with chromosome 10 loss associated with progression from low-grade histology and clonal expansion of cells, which have acquired additional genetic alterations as found in astrocytomas/glioblastomas (34 , 38 , 53 , 54) . Alternatively, low-grade regions with or without detectable chromosome 10 loss may arise from neoplastic cells diffusely infiltrating the surrounding normal tissue. Irrespective of the cause, for accurate detection of genetic changes associated with advanced gliomas, samples must be obtained from anaplastic regions of the tumor.
GBMOs have been subdivided genetically into two groups: (a) those that have losses in chromosome 10 and amplification of chromosome 7, typical of the astrocytic lineage; and (b) those that have losses in chromosomes 1p and 19q, typical of the oligodendroglial lineage (55) . In the GBMOs investigated by He et al. (56) and Kraus et al. (57) , 7 of 25 and 1 of 13 had combined losses of chromosomes 1p and 19q. In the former study, 64% had loss of 10q in tumors with or without 1p/19q loss. In contrast, none of the GBMO investigated by Kraus et al. (57) had loss of 10q. In the present study, all of the glioblastomas with regions of oligodendroglial-like histology, even the case whose second surgical sample showed only anaplastic oligodendroglial tissue, showed losses of chromosome 10, and no evidence of 1p and 19q loss. Small cells with round nuclei in glioblastomas are notoriously difficult to distinguish from true oligodendroglial cells, and there is considerable morphological overlap between GBMOs and small cell glioblastomas (58, 59, 60) . The small cell phenotype in glioblastomas has been associated with EGFR amplification (60) . Interestingly, the glioblastomas with regions of oligodendroglial histology in our study showed immunocytochemical overexpression of EGFR. Together with the loss of chromosome 10, this suggests that these tumors are likely to be glioblastomas with a "small cell" component, rather than oligodendroglial. The cortical infiltrating edge of otherwise classical glioblastomas often contain clusters or isolated, less anaplastic cells with oligodendroglial appearances. The demonstration in such tumors of a shared high-grade phenotype (loss of heterozygosity 10p and 10q), and absence of 1p and 19q loss in both the solid anaplastic glioblastoma areas and the infiltrating cortical "oligodendroglial" component, suggests that other factors such tissue and microenvironmental effects may influence tumor cells to adopt an oligodendroglial-like appearance. Alternatively, morphological heterogeneity may reflect epigenetic phenomena.
In this study, we have developed a convenient and relatively high-throughput approach for analysis of allelic imbalance at multiple clinically relevant loci appropriate for use where the amount of normal or tumor DNA available for analysis may be limiting, as in the analysis of small glioma clinical samples, such as serial stereotactic biopsies, or when microdissection is essential to enrich the sample for either tumor or normal component. Controversy continues to surround the histopathological diagnosis of "mixed gliomas," and whether or not these tumors are mixed at the genetic level as well as the histopathological level requires additional study (36) . This issue becomes particularly important with the trend toward minimally invasive surgery and small biopsies. Stereotactic biopsies, which target potentially anaplastic tissue, even if obtained serially may under-represent other aspects of histological heterogeneity, and the impact of small sample size on the molecular classification has not been investigated extensively. Whereas histological heterogeneity is more likely to be evident in samples from large open resections, genetic analysis using extracted DNA from these tissues may not yield a true picture of genetic heterogeneity. Through laser capture microdissection, we have shown that the gliomas investigated were more homogeneous in their genotype than their phenotype, with regions of differing histology showing identical genetic losses of the markers investigated. On the basis of these results, genetic analysis of small biopsies is likely to yield a reliable molecular classification for these markers, provided the aggressive component is sampled, as in serial stereotactic biopsy. In this study, many tumors showed features of oligodendroglial differentiation in the absence of the -1p/-19q genotype. Additional investigation of factors that influence the phenotypic appearance of these tumors is essential in the interests of accurate diagnosis to aid therapeutic management.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Supported by Clatterbridge Cancer Research Trust.
1 To whom requests for reprints should be addressed, at Clatterbridge Cancer Research Trust, J. K. Douglas Laboratories, Clatterbridge Hospital, Bebington, Wirral, United Kingdom CH63 4JY. Phone: 44-151-343-4304; Fax: 44-151-343-1820; E-mail: carolw{at}ccrt.co.uk 
2 The abbreviations used are: EGFR, epidermal growth factor receptor; GBMO, glioblastomas with oligodendroglial component.
Received 1/ 7/03; revised 6/ 2/03; accepted 6/ 4/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. P. Mason and J. G. Cairncross Invited Article: The expanding impact of molecular biology on the diagnosis and treatment of gliomas Neurology, July 29, 2008; 71(5): 365 - 373. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Lavon, D. Zrihan, B. Zelikovitch, Y. Fellig, D. Fuchs, D. Soffer, and T. Siegal Longitudinal Assessment of Genetic and Epigenetic Markers in Oligodendrogliomas Clin. Cancer Res., March 1, 2007; 13(5): 1429 - 1437. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. Jenkinson, D. G. du Plessis, T. S. Smith, K. A. Joyce, P. C. Warnke, and C. Walker Histological growth patterns and genotype in oligodendroglial tumours: correlation with MRI features Brain, July 1, 2006; 129(7): 1884 - 1891. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Walker, B. Haylock, D. Husband, K. A. Joyce, D. Fildes, M. D. Jenkinson, T. Smith, J. Broome, D. G. du Plessis, and P. C. Warnke Clinical use of genotype to predict chemosensitivity in oligodendroglial tumors Neurology, June 13, 2006; 66(11): 1661 - 1667. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Jenkinson, T. S. Smith, K. Joyce, D. Fildes, D. G. du Plessis, P. C. Warnke, and C. Walker MRS of oligodendroglial tumors: Correlation with histopathology and genetic subtypes Neurology, June 28, 2005; 64(12): 2085 - 2089. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Walker, D. G. du Plessis, D. Fildes, B. Haylock, D. Husband, M. D. Jenkinson, K. A. Joyce, J. Broome, K. Kopitski, J. Prosser, et al. Correlation of Molecular Genetics with Molecular and Morphological Imaging in Gliomas with an Oligodendroglial Component Clin. Cancer Res., November 1, 2004; 10(21): 7182 - 7191. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
, no loss;
, loss of the upper allele;
, loss of the lower allele; 
, partial loss; -, noninformative; •, not done. The shaded areas illustrate genetic discrepancies between samples from within the same tumor or between tumor specimens obtained at different surgical episodes. In case 16, tumors investigated were from (a) left frontal and (b) right parasagittal locations.
