• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Flow-Cytometric DNA Analysis of Intracranial Tumors in Children

Posted on: Tuesday, 14 September 2004, 06:00 CDT

Abstract

The objective of this study was to investigate flow-cytometric DNA values of pediatric intracranial tumors, and to establish DNA analysis as a potential prognostic parameter.

Twenty-nine brain tumor specimens from 26 pediatric patients were cryo-preserved within a 3-year period. The DNA content was measured by flow cytometry.

Six of the tumor specimens had aneuploid DNA patterns. The median of the proliferation index was lower in the survivor group compared with the non-survivor group (36.4% and 47.5%, respectively). Ten of the 26 patients are still alive, eight were lost to follow up, and eight died.

Flow-cytometric DNA analysis may be a helpful tool for examining brain tumors in children. The small size of this study could not establish flow cytometry as a definite prognostic factor, but further prospective multicenter studies will evaluate the prognostic significance of flow-cytometric DNA analysis.

2004 Elsevier GmbH. All rights reserved.

Keywords: Intracranial tumors; DNA analysis; Ploidy; Flow cytometry; Children

Introduction

In Germany, the annual incidence of malignant disease in children younger than 15 years is 12-14 per 100,000 [8]. In this age group, the second most common cause of death is malignant neoplasia, most frequently leukemia, and second most frequently tumors of the central nervous system. The etiology of most tumors of the central nervous system is unknown. Previous radiation therapy increases the risk of developing a brain tumor [8].

Regardless of the malignancy of an intracranial tumor, surgery is the preferred therapy in most cases. Chemotherapy and radiation are additional important parts of treatment. Prognosis of a brain tumor depends on the type of tumor, its localization, its size, and its grade. Histological grading is often quite complicated and depends on the experience of the examining pathologist. Therefore, there is still a great need for new reliable prognostic parameters.

DNA analysis of tumors has become especially important over the last few years. In some cases, results have prognostic relevance and influence therapy [1,20]. Using flow-cytometric DNA analysis, we examined nuclei regarding DNA content and proliferation. The advantages of this method lie in the fact that many cells can be analyzed in a relatively short period of time, and that the results are reproducible. In most tumors, loss of chromosomes or doubling of whole chromosomes is relatively common due to chromosomal instability [14]. Some of these chromosomal changes could be detected by flow-cytometric analysis.

The aim of our study was to better understand the biological behavior of this heterogenous group of tumors by flow-cytometric analysis. In addition, we checked the prognostic value of the chromosomal changes and tried to correlate this with the previously most important factors, i.e., histology, clinical prognosis, localization and tumor size.

Patients and methods

The flow-cytomclric DNA analyses were conducted prospectively examining 28 tumor specimens from 25 consecutive children who were operated on in the Department of Neurosurgery of the University Hospital in Mainz between 1987 and 1990, and from which fresh, unfixed specimens were available. The samples were preserved in liquid nitrogen directly after surgery.

Eleven patients were female, 14 were male, and age ranged from 4 months to 15 years (median 4.8 years). The follow-up of the patients is 12-15 years.

DNA measurement

Preparation of the tumor

The samples were thawed, cut, stained, and examined pathologically to make sure that the sample contained tumor cells. A section of the tissue sample was placed in phosphate buffered saline (PBS), and then passed through a metal sieve. After washing two times in PBS, the samples were ccntrifuged, and the suspension was dissolved in 5ml 0.1% (v/v) Triton-X-100 in 0.1% (m/v) citric solution. The nuclei were prepared by repeatedly loading the suspension into one way syringes (10 ml). The generated suspension of the nuclei was finally passed through a 20 m nylon gauze to filter out larger particles.

DNA staining

For staining, we incubated 100 l of the nuclei suspension with 250 l RNAse (500 Kunitz units) in a water bath for 15 min. Afterwards, 300 l propidiumiodide (PI) solution was added, generating a final concentration of 50 g/ml. Within 4h, we analyzed the nuclei suspensions in the flow cytometer. From each tumor sample, we analyzed 20,000-40,000 nuclei. To standardize the DNA measurement, we prepared nuclei from chicken erythrocytes and calibrated them with nuclei from normal brain tissue (the DNA content of the nuclei from chicken erythrocytes amounts to 35% of the DNA content from human nuclei). Every tumor sample was measured both without nuclei from chicken crythrocytes for cell-cycle analysis, and with nuclei from chicken erythrocytes to evaluate ploidy. Cell-cycle analysis was conducted with commercial software (Cellfit, Bekton Dickinson).

Results

Description of the patients

Table 1 shows demographical data, histological diagnosis, grading, DNA ploidy, and the number of cells in the cell cycle (G^sub 0/1^, S-, G^sub 2^/M-Phase). Tumor identification was done as follows: astrocytoma (n = 10), ependymoma (n = 6), medulloblastoma (n = 3), craniopharyngioma (n = 2), hamartoma (n = 2), and 1 gangliocytoma, 1 ganglioglioma, 1 ganglioneuroma, 1 malignant orbital tumor, 1 mesenchymal tumor, and 1 oligodendroglioma.

Astrocytoma

At the time of surgery, the median age of the patients with an astrocytoma was 8.3 years, with the youngest patient 2.2 and the oldest patient 16.1 years old. Fig. 1 shows the DNA histogram of the euploid and the three non-euploid astrocytomas. One tumor showed two aneuploid cell lines (Fig. 1B).

Medulloblastoma

Of the three medulloblastomas examined, one showed an euploid pattern, whereas the others had a tetraploid DNA pattern. The proliferation index of the euploid tumor was 32%, which indicates high proliferative activity. Fig. 2 shows the DNA histograms of the three medulloblatomas.

Ependymoma

A total of six ependymomas were examined. The median age of the patients at the time of surgery was 3.9 years (1.6 11.5). All ependymomas showed an euploid DNA pattern. Three of them had an S- phase-portion of more than 25%, which indicates high proliferative activity.

Craniopharyngioma

Two girls aged 3 and 7 years suffered from a craniopharyngioma. One of the tumors showed an S-phase-portion of 62.9%, which indicates very high proliferative activity.

Orbital tumor

The orbital tumor of a 4-month-old girl was a malignant and histologically undifferentiated tumor that could not be differentiated any further by our Department of Neuropathology. We found a high proliferation index of 47.5% (sum of S- and G^sub 2^M- phase), which is consistent with the histologically documented malignancy of the tumor.

Table 1. Histology, tumor grade, age at diagnosis, and flow cytometric parameters of the tumors

Mesenchymal tumor

The mesenchymal tumor was a non-neuroectodermal tumor of a 15- year-old girl. Because of the proximity to the Plexus choroideus, the tissue was similar to a meningioma of intraventricular origin. We found an euploid DNA pattern with a low proliferation index.

Oligodendroglioma/ganglioglioma

Both tumors originated from the same patient. He first underwent surgery at the age of 3 years. The histological examination revealed a cystic oligodendroglioma with malignancy grade II. Also, a central neuroblastoma was considered. The flow-cytometric analysis revealed a non-euploid tumor with a small portion of nuclei with aneuploid DNA content. Eleven months after the first surgery, the patient relapsed, developing a tumor that was very heterogene us histologically. In some parts, a neuronal (specifically a gangliocytic) component predominated, whereas in other parts, we found glial components, partly astrocytic and partly similar to an oligodendroglioma, with small multiple calciferous hardenings. The grade of malignancy was II, and also partly grade III. The flow- cytometric analysis revealed a tetraploid tumor with a predominance of tetraploid cells. The cell line with the higher euploid DNA content (1.23 fold) was no longer delectable.

Fig. 1. DNA histogram of an euploid (A) and three aneuploid astrocytomas (B-D). G^sub 0/1^=cells in G^sub 0/1^-phase of cell cycle, G^sub 2^/M = cells in G^sub 2^/M-phase of cell cycle, -d = diploid cell line, -a = aneuploid cell line, -a^sub 1^ = aneuploid cell line 1, -a^sub 2^ = aneuploid cell line 2, and FI = fluorescence intensity.

Fig. 2. DNA histogram of three medulloblastomas: (A) diploid tumor, (B) and (C) tetraploid tumors, G^sub 0/1^= cells in G^sub 0/ 1^-phase of cell cycle, G^sub 2^/M = cells in G^sub 2^/M-phase of cell cycle, -d= diploid cell line, -t = tetraploid cell line, and FI = fluorescence intensity.

Hamartoma

The sample originated from a 3-year-old girl undergoing surgery for the first time. The DNA histogram showed an euploid DNA pattern with two peaks and a proliferation index of 24.9%. Eight months later, she was operated on for a relapse. Its histogram was nearly identical, with a size almost similar to that of the S- and G^sub 2^M-phase.

Gangliocytoma

This tumor originated from a 4-month-old boy with right-sided hemimegalencephaly. The tumor showed an euploid DNA pattern, and the proliferation index was only 5.9%, which is the lowest value ofall tumors measured. This is consistent with the fact that during the histological examination, only a few cells were considered to be malignant.

Clinical data

Ten of the 26 patients are still alive; eight have died, and the other eight patients have been lost to follow-up. The median proliferation index in the group of deceased patients is 47.5%, and that in the followed-up group of survivors is 36.4%. The difference is not statistically significant. There was no patient with a non- cuploid tumor in the group of the followed up survivors, whereas two of the dead children had such a tumor.

Discussion

Flow-cytometric DNA analysis is a reproducible, semi- quantitative method for investigating cell characteristics. Numerous tumors have been studied using this method for determining DNA content. Some types of genetic instability, i.e., loss or doubling of chromosomes could be detected rapidly and without great effort [2]. Since the publication of Hedley et al. in 1983 [9], retrospective studies based on paraffin-embedded tumor material stored in archives have often been undertaken. Rainwater et al. [17] showed that DNA content is a prognostic parameter regarding estimated failure of chemotherapy in grade III-IV Wilms-tumors. Schmidt et al. [19] gave evidence of a tendency towards low survival rate in patients with nephroblastomas, when the tumors had aneuploid cell lines or high proliferation indices. However, these results could not be confirmed statistically. Feichter et al. [3] reported that the DNA content of the 500 different adult tumors they measured constituted additional important malignancy criteria. A study conducted by Kiss et al. [12] documented a connection between ploidy and the cell colony structure of glioma in vitro.

Until now, intracranial tumors, especially in childhood, have been examined only on a small scale [1,5,15,18] often together with a series of adult intracranial tumors [5,7]. This is primarily due to the small number of cases of different brain tumors in children. In this study, we exclusively analyzed data from patients of our clinic. Thus, in the observation period of over 3 years, only one heterogeneous group of brain tumors could be analyzed.

Statistical calculations of patient survival rates were not performed owing to the relatively small number of cases of tumors in individual groups.

Three parameters could be examined by flow-cytometric analysis: aneuploidy, DNA index, and proliferation index [10].

The fiow-cytometric DNA analysis of 10 astrocytomas from children showed a total of three aneuploid tumors, a somewhat higher portion than in the study of Fitzgibbons et al. [4], in which three out of 13 astrocytomas were defined as aneuploid. However, the authors used tissue samples stored in paraffin, applying a method that has a lower capability of discriminating cell populations in close proximity. Therefore, it is possible that this study may have overlooked some aneuploid tumors. Forsyth et al. [5] also conducted a retrospective study with paraffin-embedded tumors. No association was found between pathological features, such as histological grade or flow-cytometric characteristics and patient survival. Investigating 57 fresh glial tumors, Mathew et al. used flow- cytometric analysis and found 46 diploid, four near diploid, and seven hyperdiploid tumors. They concluded that "ploidy possesses no independent prognostic significance when survival is stratified by tumor grade" [15].

During the study period, only three medulloblastomas were available for flow-cytometric analysis. In the medulloblastomas they investigated, Frederiksen et al. [6] detected an aneuploid DNA arrangement. Hoshino et al. [11] analyzed two medulloblastomas, one of which was aneuploid and the other euploid. In one of their series of 281 brain tumors, Garcia et al. found three medulloblastomas, two of which were aneuploid and one diploid [7]. Gajjar et al. studied 34 consecutive fresh medulloblastoma specimens. They found that ploidy and the clinical risk group are important prognostic factors. The authors stated that due to the low number of cases, further conclusions are speculative. Zerbini et al., investigating 58 cerebellar medulloblastomas, concluded that the adequacy of radiation dose and DNA ploidy were the most important prognostic factors in their series [24]. In another study, paraffin sections from 32 patients with medulloblastoma were analyzed by flow cytometry [21]. In this series, no prognostic value of ploidy or proliferation index could be found. The extent of resection and DNA ploidy together had some prognostic value in another series of 53 medulloblastomas [23] and in a series of 26 infants [22]. Together, these publications show heterogeneous findings in terms of prognostic value of llow-cylometric DNA analysis. The heterogeneity of patient groups and methods used to perform these studies may be responsible for these results.

All of the ependymomas studied here contained a diploid DNA arrangement, regardless of their grade of malignancy. This could be evidence that grades of ploidy cannot be used as prognostic criteria for ependymomas. By contrast, Kotylo el al. reported that seven out of 17 paraffin-embedded tumors had aneuploid DNA content [13], and a relationship between euploidy and poor prognosis could be shown. However, Reyes-Mugica et al. [18] found no statistical correlation regarding DNA content, survival rate, and histology in the 22 ependymomas studied.

Reviewing all tumor data, we noted that patients with a low survival rate tended to have a higher proliferation index. Similarly, the deceased patients had a higher proportion of aneuploid tumors. Owing to the low number of cases and the heterogeneity of the tumors analyzed, DNA analysis cannot be designated as a clear prognostic method for intracranial tumors at present [16]. Further multicenter prospective studies are necessary to verify whether or not llow-cylometric DNA analysis is a reliable method to examine common CNS tumors in children.

References

[1] M.W. Ben Arush, S. Linn, O. Ben-Izhak, R. Levy, M.P. Nahum, T. Tsuk-Shina, J.N. Guilbord, R. Elhasid. S. Postovski, Prognostic significance of DNA ploidy in childhood astrocytomas, Pediatr. Hematol. Oncol. 16 (1999) 387-396.

[2] T.L. Chan, L.C. Curtis, S.Y. Leung, S.M. Furrington, J.W. Ho, A.S. Chan, P.W. Lam, C.W. Tse, M.G. Dunlop, A.H. Wyllie, S.T. Yuen, Early onset colorectal cancer with stable microsatellite DNA and near-diploid chromosomes, Oncogene 20 (2001) 4871-4876.

[3] G.E. Feichter, K. Goerttler, D. Haag, J. Heep, W.W. Hopker, M. Kaufmann, K.L. Kramer, F. Kubli, W. Kuhn, S. Kunze, et al., DNA measurement of malignant tumors by impulse cytophotometry. The principles and importance for assessing growth behavior and the degree of abnormality, Dtsch. Med. Wochenschr. 109 (1984) 738-744.

[4] P.L. Fitzgibbons, R.R. Turner, A.J. Appley, P.C. Bishop, P.W. Nichols, A.L. Epstein, M.L. Apuzzo, P.T. Chandrasoma, Flow cytometric DNA and nuclear antigen content in astrocytic neoplasms, Am. J. Clin. Pathol. 89 (1988) 640-644.

[5] P.A. Forsyth, E.G. Shaw, B.W. Scheithauer, J.R. O'Fallon, D.D. Layton Jr., J.A. Katzmann, Supratentorial pilocytic astrocytomas. A clinicopathologic, prognostic, and flow cytometric study of 51 patients, Cancer 72 (1993) 1335-1342.

[6] P. Frederiksen, E. Reske-Nielsen, P. Bichel, Flow cytometry in tumors of the brain, Acta Neuropathol. (Berl.) 41 (1978) 179- 183.

[7] R. Garcia, A. Bueno, S. Castanon, P. Ruiz-Barnes, J. Maria de Campos, E. Kusak, J.R. Fortes, F. Orliz, J.L. Sarasa, Study of the DNA content by flow cytometry and proliferation in 281 brain tumors, Oncology 54 (1997) 112-117.

[8] P. Gutjahr, Tumoren des Zentralnervensystems, in: P. Gutjahr (Ed.), Krebs bei Kindern und Jugendlichen, 4th Edition, Deutscher rzte-Verlag, Kln, 1999, pp. 320-369.

[9] D.W. Hedley, M.L. Friedlander, I.W. Taylor, C.A. Rugg, E.A. Musgrove, Method for analysis of cellular DNA content of paraffin- embedded pathological material using flow cytometry, J. Histochem. Cytochem. 31 (1983) 1333-1335.

[10] C.J. Herman, Cytometric DNA analysis in the management of cancer. Clinical and laboratory considerations, Cancer 69 (1992) 1553-1556.

[11] T. Hoshino, K. Nomura, C.B. Wilson, K.D. Knebel, J.W. Gray, The distribution of nuclear DNA from human brain-tumor cells, J. Neurosurg. 49 (1978) 13-21.

[12] R. Kiss, I. Camby, I. Salmon, P. Van Ham, J. Brotchi, J.L. Pasteels, Relationship between DNA ploidy level and tumor sociology behavior in 12 nervous cell lines, Cytometry 20 (1995) 118-126.

[13] P.K. Kotylo, P.B. Robertson, N.S. Fineberg, B. Azzarelli, R. Jakacki, Flow cytometric DNA analysis of pediatric intracranial ependymomas, Arch. Pathol. Lab. Med. 121 (1997) 1255-1258.

[14] C. Lengauer, K.W. Kinzler, B. Vogelstein, Genetic instabilities in human cancers, Nature 396 (1998) 643-649.

[15] P. Mathew, T. Look, X. Luo, R. Ashmun, M. Nash, A. Gajjar, A. Walter, L. Kun, R.L. Heideman, DNA index of glial tumors in children. Correlation with tumor grade and prognosis, Cancer 78 (1996) 881-886.

[16] A.R. Rabson, Flow cytometry in the diagnosis of brain tumors, Neurosurg. Clin. N. Am. 5 (1994) 135-146.

[17] L.M. Rainwater, Y. Hosaka, G.M. Farrow, S.A. Kramer, P.P. Kelalis, M.M. Lieber, Wilms tumors: relationship of nuclear deoxyribonucleic acid ploidy to patient survival, J. Urol. 138 (1987) 974-977.

[18] M. Reyes-Mugica, P.M. Chou, M.M. Myint, C. Ridaura-Sanz, F. Gonzalez-Crussi, T. Tomita, Ependymomas in children: histologic and DNA-flow cytometric study, Pediatr. Pathol. 14 (1994) 453-466.

[19] D. Schmidt, B. Wiedemann, W. Keil, E. Sprenger, D. Harms, Flow cytometric analysis of nephroblastomas and related neoplasms, Cancer 58 (1986) 2494-2500.

[20] H. Struikmans, D.H. Rutgers, G.H. Jansen, C.A. Tulleken, I. van der Tweel, J.J. Battermann, Prognostic relevance of cell proliferation markers and DNA-ploidy in gliomas, Acta. Neurochir. (Wien.) 14\1 (1998) 140-147.

[21] D.M. Tait, R.A. Eeles, R. Carter, S. Ashley, M.G. Ormerod, Ploidy and proliferative index in medulloblastoma: useful prognostic factors? Eur. J. Cancer. 29A (1993) 1383-1387.

[22] T. Tomita, M. Yasue, H.H. Engelhard, D.G. McLone, F. Gonzalez-Crussi, K.D. Bauer, Flow cytometric DNA analysis of medulloblastoma. Prognostic implication of aneuploidy, Cancer 61 (1988) 744-749.

[23] M. Yasue, T. Tomita, H. Engelhard, F. Gonzalez-Crussi, D.G. McLone, K.D. Bauer, Prognostic importance of DNA ploidy in medulloblastoma of childhood, J. Neurosurg. 70 (1989) 385-391.

[24] C. Zerbini, R.D. Gelber, D. Weinberg, S.E. Sallan, P. Barnes, W. Kupsky, R.M. Scott, N.J. Tarbell, Prognostic factors in medulloblastoma, including DNA ploidy, J. Clin. Oncol. 11 (1993) 616- 622.

P. Habermehl(a,*), M. Knuf(a), M. Schwarz(b), J. Bohl(c), U. Bartels(a), P. Gutjahr(a), K. Hohenfellner(a)

a Pediatric Hospital and Ambulant Clinic of Johannes Gutenberg- University Mainz, Germany

b Clinic and Ambulant Clinic of Neurosurgery of Johannes Gutenberg-University Mainz, Germany

c Department of Neuropathology of Johannes Gutenberg-University Mainz, Germany

Received 11 April 2003; accepted 7 January 2004

* Corresponding author. Universittskinderklinik, Johannes Gutenberg University Mainz, Langenbeckstrasse 1, D-55101 Mainz, Germany. Tel.: +49-(0)6131-172768; fax: +49-(0)6131-173918.

E-mail address: pirmin.habermehl@uni-mainz.de (P. Habermehl).

Copyright Urban & Fischer Verlag 2004

More News in this Category