A Cytogenetic Study of 19 Recurrent Gliomas Eric Pruchon, Laurent Chauveinc, Laure Sabatier, Anne-Marie Dutrillaux, Michelle Ricoul, Jean-Yves Delattre, Felipe Vega, Michel Poisson, Frederic Hor, and Bernard Dutrillaux ABSTRACT: A cytogenetic analysis was performed on 19 recurrent gliomas all of which had been treated by radiotherapy. All cases exhibited clonal chromosomal anomalies, the tumors were classified into four categories in relation to their mona- or polyclonality and to the presence or absence of a clonal evolution. Polyclonal tumors without clonal evolution had a delay of recurrence significantly longer than monoclo- nal or polyclonal tumors with clonal evolution. This difference could be related to the presence of clones with different malignant potential, which could be differentiated by their pattern of chromosomal aberra- tions. The malignant potential of "highly malignant" clones resulted from the juxtaposition of imbalances, such as monosomy 10, as in high-grade primary gliomas, and presumably radiation-induced structural rearrangements. That of clones of law malignancy was almost limited to the presence of multiple balanced structural rearrangements, probably induced by radiation. INTRODUCTION Except for a few cases of translocations or deletions which could be related to either the alteration of an oncogene or loss of an antioncogene, the biologic meaning of chromo- some alterations occurring in solid tumors remains specula- tive. The number of chromosome alterations varies from case to case, possibly in relation to the stage of progression of a given tumor type. It also varies with tumor type. For instance, breast cancer cells are frequently highly rearranged, whereas other epithelial tumors, such as endometrial adenocarci- noma, maintain less aneuploid karyotypes [1, 2]. Considering that chemo- or radiotherapy might induce chromosome alterations, most authors have selected un- treated patients for their studies. Thus, few data o~xist on chro- mosome alterations in recurrencies. However, in tumors that relapse frequently, it may be of interest to study the effects of therapy, the distribution of chromosome alterations capa- ble of giving information about clonality, the stage of progres- sion of surviving cells, their resemblances to and differences from untreated tumors at equivalent stages, and an eventual nonrandom induction of rearrangements, possibly in rela- From the Laboratoire de Cytog~n6tique et G~n~tique (E. P., L. C., L. S., M. R.), Fontenay-aux-Roses, France; UBA 620 C.N.R.S.- Institute Curie, Section de Biologie (A-M. D., B. D.), Paris, France; Clinique Neurologique (J-Y. D., E V., M. E), HOpital de la Salp6tri~re, Paris, France; and Service de Neurochirurgie (E H.), Hopital du Val de G~ce, Paris, France. Address reprint requests to: Dr. B. Dutrillaux, URA 620 C.M.R.S.- Institut Curie, Section de Biologie, 26 rue d'Ulm, 75231 Paris Cedex 05, France. Received September 30, 1993; accepted January 11, 1994. tion to the acquisition of new characteristics of the relaps- ing tumors. To approach these questions, glial tumors constitute an interesting model. The karyotypes of untreated tumors are fairly simple and well characterized, and post-therapeutic relapses occur frequently. Here, we report the cytogenetic study of a series of recurrent glial tumors and compare the results with those obtained in untreated patients [3-10]. Different patterns of chromosome alterations are observed among the recurrent tumors, which appear to be related with the precocity of the recurrence. MATERIALS AND METHODS Biopsies from recurrent tumors were obtained from 19 pa- tients. Histologic diagnoses were performed according to the brain tumor classification proposed by the WHO [11] and Burger et al. [12] for the fibrillary astrocytic neoplasms. A summary of clinical findings is given in Table 1. One part of the tumor sample was immediately frozen in liquid nitrogen for RNA sampling. Another part was preserved in culture medium for karyotyping and, whenever enough material was available, a last part was transplanted in athymic mice for karyotyping and further analyses through serial subcutaneous passages. The primary cultures were per- formed as follows: biopsies were aseptically finely minced and seeded in F12 medium supplemented with fetal calf se- rum, epidermal growth factor, glutamine, and antibiotics. Metaphases were harvested after 10-30 days' culture. Prepa- rations and R-banding were performed according to our usual procedures [13]. Karyotypes and cytogenetic clones were de- scribed according to the ISCN [14, 15]. © 1994 Elsevier Science Inc, 655 Avenue of the Americas, New York, NY 10010 85 Cancer Genet Cytogenet 76:85-92 (1994) 0165-4608/94/$07.00