Clinical Investigation of Neuroblastoma with Partial Deletion in the Short Arm of Chromosome 11

Several loci on the short arm of chromosome 1 (ip) have been reported as the consensus deleted regions for the putative suppressor genes of neurobiastoma by deletion mapping. The significance of deletion in ip on the clinical features of neuroblastoma remains controversial. To clarify the relationship between the clinical features of neuroblastoma cases and genetic status of ip, we performed deletion mapping on ip on samples obtained from 58 cases with neuroblastoma using 12 highly polymorphic microsatellite or minisatellite loci. Loss of heterozygosity of ip was detected in 19 cases (33% ) of primary tumors and in 21 cases (36% ) when metastatic and recurrent sites were included. They were classified into two groups according to the ip deletion pattern: interstitial deletion (group I, n 1 1) and terminal deletion (group T, ii = 10). The shortest region of overlap in group I ranged between FGR and DJSJ 70 (lp36.l-2). Clinically, all group I cases survived disease free, and none of these cases showed MYCN amplification. However, in group T, eight (80% ) cases showed a large terminal deletion from D1S162 (1p32-pter), including the shortest region of overlap of group I, and two (20% ) showed a very terminal deletion from DISI6O (ip 36.3). Of the group T cases, only two survived disease free, and seven (70%) showed MYCN amplification. Thus, the candidates for the locations of neuroblastoma suppressor genes on ip may involve at least two regions, which demonstrate different clinical features. INTRODUCTION Neuroblastoma. a childhood malignant tumor arising from neural crest. shows various clinical behaviors from spontaneous regression to unfavorable prognosis due to aggressive growth Received 7/16/96; revised 1/28/97: accepted 4/2/97. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked adt’ertiseme,mt in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. This work was supported by a Grant-in-Aid for Cancer Research from the Japanese Ministry of Health and Welfare and from the Ministry of Education. Science, Sports and Culture. 2 To whom requests for reprints should be addressed. Phone: 8 1-82257-52 16: Fax: 8 1 -82-257-52 19: E-mail: [email protected]. despite multimodal therapy. The prognosis mainly depends upon the biological characteristics of the tumor cells. especially amplification of the MYCN gene (I. 2, NTRK/ expression (3-7), telomerase activity (8). and DNA ploidy pattern (9-I I). Alterations of chromosome Ip have been reported in 30 80% of clinical samples and cell lines that have been successfully kanyotyped (12-16). Studies of the LOH3 on chromosome lp. using RFLPs and/on microsatellite markers. have found that Ip LOH occurs in 30-40% of neuroblastoma tumor specimens ( I 7-23) and that the SRO of I p deletions maps to the chromosomal region 1p36.1-3. These data suggest that suppressor genes associated with neuroblastoma development may be located in this region, whereas recent more detailed analyses of Ip deletions have suggested the involvement of multiple different neunoblastoma suppressor genes on lp (24-27). Although sevenal prognosis-associated factors have been reported. data on the relationship between Ip deletion and patient prognoses have remained contradictory ( I 8, 25, 26. 28-31). Mienosatellites and simple sequence repeats are ideal markers for highly informative DNA polymorphisms. because they are widely dispersed throughout eukaryotic genomes 32 and are easily detectable using the PCR (33-36. thus enabling a representative survey of the genome to be studied. In the present study. detailed deletion mapping of chromosome Ip. mainly using microsatellite loci. was carried out for neuroblastoma tumor specimens resected from 58 patients, and the clinical outcomes of these patients and biological behaviors of the tumors were analyzed. MATERIALS AND METHODS Patients. In the patients who underwent operation or surgical biopsy between 1984 and 1995 in our hospital on other institutions in Japan (listed in the “Acknowledgments”). both primary neuroblastoma tissue and normal tissue samples could be obtained from 58 patients at the time of the operation with informed consent. We enrolled these 58 patients in this study and also investigated eight metastatic samples (one bone marrow, three liven, and four lymph node metastatie lesions) and two relapse samples (one bone marrow and one intnaabdominal relapse sample). Seven of the primary tumor samples were obtained after chemotherapy. and the remaining samples were neseeted before chemotherapy. All tum(ir samples contained more than 60% of tumor cells in specimens examined. In each case, normal tissue was sampled from peripheral blood, normal adrenal gland. or appendix to obtain constitutional DNA. These tumors and normal tissues were stored at -80#{176}Cuntil DNA isolation was performed. In these 58 patients, the ages at diagnosis ranged between 2 months and 20 years, and 36 patients Research. on October 23, 2017. © 1997 American Association for Cancer clincancerres.aacrjournals.org Downloaded from 1222 I p Deletiomi iii Neuroblastoma were detected by a mass screening program for neunoblastoma that has been preformed for infants aged 6 months in Japan. The disease staging was classified using the standard clinical and pathological criteria of Evans ci a!. (37). Stage III patients who were I year or older when their tumors were diagnosed and all stage IV patients were treated postoperatively with the protocol (if the Study Group of Japan: they received cyclophosphamide, vincnistine, doxorubicin, cis-dichlorodiamine platinum, and etoposide (38). Other patients underwent an operation and received postoperative chemotherapy consisting of eyclophosphamide, vincnistine, and doxonubicin Ethe James’ (39) or Saint-Jude protocol (40)]. Isolation of DNAs. High molecular genomie DNA was isolated from tumor and normal frozen tissue specimens using proteinase K and SDS. followed by phenol/chloroform extractions as described previously (41). Genomic DNA was isolated from constitutional blood samples using a DNA Extractor WB kit (Wako, Osaka, Japan). Matched tumor and constitutional DNAs were used as templates for PCR amplification. PCR Conditions. PCR procedure was performed with specific sets of primers at two minisatellite loci (D/S76 and !)/S80. lp36.3: Ref. 21), nine microsatellite loci (DJSI6O, lp36.3: D/S170, lp3#{244}.2: FGR, lp36.2-l: CRTM and D1S168, 1p35:MYCL/. 1p32-34:DISJ62andDJSI88. 1p32;andFl3B. lq3l-32: Refs. 42-45), and one single strand conformational polymorphism locus (TNFR2, lp36.2; Ref. 46). The PCR procedune for minisatellite loci was performed with 2-10 ng of template DNA in a reaction mixture of 50 i1 containing 200 i.si of each deoxynucleotide triphosphate, I .5 msi MgCl2, 0.25 p.M of each primer. and I unit of Taq DNA polymenase (Wako). PCR amplification protocol was performed using a Program Temp Control System PC-800 (Astee, Fukuoka, Japan) for 35 cycles of 60 s at 95#{176}C,70 s at 63#{176}Cor 57#{176}Cfor D1S76 and I)1S80. respectively. and 180 s at 74#{176}C.Extension during the fimial cycle was continued for 10 mm. The amplified fragments were examined by 2% agarose gel eleetrophoresis and ethidium bromide staining. In analysis for microsatellite loci, 0.25 p.M of one of the matched primers was labeled in 5 p.1 using 370 kBq of I”t’-32P]ATP and I unit ofT4 kinase (Life Technologies, Inc., Gaithersburg. MD) for 20 mm at 37#{176}C and S mm at 54#{176}C before PCR. The PCR amplifications, except for MYCLI locus, were perfcwmed for 30-35 cycles of 60 s at 95#{176}C,70 s at 53-63#{176}C, and 70 5 at 74#{176}C in a reaction mixture described for minisatellite loci. The PCR amplification for MYCLJ locus was performed for 45 cycles. After denatunation in loading dye (95% fonmamide) for 5 mm at 95#{176}C,the PCR products were loaded onto a 0.35-mm thick 5% denaturing polyacrylamide gel containing 7 M urea. After eleetrophoresis for 3-4 h at 40 W, the gels were dried and subjected to autoradiography. PCR products of the single-strand confonmational polymorphism locus TNFR2 (46) were loaded onto a 0.35-mm thick 4.5% nondenatuning acrylami ide gel. run at 4#{176}C and 40 W constant power for 3-4 h, and subjected to autoradiognaphy. Autoradiography was performed using both X-ray film and a Bio-Imaging analyzer (BAS-2000; Fuji, Tokyo, Japan). The radioactivity of the PCR products was compared between tumor and constitutional DNA in each case. We measured the intensity of the band amplified from each allele and used the intensity ratio of two alleles in constitutional DNA as the control of hetenozygosity. We diagnosed LOH only when the observed intensity of one allele in the tumor DNA was less than 70% of expected intensity, which was calculated from the control, to minimize the effects of aneuploidy. Southern Blot Hybridization. MYCN amplification and LOH of the MYCLJ (RFLP) locus were investigated by Southem blot analysis as reported previously (8). Briefly, 2 p.g of genomic DNA were digested to completion with 10 units of the restriction enzyme EcoRI. The DNA fragments were separated by electrophonesis on 0.8% agarose gels and then blotted onto nitnocellulose filters. The filters were hybridized with an MY N probe (PN-,nvc. 1 : Oncor Inc., Gaithersburg, MD) or an MY L/ probe (JCRB CO049) and subjected to autoradiognaphy. Flow Cytometric Analysis of the Cellular DNA Content. Cellular DNA content was analyzed as reported previously (47-49). Briefly, tumor samples were thawed to 20#{176}Cand disrupted mechanically using scissors. After treatment with trypsin to isolate bare nuclei, the samples were stained with propidium iodide using a modified method of Vindel#{248}v et a!. (50) and analyzed with a FACScan flow cytometer (Beeton Dickinson, San Jose, CA). Results for 20,000 nuclei were plotted as histograms. The DI was determined by calculating the ratio of modal channel numbers of the tumor G0-G1 peak to that for normal diploid cells. Hence, the DI of diploid neunoblastoma cells was I .0; tumors with a distinct population from DI = I .0 were defined as aneuploid. Tumors were considered tetnaploid if a peak occurred with a DI of I .90-2.10 in at least 20% of the analyzed cells and if a peak corresponding to a G2-M tetraploid cell population was also present. When more than two different aneuploid G0-G1 cell populations were present, the tumor was classified as polyploid. Statistical Analysis. Correlations between lp LOH and each of the other factors were analyzed using x2 or Fisher’s exact test where appropriate. The overall survival curves for each group of patients were estimated by the Kaplan-Meien method, and the resulting curves were compared using the Cox-Mantel test. Differences were considered significant at P 0.05. RESULTS Deletion Mapping of Chromosome ip. All 70 tumors including recurrent and metastatie tumors from 58 cases were informative for more than 3 of the 12 loci on chromosome I p. The observed heterozygosity indices of 10 microsatellite markens ranged from 0.54 to 0.82, and those of two minisatellite markers were 0.52 and 0.68, respectively, which were similar to the reports described previously (21, 36, 42-46). Among the 58 primary tumors. including 7 samples obtained after chemotherapy, 19 (33%) showed LOH on at least one locus. In these primary tumor samples with LOH, only one (Nl45) was obtamed after chemotherapy. Moreover, two cases (N30 and Nl99) showed LOH in recurrent on metastatie sites without detectable LOH in the primary tumor. Thus, Ip LOH was detected in 21 of 58 cases (36%), and representative LOH patterns were shown in Fig. 1 . Results of deletion mapping on lp in these 21 eases were shown in Fig. 2. On the basis of the deletion patterns, we classified these cases into two groups: interstitial deletion (group I, n = 1 1 ) and terminal deletion (group T, ii = 10). In group I, the SRO encompassed the locus Research. on October 23, 2017. © 1997 American Association for Cancer clincancerres.aacrjournals.org Downloaded from Ni 42 TC Ni 43 TC N30