Aneuploidy and Aneusomy of Chromosome 7 Detected by Fluorescence in Situ Hybridization Are Markers of Poor Prognosis in Prostate Cancer'

Home , Centromere, Chromosome 7, Tetrasomy

[CANCERRESEARCH54,3998-4002,August1, 19941 Advances in Brief

Antonio Alcaraz, Satoru Takahashi, James A. Brown, John F. Herath, Erik J- Bergstralh, Jeffrey J. Larson-Keller, Michael M Lieber, and Robert B. Jenkins2 Depart,nent of Urology [A. A., S. T., J. A. B., M. M. U, Laboratory Medicine and Pathology (J. F. H., R. B. fl, and Section of Biostatistics (E. J. B., J. J. L-JCJ, Mayo Clinic and Foundation@ Rochester, Minnesota 55905

Abstract studies on prostate carcinoma samples. Interphase cytogenetic analy sis using FISH to enumerate chromosomes has the potential to over Fluorescence in situ hybridization is a new methodologj@which can be come many of the difficulties associated with traditional cytogenetic used to detect cytogenetic anomalies within interphase tumor cells. We studies. Previous studies from this institution have demonstrated that used this technique to identify nonrandom numeric chromosomal alter ations in tumor specimens from the poorest prognosis patients with path FISH analysis with chromosome enumeration probes is more sensitive ological stages T2N@M,Jand T3NOMOprostate carcinomas. Among 1368 than FCM for detecting aneuploid prostate cancers (4, 5, 7). patients treated by radical prostatectomy, 25 study patients were ascer We designed a case-control study to test the hypothesis that spe tamed who died most quickly from progressive prostate carcinoma within cific, nonrandom cytogenetic changes are present in tumors removed 3 years of diagnosis and surgery. Tumors from 25 control patients who from patients with prostate carcinomas with poorest prognoses . We survived for more than 5 years and who were matched for age, tumor identified 25 patients with clinically localized prostate carcinoma histological grade, and pathological stage also were evaluated. The tumors treated by radical prostatectomy who died from metastatic prostate from all 25 (100%) poor prognosis patients and from 11 of 25 (44%) cancer within 3 years after treatment. These patients were then control patients were found to be aneuploid by fluorescence in situ hy matched for age, pathological stage, and tumor histological grade with bndization (P < 0.0001). Alterations ofchromosome 7 were observed in 24 a group of patients who survived for more than 5 years after surgery. of the tumors (96%) from the poor prognosis patients versus 3 tumors (12%) from the control group (P < 0.0001). Moreover, a characteristic FISH analysis was then carried out using chromosome enumeration aneuploidy pattern with multiple abnormal chromosomes and a hypertet probes for 12 different chromosomes. rasomic population was generally found in tumors from the poor prog Patients and Methods nosis patients. This preliminary study suggests that fluorescence in situ hybridization studies of prostate cancer specimens may help to identify Study Design. Among 1386 patients with pathological stage T2NOM@Jand those patients at highest risk for early cancer death. T3N@M0prostatecancer who underwent a radical prostatectomy at the Mayo Introduction Clinic from 1966 to 1990, 31 died of prostate cancer between 6 months and 3 years after surgery. These patients (30 stage T3N@MI@and1 stage T2N@,M,@) Prostate cancer is the most common cancer in United States men, were matched for age, pathological stage, and tumor grade with patients with approximately 200,000 newly diagnosed cases and 38,000 deaths surviving for more than 5 years after surgery. For those poor prognosis cases due to prostate cancer per year (1). The distinction between those resected prior to 1988, the control cases were also matched for date of surgery. Six cases were excluded in both groups because the paraffin-embedded tissue cases of prostate cancer destined to progress rapidly to lethal meta blocks did not contain residual tumor, or the cells were poorly preserved. Ten static disease and those with little likelihood of causing morbidity and BPH surgical samples, from the same time period, were studied to obtain mortality is a major goal of current prostate cancer research. normal value information. Factors such as clinical and pathological stage and histological In Situ Hybridization with Chromosome Enumeration Probes. After grade are conventionally used to help predict prognosis for patients pathological confirmation, six adjacent 50-sm tumor tissue sections were used with localized prostate carcinoma (2). Among newer techniques, for FCM and FISH analyses. Tissue deparaffmization was performed. Sections ploidyIDNA analysis using FCM3 may help refine prognostic risk for were washed for 10 mm with 2 ml Histo-Clear (National Diagnostics, Atlanta, patients with tumors of common histological grades and stages (3). GA) three times. The tissue was then dehydratedin 100% ethanol (5 min, However, it is widely recognized that FCM ploidy analysis cannot twice) and then rinsed in water (5 mm, twice). Tissues were digested in pepsin detect small changes in DNA content or chromosome number. Rou solution (Sigma P.7012; Sigma Chemical Co., St Louis, MO; 4 mg/mI in 0.9% NaC1,pH 1.5) for 2.5 h at 37%. After filtering, isolated nuclei were washed tine cytogenetic studies of solid tumors are a more precise approach to twice with phosphate-buffered saline and vortexed. The resultant nuclear detect numerical and/or structural defects. However, because met suspension was applied to Superfrost slides, and the slides were oven-dried at aphase cells are required for analysis, cytogenetic studies are ham 65Â°Cfor 10 min. FISH probes, labeled with SpectrumOrange or Spec pered by the difficulties of specifically stimulating tumor cells to trumGreen, specific for the centromere region of 11 chromosomes (4, 6â€”12, divide. It has been particularly difficult to perform routine cytogenetic 17, 18, and X) and for the mid-distal Yq (Yq12) region, were obtained from Imagenetics (Framingham, MA). Only single-probe hybridizations were per Received 4/6/94; accepted 6/14/94. formed. All of the currently available Imagenetics chromosome enumeration The costs of publication of this article were defrayed in part by the payment of page probes were used in the study. charges. This article must therefore be hereby marked advertisement in accordance with Slides containing isolated nuclei were dehydrated in an ethanol series (70, 18 U.S.C. Section 1734 solely to indicate this fact. I Supported in part by a grant from Imagenetics Inc., Naperville, IL (to R. B. J.); Grant 85, and 100%)for 2 mmeach. DNA was denaturedbyincubatingtheslides in CA 58225 from NIH (to R. B. J., M. M. L., S. T.); and by Grant FIS 93/5326 from the 70% formamide/2X SSC (300 mM sodium chloride and 30 mM sodium citrate) Hospital ainic de Barcelona and Fondo de Investigaciones Sanitarias (to A. A.). at 75Â°Cfor 4 mm, followed by dehydration in an ice-cold ethanol series. 2 To whom requests for reprints should be addressed, at Cytogenetics Laboratory, Simultaneously, the probe solution (7 @lHybridizationMix II; Imagenetics; Mayo Clinic, 200 First Street SW, Rochester, MN 55905. 2 @.tlprobe) was denatured for 6 min at 75Â°C and kept on ice until it was added 3 The abbreviations used are: FCM, flow cytometry; FISH, fluorescence in situ hybridization; BPH, benign prostatic hyperplasia; SSC, standard sodium citrate; 2SD, two to the slides. After sealing, the slides were then incubated overnight at 37Â°C. standard deviations. Following hybridization, the slides were washed in 50% formamide/2X SSC 3998

for 10 s, 2X SSC for 1 aria,and2X SSC/NonidetP-40 for 1 mmat 37Â°C.The most likely the result of homologous pairing (7). The mean + 2SD for counterstain 4,6-diamidino-2-phenylindole with antifade compound p-phe monosomy (for autosomes) and nullisomy (sex chromosomes) was nylenediaminewas addedpriorto analysis. lower than 10% for all chromosomes. All of the autosomes had an Analysis of Interphase in Situ Hybrkllzatlon. For each probe hybridiza average trisomy or tetrasomy level of less than 1.5% (Fig. 1B). For tion, signals from 300 nuclei were counted. For this number of enumerated each of the autosomes, mean + 2SD for trisomy was less than 3% and nuclei, the 95% confidence limits ofthe 0, 1, 2, 3, 4 and 5 signal proportions for tetrasomy less than 4%. The percentage nuclei with 4 signals for (p) were estimated as Â±1.96V@j@00 using the binomial distribution and were at most Â±0.057atP 0.5. each autosome was averaged to estimate the normal percentage of Flow Cytometry. The method for DNA content analysis of paraffin tetraploid cells with prostate tissue. The mean for all BPH specimens embedded tissues has been described previously (4). was 2.58% with an SD of 0.62%. Statistical Analysis Comparisons of age and tumor characteristics be Because of relative loss of chromosomes, inefficient hybridization, tween the poor prognosis (cases) and control groups were made using the and/or homologous pairing, it was also necessary to establish the @ rank-sum and tests as appropriate. Matched analyses were not performed as normal proportion of trisomic to tetrasomic cells in apparently pure the primary goal of the matching was to obtain similar groups with respect to tetraploid tumors. Several tumors in the control group and in our tumor stage and grade. Age, which is not a risk factor for cause-specific previous studies of two series of unselected prostate carcinomas (5, 7) survival, was used as a convenient way to select a specific control. contained a significant population of cells with 4 signals for all Within the case group, the generalized estimating equations of Liang and autosomes. In the control tumors from the current study, the tetra Zeger (6) were used to model aneusomyfrequencyas a functionof chromo some, taking into account within patient correlations. All tests were two-sided somic populations averaged 13.3% of the cells enumerated (range, 7.5 with a = 0.05. to 21%, eight tumors). These tumors contained no apparent aneusomic chromosomes and were thus apparently truly tetraploid (we cannot Results rule out aneusomies of any untested chromosome). We used the apparently true tetraploid tumors from the control group and estab Normal Value Studies. Mean and positive 2SD of the number of lished that, for autosomes, the mean trisomy:tetrasomy ratio was less copies for each chromosome in the 10 cases of benign prostate tissue than 0.51 for each chromosome. The mean + 2SD of the trisomy: studied is illustrated in Fig. 1. All of the chromosomes had an average tetrasomy ratio was less than 0.59 for each chromosome, except for monosomy (autosomes) or nullisomy (sex chromosomes) of less than chromosome 17, which was less than 0.82. Similar-sized tetrasomic 5%, except for chromosome 17, which had an average monosomy rate populations and similar mean trisomy:tetrasomy ratios were also of 7% (Fig. IA). We have previously shown that this higher rate is observed in the apparently purely tetraploid tumors of our previous studies (5, 7). For the BPH specimens, the mean + 2SD of nuclei with 2 signals A for chromosomes X and Y was less than 10% (Fig. 1B). There was no evidence of gain or loss of sex chromosomes in the BPH specimens. The population of cells with 2 signals for the sex chromosomes 10 0 @ionosom@'INulisomy includes both cells with the gain of one chromosome (aneusomy) and cells with twice the total chromosome content (tetraploidy). In order 8 to distinguish sex chromosomal aneusomy from tetraploidy, we ana .50 lyzed the ratio of the sex chromosome 2-signal percentage to the average autosorne 4-signal percentage in the apparently true tetraploid a2 tumors. The X and Y chromosomes had mean ratios of 1.23 and 1.30, 4 respectively. The mean + 2SD ratios for chromosomes X and Y were 1.95 and 2.08, respectively. 2 Abnormal Criteria. Based on the normal value studies described 0â€”@ @@â€”@â€¢â€”-â€”â€”@ ______above and an inspection of the FISH data from the poor prognosis and 4 6 7 8 9 10 11 12 17 18 X V control groups, we established the following conservative cut-off Chromosome criteria: (a) a tumor was classified as tetraploid if the autosomal B average of percentage nuclei with 4 signals (or the average autosomal tetrasomy) was 6%; (b) an abnormal monosomy (or nullisomy for 12 sex chromosomes) required 12% of nuclei contain 1 (or 0) FISH signals. Chromosomes that were disomic on a tetraploid background 10 0 Tnsomy were also classified as lost; (c) an abnormal autosomal trisomy was required to fulfill two criteria: (1) 7% nuclei contain 3 FISH signals 8 @TetraSOmy .50 and, because of the incidence of tetraploidy, (2) the ratio of 3-signal nuclei to 4-signal nuclei be 1 for chromosome 17 and Â½for the a2 remaining autosomes; (d) sex chromosome gain or tetraploidy re quired 12% nuclei with 2 FISH signals; (e) to distinguish sex chromosornal gain from tetraploidy, the former required that the ratio of the percentage nuclei with 2 sex chromosome signals to the average autosomal 4-signal percentage be 2.1; (f) abnormal hypertetrasomy required that the sum of the percentage nuclei with 5 signals for any 4 6 7 8 9 10 11 12 17 18 X V autosome or 3 signals for any sex chromosome be 6%; and (g) Chromosome abnormal monosorny/nullisomy, trisomy, or hypertetrasomy defined a Fig. 1. FISH signal distribution in BPH. Data are expressed as mean + 2SD for each tumor as aneuploid. Otherwise, the tumor was classified as apparently chromosomestudied.A,monosomyforthe autosomesandnullisomyforthe sex chro mosomes. B, trisomy and tetrasomy for the autosomes. For the sex chromosomes, a single diploid or apparently pure tetraploid, based upon the overall average column illustrates the percentage of cells with two nuclear signals. autosomal tetrasomy. 3999

No BPH specimens contained enumeration values which exceeded distribution, and the FISH ploidy distribution for the poor prognosis these criteria. and control groups after the exclusion of the 6 patients from both FCM and FISH Ploidy Evaluation. Table 1 shows typical results groups (see â€œPatientsandMethodsâ€•).No significant differences were from a FISH-diploid, FISH-tetraploid, and two FISH-aneuploid tu found between the two groups for matched clinical variables: age, mors. The FISH-diploid tumor (control tumor 22) had an average pathological stage, and Gleason grade. The patients from the poor autosomal tetrasomy of 3.1% and, based on the criteria listed above, prognosis group were more likely to receive adjuvant therapy contained no apparent chromosome centromere anomaly. The FISH (P = 0.047). There was no significant difference in the FCM DNA tetraploid tumor (control tumor 17) had an average autosomal tetra ploidy distribution between two groups (P 0.14). By FISH analysis, somy of 15.5% and, except for the increased tetrasomy for each all 25 tumors in the poor prognosis group were found to be aneuploid, autosome and increased disomy for each sex chromosome, contained while 11 tumors in the control group were found to be aneuploid, 8 no apparent chromosome centromere anomaly. The first FISH aneu tetraploid, and 6 diploid. This difference in FISH ploidy distribution ploid tumor (poor prognosis tumor 22) had an average autosomal was highly statistically significant (P < 0.0001). tetrasomy of 4.3% and, based on the criteria listed above, was ob Specific FISH Results. Table 3 summarizesthe aneusomic chro served to have abnormal centromere signal distributions for chromo mosomes from the poor prognosis and control groups. The aneuso somes 4, 7, 18, and Y. The second FISH aneuploid tumor (poor mies are listed as either a gain or loss relative to a reference ploidy prognosis tumor 5) had an average autosomal tetrasomy of 12.9% and level inferred from the overall FISH results. Chromosomes with was observed to have abnormal centromere signal distributions for hypertetrasomic populations 6% were defined as a gain. Since the chromosomes 7, 9, and 10 relative to tetraploid. poor prognosis cases were often observed to have a complex distri Table 2 summarizes the clinical characteristics, the FCM ploidy bution of centromere enumeration counts, in some cases, the inferred

TablecasesChromosomeCentromere 1 Centro,nerecopy number ofselected poorprognosis and control number(%)Â°0123 copy

4Control 22 (diploid) 4 0.0 2.0 91.7 1.7 4.7 0.0 6 0.0 3.3 92.0 1.3 3.3 0.0 7 0.0 3.7 86.0 4.0 63 0.0 8 0.0 4.7 88.3 3.0 4.0 0.0 9 0.0 3.3 94.3 1.0 1.3 0.0 10 0.0 4.7 92.7 0.7 2.0 0.0 11 0.0 2.0 95.0 1.0 2.0 0.0 12 0.0 3.0 93.0 1.3 2.7 0.0 17 0.0 7.0 89.3 1.7 2.0 0.0 18 0.0 1.0 94.7 2.0 2.3 0.0 x 2.7 91.3 6.0 0.0 0.0 0.0 Y 5.3 89.3 5.3 0.0 0.0 0.0

Control 17 (tetraploid) 4 0.0 3.0 81.0 2.3 13.7 0.0 6 0.0 1.0 78.7 53 15.0 0.0 7 0.0 2.7 79.3 4.3 13.7 0.0 8 0.0 5.0 75.0 4.0 16.0 0.0 9 0.0 3.7 82.3 2.3 11.7 0.0 10 0.0 2.0 76.0 2.7 19.0 03 11 0.0 5.0 71.0 4.0 20.0 0.0 12 0.0 1.3 80.3 3.3 15.0 0.0 17 0.0 5.0 76.0 4.7 143 0.0 18 0.0 4.7 76.0 3.0 16.3 0.0 x 1.7 73.7 24.3 03 0.0 0.0 Y 1.3 75.0 23.0 0.7 0.0 0.0

Poor Prognosis 22 (aneuploid) 4 0.0 4.0 78.0 9.3 83 0.3 6 0.0 6.7 87.0 4.0 23 0.0 7 0.0 3.0 79.3 13.7 4.0 0.0 8 0.0 3-3 90.3 1.7 4.7 0.0 9 0.0 2.3 87.7 4.3 5.7 0.0 10 0.0 4.0 89.3 2.7 4.0 0.0 11 0.0 3.7 91.7 0.3 4.0 03 12 0.0 3.7 87.7 2.3 63 0.0 17 0.0 11.7 82.7 2.0 33 0.3 18 0.0 48.0 50.3 1.0 0.7 0.0 x 3.7 88.3 7.0 1.0 0.0 0.0 Y 5.1 73.1 18.6 2.4 0.7 0.0

Poor Prognosis 5 (aneuploid) 4 0.0 1.3 79.0 4.3 14.7 0.7 6 0.0 6.0 71.0 8.0 14.7 0.3 7 0.0 0.3 23.0 48.7 143 13.6 8 0.0 2.0 81.3 6.7 9.0 1.0 9 0.0 2.7 68.0 14.3 14.0 1.0 10 0.0 1.3 82.0 73 8.7 0.7 11 0.0 4.0 74.3 5.7 16.0 0.0 12 0.0 4.3 77.3 4.0 143 0.0 17 0.0 2.3 82.7 2.0 13.0 0.0 18 0.0 6.3 79.7 4.0 10.0 0.0 x 2.7 80.0 16.3 1.0 0.0 0.0 Y 6.3 74.7 17.3 1.7 0.0 0.0 a Percentageof 300 nuclei with indicated centromere copy number. 4000

FISH MARKERS OF POOR PROGNOSIS IN PROSTATE CANCER

reference ploidy level may be incorrect (e.g., diploid and not tet 25- @ raploid or vice versa). However, this does not change the final I . o.oi< <0.05 PoorPrognosis @ conclusion that a particular chromosome was abnormal (e.g., alters an ! 20- . 0.001

Table 2 Clinical characteristics and ploidy results of poor prognosis and control groups zg Poor prognosis 0 (n25)PÂ°Age patients (n25)Control patients 4 I6â€˜7â€˜8â€˜9â€˜10â€˜11â€˜12â€˜17â€˜18â€˜Xâ€˜V atsurgeryâ€•0.67<604660â€”701515>7064Mean65.163.2Pathological Chromosome Fig. 2. Distribution of aneusomic chromosomes in poor prognosis (study) and control groups. Ps have been calculated for each chromosome comparing the poor prognosis and control groups. Twenty-four tumors were evaluated in the poor prognosis group for chromosomes 4, 10â€”12,and 17. For the remaining seven chromosomes in the poor stageâ€•0.99pT2NOMO10pT3NOMO2425Gleason prognosis group, all 25 tumors were evaluated. All 12 chromosomes were evaluated in all 25 tumors of the control group.

gradeâ€•0.162â€”4005â€”6377â€”102218Adjuvantpathological alterations. Twenty ofthe 25 had a tetraploid-inferred stemline ploidy. No control tumors were aneusomic for six or more chromosomes, and 8 of the 11 aneuploid tumors in the control group were found to have a single aneusomic chromosome. Twelve of the control cases had a tetraploid therapycÃ˜â€¢@47No39Yes2216FCM inferred stemline ploidy. Hypertetraploidy was commonly found in tu mors from the poor prognosis patients; 12 cases were found to have ploidy0.14Diploid56Tetraploid710Aneuploid135Not hypertetrasomy of one or more chromosomes, compared to zero cases in the control group (P < 0.0001). As can be seen in Fig. 2, for each of the 12 chromosomes, the classifiable04FISH frequency of aneusomy in the tumors from the poor prognosis patients was higher than in the tumors from the control patients. The largest ploidy<0.0001Diploid06Tetraploid08Aneuploid2511 differences in frequency were seen for chromosomes 7 (@= 35.5; P < 0.0001), 4 (x2= 22.5; P < 0.0001), and 18 (f 17.0; P < 0.0001). a Comparing the distribution of each parameterfor the poor prognosis group with that Within the poor prognosis group, we found a significant difference of the control group. in the frequency of aneusomy between chromosomes (x@ = 28.8 on b Matching variables. C Radiation or hormonal therapy within 3 months after surgery. 11 degrees of freedom; P < 0.0025). To assess which chromosomes

groupsPoorTable 3 Summary of aneusomic chromosomesin the poor prognosis and control prognosis groupControl groupInferredPresence ofTumorAneusomiesÂ°stemline ofInferredPresence ploidy@'hypertetrasomy2â€”6X3,â€”7,â€”8,â€”9,â€”10,â€”11,â€”12,â€”YX2Tâ€”1+11X2Dâ€”3â€”4,+6,â€”7,--8x3,+10,--18,â€”Yx2T+2Tâ€”4+4,â€”6X3,â€”7,â€”12,â€”YX2T+4Tâ€”5+7,â€”9,â€”10T+5â€”17Dâ€”6â€”11,â€”12,â€”17,â€”18,+XT+7Tâ€”7â€”7,â€”18Tâ€”8Tâ€”8â€”4,â€”7Tâ€”10â€”18Dâ€”10â€”4,â€”6,â€”7,â€”8,â€”9,â€”10,+11,â€”12,â€”17,â€”18,+X,+Y,â€”YT+11Tâ€”11â€”4,â€”6,â€”7,â€”11,+17,â€”18,+X,+YT+12+7Dâ€”13â€”7,â€”8X3,â€”12X3,â€”18X3Tâ€”13+6,+8,+9,+18,+YDâ€”14â€”6,â€”7,+8,--17,â€”18T+14Dâ€”15+4,â€”6,+7,â€”10,â€”11,â€”YX2T+16Dâ€”16â€”7,+8,+11,â€”YX2T+17Tâ€”17â€”4X3,+6,+7,+8,+9,+17,+18,+X,+Y,--YX2T+18T-18â€”4,â€”6,â€”7,â€”8,â€”10,â€”11,â€”12,â€”18X3,â€”XX2,+YTâ€”20â€”YX2Tâ€”20â€”4,â€”6,â€”7,â€”8X3,â€”9,â€”10X3,â€”11,--12,-â€”17X3,--18,+X,+YTâ€”21â€”8Dâ€”21+4,+7Dâ€”22Dâ€”22+4,+7,â€”18,+YDâ€”23+7,+12Dâ€”23+4,â€”6X3,+7,â€”8X3,â€”9,+10,â€”11,â€”12,+17,--18T+24Dâ€”25+7,-17Dâ€”25Dâ€”26â€”4,â€”6x3,--7,â€”8X3,â€”9X3,â€”10,â€”11,â€”12,â€”17,â€”18Tâ€”29â€”6X3Tâ€”27+4,â€”7,+11,+17,+18T+30Dâ€”29+7,+X,+YDâ€”31â€”6T-30+7,â€”17,--18,+YDâ€”32Tâ€”31â€”4,â€”6,â€”7,â€”9,â€”10,â€”11,â€”12,--17,â€”18,+YTâ€”33â€”6,â€”7,â€”11,â€”12Tâ€”ploidyâ€•hypertetrasomyTumorAneusomiesÂ°stemline

a Losses (â€”)andgains (+) of chromosomes are expressed relative to the inferred stemline ploidy level. For example, â€”6X3,â€”11, and +7 in poor prognosis tumor 23 refer to significant 1-, 3-, and 5-signal populations, respectively, relative to a tetraploid stemline ploidy. b Ploidy level inferred on inspectionof the distributionof FISH signalsfor all twelve chromosomesanalyzed.D, diploid; T, tetraploid. 4001

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1994 American Association for Cancer Research. FISHMARKERSOFPOORPROGNOSISINPROSTATECANCER were producing this significant difference, each chromosome was and are not of etiological significance. However, there is evidence tested using the generalized estimating equations model to determine that human chromosome 7 is necessary and sufficient for both the if it had a higher (or lower) rate of aneusomy than all the other establishment and maintenance of invasiveness and metastatic poten chromosomes combined. In view of the multiplicity of tests (8), the tial. For example, malignant mouse-human somatic cell hybrids retain resulting Ps were multiplied by 12 (Bonferroni correction). The only their metastatic potential after losing all human chromosomes except chromosome in the tumors from the poor prognosis group with a chromosome 7 (13). significantly higher aneusomy rate than all the other combined chro Several authors have criticized the significance of the presence of mosomes was chromosome 7 (adjusted P = 0.020). The independent trisomy 7 in human cancer cells (14). The most consistent criticism is aneusomy rates for chromosomes 4 and 18 were statistically similar to that the observed alterations in chromosome 7 number are found in the the aneusomy rates of the other combined chromosomes (adjusted infiltrating lymphocytes and not in the tumor cells (15). We have P > 0.20 for both chromosomes). performed simultaneous FISH and immunohistochemical staining and have observed trisomy 7 in prostate epithelial tumor cells but not in Discussion infiltrating leukocytes (5). To the best of our knowledge, this is the first study of matched poor In conclusion, we have found a characteristic FISH pattern of prognosis and good prognosis patients to evaluate the association of aneuploidy defined by multiple aneusomies, hypertetrasomic cells, numeric chromosome alterations ascertained by FISH with clinical and the consistent presence of alterations of chromosome 7 to have a behavior in prostate cancer or any solid tumor. We targeted high grade strong association with early death in patients with localized prostate T3NOMIJ prostate cancers treated by radical prostatectomy, tumors cancer. These fmdings may be of important practical use, since we whose rate of clinical progression is difficult to predict. Our results have developed the methodology to perform rapid FISH pretreatment indicate that an abnormal tumor FISH ploidy is significantly associ analysis on prostate biopsy core specimens that can subsequently be ated with a poor prognosis in patients with prostate cancer. The used for routine histopathological examination (5). If the results from qualitative analysis of the aneuploidy pattern is of particular impor the preliminary study reported here can be extended and confirmed, tance. The typical aneuploid tumor found in the poor prognosis the pretreatment detection of cytogenetic markers of poor prognosis patients was defined by aneusomies of multiple chromosomes relative described above may be of help in selecting patients who would to a basic tetraploid population. Hypertetrasomic cells were frequent. benefit from a most aggressive therapeutic approach. In contrast, tumors in patients with a good prognosis are often diploid or tetraploid, and their aneusomies, when present, tended to be mono References somies of a single chromosome. These results suggest that tumor cells 1. Boring, C. C., Squires, T. S., Tong, T., and Montgomery, S. Cancer statistics 1994. in the poor prognosis group are cytogenetically unstable and may CA-Cancer J. Clin., 44: 7-26, 1994. randomly lose and gain chromosomes, resulting in the different aneu 2. Gittes, R. F. Carcinoma of prostate. N. Eng. J. Med., 324: 236â€”245,1991. 3. Shankey, V., Kallionemi, 0., Koslowski, J. M., Lieber, M. M., Mayall, B. H., Miller, somies. Such cells also had likely developed structural abnormalities G., and Smith,G. J. Consensusreviewof the clinicalutility of DNA content in the retained chromosomes (9). cytometry in prostate cancer. Cytometry, 14: 497â€”500,1993. Except for loss of a Y chromosome and gain of chromosome 7, 4. Persons, D. L, Gibney, D. J., Katzmann, J. A., Lieber, M. M., Farrow, G. M., and Jenkins, R. B. Use of fluorescent in situ hybridization for deoxyribonucleic acid previous routine cytogenetic studies of metaphase cells have not ploidy analysis of prostatic adenocarcinoma. J. Urol., 150: 120â€”125,1993. clearly demonstrated nonrandom, whole chromosome loss or gain in 5. Takahashi, S., Qian, J., Brown, J. A., Alcaraz, A., Bostwick, D. G., Lieber, M. M., prostate cancer (10). In this study, we observed that all chromosomes and Jenkins, R. B. Potential markers of prostate cancer aggressiveness detected by fluorescence in situ hybridization in needle biopsies. Cancer Res., 54: 3574â€”3579, are likely to be numerically altered in clinically aggressive prostate 1994. cancers. Because all chromosomes were more likely to be altered in 6. Liang, K-Y., and Zeger, S. L Longitudinal data analysis using generalized linear models. Biometrika, 73: 13â€”22,1986. the tumors from the poor prognosis group and because a significant 7. Brown, J. A., Alcaraz, A., Takahashi, S., Persons, D. L, Lieber, M. M., Jenkins, R. B. difference in the frequency of chromosomal aneusomies was observed Chromosomal aneusomies detected by fluorescent in situ hybridization (FISH) in within the poor prognosis group, we performed detailed statistical clinically localized prostate carcinoma. J. Urol., 152, in press, October 1994. 8. Bandyk, M. 0., Thao, L C., Troncoso, P., Pisters, L L, Palmer, J. L, von tests to determine which alteration(s) occurred nonrandomly in this Eschenbach, A. C., Chung, L W. K., and Liang, J. C. Trisomy 7: a potential group. Only chromosome 7 aneusomy had a significantly higher cytogenetic marker of human prostate cancer progression. Genes Chromosomes incidence within the poor prognosis group when compared to the Cancer, 9: 19â€”27,1994. 9. Shackney, S. E., Smith, C. A., Miller, B. W., Burtholt, D. R., Murtha, K, Giles, H. R., other chromosomes. This result, along with the almost constant pres Ketterer, D. M., and Police, A. A. Model for the genetic evolution of human solid ence (96%) of chromosome 7 aneusomy within the poor prognosis tumors. Cancer Res., 49: 3344â€”3354,1989. group, strongly suggests that aneusomy of chromosome 7 is a non 10. Macoska, J. A., Micale, M. A., Sakr, W. A., Benson, P. D., and Wolman, S. R. Extensive genetic alterations in prostate cancer revealed by dual PCR and FISH random alteration which correlates with clinically aggressive prostate analysis. Genes Chromosomes Cancer, 8: 88â€”97,1993. cancer. Especially remarkable was the presence of trisomy 7 in the 11. Trent, J. M., Meyskens, F. L., Salmon, S. E., Ryschon, K., Leong, S. P. L., Davis, J. R., and McGee, D. L. Relation of cytogenetic abnormalities and clinical outcome five FCM diploid tumors in the poor prognosis group. This result in metastatic melanoma. N. Eng. J. Med., 322: 1508â€”1511,1990. suggests that trisomy 7 may be an early change that occurs prior to 12. Waidman, F. M., Carroll, P. R., Kerschmann, R., Cohen, M. B., and Field, F. G. other chromosomal aneusomies in poor prognosis prostate carcinoma. Centromeric copy number of chromosome 7 is strongly correlated with tumor grade and labeling index in human bladder cancer. Cancer Res., 51: 3807â€”3813,1991. Chromosome 7 aneusomies have been correlated with 13. Collard, J. G., van de Poll, M., Scheffer, A., Roos, E., Hopman, A. H. M., Geurts van aggressiveness in melanoma (1 1) and bladder carcinoma (12). Kessel, A. H. M., and van Dongen, J. J. M. Location of genes involved in invasion Recently, a significantly higher rate of trisomy 7 has been observed and metastasis on human chromosome 7. Cancer Res., 47: 6666â€”6670,1987. 14. Johanson, B., Heim, S., Mandahl, N., Mertens, F., and Mitelman, F. Trisomy 7 in in advanced (T3NOMOand higher) but not in early (T2N0M@)stages nonneoplastic cells. Genes Chromosomes Cancer, 6: 199â€”205,1993. of prostate tumors (8), although neither survival information nor 15. Dal Cm, P., Aly, M. S., Delabie, J., Ceuppens, J. L, van Gool, S., van Dame, B., Baert, L, van Poppel, H., and van den Berghe, H. Trisomy 7 and trisomy 10 clinical follow-up was reported. It is possible that aneusomies of characterize subpopulations of tumor-infiltrating lymphocytes in kidney tumors and chromosome 7 are simply associated with poor prognosis tumors in surrounding kidney tissue. Proc. NaIl. Acad. Sci. USA, 89: 9744â€”9748, 1992.

4002

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1994 American Association for Cancer Research. Aneuploidy and Aneusomy of Chromosome 7 Detected by Fluorescence in Situ Hybridization Are Markers of Poor Prognosis in Prostate Cancer

Antonio Alcaraz, Satoru Takahashi, James A. Brown, et al.

Cancer Res 1994;54:3998-4002.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/54/15/3998

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/54/15/3998. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.