Cytogenetic Analysis of Urologic Malignancies: Study of Tumor Colony Forming Cells and Premature Chromosome Condensation

0022-5347 /84/1311-0146$02.00/0 Vol. 131, January Printed in U.S.A.

THE JOURNAL OF UROLOGY

Copyright© 1984 by The Williams & Wilkins Co.

CYTOGENETIC ANALYSIS OF UROLOGIC MALIGNANCIES: STUDY OF TUMOR COLONY FORMING CELLS AND PREMATURE CHROMOSOME CONDENSATION J. M. TRENT,* T. STANISIC

AND

S. OLSON

From the Department of Internal Medicine, Cancer Center, and the Department of Surgery/Urology of the University of Arizona College of Medicine, Tucson, Arizona

ABSTRACT

We have utilized a human tumor clonogenic assay to grow, and subsequently cytogenetically analyze, tumor colony forming cells from human urologic malignancies. Results following chromosome banding analysis are presented from 4 cases of transitional cell carcinoma and 1 case of renal cell carcinoma. Preliminary evidence suggests a possible association between the loss of chromosome 8 and progression or recurrence of transitional cell carcinoma. Additionally, we have utilized the technique of premature chromosome condensation to identify the interphase chromatin profile of urothelial cells obtained by cystoscopy from 8 patients with transitional cell carcinoma and compared these results to urothelial cells obtained from 7 control patients. This study demonstrates that cells obtained from urologic cancers contain a high proportion of cells in late G1 , while normal urothelial cells are usually found in the early G1 phase of the cell cycle. Statistical correlation of premature chromosome condensation analysis suggests this method may be a useful adjunct to routine histopathology in discriminating between normal and cancerous urothelium. The precise identification of the patient at risk for the development of invasive bladder cancer is not routinely possible at this time. As a screening device, routine cytopathology is of limited value, especially in low grade lesions. 1 Even when carcinoma in situ (CIS) is present, Weinstein et al. 2 have pointed out the difficulties inherent in relying solely on histologic light microscopic criteria as a prognostic indicator. As stated by Veenema, 3 "histologic look-alikes ... do not always behave alike." In well-defined studies by Falor and colleagues,4 and by Sandberg,5 cytogenetic analysis has proven of considerable utility in providing prognostic information in human bladder cancer. However, the majority of these studies have not used chromosome banding analysis 4 and few studies have examined urologic malignancies for the presence of specific chromosomal alterations. 6 We have utilized a recently described clonal assay for transitional cell carcinoma (TCC) 7 to cytogenetically analyze 4 patients with TCC and 1 patient with renal cell carcinoma (RCC). Additionally, as a corollary to banding analysis, we undertook a study of 15 patients (8 with TCC; 7 controls) by means of premature chromosome condensation (PCC) in an attempt to distinguish normal versus malignant urothelium in a manner recently described for human acute leukemia. 8 MATERIALS AND METHODS

Patient samples. Tissue and cell specimens were obtained following informed consent from patients undergoing routine clinical care on the urologic service at the Arizona Health Sciences Center and Veterans Administration Hospital, Tucson, Arizona. Bladder washings were obtained by irrigation of the bladder with a sterile saline solution (0.9 percent NaCl) at the time of cystoscopy, prior to tumor resection. Samples of 4 solid bladder tumors were obtained by routine transurethral resection under anesthesia with samples processed for culture Accepted for publication August 24, 1983. Supported by grants CA-29476, CA-17094 and CA-23074 from the Public Health Service and grant CA-28303 from the National Cancer Institute. * Requests for reprints: Department of Internal Medicine, Section of Hematology and Oncology, 6th Floor Facility, University of Arizona Health Sciences Center, Tucson, AZ 85724. 146

as previously described. 7 A single renal cell carcinoma specimen was obtained at the time of radical nephrectomy. Results of pathologic stage and grade (performed at the time cultures were initiated for cytogenetic analysis) are presented in table 1. The clinical features of 15 patients studied for expression of premature chromosome condensation (PCC) are presented in table 3. All 15 patients were studied by cystoscopy, urinary cytology, and where appropriate, random bladder biopsy. Due to the limited number of cells obtained following cystoscopic lavage,7 cytogenetic analysis and PCC studies were not performed simultaneously on the same patient's sample. Studies are underway to compare both methods on the same tumor cell population. Clinical history of patients studied by cytogenetic analysis. Staging of bladder cancers was performed as per Jewett and Strong, with grading performed by the Duke method. The results of cytogenetic and PCC analyses were correlated with cystoscopic and histologic findings. Chromosome banding analyses of all cases were performed on patients who had not received chemotherapy or radiotherapy. Patient 1 had a history of multiple TCC in 1966, and following surgery received no further treatment with no evidence of recurrence for 15 years. The patient then developed hematuria with transurethral resection of the bladder tumor revealing Stage III Grade C TCC. He was subsequently treated with external beam radiotherapy (4000 rads) but died of acute respiratory failure 1 month post resection. Cytogenetic analysis was performed on this patient at the time of resection and prior to radiotherapy. Patient 2 had multiple TCC and CIS recurrent since 1978 but was untreated at the time of cytogenetic analysis. Overall survival could not be evaluated as this patient was lost to follow-up. Patient 3 had a 1 month history of gross hematuria, and at cytoscopy demonstrated a 1 x 1 cm. bladder tumor. The patient was subsequently treated with thiotepa and is currently disease-free 9 months after surgery. Patient 4 also had a 1 month history of hematuria and was initially treated with a segmental resection of the bladder and 6,500 rads of pelvic irradiation. Postoperative recurrence of the tumor was noted and the patient underwent radical cystectomy. The urethral margins were noted as positive, and 6 weeks later the patient underwent urethrectomy. The patient is currently well

CYTOGENETICS OF BLADDER CAf'JCEP., TABLE 'io

Clinical features and co!ony growth of 5 patients with urologic n~talignancies

Sex

Tumor Type

69

M

TCC

!II

C

2

56

M

TCC

II

B

3 4

62

M

A C

5

39

M F

TCC TCC RCC

II

77

Patient

Grade

III-IV High grade metastatic

Overall Survival From Culture Date (Months)

Treatment History

Stage

Multiple superficial TCC for 15 years Multiple TCC and C!S recurrent since 7/78 Newly diagnosed Newly diagnosed Newly diagnosed

Chromosomal Marker

(+/-) +

N.A.*

186 ND 480 106

+ +

TCC = Transitional Cell carcinoma. RCC = Renal cell carcinoma. * NA = not analyzed-patient lost to follow-up. ''*=NED = No evidence of disease. +=Not done.

TABLE 2. Cytogenetic assessment of TCFUs from 5 patients with urologic malignancies Case No.

Diagnosis

Total Banded Cells Analyzed

Banding Techniques

TCC

25

G,Q

2

TCC

61

G,Q,C

3

TCC

10

G

4

TCC

27

G

5

RCC

46

G,C

Clonal Karyotypic Abnormalities

Modal Chromosome Number (Range)

Per Cent Polyploidy

43 (26-47) 43 (33-46) 44 (26-L(4) 67 (54-69)

48

11

-8,

31 (23-44)

22

-1,-2,-3,-4,-5,-6 -7, -8, -9, -11, -12 -14, -16, -19, -19 -22) -x

Numeric

Structural

-2, -5, -8

3

-8, -15, -Y

0

-8, -16

I'

+l Umar

+16, + 16, - 18*

M 1 _ 3 Umar M 4 del(l) (p22:) M,- 7 Umar M 1 - t(l;2)(ql2;p21) Mar 2 - Umar

* Substantial numeric chromosome variation (loss and gain) was observed in this patient's tumor.

12 months after initial diagnosis and shows no evidence of disease. Patient 5 was admitted for elective cholecystectomy with symptoms of right upper quadrant and shoulder pain. At surgery a mass was noted in the right and radical nephrectomy was performed. Postoperative bone scan revealed multiple metastases and the patient died within 2 months of bronchopneumonia after unsuccessful attempts at therapy with irradiation and chemotherapy. Assay for formation. The clonogenic assay utilized was that of Buick et al. 7 Briefly, cells were plated at 5 x 105 cells per ml., with cultures incubated at 37C in a 7.5 per cent CO 2 humidified atmosphere of air. Cultures were examined at regular intervals utilizing an inverted microscope, and scored for colony formation after 7 to 28 Colonies were defined as aggregates of >40 cells with clusters defined as aggregates of 8 to 40 cells. Data on formation for 1 to 5 1s '"'"'"·wcu in table L Cytogenetic of tun10r cells was performed after the technique of Trent with cultures harvested for chromosome after approximately l week of incubation. Air-dried slides were prepared and banded as described. 10 Chromosome changes listed in table 2 represent alterations following the recommendations of the Standing Committee on Human Cytogenetic Nomenclature [ISCN(l978)]. 11 When structurally abnormal "marker" chromosomes could not positively be identified, they are referred to as unidentified markers (Umar). Premature chromosome condensation. Fusion of cells for analysis was performed using Sendai virus and mitotic Chinese hamster ovary (CHO) cells as described by Hittelman and Rao. 12 Whenever possible 100 PCC were analyzed for each patient sample. G 1 PCC were graded on a scale of 1 to 6 depending on the condensation of the interphase chromatin (I most condensed, 6 least condensed) after previously described criteria. 12 • 13 Results of G 1 interphase chromatin analysis (table 3) are presented as a ratio of cells found late in the G 1 phase (stages 4 to 6), over the total number of G1 PCC observed-

called the proliferative potential index (PPI). 14 All studies were conducted in a blinded fashion, with the individual scoring the fusions being unaware of the clinical history of the patient sample. Additionally, both cancerous and noncancerous urothelium was processed at the same time (and with the same reagents), to insure similar fusion conditions. RESULTS

Cytogenetic features of TCFUs from all in table 2. The modal chromosome number TABLE

3. Clinical features of 15 patients studied by premature chromosome condensation Diagnosis cystitis

12

888 117 811

13

866

14 15 16 17

249 262 384 107

18 19

818 116

20

661

10 11

Interstitial cystitis Neurogenic bladder Left ureteral stone Prostatitis *Previous TCC No current lesion or CIS Grade IV invasive urethral carcinoma Low grade noninvasive TCC Poorly differentiated TCC Invasive Grade III TCC Recurrent Grade lII Stage A TCC Grade III Stage A TCC Adenocarcinoma of bladder (effusion) Invasive high grade carcinoma

PP! (Per cent)

Cytologic Pos. for Malignancy

0 0

0 0 39 0 0

94

+

50 53 71 87.5 and 67.5 33 83

+

50

+

+ + +

+

* This value is indicated under the Histology Negative column of figure 4 (see text). BPH = Benign prostatic hypertrophy. CIS = Carcinoma in situ. PP! = Total cells in late G 1 (stages 4-6) over total G 1 PCC times 100.

148

TRENT, STANISIC AND OLSON

between cases varied from 31 to 67 with a wide range of chromosome numbers observed (26 to 72). Four of 5 patients (cases 1 to 2, 4 to 5) demonstrated polyploid mitoses. In each case polyploid mitoses contained chromosome changes found in the near-diploid population. Chromosome banding analysis revealed a wide variety of chromosomal alteration in all 5 patients. A representative Gbanded karyotype from cases 4 to 5 is presented in figures 1 and 2. Pictoral documentation of chromosome breakage and an unidentified marker chromosome from case 1 is presented in figure 3. Monosomy for chromosome 8 was the most commonly observed clonal numeric alteration (4 out of 5 patients). Three cases (cases 1, 4, 5) contained 1 to 7 clonal marker chromosomes. Case 4 demonstrated a wide range of numeric alterations as well as several nonclonal markers. The stem-line of 1 patient

®

12

13

14

15

17

19

21

22

X

2

FIG. 2. Near-haploid G-banded karyotype from TCFUs of patient with renal cell carcinoma (RCC) (Case 5). Clonal changes included -1, -2 -3 -4 -5 -6 -7 -8 -9 -11 -12 -14 -16 -19 -19 -22 -X, M'1 - t(1;2)(q12;p21), MrUmar.' Insert sh~ws G- and C-b~ndetl confirmation of origin of marker M 1 [normal chromosome l's inverted to demonstrate breakpoints]. Nullisomy 3 and +6 in this single cell do not reflect clonal alterations.

®

FIG. 1. Documentation of substantial chromosome breakage within tumor cells of clinically untreated patient with TCC (Case 1). A, B, examples of chromatid gaps (ctg) and breaks (ctb). C, example of complex chromatid exchange (cte). D, example of quadraradial formation (qr). E, large marker chromosome (mar) from TCFUs of Case 1 compared to normal chromosome 1 (arrow).

(case 5) demonstrated simple monosomy for a majority of chromosomes producing a near-haploid modal number (31 chromosomes). The patient's tumor cells always retained 2 copies of chromosomes 10, 13, 17, 18, 20 and 21. However, both copies of chromosome 19 were missing in all tumor cells from this patient. In addition to these substantial numeric alterations, the patient contained a large metacentric marker chromosome thought to derive from a translocation between chromosomes 1 and 2 [t(1;2)(q12;p21)], and 1 unidentified small metacentric marker (fig. 2). Premature chromosome condensation. Sixteen specimens from 15 patients were analyzed by PCC. Eight patients had documented carcinoma of the bladder, while 7 patients were analyzed following referral for urologic indications other than cancer (table 3). A single normal patient (case 12) had a previous history of bladder carcinoma but currently had no evidence of tumor by cystoscopic and cytological criteria. One untreated patient with carcinoma (case 17) was studied from bladder washes separated by a 1-week interval. The results of this study are found in figure 4. As can be noted, a highly concordant relationship exists between the clinical finding (positive or negative for urothelial cancer) and a high (;;s,37 per cent) PPI (p = 0.0014). In the single patient in which sequential samples were analyzed (case 17), a good concordance of PCC values was observed. The patient had not received therapy between the 2 studies and continued to demonstrate a very high PPI (87 per cent and 67.5 per cent respectively) (fig. 4, open circles). DISCUSSION

Chromosomal studies. The success rate for cytogenetic analysis of tumor colony forming cells is approximately 60 per cent for TCC and 75 per cent for RCC. This success rate is comparable to other direct or liquid culture techniques used in our

149

CYTOGEY..1ETICS OF BLADDER CA:l"•JCER

2

8

6

13

14

19

20

5

4

3

9

15

10

12

16

21

m4 m5 mG Mt\FlKERS

22

UNIDENTIFIED

17

18

X

y

MARKERS

FIG. 3. G-banded karyotype of polyploid cell from patient with transitional cell carcinoma (Case 4). Numerous numeric chromosome alterations (gain and loss) as well as clonal (box) and nonclonal marker chromosomes were observed. (See text).

laboratory. 9 • 1 4 The major advantage to the use of a clonal assay for cytogenetic analysis of human cancers resides in the reproducible and highly selective growth of tumor cells using this method. 14- 16 In contrast, the use of nonclonal techniques for cytogenetic analysis (combined with non-cellular specific mitotic arresting agents such as colchicine) would be expected to capture all dividing cell populations (normal or malignant) present during the culture period. However, because in vitro of TCC or RCC in our clonal assay is only approximately 40 per cent, 7 the method is most useful for cytogenetic analysis of selected patients, rather than for routine of tumor u~,.u,-,,~u The vuma,v11.uu u,,u,u,;o in this on bladder although performed on a small patient sample, are in agreement with reports from other laboratories.4- 6 As expected from previous studies,4-6 the chromosomal of invasive TCC was accompanied a high degree of karyotypic "instability" with both chromosome markers, and substantial numeric alterations (cases 1, 4). Our limited study, when coupled to the much larger work of Falor and colleagues 4 and Sandberg,5·6 supports the notion that substantial karyotypic alterations (that is, production of marker chromosomes) may reflect a propensity for recurrence and/or clinical progression in TCC. The single patient with noninvasive cancer (case 3) demonstrated a near-diploid chromosome range with no marker chromosomes, a finding consistent with previous reports. 6 Of interest, our study provides evidence for a consistent alteration among 3 of 4 cases of TCC; monosomy of chromosome 8. The clinical course of all patients in our study with monosomy for chromosome 8 was associated with clinical progression (recurrence) of disease. Although this finding will indeed require further confirmation, our results may suggest an association between the loss of chromosome 8 and tumor pro-

gression in TCC. Recent studies oftumorgenesis have suggested that structural chromosome alterations may be related through transforming sequences (oncogenes) to the generation of human malignancies. 17 Because unidentified marker chromosomes are present in 3 of our patients demonstrating monosomy 8, it is possible that portions of this chromosome remain within the genome, but went unrecognized despite banding analysis. In addition to frequent alteration of chromosome 8, numeric and structural alterations of chromosome 1 (cases 4, 5), a common finding in solid tumors including TCC, 6• 14 were observed. The with RCC demonstrated a near-haploid of 31 (table 2, fig. 2). Although near-haploid vm"''"Tari in a Of LU1TI0f typeSlS-Zl this 1st description of a near-haploid tumor lacking copies of an autosome [non-sex] chromosome. 20 Specifically, loss of both chromosome 19's occurred in all cells examined from this patient. However, although it was not possible to recognize any contribution of chromosome 19 to either marker chromosome in this tumor, this possibility cannot totally be excuded. The relationship between the karyotypic findings and the clinical course in 2 patients from our study (cases 1, 5) is of interest. In case 1, substantial chromosome breakage (chromosome breaks, gaps, quadraradials) was observed in 75 per cent of this patient's tumor cells (fig. 1). This type of chromosome change is commonly observed in cells exposed to chemotherapy, radiotherapy, or other clastogenic agents. However, at the time of our cytogenetic evaluation no therapy had been administered. Also, it is unlikely that this level of chromosome damage is associated with our in vitro culture technique, as no evidence of similar breakage was observed in any other patient from this study, and also the extreme rarity of such breakage

150

TRENT, STANISIC AND OLSON

Relationship Between Pathologic Finding of Tissue Specimen (Negative or Positive for Urothelial Malignancy) and the Proliferative Potential Index (PPI*) N=16 100 0

••

-

0

•

'#. a.. 50

a..

0 Histology Negative

Histology Positive

*Ratio: Late-G 1 /Total G1 PCC p =0.0014 FIG. 4. Analysis of urothelial cells using premature chromosome condensation (PCC). Sixteen specimens from 15 patients were analyzed cytopathologically, and results compared in blinded fashion to determine cell cycle status (i.e. percentage of cells in early or late G1 ) utilizing PCC. A highly concordant relationship exists between clinical findings (positive or negative for urothelial cancer) indicating that cells obtained from urologic cancers contain high proportion of cells in late G1 while normal urothelial cells are usually found in early G1 phase of cell cycle (see text). in studies of other tumors from this laboratory. Rather, it appears that extremely high levels of chromosome breakage in tumor cells from untreated patients may reflect an increase in the malignant potential of these cells, resulting in clinical progression and short survival duration, a notion supported by our previous work. 22 Unfortunately normal tissue (peripheral blood lymphocytes [PB Ls]) was not available for study in this patient. However, in our previous studies of patients with substantial chromosome breakage in their tumor cells, a corresponding increase in chromosomal breakage in normal cellular elements has not been observed. 22 Premature chromosome condensation. Premature condensation of interphase chromatin results from the viral fusion of a mitotic cell with a nonmitotic cell, causing a "premature" condensation of chromatin (PCC) in the nonmitotic cell. 23 Cytologic analysis of cells prior to DNA synthesis (G,-phase cells) reveals chromosomes with a single chromatid, cells following DNA synthesis (Grphase cells) have 2 chromatids, and cells undergoing DNA synthesis (S-phase cells) have a mixture of both single and double chromatid segments yielding a "pulverized" appearance. 12 ' 13 Utilizing the PCC technique, Hittelman and colleagues8 have recently proposed that the temporal position of cells in the G 1 phase (in effect, early or late G 1 cells) provides a reliable measure for discriminating between normal and malignant cell populations (normal cells reside early in G,;

malignant cells reside late in G 1 ). These findings (performed principally on malignant lymphoblasts) provided the basis for investigating the usefulness of the PCC technique in urothelial cell populations. Results of the cell fusion procedure are expressed in terms of a "proliferative potential index" (PPI). The PPI is the ratio of the number of fused G 1 cells found late in the G 1 stage over the total number of G 1 PCC's observed. After the work of Hittelman et al.,8 ;;.37 per cent was considered positive evidence of malignancy. As shown in figure 4, a highly concordant relationship was observed between the clinical finding (positive or negative for urothelial cancer) and the PPI (p = 0.0014). Our preliminary study strongly suggests that normal versus cancerous urothelium may be accurately identified by PCC analysis. Two exceptions (cases 10 and 18, figure 4, solid arrows) were observed in our limited study. Case 18 demonstrated a PPI of 33 per cent (consistent with the "normal" range) although the patient had a pathologically diagnosed Grade III Stage B TCC. One explanation for these results could reside in the removal of the patient's single 2 cm. mass prior to bladder barbatage, possibly resulting in a lack of malignant cells for PCC analysis. In case 10 a PPI of 39 per cent was observed in a patient with a left ureteral stone demonstrating no evidence of malignancy. This again represents a value of borderline significance. However, it should be noted that the threshold value of 37 per cent established for prediction of relapse in human adult leukemias 8 and utilized in this report, may not represent a cutoff value relevant to human urologic malignancies. Clearly, additional studies will be required before confidence levels can be accurately assigned to normal versus malignant human urothelium. It is of interest to note that the PPI values for normal bladder specimens were consistently near zero. These values represent both cases with a complete lack of G 1 fusions, as well as cases in which no late G1 fusions could be observed. The total lack of fusions from some normal bladder specimens is not thought to represent an artifact of the technique itself, because cancer specimens run simultaneously produced sufficient fusions for analysis. Additionally, difficulties in the ability to fuse appear to be an inherent property of normal cells (personal communication, A. Sandberg, Roswell Park, NY). Although our results with the PCC technique are limited by small patient numbers, PCC may provide a useful method to aid in discriminating between cancerous and noncancerous urothelial lesions. However, it should be stressed that as currently performed this method is labor-intensive, requiring significant time for sample preparation and technical manipulation, as well as substantial time for cytologic screening. Our experience with the PCC technique (as currently performed) suggests its usefulness in a research setting but would seem to preclude its use for large scale screening of early bladder cancers. However, as modifications are made to increase the sensitivity and decrease the time and expense required for this procedure, the PCC technique may provide a useful adjunct to routine histopathology. Acknowledgments. We would like to acknowledge the excellent technical assistance of Ms. K. Massey, Mr. F. Thompson, and Ms. B. Owens. REFERENCES 1. Frable, W. J., Paxson, L., Barksdale, J. A. and Koontz, W.W. Jr.: Current practice of urinary bladder cytology. Cancer Res., 37: 1800, 1977. 2. Weinstein, R. S., Miller, A. W. and Pauli, B. V.: Carcinoma in situ:

comments on the pathobiology of a paradox. Urol. Clin. North Am., 7: 533, 1980. 3. Veenema, R. J.: Comments on recognition of "early" bladder cancer and premalignant epithelial changes. Cancer Res., 37: 2836, 1977.

4. Summers, J. L., Falor, W. H. and Ward, R.: A 10-year analysis of chromosomes in non-invasive papillary carcinoma of the bladder. J. Urol., 125: 177, 1981.

151 A.: Chrornosorne H1arkers ar-1d progression of bladder cancer. Cancer Res., 37: 2950, 1977. Sandbei-g, A. A.: Chromosome studies in bladder cancer. In: Carcinoma of the Bladder. Edited by J. G. Connolly. New York: Raven Press, p. 127, 1981. Buick, R. N., Stanisic, T. H., Fry, S. E., Salmon, S. E., Trent, J. M. and Krasonich, P.: Development of an agar-methyl cellulose clonogenic assay for cells in transitional cell carcinoma of the human bladder. Cancer Res., 39: 5051, 1979. Hittelman, W. N., Lowanda, B. C., Dosik, M. D. and McCredie, K. B.: Predicting relapse of human leukemia by means of premature chromosome condensation. N. Engl. J. Med., 303: 479, 1980. Trent, J.M. and Salmon, S. E.: Human tumor karyology: marked analytic improvement via short term agar culture. Br. J. Cancer, 41: 867, 1980. Trent, J. M.: Protocols of procedures and techniques in chromosome analysis of tumor stem cell cultures in soft agar. In: Cloning of Human Tumor Stem Cells, Edited by S. E. Salmon. New York: Alan R. Liss, Inc., pp. 345-349, 1980. International System for Human Cytogenetic Nomenclature (ISCN). Cytogenet. Cell Genet., 21: 309, 1978. Hittelman, W. N. and Rao, P.: Mapping G 1 phase by the structural morphology of the prematurely condensed chromosomes. J. Cell Physiol., 95: 333, 1978. Trent, J. M: Recent advances in cancer cytogenetics. Arizona Med., 38: 836, 1981. Trent, J. M.: Specificity and significance of chromosome change in human gynecologic malignancies. In: Cancer of the Female

5. Sandberg

6. 7.

8. 9. 10.

11. 12.

13.

14.

Reproductive

1

15.

16,

17.

18.

19.

20.

21. 22.

23.

~f. "\Nhitehouse a:1d

VVil-

lia::ns. Sussex: Ltd. (In prsss). Carney D. N.J Gazdar, A. Bunn, P.A. and Guccion J.: Demonstration of the stern-cell nature of clonogenic tumor cells from iung cancer patients. J. Stem Cells, 1: 149, 1981. Dow, L. W., Bhakta, N. and Williams, J.: Clonogenic assay for Wilm's Tumor: improved technique for obtaining single-cell suspension and evidence for tumor cell specificity. Cancer Res., 42: 5262, 1982. Yunis, J. J.: The chromosomal basis of human neoplasia. Science, 221: 227, 1983. Trent, J. M. and Salmon, S. E.: Karyotypic analysis of human ovarian carcinoma cells cloned in short term agar culture. Cancer Genet. Cytogenet., 3: 279, 1981. Sekine, S.: Cytogenetic observations in tumors of the urinary tract and male genitals. Jap. J. Urol., 67: 452, 1976. Pathak, S., Strong, L. C., Ferrell, R. E. and Trindale, A.: Familial renal cell carcinoma with a 3;11 chromosome translocation limited to tumor cells. Science, 217: 939, 1982. Sandberg, A. A.: The Chromosomes in Human Cancer and Leukemia. New York: Elsevier, p. 511, 1980. Trent, J. M. and Salmon, S. E.: Preferential breakage of human tumor chromosomes: expression of "fragile sites" in malignancy. Proc. Am. Assoc. Cancer Res., 21: 39, 1980 (Abstract). Rao, P. N.: The phenomenon of premature chromosome condensation. In: Premature Chromosome Condensation-Application in Basic, Clinical, and Mutation Research. Edited by Potu N. Rao. New York: Academic Press, Inc., p. 2-41, 1982. 1

1