Discrepancies of DNA Content of Various Solid Tumours before and after Culture Measured by Image Analysis

Path. Res. Pract. 189, 1161-1168 (1993)

Discrepancies of DNA Content of Various Solid Tumours before and after Culture Measured by Image Analysis Comparison of Cytogenetical Data M. Lorenzato1, M. Doco2 , B. Visseaux-Coletto1, D. Ferre2 , H. Bellaoui 1, G. Evrard 1 and J. J. Adnet 1 lLaboratoire Pol Bouin and INSERM U 314, Reims, 2Laboratoire de cytogenetique, hOpital Maison Blanche, Reims, France

SUMMARY Parallel cytophotometric ploidy studies and cytogenetic analysis were performed on 15 various human solid tumours. The quantification of DNA by image analysis was carried out on cytological imprints of fresh tumours and on smears obtained after cell culture. The results obtained by both sets of calculations were compared with each other and with the cytogenetic results. 6 cases (40 %) showed concordance between the 3 techniques. One case was aneuploid for both DNA image analysis measurements but the cytogenetic data showed only a diploid stem line. In 3 cases outof15 (20 %), smears DNA analysis and cytogenetic results were concordant: in 2 tumours, the culture step failed to preserve aneuploid stem lines that were present in the imprint analysis. In the third one, a minority tetraploid peak observed after culture was absent on the imprint slide. Concordance between imprints and cytogenetic data and discordance with smears' analysis was observed in 3 cases (20 %). These 3 cases were diploid or near diploid but the DNA analysis on the smears after culture showed an aneuploid stem line in each case. The last 2 cases showed a total disagreement between the 3 techniques. By measuring the DNA content with an image analyser, the observer can ensure that only tumoral cells are taken into account. The present study revealed that cytogenetic data represent only about 60 % of the population that is effectively present in the culture dish and that the cultured population represents only 47 % of the population present on the fresh tumour imprint.

Introduction There are at least two techniques for revealing the chromosome content of tumoral cells. The first one uses cytometry after DNA staining and measures the DNA content and the second consists of a cytogenetic analysis and needs a period of culture. It offers details of the structure and number of abnormalities in chromosomes. Both techniques are of interest and complement each other. However, many authors have reported conflicting ©

1993 by Gustav Fischer Verlag,

Stuttgart

results when comparing DNA quantification (most often by flow cytometry)7,9, 12, 13, 14, 19 and cytogenetic data. We also observed such discrepancies, and we suspected that this might be due to the culture step which selects or eliminates some sub-populations. In order to test this hypothesis, we quantified the DNA content of 15 various solid tumours by image analysis on cytological imprints of fresh tumours and on smears after 0344-0338/93/0189-1161$3.50/0

1162 . M. Lorenzato et al.

culture. These results were compared to each other and to the cytogenetic results.

Results Description of the Histograms Obtained

Material and Methods We analysed the DNA content both on imprints of the fresh tumours and on smears after culture of 15 surgical specimens of various tumour types.

DNA Quantification Nuclear DNA content was determined using the CAS 200 image analysis system (BECTON DICKINSON) with the ploidy measurement software2 • Air-dried imprints of the tumours were stained according to the Feulgen technique after a 30-minute period of fixation in buffered formalin and a 1-hour 5N HCI hydrolysis at room temperature. The calibration slides were imprints of fresh rat liver, as recommended by the CAS. These slides were air-dried, and fixed 30 mn in buffered formalin and dried until being stained in the same bath as the slides that had to be analysed. For the ploidy measurement software, at least 20 calibration cells are necessary to be authorized to accede to the quantification section. The cells are chosen by the computer (tetraploid hepatocytes), and the user can accept or reject each cell. To have the calibration as precise as possible, the coefficient of variation of the calibration histogram was always less than 3 %. For DNA quantification, where possible, we measured at least 200 cells. The results are expressed as the measured DNA Index (DI) of the GO/G1 population, since by definition the DNA index of normal cells is 18 • In our study we worked with an image analyser (CAS 200) and as described by Taylor et al. we consider that a DNA diploid population can have an index varying between 0.92 and 1.08, and in the same way, the index of a tetraploid population can vary from 1.92 to 2.08.

Cytogenetic Analysis A sample of the surgical specimens was collected in HAMIDMEM medium supplemented with 10 % foetal calf serum. The samples were then submitted to a mechanical and enzymatic dissociation (collagenase II 1.4 mg (SIGMA) and DNAse 1.3 mg (SIGMA) for 10 ml HAM F12 medium) for 1-2 hours at

3rc.

The cells were placed into culture dishes at 37°C and incubated (025 %, C02 5 %, N2 90 %). After 48 hours of incubation, the supernant containing necrotic cells was discarded and the culture dishes were re-incubated after the addition of fresh medium. The dishes were observed daily. At 70 % confluence, they were treated for karyotyping. At this step, an aliquot of the culture was taken to obtain smears. Mitotic cells were blocked at metaphase with colchicine (0.02 g/ml, for 2 hours), and trypsinized and centrifuged for 10 min at 1 000 tr.lmin. To disperse the chromosomes, the cells were submitted to a hypotonic shock with a medium containing 50 % distilled water, foetal calf serum and KCL (0.075 M) for 10 min at 3rC, followed by fixation (acetic acid, ethanol 113 v/v). After centrifugation, the cells were spread on cold slides and dried overnight. To obtain the karyotype, at least 30 cells were cytogenetically analyzed when possible.

The nuclei measured are classified on a histogram according to their DNA content. On the abscissa, the DNA mass of the nuclei and on the ordinate, the number of nuclei are shown. The letters ABCD on the abscissa are used to classify the measured population of cells into four separate categories. The percentage of each area A, Band C is given on the right lower side of the result figure. The letter D is used as an outlier classification causing the program to ignore the cells labelled in that manner. For each peak, for example for the main peak, three values are given, i.e. the mode of the DNA mass of the peak in picograms, the DNA index of the peak and the average area of the nuclei in the modal bin of the peak. The second peak values are calculated similarly to the values of the main peak; the second peak is chosen by the user. We obtained DNA quantification on imprints and culture smears with parallel cytogenetic results in 15 cases from various tumour types. Table 1 summarizes the clinical data and the different results obtained for each tumour. The cytogenetic index is calculated according to Mendelsohn 11. Agreement Between the Three Methods The three methods were in agreement in 6 cases (40 %) with respect to the presence or absence of DNA aneuploidy. All these cases were diploid or both diploid and tetraploid. Two tumours were benign and four were malignant (Fig. 1). Concordance Imprint/Smear, Discordance with Cytogenetics One case showed similarities between imprints and smears but contradicted with the cytogenetic result. Both Feulgen quantifications showed an aneuploid peak (ID 1.89 and 1.84 respectively), whereas the cytogenetic modal chromosome number was 49 (+4, +18, +2), which corresponds to a cytogenetic index of 1.09 (Fig. 2). This population was in a minority in the culture which was mainly composed of stromal cells. The diploid peak present in the smear histogram corresponds to the stromal cells. The near diploid population is present in a very small proportion (12 %) on the fresh tumour imprint and in a much higher percentage (40 %) on the smear after culture. Discordance Imprint/Smear, Concordance Smear/Cytogenetics Cases 8 and 10 (serous papillary cystadenoma of the ovary and pheochromocytoma respectively) were most interesting because of the similarity between smears and cytogenetic results, but there was discordance with the imprint results. In cases 8 and 10, the culture step failed to preserve the aneuploid population present in the fresh tumour, which was revealed by the analysis of the imprint (Fig. 3).

DNA Content by Image Analysis and Cytogenetics· 1163 Table 1. Clinical, cytophotometric and cytogenetic data of the 15 tumours analyzed. DI: DNA Index, min: population in the minority, maj: population in the majority D1 Imprint

01._

I

t}1)eJlf tho IWDOIIr Coocotdaoce betweeII the 3 IeduIl

III

107

2

LI VER foc:al nodu1ar byperplasla

096

)

5

Neuroblastoma LUNG leiomyosarcoma LUNG epldermc)ld can:tnoma

103 2.02 0.94 099 III

6

ELBOW . synovIaJosan:oma

7

Conoordance IJDpnotIlIIDCIIt, dwordance ,.ith CJ\O&enetict BREAST Invasive dIIctaI carc:!ooma (SBRl)

103 2 13

N-

4

Ditcordanc:c Imprinll_ ,

cltmmoeome ItUIIIber

IDdt:x

46 XX (~) 45(022) (min) 46XX(min) 92 XXXX (min) 46XX 46XY 46XY (maj) 47XYY(mIn) 47XY(+12) (min) 47XY(+3) (nun)

I 0.99 I 2 I I I 1.009 ( 024 1035

(0) 2

46XX (maj) 92XXXX (min)

I 2

1.89 1.12 (12.1%)

1.84 1.02 (3W.)

49 (+4,+18,+2)

1.092

( 29

0.96

46XX

I

101 191 0.97

46XY (rnaJ) 92XXYY (min) 46XY

I 2 I

46XX(~)

I 1.035 I 0.975 I 2

10) II 104

CXICICOI"dance

StnU1~

8

OVARY serous paptllary cystadenoma

9

INTESTINE hyperoeJlular leiomyoma

094

10

Pheochromocytoma (relapse)

088 182

2.58

Conoordanc:e imprinWcytoaenetiCli. diJc:oIdancc wu.b II

smears OVARY : teratoma

112

12

KIDNEY polycystic

112

13

THYROID veJlcular adenoma

103

14

DiJcordancc betwccn the 3 mctbods BREAST mucosal colloid wcinoma

IS

LUNG epidermoid c;arcinorna

125 242 I 12 2.16 IJS

196-3.94

121 2.4 1.23 2.83

c e 1

diffuse aneuploidy

(41 586On80818283 8S) 46XY(maj) 4SXY(-7) (min)

1.09 2.04

70

Distribution of DNA Mass 1

a

1..2

C

28

u n t

11..

pg. 6.2 Mass .91..1.. Index 59.12 \lJ1l2 Area Second Peak 13.1 pg. Mass 1. 98 Index Area 172.6 pm2

O~===r====j==="'f"====r

a l l

DDDDDDDDDAAAAAAABBBBBBBBBB(BBBBB DNA Mass (Picograms)

100 C

e

1 1

Fig. 1. Concordance between the 3 techniques. Case 3. Neuroblastoma. a: imprint analysis, b: smear analysis, cytogenetic: 46XX.

h

u n

20 O~====F===~F====9==~==r

0 1 2

Count Total 206 Shown(ABC) 201 OUtlier(D) 2 Off scale 3 AreaA 89.55 AreaB 10.1..5 .0 AreaC

%

0;"

~~

Main Peak pg. 7.0 Mass 1.03 Index 38.86 1l.m2 Area Second Peak 1«,.5 pg. Mass 2.13 Index 78.2(0 \lJ1l2 Area


60 «,0

t

2

8

c o

Distribution of DNA Mass 1

I

Main Peak

2

56

1

o

4?XX (+3) (min) 46XX(maj) «XX(mili) 46XX (rnaj) 92XXXX (min)

3

DDDDDAAAAABBBB(BBBBBBBCCCCCCCCCC DNA Mass (Picograms)

Count 201 Total Shown(ABC) 200 OUtlier(D) 1 a Off scale 88.5 ., AreaA 9.5 AreaB '1" 2.0 AreaC

.'.

1164 . M. Lorenzato et al.

C

e

5

c

28

u n

110

1 1

o

Distribution of DNA Mass 1 2 10

70

I

a

0~=====T======T======r======T

t

o

123

DDDDAAA~AABBBBBBBCCCCCCCCCCCCCCC

DNA Mass (Picograms)

C

e

t,.o

1

30

c

20

u

10

1

o

n t

Distribution of DNA Mass 1 2 10

h

50

0~~~9====='=====~====~

o

123 DDDDAAAAABBB(BBCCCCCCCCCCCCCCCCC DNA Mass (Picograms)

Distribution of DNA Mass 101 2

a

70 C

e 1 1

c o

u n t

O~==~~~~~====~====~

o

123

DDDDDAAA~AAABBBBBBBBBBCCCCCCCCCC

DNA Mass (Picograms)

c

50

e 1

4-0

1

30

C

20

u

10

o

n t

Distribution of DNA Mass

T

2

h

t,.

O~==~~~~~====F===~ o 1 2 3

AAAAAAAAABBBBB>CCCCCCCCCCCCCCCCC DNA Mass (Picograms)

Main Peak Mass 12.5 pg. Index 1. 89 Area 137.5 Second Peak Mass 7.10pg. Index 1.12 Area 70.92 Count 'fotal 214 Shown(ABC) 212 Outlier(D) 1 Off scale 1 AreaA 12.7(0 % AreaB 69.3(0 % AreaC 17.92 '" Main Peak Mass 6.0 pg. Index .887 Area 36.10-6 pm2 Second Peak Mass 12 . 5 pg. Index 1. 81,0 Area 69.82 urn ~ Count 205 Total 204 Shown(ABC) Outlier(D) 1 o Off scale 39.22 AreaA AreaB 57.35 . ". 3.1,03 " AreaC

Fig. 2. Concordance imprint/smear, discordance with cytogenetics. Case 7. Breast carcinoma (SBR 3). a: imprint analysis, b: smear analysis, cytogenetic: 49 (+4, +18, +2).

Main Peak Mass 17.0 pg. Index 2.58 Area 90.12 urn: Second Peak Mass 8.5 pg . Index 1.29 Area 62.57 urn ~ Count 253 Total 2(0(' Shown(ABC) o Outlier(D) 7 Off scale 310.96 ,_ AreaA 61. 38 % AreaB 3.66 % AreaC

Main Peak Mass 6.5 Index .96 33.89 Area Second Peak 11,0.2 Mass 2.09 Index 75.55 Area

pg.

urn: pg .

urn :

Count 160 Total 159 Shown(ABC) 0 Outlier(D) 1 Off scale 63.52 " AreaA 22.6(0 ~/. AreaB 13.810- ... AreaC

Fig. 3. Discordance imprint/smear, concordance smear/cytogenetics. Case 8 Ovary: serous papillary cystadenoma. a: imprint analysis, b: smear analysis, cytogenetic: 46XX.

DNA Content by Image Analysis and Cytogenetics· 1165

e

1 1

C 0

2

80 60 ~O

u

20

t

0

n

a

1

100 C

Area(ABC) A Main Peak pg. 6.2 Mass .9<'<' Index 71. 2<' \llII 2 Area Second Peak 6.0 Mass PI'· .91<. Index 71. 58 \llII" Area

Distribution of DNA Mass

0

1

1

AAAAAAAAAAA' AAAAAAAAAAAAAAA

DNA Mas:! (Picosrams)

30 ('

Fig. 4. Disordance imprint/smear, concordance smear/cytogenetics. Case 9. Intestine: hypercellular leiomyoma. a: imprint analysis, b: smear analysis, cytogenetic: 46XY (maj), 92XXYY (min).

u n t

1

Distribution of DNA Mass ~ 2 <.

b

21'i

:~1

· .. 0~~~!!!!1~~~~2~L~!!=32 DDAA~.AAABBB(BBBCCCCCCCCCCCCCCCC

DNA Mass (Picograms)

C

e 1

200

120

C

80

u

100

n t

I

160

1

o

1

a

2

Main Peak Mass 6.9 pg . Index :)..01 Area 109.102 ).lIl\2 Second Peak Mass 13.0 pg. Index 1. 91 Area 67.31 ).lin" Count 150 Total 1I.t7 Shown(ABC) 1 Outlier(D) 2 Off scale 46.26 AreaA 36 . 05 AreaB 17 .69 AreaC

Mass

Index Area

O~===-~-===~======T=====~

o

123

Distribution of DNA Mass

10

1

b

2

% % %

79.37

pg.

\llII 2

Mass

8.9

pg.

e 1 1

80

C

ItO

u

2

Count

o

n t

60

o~~~~~~~~ o

1

~(.

Count Total 32<' Shown(ABC) 321 Outlior(D) 3 o Off scale 91. 28 % AreaA 8.1 % AreaS .62 % AreaC

Main Peak

It

13.1 1. 98

Index 1. 35 Area 71t.39 \llII 2 Second Peak Mass 17.2 pg. Index 2.61 Area 91t.86 ).llm2

C

0/, ~.

pg. 6.8 Mass 1.03 Index 59.18 \llII 2 Area Second Peak

DDDDAAAABBBBB(BBCCCCCCCCCCCCCCCC DNA Mass (Picoarams)

Fig. 5. Concordance imprint/cytogenetics, discordance with smear. Case 13. Thyroid vesicular adenoma. a: imprint analysis, b: smear analysis, cytogenetic: 46XX (maj), 92XXXX (min).

Count 206 Total 206 Shown(ABC) 0 Outlier(D) 0 Off scale 100.0 AreaA .0 AreaB .0 AreaC

2

3

DDDDDDAAAAABBBBBB(BSCCCCCCCCCCCC DNA Mass (Picoarama)

Total

224-

221 Shown(ABC) 1 Outlier(D) 2 Off scale 76.4-7 AreaA 20.36 AreaB 3.17 AreaC

% % %

1166 . M. Lorenzato et al.

c e 1 1

102

C

28

o

Distribution of DNA Mass

70

1

2

a

~

56

u

n t

0~====~=====r=====T=====T o 1 2 3

DDDDAAAAAAAABBBBB
20

1

(:

1

16

1

12

c

8

Distribution of DNA Mass 2

I

~

I

h

8

u n t

O ~====9=====~======~====~

o

1 3 ~ 6 DDDDAAAAABBBB
DNA Mass CPicograms)

AreaCABC) A Main Peak Mass 8.7 pg. Index 1. 21 Area 8~.83 prn2 Second Peak Mass 17.2 pg. Index 2.10. Area 11o~.5 prn2 Count Total 220 ShownCABC) 207 OutlierCO) 12 Off scale 1 AreaA 70.53 % AreaB 26 . 09 % AreaC 3.38 % AreaCABC) A Main Peak Mass 13.3 pg. Index 1.96 Area 73.910 )lJn 2 Second Peak Mass 26 . 8 pg. 3.910. Index 128.0 p.m 2 Area Count Total 113 110 ShownCABC) OutlierCO) 3 o Off scale ~8.18 % AreaA 1.2.73 % AreaB 9.09 % AreaC

Fig. 6. Discordance between the 3 techniques. Case 14. Breast: mucosal colloid carcinoma. a: imprint analysis, b: smear analysis, cytogenetic: diffuse aneuploidy.

In case 9, the results were similar between the three techniques concerning the main population. However, after culture, the presence of a minority sub-population was revealed. This .second population did not appear on the imprint histogram (Fig. 4).

distributed between hypodiploidy (41 chromosomes) and hypotetraploidy (85 chromosomes) (Fig. 6).

Concordance Imprint/Cytogenetic Discordance with Smears

To our knowledge, there is no other study which compares by flow cytometry or by image analysis the DNA content of a tumoral population before and after culture, to the cytogenetical data. The three techniques used were concordant in 40 % of the cases. These results are in agreement with other studies such as those of Coon et al. 4 or Shackeney et al.14 who worked on 30 and 20 tumours respectively. Some authors obtained better results i.e. a better percentage of concordance, but they studied a large number of cases of only one tumoral type (147 childhood leukaemia 10 and 283 adult leukaemia3 ). The majority of the discrepancies described were when the cytogenetic analysis showed hypodiploid metaphases and flow cytometry a diploid population 14 • This could be linked to the capacity of resolution of the cytometer. Concerning the discordant cases we obtained, only one case showed concordance between both DNA quantifications (DNA aneuploidy + near diploidy), but disordance with the cytogenetical result (unique diploid clone). The absence of the DNA aneuploid stem line in the cytogenetic analysis can be easily explained by the fact that too few metaphases were obtained for karyotyping. It must also be

In three cases, there was discordance between imprints and smears, but close similarities between imprints and cytogenetics. Cases 11, 12 and 13 clearly identified diploid or near diploid peaks on the imprint analysis. The cytogenetic study revealed only diploid, or diploid and tetraploid metaphases (Fig. 5), but in each case, the smear analysis revealed an abnormal population. An aneuploid peak was present in cases 11 and 13 and a near diploid/near tetraploid population appeared in case 12.

Discordance Between the Three Techniques Cases 14 and 15 showed a total disagreement between the three techniques. Case 14 was aneuploid on the imprint analysis, but multiploid in both smears and cytogenetic analysis. The smear analysis identified two main peaks (ID 1.96 and 3.94) which correspond to a near tetraploid and near octaploid population. These two populations were not present in the cytogenetic analysis. The cytogenetic study revealed a diffuse aneuploidy with various metaphases

Discussion

DNA Content by Image Analysis and Cytogenetics· 1167

noticed that in that case, the culture step had favoured the development of the near diploid stem line (higher proportion on the smear slide). In three other cases, the culture modified the DNA content of a tumoral cell population i.e. disappearance of a DNA aneuploidy in two cases and polyploidization of the base stem line in the third one (case 9). This phenomenon has already been described by Atkin et al. 1, in a study which compared Feulgen staining results with cytogenetic data. Polyploidization was also observed by Deaven et al. 6 • They explained that when the stem line stabilizes as a consequence of exponential growth, it tends to contain invariant DNA values close to multiples of the haploid genome. It is more difficult to explain the discrepancies of the cases 11, 12 and 13 where imprints agreed with the cytogenetic data but not with the smears' results. It seems likely that in cases 11 and 12, the discordance can be explained by the heterogeneity of the tumour (ovarian teratoma and polycystic kidney) which is likely to contain various stem lines. The presence of a minor DNA aneuploid population selected during the culture and probably submitted to polyploidization is one explanation for this discordance. However, this DNA aneuploidy was not detected in the metaphases examined for karyotyping. Polyploidization can also be invoked to explain the discrepancies between imprint and smear analysis in case 14. However, the cytogenetic analysis showed a diffuse aneuploidy, ranging from 41 to 85 chromosomes, which does not correspond to the DNA histogram obtained. Such a diffuse aneuploidy had already been described by Shackeney et al.1 4 . On the other hand, case 15 showed a multiploid histogram for the imprint and a polyploid (diploid + tetraploid) one on the smear. In some cases, as described by Siegfried et al. 15 , multiple populations can exist in the tumour which do not proliferate in vitro. Some cells will adapt to the culture. These cells come from the major DNA aneuploid or near diploid component of the solid tumour in most cases. This is what probably happened in case 15. Our study attempts to demonstrate that cytogenetic data represent only 60 % of the population that is present in the culture dish, and that the cultured population represents only 47 % of the population present on the fresh tumour imprint. Image cytometry is very useful for ensuring that only tumoral cells are taken into account. We did not have the problem of hypodiploid abnormal karyotypes associated to a diploid DNA content like other authors 3,4,16 who analyzed DNA by flow cytometry. By this technique, no difference can be made between tumoral and stromal cells, and if the non-tumoral population is important, no distinction can be made between the diploid peak representing those cells and an eventual hypodiploid tumoral population. The aim of the DNA quantification is to estimate the prognosis (overall or disease-free survival). Unfortunately we do not yet know any further progression or relapse in any of the 15 tumours we have studied. Some authors like Crissman5 emphasized that abnormalities of the karyotype

in case where flow cytometry only shows diploid stem line, can predict a more aggressive disease. Moreover, as suggested by Teyssier 18 , 37 % of the non-malignant tumours have clonal chromosome changes (numerous deviations and/or structural rearrangement). Their analysis may provide information about genetic events which change benign tumour cells into those showing malignant potential.

References 1 Atkin N], Mattinson G, Baker MC (1966) A comparison of the DNA content and chromosome number of fifty human tumours. Br] Cancer 20: 87-101 2 Bacus S, Grace L] (1987) Optical microscope system for standardized cell measurements and analyses. Appl Optics 26: 3280-3293 3 Barlogie B, Raber MN, Schumann], Drewinko B, Schwartzendruber DE, Gohde W, Adreeff M, Freireich E] (1983) Flow cytometry in clinical cancer research. Cancer Res 43: 3982-3997 4 Coon]S, Schwartz D, Summers]L, Miller AW, Weinstein RS (1986) Flow cytometric analysis of deparaffinized nuclei in urinary bladder carcinoma. Comparison with cytogenetic analysis. Cancer 57: 1594-1601 5 Crissmann]D, Zarbo R], Tyoku SM (1988) DNA analysis of urothelial neoplasia by flow cytometry and chromosomal analysis. Cytometry 9: Suppl 2, p20A 6 Deaven LL, Cram LS, Wells RS, Kraemer PM (1981) Relationship between chromosome complement and cellular DNA content in tumorigenic cell population. In: Genes, Chromosomes and Neoplasia. Arrighi FE, Rao PN, Strubblefield E (Eds), Raven Press, NY, 419-448 7 Dooley WC, Allison DC, Robertson] (1991) Discrepancies among the metaphase, telophase and the GO/G1 and G2 DNA peaks of heteroploid cell lines. Cytometry 12: 99-106 8 Hiddeman W, Schumann], Andreeff M (1984) Convention of nomenclature for DNA cytometry. Cytometry 5: 445-446 9 Lee GM, Youngberg GA (1986) A misleading flow cytometric analysis of DNA in an adenocarcinoma: a comparative flow cytometric and cytogenetic study.] Surg Onco131: 275-278 10 Look AT, Robertson PR, William DL, Rivera G, Bowman WP, Pui CH, Ochs ], Abromowitch M, Kalwinsky D, Dahl GV, George S, Murphy SB (1985) Prognostic importance of blast cell DNA content in childhood lymphoblastic leukaemia. Blood 65: 1079-1086 11 Mendelsohn ML, Mayall BH, Bogart E, Moore DH, Perry BH (1973) DNA content and DNA based centromeric index of the 24 human chromosomes. Science 179: 1126-1129 12 Petersen SE, Friedrich U (1986) A comparison between flow cytometric ploidy investigation and chromosome analysis of 32 human colorectal tumours. Cytometry 7: 307-312 13 Remvikos Y, Gerbault-Seurreau M, Vielh P, Zafrani B, Magdelenat H, Dutrillaux B (1988) Relevance of DNA ploidy as a measure of genetic deviation: A comparison of flow cytometry and cytogenetics in 25 cases of human breast cancer. Cytometry 9: 612-618 14 Shackney SE, Burholt DR, Pollice AA, Smith CA, Pugh RP, Hartsock R] (1990) Discrepancies between flow cytometric and cytogenetic studies in the detection of aneuploidy in human solid tumours. Cytometry 11: 94-104 15 Siegfried]M, Ellison D], Resau]H, Miura I, Testa]R (1991) Correlation of modal chromosome number of cultured non-small

1168 . M. Lorenzato et al. cell lung carcinomas with DNA index of solid tumour tissue. Cancer Res 51: 3267-3273 16 Smeets AWGB, Pauwels RPE, Beck JLM, Geraedts JPM, De Bruyne FM], Laarakkers L, Feitz WF], Voous GP, Ramaaekers FCS (1987) Tissue specific markers in flow cytometry of urological cancers. III Comparing chromosomal and flow cytometric DNA analysis of bladder tumors. fnt] Cancer 39: 304-310 17 Taylor SR, Titus-Ernstoff L, Stitlely S (1989) Central values and variations of measured nuclear DNA content in imprints of

normal tissues determined by image analysis. Cytometry 10: 382-387 18 Teyssier ]R, Ferre 0 (1989) Frequent clonal chromosomal changes in human non malignant tumours. Int ] Cancer 44: 828-832 19 Tribukait B, Granberg-Ohman I, Wijkstrom H (1986) Flow cytometric DNA and cytogenetic studies in human tumours: A comparison and discussion of the differences in modal values obtained by the two methods. Cytometry 7: 194-199

Received February 1, 1993 . Accepted in revised form ]uli 23 , 1993

Key words: DNA - Image analysis - Cell culture - Cytogenetics - Solid tumours Dr. M. Lorenzato, Laboratoire Pol Bouin, Hopital Maison Blanche, 45 rue Cognacq-]ay, 51092 Reims, Cedex, France