Correlation of Karyotype with Patient Sex and Age in Acute Myeloid Leukemia

You-Sheng Li

ABSTRACT: Karyotypes were determined in 109 patients with acute myeloid leukemia. The propor- tions of patients with nonclonal chromosomal abnormalities, with numeric changes only, and with severe chromosomal aberrations were all found to increase to a statistically significant degree with patient age. Most patients in whom less than half of all cells were abnormal were elderly males. These results indicate that the cell karyotype was less stable in elderly patients than in younger patients. A higher frequency of nonclonal chromosomal changes was found in patients with clonal abnormalities compared with those without such abnormalities. Male patients tended to gain chromosomes and had more hyperdiploid abnormalities than female patients, who tended to lose chromosomes and had more hypodiploid abnormalities. This trend of chromosomal gain in males and loss in females mainly involved chromosomes sim- ilar in size to the sex chromosomes. Three female patients with trisomy 8 and one with 7q + and t(8;21) showed an X chromosome twisted into a spiral shape. The results indicate that the initial karyotype effects the formation of some numeric changes. These findings are dis- cussed in relation to possible secondary chromosomal changes and karyotypic instability.

INTRODUCTION

Although a number of disease- and disease-subtype-specific chromosomal changes have been well established, they remain scanty among the numerous abnormalit ies revealed in human malignancy. The cause and biologic significance of such numer- ous abnormalit ies have been far from fully studied. Interestingly, several correla- tions of chromosomal changes with patient sex and age have emerged from this study, which allows further understanding of the great diversity of chromosomal changes in human malignancy.

MATERIALS AND METHODS

Karyotypes of bone marrow or peripheral blood ceils were determined in 109 pa- tients with acute myeloid leukemia (AML) seen at the Universi ty of Cambridge be- tween March 1980 and September 1982; 66 of these patients have been discussed previously [1]. Of the 109 patients with AML, 103 were studied at the time of di- agnosis and 6 during relapse. As the patterns of evolut ionary chromosomal changes revealed in a small percentage of patients with AML are essentially the same as those observed at the time of diagnosis [2, 3], the few patients studied during re-

From the Division of ttematology, Department of Medicine, Capital Hospital, Chinese Academy of Med- ical Sciences. Peking, China.

Address requests for reprints to Dr. You-Sheng Li, Department of Medicine, Capital Hos- pital, Peking, China.

Received July 20, 1983; accepted November 9, 1983.

73

© 1985 by Elsevier Science Publishing Co., Inc. Cancer Genetics and Cytogenetics 14, 73 81 (1985) 52 Vanderbilt Ave, New York, NY 10017 0165-4608/85/$03.30

7 4 Y . s . Li

lapse were inc luded. In addi t ion, to confirm certain conclusions, data on 209 cases of AML with chromosomal ly abnormal clones were collected from the l i terature [1, 4-12].

A convent ional cell cul ture technique and a s tandard G-banding technique were used [13]. The observat ion of at least two pseudod ip lo id or hype rd ip lo id cells or three hypod ip lo id cells showing the same abnormali t ies was required to establish the presence of an abnormal clone, and only patients wi th such clones were clas- sified as abnormal. When chromosomal ly abnormal cells were observed that did not meet the requirement for acceptance as clonal abnormali t ies , they were re- corded as nonclonal abnormali t ies . Occasional monosomies , chromosomal breaks, and delet ions were not regarded as nonclonal abnormali t ies , however, as they may have arisen during the preparatory process. At least 10, but usual ly 20 or more metaphases, were examined for each patient. The chromosomes were ident if ied ac- cording to the Internat ional System for Human Cytogenetic Nomencla ture [14].

All karyotypes were de te rmined by the study of cul tured cells, a l though the di- rect method as well as the cul ture method was appl ied in a few patients.

The ratio between chromosomal ly different cells was de te rmined by the s tudy of both Leishmansta ined and G-banded preparat ions. The combinat ion of analyses of Leishmansta ined and G-banded preparat ions a l lowed the s tudy of more cells. It was possible to analyze every cell that entered into the field of vis ion in most Leish- mansta ined preparat ions and thus avoid bias toward certain cells. This approach y ie lded more accurate results for establ ishing the ratio be tween chromosomal ly dif- ferent cells.

The patients were d iv ided into four groups according to the ratio between karyo- typica l ly normal and abnormal cells, taking only clonal abnormal i t ies into account: NN (normal cells only), NA (less than half abnormal cells), AN (less than half nor- real cells), and AA (abnormal cells only). Each of these four groups was further d iv ided into two subgroups according to the absence {O) or presence (P) of non- clonal abnormal i t ies to produce a total of eight subgroups: NNO, NNP, NAO, NAP, ANO, ANP, AAO, and AAP.

According to Mite lman and Levan [15, 16], chromosomal changes are of two types: pr imary and secondary. Secondary changes can be accumula ted pass ively at random, but cell selection may confer a survival advantage. Pr imary chromosomal changes are closely related to the genesis of neoplasms, whereas the numerous sec- ondary changes s imply reflect the karyotypic instabi l i ty of mal ignant cells. There- fore, the fol lowing parameters were used to s tudy the possibi l i ty of secondary chro- mosomal changes and karyotypic instabi l i ty in AML: (a) The presence or absence of nonclonal chromosomal abnormali t ies; (b) the propor t ion of pat ients with nu- meric changes only; (c) the percentage of karyotypica l ly normal cells in pat ients with clonal abnormali t ies; and (d) the propor t ion of patients with severe aberrat ions (i.e., five or more aberrant chromosomes and/or unident i f ied marker chromosomes).

R E S U L T S

One pat ient was exc luded because his age was not known. The remaining 108 pa- tients in this series were tabulated according to ages and karyotypic pat terns in Table 1, which shows that the propor t ion of patients wi th nonclonal chromosomal abnormali t ies increased with age. In the pat ients with nonclonal abnormali t ies , 55% (18 of 33) were 60 or more years old and 9% were less than 30 years old, whereas 29% (22 of 75) of patients wi thout nonclonal abnormal i t ies were 60 or more years old and 17% were less than 30 years old (p < 0.025). Thus, nonclonal abnormali t ies occurred more frequently in elderly patients. Nilsson et al. [17] pointed out that e lder ly pat ients were more l ikely to have a mixture of karyotypi-

Tab

le 1

. K

aryo

typi

c pa

tter

ns a

nd p

atie

nt a

ge i

n 10

8 pa

tien

ts w

ith

acu

te m

yelo

id l

euk

emia

NN

N

A

AN

A

A

Ag

e (y

r}

NN

O

NN

P

NA

O

NA

P

AN

O

AN

P

AA

O

AA

P

To

tal

NC

O

N

SV

(%

) (%

) (%

)

0-2

9

5 2

1 2

6 3

0-5

9

20

6 1

1 9

3 8

60

-89

16

3

3 3

2 5

3

To

tal

52

9 21

26

NC

(%

) 21

57

38

35

ON

(%

) 67

52

23

S

V (

%)

22

19

31

16

19

6 6

2 50

24

18

8

7 42

43

31

21

10

8

Abb

revi

atio

ns:

NN

, no

rmal

cel

ls o

nly;

NA

, le

ss t

han

half

abn

orm

al c

ells

; A

N,

less

th

an h

alf

norm

al c

ells

; A

A,

abn

orm

al c

ells

onl

y; O

, no

nclo

nal

abno

r-

mal

itie

s ab

sent

; P,

non

clon

al a

bn

orm

alit

ies

pres

ent;

NC

, n

on

clo

nal

abn

orm

alit

ies;

ON

, n

um

eric

cha

nges

onl

y; S

V,

seve

re a

berr

atio

ns.

7 6 Y.S . Li

cally normal and abnormal cells. Such a t rend was also present in this series of patients. Of nine NA patients, six were 60 or more years old, whereas only 36% (36 of 99) of the remaining patients were 60 or more years old (p - 0.06). However, AN patients d id not appear to be older in compar ison with the remaining patients. As shown in Table 1, the propor t ion of pat ients wi th numeric abnormal i t ies only

Figure 1 A metaphase showing tr isomy 8 and a twis ted X chromosome. The latter is spiral in shape, whereas the remaining chromosomes are well spread.

R

xl

¢

Correlation of Karyotype with Sex and Age in AML 77

and of those with severe chromosomal aberrations also increased with age. Both differences were statistically significant.

We have noted a significantly lower frequency of chromosomal abnormali t ies in female patients compared with male patients [1]. In this study, the frequency of clonal abnormalit ies was 45% (24 of 53} in female patients as compared with 59% (33 of 56) in male patients. A further study of differences between the sexes re- vealed a trend in numeric changes of chromosomal gain in males and loss in fe- males. When patients with hypodiploid karyotypes caused by changes in sex chro- mosomes were excluded, the sex ratio (male/female) was 2.8 (17/6) in hyperdiploid patients and 0.7 (5/7) in hypodiploid patients (p = 0.05). When patients with clonal abnormalities were classified as NA, AN, and AA, this trend was found to become more distinctive with the increase of normal cells in different groups, f.e., the most distinctive in NA and the least in AA patients. Eight of the n ine NA patients were male, and five of them had either trisomy 8 or trisomy 21. The sex ratio in patients with trisomy 8 was 4/0, 1/1, and 0/3 in NA, AN, and AA patients, respectively (p

0.03). No difference between the sexes was apparent in 11 patients with t(8;21) or t(15;17) (presumably primary changes}. In patients with structural changes, the sex ratio was 2/0, 3/4, and 8/6 in NA, AN, and AA patients, respectively. The fre- quency of t(8;21) plus t(15;17) was 0 (0 6f 9), 14% (3 of 21), and 30% (8 of 27) in NA, AN, and AA patients, respect iwly.

In this series three female patients with trisomy 8 and one with 7q+ and t(8;21) showed in some cells an X chromosome twisted into a spiral shape (Fig. 1). A thorough reexaminat ion of the G-banded preparations from other patients failed to reveal any similar configurations. It is very unl ikely that these twisted X chromo- somes were ring chromosomes or a result of other abnormalities, as they consis- tently appeared as right-handed spirals. All four patients showed a relatively better quality of metaphase chromosomes than the remaining patients.

To confirm the presence of this trend of chromosomal gain in males and loss in females, data on 209 cases of AML showing clonal abnormalities were collected from the literature, and consistent results were revealed. Findings in the 266 pa- tients, 57 from the present series and 209 from the literature, are summarized in Table 2. A statistical difference between the sexes was established in chromosomes #7 and #21 in addit ion to chromosome #8. Chromosomes #7, #8, and #21 are all similar in size to the sex chromosomes (p = 0.04). As this trend was more distinctive in NA patients and in patients with numeric chromosomal changes only, 28 NA patients with numeric changes only are presented diagramatically in Fig- ure 2.

Table 2. Karyotypic patterns and sex ratio (male/female) in 266 patients with acute myeloid leukemia

Karyotypic pattern NA AN AA Total

Hyperdiploid 5.7 (17/3) 1.9 (21/11) 1.4 (21/15) 2.0 (59/29) Hypodiploid 0.63 (5/8) 0.71 (15/21) 1.3 (17/13) 0.88 (37/42) Monosomy 7 0.25 (1/4) 2.3 (9/4) 4.0 (8/2) 1.8 (18/10) Trisomy 8 10 (10/1) 1.3 (8/6) 0.9 (9/10) 1.6 (27/17) Trisomy 20 (0/0) (1/0) 4.0 (4/1) 5.0 (5/1) Monosomy 20 0.3 (1/3) 2.0 (2/1) 1.5 (3/2) 1.0 (6/6) Trisomy 21 3.0 (3/1) (2/0) 2.0 (4/2) 3.3 (10/3) Monosomy 21 0.3 (1/3) 1.3 (4/3) 1.5 (3/2) 1.0 (8/8)

Abbreviations: NA, less than half abnormal cells; AN. less than half normal cells; AA, abnor- mal cells only.

78 Y.s . Li

0

0 ~b

@

c+

+i0

+5

+4

+3

+2

+i

0

-I

-2

H Male

+I

0

-i

-2

-3

-4

q L] Female

v • • • ! i 1 , i r i f ~ I i r ! 1 v v

1 1 1 1 1 1 1 1 1 1 2 2 ? i 2 3 4 5 6 7 i~ 9 o i 2 3 4 5 6 7 8 9 o I ~ x Y c

Figure 2 Distr ibution of numer ic chromosomal changes in 28 pat ients (19 male, 9 female) in whom less than half of all cells are abnormal.

DISCUSSION

General ly speaking, nonclonal changes are secondary rather than pr imary. Theoret- ically, nonclonal changes occur in single cells and are incompat ib le wi th cell pro- l iferation before harvest for s tudy of the chromosomes. They must have ar isen dur- ing the last cell cycle before harvest. Of course, the nonclonal changes ident if ied by s tudy of a l imi ted cell popula t ion might not be genuinely nonclonal in nature.

Despite cont inuous controversy regarding the mechanism of mal ignant transfor- mation, it seems l ikely that gone muta t ion or gone t ransposi t ion [18, 19] is the basis of the genesis of neoplasms and that the muta ted or t ranslocated genes are most ly dominant . Only structural aberrat ions of chromosomes can cause gone muta t ion or

Correlation of Karyotype with Sex and Age in AML 79

rearrangement of genes. The contr ibut ion of numeric changes to the deve lopment of cancer might be to increase the number of mutated or t ranslocated genes or to e l iminate normal genes, i.e., to change the balance between normal and abnormal genes in favor of mal ignant viabil i ty. As a result, numeric changes are l ikely to be secondary, and structural changes are more l ikely to be primary. Thus, the presence of numeric changes only might imply karyotypic instabi l i ty of cells.

It has been suggested that a progressive acquisi t ion of chromosomal abnormali- ties occurs during the course of acute leukemia [17]. Some chromosomal ly normal cells coexist ing with abnormal cells may be nonleukemic in some cases, especia l ly when the percentage of normal cells is small. However, the p redominance of karyo- typica l ly normal cells over abnormal ones in NA patients ahnost certainly indicates the emergence of new cell l ines and, therefore, suggests karyotypic ins tabi l i ty of cells.

The mul t ip le-s tep genesis of neoplasm seems to be the theory most favored, but it is very unl ike ly that detectable chromosomal changes are involved in all of these steps. In pat ients wi th five or more aberrant chromosomes, secondary changes must have occurred. Unlike the changes in congenital chromosomal abnormali t ies , un- identif ied marker chromosomes often occur in malignancy. The latter chromosomes may arise via different mechanisms, e.g., homogeneous ly staining regions, which are restr icted to mal ignant cells.

Str ikingly enough, all four of the parameters favoring secondary changes and karyotypic instabi l i ty were found to increase with pat ient age. This observat ion may indicate that the cell karyotype was less stable, as it showed more changes in el- der ly pat ients than in younger patients. This hypothes is is consis tent wi th the de- creased capaci ty of DNA repair [20, 21] and increased frequency of chromosomal aberrat ions [22] found in cells from elder ly people.

The significantly higher frequency in the present s tudy of nonclonal abnormali- ties in pat ients with clonal abnormal i t ies indicates the karyotypic ins tabi l i ty of karyotypica l ly abnormal cells compared with normal cells. According to Deaven et al. [23], the cell karyotypes of some cell l ines tend to stabil ize at a certain DNA content per cell, namely, mul t ip les of the haplo id genome. It is unders tandable that most chromosomal changes, except reciprocal t ranslocations, may cause distur- bance and instabi l i ty of karyotypes. None of the 11 pat ients wi th t(8;21) or t(15;17) had nonclonal abnormali t ies , a f inding significantly different from that in the re- maining pat ients (p = 0.02). This observation seems to be in agreement with the results of Deaven et al [23]. However, patients wi th t(8;21) tended to lose a sex chromosome, which is a lways clonal in nature.

The rules by which the architecture of chromosomes in nuclei is formed are little known. The frequency of d i sp lacements for each chromosome was roughly in- versely propor t ional to the size of each chromosome [24]. Bennett recent ly ad- vanced a model for depos i t ion of chromosomes in which the chromosomes are ar- ranged in the nuclei according to the length of the chromosomal arms [25]. Thus, the size of chromosomes appears to play an impor tant role in their arrangement in nuclei and may shed some light on the unusual results of this study. The sex chro- mosomes might be close to or mis taken for the autosomal chromosomes, which are s imilar in size. As the genome of males is quant i ta t ively smal ler than that of fe- males, a tentative explanat ion is that gains and losses of certain chromosomes in male and female patients, respect ively, may lead to more stable karyotypes and offer a growth advantage to cells under certain circumstances. Al though an expla- nat ion of the t rend of chromosomal gain in males and loss in females would be premature, it is clear that nei ther the convent ional theory of natural select ion against random chromosmnal aberrat ions nor the genes pos tu la ted to govern chro- mosomal behavior can expla in this trend.

8 0 Y.S. Li

The determinat ion of numeric changes in chromosomes #7, #8, and #21 by the original consti tutional karyotypes indicates that they are secondary to a leukemic state. This finding is further confirmed by the observation that the trend of chro- mosomal gain in males and loss in females, primarily involving chromosomes #7, #8, and #21, was most pronounced in NA patients. As noted previously, the chro- mosomal abnormalit ies in NA patients are almost certainly secondary in nature. The increasing frequency of patients with numeric chromosomal changes only from AA and AN to NA in the present series [Table 1) favors the view that such changes may be secondary. In contrast, the frequency of t(8;21) plus t(15;17) increased from NA and AN to AA.

The trend of gains in males and losses in females in numeric chromosomal changes is based only on the analysis of patients with AML, and thus is not neces- sarily applicable to other disorders. However, the predominance of males with the congenital chromosomal abnormalit ies of tr isomy 8 [26] and trisomy 21 [27] implies that this trend might not be restricted to AML.

Some numeric changes of chromosomes #7, #8, and #21 have been related to patient age and exposure to potentially carcinogenic/mutagenic agents [28]. The mechanism for the clustering of numeric changes in these chromosomes has not been fully explored and thus is open to further study.

The karyotypic analysis of the present 109 patients with AML was done by the author at the Department of Haematological Medicine, University of Cambridge, England, between March 1980 and September 1982. The author wishes to thank Professor F. G. J. Hayhoe and his colleagues at Cambridge for encouragement and for supplying specimens and clinical data.

REFERENCES

1. Li YS, Khalid G, Hayhoe FGJ (1983): Correlation between chromosomal pattern, cytologi- cal subtypes, response to therapy, and survival in acute myeloid leukemia. Scand I Hae- matol 30:265-277.

2. First International Workshop on Chromosomes in Leukaemia (1978): Chromosomes in acute nonlymphocytic leukaemia. Br J Haematol 39:311-316.

3. Testa JR, Mintz U, Rowley JD, Vardiman JW, Golomb HM (1979): Evolution of karyotypes in acute nonlymphocytic leukemia. Cancer Res 39:3619-3627.

4. H6gstedt B, Nilsson PG, Mitelman t ~" (1981): Micronuclei in erythropoietic bone marrow cells: relation to cytogenetic pattern and prognosis in acute non-lymphocytic leukemia. Cancer Genet Cytogenet 3:185-193.

5. Yunis II, Bloomfield CD, Ensrud K (1981): All patients with acute nonlymphocytic leuke- mia may have a chromosomal defect. N Engl J Med 305:135-139.

6. Mitelman F, Nilsson PG, Levan G, Brandt L (1976): Non-random chromosome changes in acute myeloid leukemia. Chromosome banding examination of 30 cases at diagnosis. Int J Cancer 18:31-38.

7. Oshimura M, Hayata i, Kakati S, Sandberg AA (1976): Chromosomes and causation of human cancer and leukemia. XVII. Banding studies in acute myeloblastic leukemia (AML). Cancer 38:748-761.

8. Rowley J, Potter D (1976): Chromosomal banding patterns in acute nonlymphocytic leu- kemia. Blood 47:705-721.

9. Shiraishi Y, Taguchi H, Niiya K, Shiomi F, Kikukawa K, Kubonishi S, Ohmura T, Hama- waki M, Ueda N (1982): Diagnostic and prognostic significance of chromosome abnormal- ities in marrow and mitogen response of lymphocytes of acute nonlymphocytic leukemia. Cancer Genet Cytogen~t 5:1-24.

10. Philip P, Jensen MK, Kilhnann SA, Drivsholm A, Hansen NE (1978): Chromosomal band- ing patterns in 88 cases of acute nonlymphocytic leukemia. Leuk Res 2:201-212.

11. Borgstr6m GH, Teerenhovi L. Vuopio P, de la Chapella A, Berghe HVD, Brandt L, Golomb

Cor re l a t i on of K a r y o t y p e w i t h Sex a n d Age in A M L 81

HM, Louwagie A, Mitelman F, Rowley JD, Sandberg AA (1980): Clinical implications of monosomy 7 in acute nonlymphocyt ic leukemia. Cancer Genet Cytogenet 2:115-126.

12. Hagemeijer A, Hfihlen K, Abels J (1981): Cytogenetic follow-up of patients with nonlym- phocytic leukemia. II. Acute nonlymphocyt ic leukemia. Cancer Genet Cytogenet 3: 109-124.

13. Li YS, Hayhoe FGJ (1982): Cytogenetic study in acute myeloid leukaemia using peripheral blood samples sent by post. J Clin Pathol 35:861-865.

14. The Standing Committee on Human Cytogenetic Nomenclature (1978): An international system for human cytogenetic nomenclature. Cytogenet Cell Genet 21:313-404.

15. Mitelman F, Levan G (1978): Clustering of aberrations to specific chromosomes in human neoplasms. III. Incidence and geographic distribution of chromosome aberrations in 856 cases. Hereditas 89:207-232.

16. Mitelman F, Levan G (1981): Clustering of aberrations to specific chromosomes in human neoplasms. IV. A survey of 1,871 cases. Hereditas 95:79-139.

17. Nilsson PG, Brandt L, Mitehnan F (1977): Relation between age and chromosomal aber- rations at diagnosis of adult nonlymphocytic leukemia. Leuk Res 1:385-386.

18. Cairns J (1981): The origin of human cancer. Nature 289:353-357.

19. Klein G (1981): The role of gene dosage and genetic transpositions in carcinogenesis. Na- ture 294:313-318.

20. Squires S, Johnson RT, Collins ARS (1982): Initial rates of DNA incision in UV-irradiated human cells: differences between normal, xeroderma pigmentosum and tumour cells. Mu- rat Res 95:389-404.

21. Lambert B, Ringborg V, Swanbeck G (1977): Repair of UV-induced DNA lesions in periph- eral lymphocytes from healthy subjects of various ages, individuals with Down's syn- drome and patients with actinic keratosis. Mutat Res 46:133-134.

22. Hedner K, H6gstedt B, Kolnig A-M, Mark-Vendel E, Strombeck B, Mitelman F (1982): Relationship between sister chromatid changes and structural chromosome aberrations in relation to age and sex. Hum Genet 62:305-309.

23. Deaven LL, Cram LS, Wells RS. Kraemer PM (1981): Relationships between chromosome complement and cellular DNA content in tumorigenic cell populations. In: Genes, Chro- mosomes, and Neoplasia. FE Arrighi, PN Rao, E Stubblefield, eds. Raven Press, New York.

24. Ford JH, Lester P (1982): Factors affecting the displacement of human chromosomes from the metaphase plate. Cytogenet Cell Genet 33:327-332.

25. Bennett MD (1982): Nucleatypic basis of the spatial ordering of chromosomes in eukary- otes and the implications of the order for genome evolution and phenotypic variation. In: Genome Evolution. GA Dovcr, RB Flavell, eds. Academic Press, New York, pp. 239-260.

26. Reyes PG, Hsu LYF, Strauss L, Rose J, Hirschhorn K (1978): Trisomy 8 mosaicism syn- drome: report of monozygotic twins. Clin Genet 14:90-97.

27. Zellweger H, Simpson J (1977): Chrmnosomes in Man. Spastic International Medical Pub- lications, London.

28. Rowley JD, Alimena G, Garson OM, Hagemeijer A, Mitelman F, Prigogina EL (1982): A collaborative study of the relationship of the morphological type of acute nonlymphocyt ic leukemia with patient age and karyotype. Blood 59:1013-1022.