Prognostic significance of karyotype in a twelve-year follow-up in childhood acute lymphoblastic leukemia

Prognostic Significance of Karyotype in a Twelve-Year Follow-up in Childhood Acute Lymphoblastic Leukemia Nicole Dastugue, Alain Robert, Catherine Payen, Daniele Cl6ment, Alegria Kessous, C6cile Demur, Herv6 Rubie, Henri Plaisancie, Georges Bourrouillou, and Pierre Colombies

ABSTRACT: We report a follow-up of 49 children with acute lymphoblastic leukemia (ALL) diagnosed between 1972 and 1978 (follow-up 12-18 years). This series allowed us to analyze the predictive value of karyotype in a long-term follow-up. Karyotypes were abnormal in 33 cases (67.3%): pseudodiploidy

in 11 (22.4%), hyperdiploidy ~50 chromosomes in 8 (16.3%), hyperdiploidy 47-50 chromosomes in 11 (22.4%), and hypodiploidy in 3 cases (6.1%). Event-free survival (EFS) and survival studies showed that the outcome of patients was determined only by treatment and karyotype. Eleven patients have survived, nine in first remission (6 years 5 months to 15 years 2 months), and two are in second remission (3 years 8 months and 8 years 2 months). All ploidy groups are represented in these patients. Late relapses can occur in the hyperdiploid ~50 group, thus accounting for shorter EFS than expected, but because of the unusually long second remission of one patient, the rate of surviving patients was higher for this ploidy group than for all other ploidy groups together. Conversely, patients with only numerical abnormalities (no matter which ploidy group they belonged to), had a better outcome than did patients with structural changes or normal karyotypes and no discrepancy between EFS and survival curves was observed in this chromosomal group. Thus, our results suggest that numerical changes only should be considered an indicator of low risk factor, but our results, based on partially banded karyotypes, need to be verified by a current method and therapy.

INTRODUCTION

MATERIALS AND METHODS

Prognosis in childhood ALL is significantly influenced by the initial hematologic data and also by the karyotypes of blast cells. Ploidy has been recognized as a prognostic factor [1, 2], and hyperdiploidy (~50 chromosomes) is associated with the most favorable prognosis [3, 4]. Structural abnormalities [3, 5], particularly translocations [6], are strong prognostic factors because they might provide additional means for identifying the high-risk treatment failure group. Few reports on long-term follow-up have yet been published [7, 8]. We report 49 children with follow-up of 12-18 years, including clinical, hematologic, and chromosomal data, and describe their prognostic influence.

Patients We report 49 children with ALL aged ~15 years, diagnosed between 1972 and 1978 (follow-up 12-18 years) and referred to the Department of Pediatric Hematology, Unit6 d'H6matologie Infantile, Professeur R6gnier, CHU H6pital Purpan, Toulouse, France. Nine of the patients are still in first complete remission, and two others are still in prolonged second remission (Table 1). Only patients who had undergone successful cytogenetic analysis were included in this study. Diagnosis of ALL was based on blast cell morphology and on negativity of myeloperoxidase staining, since immunologic markers were available only after the end of this study period. Morphology was reassessed from the initial slides to classify the patients according to the French-American-British (FAB) criteria. Tumoral syndrome was considered to exist when at least one of the following criteria was observed: adenopathies with a diameter ~-3 cm or splenomegaly or hepatomegaly 8 cm beyond the costal margin or mediastinal mass. As therapy administered during this 7-year period evolved, two periods could be considered schematically: before 1974 (treatment A) with mild therapies, and after

From the Laboratoire Central d'H6matologie et de G~n~tique (N. D., C. D., H. P., P. C.), Unit6 d'H~matologie Infantile, Pediatrie B (A. R., H. R.), and Laboratoire d'Informatique M~dicale (C. P.), CHU Purpan, Toulouse, France (A.K., D.C.), Inserm UIO0, Chu Purpan, Toulouse, France. Address reprint requests to: Dr. Nicole Dastugue, Laboratoire Central d'H6matologie et de G~n~tique, Centre R~gional de Transfusion Sanguine, C.H.U. Purpan, 31052, Toulouse C6dex France. Received December 12, 1991; accepted June 29, 1992.

49 © 1992 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010

Cancer Genet Cytogenet 64:49-55 (1992) 0165-4608/92/$05.00

50

Table 1

Karyotypes and hematologic

P a t i e n t / s e x / a g e (yr) Normal karyotypes 1/M/6.3 2/M/2.8 3/M/3.7 4/M/3.2 5/M/0.5 6/M/2.5 7/M/3 8/M/12.6

9/F/4.5 10/F/2.2 11/F/8 12/F/12.5 13/F/12.8 14/F/11 15/F/6 16/F/1.5 Pseudodiploidy 17/M/11.5 18/F/2 19/M/10.5 20/F/4.3 21/M/2.4 22/F/9.6 c 23/M/9.3 24/M/8.5 25/M/11.5 26/M/2.5 27/M/9.5 Hyperdiploidy >50 chromosomes 28/M/2 29/M/1.7 30/M/3.8 31/F/3.3 32/F/4.7 33/F/3 34/M/2.7 35/F/2.8 ~ Hyperdiploidy 47-50 chromosomes 36/F/6 37/F/4.5 38/F/12.1 39/M/3 40/F/4.7 41/F/0 .5 42/F/3 .2 43/M/14 44/M/12 45/F/3.6 46/M/6.7 Hypodiploid karyotypes 47/F/4.4 48/M/5 49/M/10.6

d a t a of 4 9 p a t i e n t s WBC ( x 109/L)

EFS (too]

Karyotype

1.8 14.9 48 36 97 2.1 3.5 850 7.6 17 18.7 31.6 58.3 20O 55 127

8 6 2O 6 21 21 22 11 10 21 2 13 17 6 78 b 24

46,XY 46,XY a 46,XY 46,XY 46,XY Q 46,XY ° 46,XY ° 46,XY 46,XX ° 46,XX 46,XX ~ 46,XX ° 46,XX 46,XX ° 46,XX o 46,XX,t(4;8)(q32;q22)c

4.5 7.1 170 11 94 9.9 94 135 75 32.4 12.4

7 0 7 49 7 46 16 38 15 174 b 5

46,XY, + C, - G 46,XX,-D,-F,+ G x 2 46,XY, - B, - C, + 2 m a r 46,XX,del(6)(q13q23) 46,XY,del(6)(q13q23),der(16)t(16;?)(p13;?) 46,XX,del(6)(q13q23),der(10)t(10;?)(q24;?) 46,XY,i(10q),t(11;14)(p13;q11) 46,XY,t(7;13)(q11;q34) 46,XY,der(4)t(4;?)(p16;?),der(5)t(5;?)(q35;?) 46,XY,del(12)(q14),der(12)t(12;?)(p12;?),der(13) 47,XYY~, - C, + m a r

8.3 74 34 14.8 2.6 4.7 11.7 2.8

7 40 145 ~ 165 h 14 135 b 14 96

52,XY,+B,+C x 2,+D,+E,+G 55,XY,+A,+C x 4,÷D,+E,+G x 2 5 5 , X Y , + B , + C x 3, + D , + E , + G x 2 , + m a r 5 5 , X X , + 4 , + 6 , + 7, + 14 x 2 , + 1 8 , + 2 0 , + 21 x 2 55,XX, + X, + 4, + 6, + 8, + 10, + 14, + 21 x 2 , + i ( 1 7 q ) 5 6 , X X , + B , + C x 4, + D , + E x 2 , + G x 2 57,XX,+4,+5,+6,+8,+10,+14,+17,+18,+21×2,+22 59,XX,+A,+B x 2,+C x 4,+D,+E x 2,+F,+G x 2

5.4 42 166 2.1 12 18 7.5 30 9.1 9.8

183 b 156 b 18 11 144 b 0 17 5 0 13 4

47,XX, + D 47,XX, + 21 48,XX,+ 8 , ÷ 21 48,XY, + 14, + 21c 4 8 , X X , + X , + 21c 47,XX,de|(12)(p12) + 19 48,XX, + 10,del(12)(p12) ÷ 18 47,XY,t(2;3)(q31;?p25) + 10,del(6)(q13q23) 47,XY,t(11;14)(p13;q11), + m a r 48,XX, + 11,i(17q], + 21 5 0 , X Y , - A , + B x 2 , ÷ C x 3 , - F , - G,÷ 2mar

120 22.9 3.3

7 0 184 ~'

45,XY, - G 45,X,-Y

27,X,+ X , + 1 0 , + 1 8 , + 21/54,XX,+ X x 2 , + 1 0 x 2 , + 1 8 x 2 , +

Abbreviations: EFS, event-free survival. ° Banded karyotype. Still in first remission. Patients 22 and 35 have been in second complete remission for 98 and 74 months, respectively.

21x2

Karyotype and Follow-up in Childhood ALL 1974 (treatment B) with more aggressive therapies. In treatment A, induction chemotherapy was based on daily administration of prednisone and weekly administration of vincristine, whereas the combination 6-mercaptopurinemethotrexate was used as maintenance therapy. Central nervous system (CNS) prophylaxis with radiation and intrathecal methotrexate was used in most patients. Treatment B corresponded to the 0.8 LA 74 protocol (HSpita] St. Louis, Paris). Induction therapy included prednisone, vincristine, daunorubicin, and cyclophosphamide. Asparaginase was administered as consolidation treatment for 10 days. CNS prophylaxis was used systematically with intrathecal methotrexate combined with CNS radiation. Maintenance therapy included methotrexate with 6-mercaptopurine. Reinductions were administered monthly for 6 months, then every 3 months for 1 year, and then every 6 months for 3-5 years. Cytogenetic Studies Karyotypes were established at diagnosis in all patients. Cytogenetic analysis was performed in each case from bone marrow (BM) by a direct technique. BM samples were directly placed in isotonic medium already containing colchicine (0.5/~g/ml) and harvested after 30-minute incubation at 37°C in hypotonic treatment (KC1 0.075 M) and fixation with acetic acid : ethanol (3 : 1). Cells were spread on wet slides, and RHG banding was attempted in each case. A minimum of five mitoses was required if a patient was to be included in the study. When banding failed, chromosomes were classified according to the Denver nomenclature and a clone was defined by the presence of identical markers in at least two mitoses and, in aneuploidy, by the same pattern of chromosome groups in at least three mitoses. When banding was obtained, chromosomes were reclassified according to the International System for Human Cytogenetic Nomenclature [9]. Each karyotype received a twofold classification. The first classification corresponded to the standard ploidy groups: normal, pseudodiploidy, hyperdiploidy >50 chromosomes (hyperdiploid >50), hyperdiploidy 47-50 chromosomes (hyperdiploid 47-50), and hypodiploidy. The second classification was based on the presence of numerical and/or structural abnormalities. A patient was classified as having only numerical abnormalities when the chromosome number was ~46 and when no structural abnormality was detected. A patient was classified as having structural abnormalities when a structural abnormality (marker, translocation, deletion, isochromosome, or derivative) was found alone or was associated with a numerical change. Constitutional karyotypes were established in each case from peripheral blood cultured for 72 hours and stimulated by phytohemagglutinin. Statistical studies The groups were compared by chi-square tests and analysis of variance (ANOVA). Prognostic influence of factors was determined by log-rank test and by a Cox regression model (univariate and multivariate analyses). EFS corresponded to the first complete remission duration, and curves were constructed by the Kaplan-Meier procedure.

51 RESULTS Complete results are shown in Table 1. A banded karyotype was obtained in 32 patients. Distribution according to ploidy groups was normal karyotype in 16 cases (32.6%), pseudodiploidy in 11 cases (22.4%), hyperdiploidy >50 in eight cases (16.3%), hyperdiploidy 47-50 in 11 cases (22.4%), and hypodiploidy in three cases (6.1%). Numerical versus structural classification showed 14 cases (28.5%) with only numerical abnormalities (six hyperdiploid >50 cases, five hyperdiploid 47-50 cases, three hypodiploid cases) and 19 cases (38.7%) with structural abnormalities (11 pseudodiploid cases, two hyperdiploid >50 cases, and six hyperdiploid 47-50 cases). The modal numbers showed a standard bimodat distribution, with one peak occurring at 46 and a second peak occurring at 55 chromosomes. In the hyperdiploid >50 group, the most frequent extra chromosomes were chromosomes 4, 6, 14, and 21. These gains corresponded to the pattern + B, + C, + D, + G without banding. The hyperdiploid 47-50 group differed from the hyperdiploid >50 group in having greater variability of added chromosomes, except for chromosome 21 which was observed in 3 of 11 cases. In the hypodiploid group (three cases), one case was a near-haploid karyotype (27,X,÷X,+C,+18,+21) with a sideline of 54 chromosomes, an exact duplication of the nearhaploid stemline (previously reported [10]). A translocation was identified in eight banded karyotypes. In case 19, a t(4;11) was evoked by the morphology of blast cells and the cytogenetic aspect was consistent with this change: loss of one chromosome in B and C groups and the presence of markers consistent with 11q+ and 4 q - . Four children (6.8%) had a constitutional abnormality (two Down syndromes, one 47,XYY karyotype, and one balanced t(4;8)(q32;q21). Three of the children acquired extra changes. The 47,XYY case showed structural aberrations, whereas the two children with Down syndromes acquired new trisomies (14 in one, and X in the other). Correlations of clinical and hematologic data with chromosome groups Age, sex, tumoral syndrome, leukocyte count, blast cell percentage, and FAB classification were compared according to ploidy groups (normal, pseudodiploid, hyperdiploid >50, hyperdiploid 47-50, hypodiploid) (Table 2). No significant relationships were noted, but some trends can be described. The hyperdiploid >50 group tended to have a younger median age and to have fewer tumoral syndromes and lower leukocyte counts, whereas the pseudodiploid group showed higher male percentages, the highest median age, and more frequent tumoral syndromes. EFS and survival rate When EFS and survival were evaluated by actuarial methods, the only factors showing a significant predictive value were treatment and karyotype. Only 5% of the children treated before 1974 (treatment A) as opposed to 32% of the children treated between 1974 and 1978 (treatment B), still survive (p < 0.01).

52

N. Dastugue et al.

Table 2

Clinical and hematologic data according to chromosome groups

No. of patients Sex (no. of males) Age (yr) Median Range Mean Tumoral syndrome (no. of patients) WBC count ( x 109/L) Median Range Mean FAB L1 L2 L3 Complete remission (no. of patients)

Total sample (49)

Normal (16)

Pseudodiploid (11)

Hyperdiploid >50 (8)

Hyperdiploid 47-50 (11)

Hypodiploid (3)

26

8

8

4

4

2

4.5 0.5-14 5.1

3.7 0.5-12.8 5.8

8.9 2-11.5 7.4

3 1.7-4.7 3.1

4.6 0.5-14 6.4

5 4.4-10.6 6.7

28

9

9

2

7

1

17 1-850 57.2

31 1.8-850 97.8

22 4.5-170 58.4

8 2.6-74 18.6

9 2.1-166 27.4

22.9 3.3-120 48.3

36 11 1

13 3 0

7 2 1

6 2 0

8 3 0

2 1 0

45

16

10

11

9

2

Abbreviations. WBC,white blood cell; FAB. French-American-Britishclassification.

W h e n EFS was studied according to ploidy groups (normal, pseudodiploid, hyperdiploid >50, hyperdiploid 47-50, hypodiploid), no significant difference was noted. Even w h e n t h e hyperdiploid >50 group was compared directly with the four other groups together (Fig. 1), its outcome was no better at follow-up of 18 years. The curve clearly showed that late relapses can occur in the hyperdiploid >50 group: one patient relapsed after 3 years 3 months and another after 8 years. Both patients relapsed in ALL. The patient who relapsed after 8 years has been in second complete remission for 6 years, indicating that the survival of patients with hyperdiploidy >50 was better than their EFS (p < 0.05) (Fig. 2). Children with normal karyotypes and those with structural abnormalities who

shared a similar outcome (overlapping of actuarial curves) were combined and compared with the group with only numerical changes. Outcome was significantly better for the group with only numerical changes, since 46% of patients are still in first complete remission at 18-year followup (Fig. 3) and 50% of the children in this category survive (six in first remission and one in second remission). In surviving patients with only numerical changes, not only three hyperdiploid >50 (cases 31, 33, and 35) but also three hyperdiploid 4 7 - 5 0 cases (cases 36, 37, and 40), and one hypodiploid (loss of Y chromosome) case (case 49) were included. The chromosome most often involved in numerical changes was chromosome 21. A n added chromosome 21 was formally identified in eight b a n d e d karyotypes, and

Figure 1 Event-free survival (EFS) of hyperploidy >50 chromosomes as compared with EFS of all other ploidy groups. Owing to late relapses, the final outcome of the hyperdiploid > 50 group was not significantly better at follow-up of 18-years. EFS lOO

(%)

75

I I i I i

NS

,I

SO. .

.

.

.

.

.

.

.

.

.

i

. . . . . . . . . .

LI.

. . . .

I

Hyperploidg >50

(n=8)

2S. I

IIII

, Other ( n = 3 7 )

Q '10

'15

~0

Years

Karyotype and F o l l o w - u p in C h i l d h o o d ALL

53

Survival (,~)

loo ...... i

i 7550.

,.,1

. ,

p
-JJ

,

Hyperploidy>50(n=8) 25. ' Other (n=41)

IIII

o

'o

~

'Io

'is

'20

Years

Figure 2 Survival of patients with hyperdiploidy >50 chromosomes as compared with survival of other ploidy groups studied together. Patients with hyperdiploidy >50 chromosomes had longer survivals than patients in other ploidy groups. The survival rate of the hyperdiploid >50 group is higher than event-free survival shown in Figure 1 because one patient (patient 35) has been in second complete remission for 8 years. Her survival is 14 years.

count, tumoral syndrome, FAB classification), proved to be of prognostic interest after 18-year follow-up.

an a d d e d G was noted in five n o n b a n d e d h y p e r d i p l o i d >50 cases. Because of the close (if not absolute) association of trisomy 21 w i t h the h y p e r d i p l o i d > 5 0 group already reported [6, 11, 12], we can assume that + G was related to trisomy 21 in all five cases. EFS was not significantly longer for patients with an acquired extra chromosome 21 at diagnosis, but survival was slightly longer (p < 0.05). Finally, cured patients were observed in all p l o i d y groups: one patient in the normal karyotype group, two patients in the p s e u d o d i p l o i d group, four patients in the h y p e r d i p l o i d > 50 group, three patients in the h y p e r d i p l o i d 4 7 - 5 0 group, and one patient in the h y p o d i p l o i d group. No other standard prognostic factor tested (age, leukocyte

Relative Importance of Different Prognostic Factors Cox regression analysis allows testing of each factor either i n d i v i d u a l l y or by taking other factors into account. In univariate analysis of survival, the variables significantly improving prognosis were treatment B, the presence of only numerical changes, and high m o d a l numbers. Significance was stronger for treatment and only n u m e r i c a l changes (p < 0.01) than for m o d a l numbers (p < 0.05). To analyze their prognostic impact further, the factors of treatment and only numerical changes were tested w i t h multivariate

Figure 3 Event-free survival of patients with only numerical abnormalities as compared with normal karyotypes combined with structural abnormalities (alone or associated with numeric abnormalities). The proportion of patients with numerical abnormalities was higher at an 18-year follow-up (p < 0.05). EFS (~)

IOO 75

,

p
|

r i

", !

. - -

50. .........

J . ~ . . . ~ - L - J Nurnerlc enly (n=l 3)

25. / Q

'

'I0

, ,

Other (n=32)

'15

~0

Years

54 analysis. These two factors maintained the same predictive value in this test (p ~ 0.001). Thus, in our study, the strongest cytogenetic factor in favor of good outcome was the presence of only numerical abnormalities with no structural changes. Because the number of patients was too few, the significance of this factor could not be tested for each treatment. DISCUSSION

The follow-up of childhood ALL we describe (range 12-18 years), is to our knowledge the longest ever reported in the literature. Cytogenetic results are in agreement with those of other series studied during the same period. Clonal abnormalities detected (67.3%) and distribution of ploidy groups (pseudodiploid, 22.4%; hyperdiploid ~50, 16.8%; hyperdiploid 47-50, 22.4%; hypodiploid, 6.1%) are in keeping with those reported by Secker-Walker [4, 8] and at the Third International Workshop on Chromosomes in Leukaemia [3] except for the hyperdiploid 47-50 group, which was slightly overrepresented in our series. Structural abnormalities include nonrandom changes such as deletions of the long arm of chromosome 6 and of the short arm of chromosome 12 and translocations such as t(11;14)(p13;q11). One karyotype was consistent with a t(4;11), and clinical data were in agreement with this abnormality. Thus, we can assume that other nonrandom translocations were not formally detected but might have been present in nonbanded karyotypes and classified as structural changes. Each ploidy group shows the hematologic features already reported in the literature, i.e., a concentration of adverse prognostic factors in patients with pseudodiploidy and an opposite trend in those with hyperdiploidy ~50 chromosomes. Therapies used in this series achieved complete remission in most patients regardless of the ploidy group to which they belonged but did not prove effective in obtaining long remissions as compared with advanced therapies. The role of treatment in the final outcome is clearly demonstrated by the increasing rates of patients who ultimately recovered. Only 5% of the patients treated before 1974 (treatment A) survived, as compared with 32% of patients treated with the therapy protocol used between 1974 and 1978 (treatment B) (p ~ 0.01). This accounts for the lower incidence of cured patients in our series (21%) as compared with all the other series that included patients after 1978. With therapies used since 1980, 50% of patients with childhood ALL recover [13]. It is now claimed that with present progress in treatment, most children will recover [14, 15]. The only predictive factor in our series other than treatment is karyotype. In our 18-year follow-up, the hyperdiploid ~50 group, known as the group with the highest likelihood for recovery [7, 8], is characterized by relatively longterm first remissions (48.5% are disease-free at 3-year follow-up--a high percentage given the low effectiveness of the therapy used), and by risk of late relapses, occurring at 3 years 3 months and 8 years after diagnosis. Such an occurrence of late relapses has never been reported in hyperdiploidy ~50 chromosomes. In the long follow-up peri~

N. Dastugue et al. ods described by Secker-Walker [8] and Pui [16] and at the Sixth International Workshop on Chromosomes in Leukemia [7], 75% of the patients in the hyperdiploid ~50 group were disease-free after 6 years. However, since treatment has improved markedly in the past 10 years, late relapses may not occur even with longer follow-up periods. Only long follow-up of ALL treated with current therapies will show whether such therapies can prevent late emergence of residual blast cells. In the group with only numerical changes, unlike in the hyperdiploid ~50 group, 50% of patients survived after 12year follow-up owing to the presence of three hyperdiploid 47-50 karyotypes and of one hypodiploid karyotype (loss of Y chromosome) besides three hyperdiploid ~50 karyotypes in the group with only numerical changes that survived. These findings suggest that the good prognostic group might include not only the hyperdiploid ~50 karyotypes but also the hyperdiploid 47-50 karyotypes without structural changes. This assumption is consistent with the results of the Sixth Interntional Workshop on Chromosomes in Leukemia [7]. In this follow-up of patients registered in 1980, the hyperdiploid 47-50 group shows intermediate prognosis (50% alive at 6-year follow-up). This relatively good outcome may result from the high proportion of numeric-only abnormalities (12 of 28 cases). An added chromosome 21 was the most frequent change. Although EFS was not significantly longer for patients with an acquired trisomy 21, a high proportion of patients who finally recovered (5 of the 11 cured patients) had an acquired trisomy 21 at diagnosis. This chromosome gain may be a determining factor in outcome. A suppressive activity, able to control tumorigenicity, has been evidenced for several chromosomes and suggested for chromosome 21 [17]. Thus, the acquired extra chromosome 21 in ALL may be a favorable factor in outcome. Our long-term follow-up contributes new data on the significance of karyotype in childhood ALL: risk of late relapses in patients with hyperdiploidy ~ 50 chromosomes, a suggestion that hyperdiploidy ~46 chromosomes without structural changes be included in the good prognosis group, and the possibility of a favorable role of chromosome 21. Because our results are based on partially banded karyotypes and on nonoptimum therapies of the past, however, their relevance should be reassessed in fully analyzed karyotypes and in light of present therapies. The authors thank Professor Roland Berger for critical reading of the manuscript, Mariana Titorov for editorial assistance, and Laurence Pinon for typing the manuscript. This study was supported by grants from "la ligue contre le cancer" and "la Fondation de France." REFERENCES

1. Secker-WalkerLM, Lawler SD, Hardisty RM (1978): Prognostic implications of chromosomal findings in acute lymphoblastic leukaemia at diagnosis. Br Med J 2:1529-1530. 2. Secker-Walker LM, Swansbury GJ, Hardisty RM, Sallan SE, Garson OM, Sakurai M, Lawler SD (1981): Cytogenetics of acute lymphoblastic leukaemia in children as a factor in the prediction of long-term survival. Br J Haematol 52:389-399.

Karyotype and Follow-up in Childhood ALL

3. Third International Workshop on Chromosomes in Leukemia, 1980 (1981). Cancer Genet Cytogenet 4:96-137. 4. Williams DL, Tsiatis A, Brodeur GM, Look AT, Melvin SL, Bowman WP, Kalwinsky DK, Rivera G, Dahl GV (1982):Prognostic importance of chromosome number in 136 untreated children with acute lymphoblastic leukemia. Blood 60:864-871. 5. Bloomfield CD, Goldman AL, Alimena G, Berger R, Borgstr6m GH, Brandt L, Catovsky D, de la Chapelle A, Dewald GW, Garson OM, Garwicz S, Golomb HM, Hossfeld DK, Lawler SD, Mitelman F, Nilsson P, Pierre RV, Philip P, Prigogina E, Rowley JD, Sakurai M, Sandberg AA, Secker-Walker LM, Tricot G, Van Den Berghe H, Van Orshoven A, Vuopio P, WhangPeng J (1986): Chromosomal abnormalities identify high-risk and low-risk patients with acute lymphoblastic leukemia. Blood 67:415-420. 6. Williams DL, Harber J, Murphy SB, Look AT, Kalwinsky DK, Rivera G, Melvin SL, Stass S, Dahl GV (1986): Chromosomal translocations play a unique role in influencing prognosis in childhood acute lymphoblastic leukemia. Blood 68:205-212. 7. Bloomfield CD, Secker-Walker LM, Goldman AI, Van Den Berghe H, De La Chapelle A, Ruutu T, Alimena G, Garson OM, Golomb HM, Rowley JD, Kaneko Y, Whang-Peng E, Prigogina E, Philip P, Sandberg AA, Lawler SD, Mitelman F (1989): Sixyear follow-up of the clinical significanceof karyotype in acute lymphoblastic leukemia. Cancer Genet Cytogenet 40:171-185. 8. Secker-Walker LM, Chessels JM, Stewart EL, Swansbury GJ, Richards S, Lawler SD (1989): Chromosomes and other prognostic factors in acute lymphoblastic leukaemia: A long-term follow-up. Br J Haematol 72:336-342. 9. ISCN (1985): An International System for Human Cytogenetic Nomenclature, Harnden DG, Klinger HP (eds.); published in

55

collaboration with Cytogenet Cell Genet (Karger, Basel, 1985); also in Birth Defects: Original Article Series, Vol. 21, No. 1 (March of Dimes Birth Defects Foundation, New York, 1985). 10. Kessous A, Corberand J, Grozdea J, Colombi~s P (1975): Clone cellulaire /~ 27 chromosomes dons une leuc~mie aigu~ humaine. Nouv Rev Fr Hematol 15:73-81. 11. Prigogina EL, Puchkova GP, Mayakova SA (1988): Nonrandom chromosomal abnormalities in acute lymphoblastic leukemia of childhood. Cancer Genet Cytogenet 32:183-203. 12. Smets LA, Slater RM, Behrendt H, Van'T Veer MB, HomanBlok J (1985): Phenotypic and karyotypic properties of hyperdiploid acute lymphoblastic leukaemia of childhood. Br J Haematol 61:113-123. 13. Sallan SE, Weinstein HJ, Nathan DG (1981): The childhood leukemias. J Pediatr 99:676-688. 14. Champlin R, Gale RP (1989): Acute lymphoblastic leukemia: Recent advances in biology and therapy. Blood 73:2051-2066. 15. Rivera GK, Raimondi SC, Hancock ML, Behm FG, Pui CH, Abronmovitch M, Mirro Jr J, Ochs JS, Look AT, Williams DL, Murphy SB, Dahl GV, Kalwinsky DK, Evans WE, Kun LE, Simone JV, Crist WM (1991): Improved outcome in childhood acute lymphoblastic leukaemia with reinforced early treatment and rotational combination chemotherapy. Lancet 337:61-66. 16. Pui CH, Raimondi SC, Dodge RK, Rivera GK, Fuchs LAFI, Abromowitch M, Look AT, Furman WL, Crist WM, Williams DL (1989): Prognostic importance of structural chromosomal abnormalities in children with hyperdiploid (>50 chromosomes) acute lymphoblastic leukemia. Blood 73:1963-1967. 17. Saxon PJ, Srivatson ES, Stanbridge EJ(1986): Introduction of human chromosome 11 via microcell transfer controls tumorigenic expression of Hela cells. EMBO J 5:3461-3466.