Influence of M-BCR Breakpoint Sites on the Duration of

Chronic Phase in 100 Patients with Chronic

Myelocytic Leukemia

Kimio Tanaka, Tetsuo Hashimoto, Nobuo Oguma, Hiroo Dohy, and

Nanao Kamada

ABSTRACT: Rearrangements of the bcr (M-BCR) gene were studied in 100 patients with chronic myelo-
cytic leukemia (CML). To determine the significance of a chimeric gene expression in the progression of
CML, we analyzed 43 patients for bcr-ABL chimeric m R N A expression. Both DNA and RNA analyses re-
vealed a possible influence of breakpoint sites in the bcr region an the duration of the chronic phase.
Patients with the breakpoint located at about the 1-kb region between BamHI and HindlII in bcr exon 3
(region C2) had a significantly shorter chronic phase (31 months) (p : 0.028) than patients in whom the
breakpoint was located in other regions. When the bcr locus was divided into 5" and 3' regions as for the
BamHI cleavage site located near the 5' region ofbcr exon 3, the chronic phase duration in patients with
the 5' site (HindlII-BamHI) and 3' site (BamHI-EcoBI site) was 75 and 38 months, respectively. However,
the difference was nat statistically significant (p = 0.128). These results suggest that only the breakpoint
site at C2 on the bcr locus, rather than breakpoint sites in other regions, has an important role in the
progression of CML.

INTRODUCTION
Chronic myelocytic leukemia (CML) is a clonal disease with
a specific chromosome translocation between chromosomes
9 and 22. Breakpoint on chromosome 22 is confined to a very
limited region of 5.8 kb, known as the bcr (breakpoint clus-
ter region) or M-BCB (major breakpoint clustering region)
[1, 2]. The relationship between 9;22 chromosomal translo-
cation and chimeric bcr-ABL gene in CML has been well stud-
ied. Chimeric DNA consisting of the 5' bcr and 3' c-ABL on-
cogene expresses 8.5-kb mRNA and p210 c-ABL protein
instead of 6- or 7-kb mRNA and p145 normal protein [3-5].
The bcr-ABL chimeric gene is known to have an important
role in the pathogenesis of CML. Although several recent
studies have reported the relationship between breakpoint
site and duration of the chronic phase of this disease [6-21],
the results are still contradictory. The present study focused
on the influence of the bcr in 100 patients with CML and
From the Department of Hematology (K. T., N. 0., N. K.), Re-
search Institute for Nuclear Medicine & Biology, Hiroshima Univer-
sity; Hiroshima Bed Cross Hospital &Atomic-Bomb Survivors Hos-
pital (H. D.), Hiroshima; and The Institute of Statistical Mathematics
(T. H.), Tokyo, Japan.
Address reprint requests to: Nanao Kamada, M.D., Department
of Hematology, Research Institute for Nuclear Medicine & Biology,
Hiroshima University, Kasumi 1-2-3, Minami-ku, Hiroshima 734 Ja-
pan. Fax No. 082-256-07103.
Received April 23, 1993; accepted June Z 1993.
on the effects of expression of the bcr-ABL gene on the dura-
tion of chronic phase. These latter effects were analyzed in
43 of these patients.
PATIENTS AND METHODS
Patients
One hundred randomly selected CML patients, who had
visited Hiroshima University Hospital and Hiroshima Red
Cross Hospital over the last 13 years, were examined cytoge-
netically. All the patients had the typical clinical and hemato-
logic features of CML with Philadelphia (Ph) chromosome.
Samples were collected from 41 patients in the chronic phase,
40 patients in the blastic phase, and 19 patients in both
chronic and blastic phases. Two or more major parameters
from among the following were used for the diagnosis of blas-
tic phase: the percentages of myeloblasts plus promyelocytes
in peripheral blood and bone marrow were more than 30%
and 50%, respectively, there was a rapid increase in lymph
node adenopathy, in tumor formation, and in the clonal ap-
pearance of additional chromosome aberrations with Ph
chromosome. All patients received hydroxyurea and/or
busulphan as initial therapy. Seven patients were treated with
interferon-a as secondary therapy. Patients who had been
treated by bone marrow transplantation were excluded from
this study.

39
© 1993 Elsevier Science Publishing Co., Inc. Cancer Genet Cytogenet 70:39-47 (1993)
655 Avenue of the Americas, New York, NY 10010 0165-4608/93/$06.00
40 K. Tanaka et al.

Chromosome Analysis method, using reverse transcriptase (RT-PCR), as reported

Chromosomal analyses of the bone marrow ceils were per-
formed with the same samples that were used for the DNA
study Chromosomes were analyzed after 24-hour culture and
banded with trypsin-Giemsa staining [22]. The karyotypes
were determined according to the ISCN (1985) [23].
DNA Analysis
DNA from the leukemic ceils of stored and fresh bone mar-
row/peripheral blood samples from 100 patients was extracted
by the phenol-chloroform method. The DNA was digested
with four to eight restriction enzymes, according to the
manufacturer's recommendations. Ten micrograms of DNA
fragments were separated by 0.8% agarose gel electrophore-
sis and transferred to nylon membrane filters by the South-
ern method [24]. Hybridization was performed in 50% for-
mamide, 20 mM sodium phosphate, 5 x SSCP, sonicated
salmon sperm DNA, 0.1% BI_DTT (skim milk), 5% SDS, and
with probes radiolabeled by nick-translation. The probes
used were a 1.2-kb fragment of the bcr gene treated with
HindIII-BglII fragment as 3' bcr probe (Oncogene Science Inc.)
[1]; a O.6-kb HindlII-BamlI fragment; a 2-kb BgllI-HindDI frag-
ment; and a 3-kb Hindlll-BglII fragment as 5' bcr probes (these
5' bcr probes were obtained from Dr. S. Hirosawa of Tokyo
Medical & Dental University) [25]. The restriction enzyme
map of the bcr gene and the location of the probes are shown
in Fig. 1. Three probes locating m-BCR (minor breakpoint
clustering region), where the breakpoint region in Phi-posi-
tive acute lymphocytic leukemia is usually found, were also
used to identify the breakpoint in several CML patients [25].
previously [27]. Four microliters of the amplified cDNA prod-
uct was separated on gel electrophoresis with 1% Seakem
(FMC) and 4% NuSieve (FMC) in I × TBE buffer. The filters
were hybridized with two radiolabeled oligonucleotide
probes locating the junction regions of bcr exon 2-ABL exon
II, bcr-exon 3-ABL exon II, and BCB exon 1-ABL exon IT(which
is usually found in Ph positive acute lymphocytic leukemia)
by the 5' end labeling method. The hybridization was per-
formed in mixtures containing 6 × SSC + 0.1% SDS at 60°C.
The synthesis of primers and probes was based on the pub-
lished sequences [28, 29]. RT-PCR was also performed, using
c-ABL RNA as a positive control, to detect the expression
of the c-ABL oncogene.
Statistical Analysis
Eighty-eight patients were subjected to statistical analysis.
Their histories were retrospectively reviewed, noting clini-
cal features, including age, sex, hematologic profile at diag-
nosis, and the duration of chronic phase. Boxplot distribu-
tions were used for the graphic representation [30]. Estimates
of the duration in chronic phase for different breakpoint
regions were obtained using the Kaplan-Meier product limit
method. Mantel Cox's test was used for detecting significant
differences among patient groups. Multivariate regression
methods, using Cox's proportional hazard model, were also
applied to assess the relative prognostic value of each pa-
tient's characteristics [31, 32]. The program package BMDP
was used for these analyses [33].

RNA Analysis RESULTS

Total cellular RNA was extracted from the fresh bone mar-
row cells of 43 of the 100 patients by the guanidium isothio- Ph Translocation and bcr Rearrangement
cyanate method, using CsC1 ultracentrifugation [26]. RNA Of the 100 patients, 93 had the standard type of Ph chromo-
expression was examined by a polymerase chain reaction some and seven had complex types, i.e., t(7;9;22)(p13;q34;

Figure 1 Restriction enzyme map of the bcr (M-BCR) gene and location of probes. Regions in and around the bcr
were divided into eight segments: A, B, C1, C2, D, El, E2, and F. Bg: BglII, B: BamHl, H: HindIII, E: EcoRI, K: KpnI,
S: SacI, X: XbaI, Bc: BclI.

M-BCR
(bcr)

C E
A B CII C2 I DE-'FE2 F
reoion ..................... I
I
I
! I
I
I I
!
. . . . . . .

I
! !
I I
,.~

,
I I I I t
t I
I I I I I I
i I i, i I t

5' 1
T T Bi ' l l
I I
1
I 3,
H Bg Bg H H Bg H

0.1kbHB 1.2kbHBg

~\\\\\\\\\\\\\\\\\\~1
S.Skb H Bg 2kbH BO
M-BCR Breakpoint Site in CML 41

q11), t(9;22;10)(q34;q11;p15), t(9;22;22)(q34;q11;q13), t(4;9;22)
(q13;q34;q11), t(9;22;14)(q34;q11;q21), t(9;9;22)(p15;q34;q11),
and also a masked Ph chromosome presenting der(9)t(9;22)
(q34;q11). The DNA analyses revealed gene rearrangements
of the M-BCR locus in 98 of the 100 patients. Two patients
with Ph chromosome did not show rearrangements within
the bcr (M-BCR).
Distribution of Breakpoints Within the bcr Locus
The classical 5.8-kb region of the bcr locus (M-BCR) was
divided into eight segments which were designated A, B,
C1, C2, D, El, E2, and F according to the cleavage sites of
the restriction enzyme, as shown in Fig. 1. The distribution
of the breakpoints in the 100 patients is shown in Fig. 2.
Ninety-one patients had breakpoints in the C1, C2, and D
regions; one patient had a breakpoint in the F region; and
no patients showed breakpoints in the B or E2 regions. Two
patients had breakpoints outside the bcr (M-BCR), but did
not have breakpoints at m-BCR. The breakpoints of the CML
patients were distributed in a region of about 3-kb length
within the bcr locus, more restricted than the c o m m o n l y
reported 5.8-kb region. Nine patients had a deletion of the
bcr locus at the 3' site and three had the deletion at the 5'
site. In those patients with a deletion at the 3' bcr site, rear-
rangement bands were detected by the 5' bcr probes but not
by the 3' bcr probe. On the other hand, rearrangement bands
were detected by the 3' bcr probe, but not by 5' bcr probes,
in the patients with a 5' site deletion (data not shown). No
specific breakpoint site was detected for the patients with
either 3' or 5' deletions. Eight patients with lymphoid crisis,
two with mixed lymphoid and myeloid crisis, and seven with
the complex 9;22 translocations did not have any specific
breakpoint region in the bcr.
Expression of bcr-ABL chimeric mRNA
RT-PCR analyses were carried out in 43 of the 100 patients.
The bcr-ABL chimeric mRNA expression specific for CML
was detected in 41 of the 43 patients studied. Eight of the
43 patients (18.6%) expressed mRNA derived from a junc-
tion of bcr exon 2 and ABL exon IT(b2-a), 20 (46.5 %) expressed
mRNA derived from bcr exon 3 and ABL exon II (b3-a), and
13 (30.2%) expressed mRNA derived from both types of
mRNA (Fig. 3). The other two patients with the breakpoints
"outside" the bcr locus had no b3-a or b2-a expression (Fig.
3). Both DNA and RNA analyses in the same samples from
each of the 43 patients revealed a clear correlation between
the breakpoints at the bcr locus and the types of expression
in the bcr-ABL mRNA. The majority of the breakpoints oc-
curred between bcr exon 2 and exon 3 (region C1), or between
exon 3 and exon 4 (regions C2 and D). Four patients with
breakpoints at region C1 had only b2-a expression. Two of
the four patients with breakpoints in regions C1 or C2 and
two of the 12 patients with breakpoints in region C2 showed
only b2-a expression. Seventeen patients with the breakpoint
in region D expressed only b3-a, or both b2-a and b3-a. In
the patients with the expression of both b2-a and b3-a, the
intensity of the 125-bp product band of b2-a expression was
usually faint, suggesting a lower level of expression in these
patients. These results showed that patients with breakpoints
within intron 2 (between exons 2 and 3) expressed b2-a

Figure 2 Distribution of breakpoints in the bcr (M-BCR) region in 100 patients (refer to Fig. 1 of bcr divisions and
cleavage sites). Fourteen, 30, and 43 patients had breakpoints at the C1, C2, and D regions, respectively. Breakpoints
were distributed within the 30-kb region shown by the hatched zone, covering from exon 2 to exon 4. Nine of the
100 patients (9.0%) had a deletion at the 3' bcr side and three of 100 (3.0%) had a deletion at the 5' bcr side. These
patients also had a similar breakpoint region, i.e., C1, C2, and D. A patient with a 5' bcr deletion had a breakpoint
at region F. Numbers in circles show the numbers of patients whose breakpoint region could be identified in neither
the C1 nor the C2 region. Two patients had breakpoints outside the bcr. Five exons (bcr exons 1 to 5) in and around
the bcr are shown by black boxes. Restriction enzymes, Bg: BglII, B: BamHI, H: HindIII, E: EcoRI.

bcrex°n I I 3~ I I
I 2 5
..... bcr

region , •
A .
•
B ,C1,C2,
• | |
D ,El .E2.
• • •
F • OUTSIDE;.
• |

H Bg H B H Bg E B H
No.of cases 100 0 0 14 30 43 6 0 1

with 3'bcr 9 2~4 2

deletion ( 9.0 %)
with 5'bcr 3 I I I
deletion (3.0 %)
0 Neither C| nor C2 region w a s identified.
42 K. Tanaka et al.

bcrex°n 1~ ! ~ ~1 !
bcr

region .= A = B | C1| C2 | D I E1 ." E2 ," F outside,

H Bg H B H Bg E B H total

DNA analysis 0 0 4 12 17(1) 3 0 1 (1) 2 43cases

®
RT- PCR analysis
b2-a b3-a
- 4 ~02 8 (18.6%)

-t- -H- 3 7 3 13(30.2%)

- -iF Q~ 7 10(I) !(1) 20(46.5%)
- - 2 2(4.6%)

O N e i t h e r 01 n o r C 2 r e g i o n w a s i d e n t i f i e d .
( ) : 5'bcr deletion
Figure 3 Relationship between the breakpoint region (refer to Fig. 1 for bcr divisions and cleavage sites) and ex-
pression of chimeric bcr-ABL mRNA. The number of patients in each bcr region and the types of bcr-ABL chimeric
mRNA are shown. Forty-three patients were analyzed for both breakpoint region and expression type and the relation-
ship between these factors was studied. Types of bcr-ABL chimeric mRNA detected by RT-PCR analysis were classified
into the following four categories: b2-a (RNA from the junction of bcr exon 2 and the ABL gene), b3-a (RNA from
the junction of bcr exon 3 and the ABL gene) are + . - , + + +, - + . , and - - . One plus (+), two plus (+ +),
and minus ( - ) indicate the expression levels of the chimeric b2 or b3-ABL mRNA. Numbers in circles show numbers
of patients whose breakpoint region could be identified neither in the C1 nor in the C2 region. Numbers in paren-
theses show numbers of patients with the deletion at the 5"side of the bcr. Five exons (bcr exons I to 5) in and around
the bcr are shown by black boxes. Restriction enzyme, Bg: BglII, B: BamHI, H: HindIII, E: EcoRI.

mRNA, w h e r e a s patients w i t h breakpoints w i t h i n intron 3
(between exons 3 and 4) expressed b3-a, or both b2-a and
b3-a mRNA.
Sequential observations of the chimeric bcr-ABL mRNA
during the course of the disease were studied in 11 of the
43 CML patients to check any alteration in the expression
of the bcr-ABL gene during the clinical course. O n l y two pa-
tients showed changes in this expression. One patient ex-
p r e s s e d o n l y b3-a mRNA in the chronic phase and subse-
quently showed both b2-a and b3-a mRNA in the blastic
phase, w h e r e a s the other patient expressed both mRNA (b2-
a and b3-a) in the chronic phase but only b3-a mRNA in the
blastic phase.
Expression of the BCR-ABL gene, w h i c h is a chimeric gene
resulting from rearrangement at the m-BCB region, was also
e x a m i n e d in 19 CML patients, i n c l u d i n g six w i t h l y m p h o i d
blast crisis. Only one patient in the chronic phase, w i t h a
b r e a k p o i n t in region F a n d also w i t h a deletion at 5' bcr, ex-
p r e s s e d BCR-ABL mRNA in a d d i t i o n to the b3-a mRNA.
Prognostic Importance of the Breakpoint Site Within
the bcr Locus
Twelve of the 100 patients were excluded from this analysis:
seven of these 12 d i d not present specific b r e a k p o i n t sites
(four had breakpoints in the C1 or C2 region, two had break-
points outside the bcr, and one had the breakpoint at region
F). The r e m a i n i n g five patients were in the blastic phase at
the time of diagnosis. Of the 88 patients analyzed, 54 evolved
into blastic phase, w i t h the r e m a i n i n g patients being in the
chronic phase. The chronic phase patients were censored.
Distribution of age, WBC, platelet counts, and b a s o p h i l
percentages are shown in boxplots in Figure 4. The distri-
b u t i o n of these parameters d i d not show any differences in
the four breakpoint regions (C1, C2, D, and El). The cumula-
tive percent survival of patients in the chronic phase accord-
ing to breakpoint regions is shown in Figure 5. The m e d i a n
durations of the chronic phase in patients with breakpoints
in the C1, C2, ]3, and E1 regions were 75, 31, 57, and 38 months,
respectively. Patients w i t h breakpoints in region C2 h a d the
shortest chronic phase. A statistically significant difference
was observed among regions C1, C2, D, and E1 by Mantel-
Cox's test (p = 0.028) (Fig. 5A). A l t h o u g h the p value could
be slightly higher than this value because of the m u l t i p l e
c o m p a r i s o n p r o b l e m [34], the prognosis of groups C1 and
C2 seemed to be clearly different.
Multivariate analyses based on the Cox's proportional haz-
ards m o d e l was also u s e d to assess the interactive effects of
prognostic factors on the duration of the chronic phase. The
results are summarized in Table 1. Breakpoint at C2 (p < 0.05)
and a platelet count of less t h a n 150,O00/mm 3 (p < 0.001)
were found to be significant prognostic factors w h i c h cor-
related to shorter duration of the chronic phase. These find-
M-BCR Breakpoint Site in CML 43

(yrl WBC
(xl01m~)
N IlIHI ••

400101, •

511H • II

70
T
!U
IIIII • II

40000 • II

Sill • II

60 ~'III • II

1000 • II

5 0 0 • Oil

30 200. O0

400.00

50 • O0

10 20 • O0

C1 C2 D E1 Cl C2 D E1
(X l(~/mm 3 ) lILT ('/,), BASO
250 20.0

290 X

15.0

I0.0
A
I00

6.0
60

C1 C2 D E1 Cl C2 D E1
Breakpoint regions Breakpoint regions

Figure 4 Baxplots[30] showing the distribution of A) Age, B) WBC, C) PLT, and D) BASOfor each breakpoint re-
gion (C1, C2, D, and El). Median and 95% confidence limits are shown by the bar and diamond shape within the
box, respectively. Numbers of patients in regions C1, C2, D, and E1 were 13, 29, 40, and 6, respectively. WBC; white
blood cell counts, PLT; platelet counts, BASO; basophil percent in the peripheral blood.

ings clearly demonstrate that the duration of the chronic
phase in patients with breakpoints at the C2 region was sig-
nificantly shorter than that in patients with breakpoints in
the C1 and D regions. To evaluate precisely the role of bcr
exons 2 and 3, the bcr was divided into two regions by two
approaches. In the first approach, the bcr locus was divided
into two broad regions at the BamH1 cleavage site located
just before exon 3, resulting in the 5' site having the C1 re-
gion and the 3' site having the C2 + D + E1 regions. The
median duration in the chronic phase of patients with break-
points in the C1 region and the C2 + D + E1 group was 75
and 38 months, respectively. In the second approach, the
bcr locus was divided into two regions at the HindIII cleav-
age site located in intron 3, resulting in the 5' site having
the C1 + C2 regions and the 3' site having the D + E1 regions.
The median durations at C1 + C2 and D + E1 were 55 and
67 months, respectively (Fig. 5C). The categorization of the
breakpoint regions, C2 + D + E1 versus C1 and D + E1 versus
C1 + C2, was also confirmed independently by multivari-
ance analysis. The estimated hazard ratios for C2 + D + E1
versus C1 and D + E1 versus C1 + C2 were 1.9168 and 0.8143,
respectively (Fig. 5B). However, no prognostic significance
for these breakpoint regions was detected in either of the
above two categorizations.
The correlation between the type of bcr-ABL expression
and the duration of the chronic phase was also analyzed in
43 patients. To date, only four patients with b2-a expression
and 15 with b3-a or b2-a and b3-a expression have entered
the blastic phase. The median duration of the chronic phase
in the b2-a group and the b3-a or b2-a and b3-a group was
62 and 37 months, respectively. However, at the time of writ-
ing, Mantel Cox's test has not shown any statistically signifi-
cant difference between the groups showing the two types
of expression (p = 0.71).
DISCUSSION
To detect the precise breakpoint site, multiple analyses were
performed with several restriction enzymes and four bcr
probes. This study design remarkably reduced the chance
of misidentification of the breakpoint sites and also mini-
mized the problems usually engendered by the presence of
44 K. Tanaka et al.

!"
A iN [35]. We illustrate that the breakpoints of 98 patients with
the bcr rearrangement occurred in a more restricted region
of about 3 kb in the center of the bcr locus than that of the
commonly reported 5.8-kb region [1].
Shtivelman et al. [36] proposed alternative splicing mech-
n anisms in the bcr-ABL junction in the case of the simultane-
ous expression of b2-a and b3-a mRNA. The RT-PCR analy-
ses carried out in this study in 43 patients revealed the
tl
c,L expression of b2-a to be only 18.6% and the expression of

t- 7_7
" . . . . I
I
. . . . .

~°°°~ ]
b2-a and b3-a together to be 46.5%. These frequencies were
almost the same as those estimated previously [37].
Our sequential analysis of 19 patients selected at random
revealed no change in the breakpoint region during chronic
0 ~o ~, ,i0 iu ,~ and blast phases. This absence of any alteration indicates that
~aUon of ~ l~OO (mona)
our study of 100 patients (considering no change in the break-
g iN point site in any of the patients) was of great value for the
precise statistical estimation of prognosis, irrespective of
stage. The sequential RNA study also showed a very low fre-
quency of alterations in b2-a and b3-a expression types by

|,, phases.
Mills et al. [38] recently proposed the importance of pres-
ence of bcr exon 3 in hybrid mRNA for determining progno-
sis to blast crisis in CML. In the present study, the influence
of the bcr breakpoint sites on the duration of the chronic
phase was analyzed using both DNA and RNA samples from

J, ......;
the same patients. Those patients with a breakpoint in the
C2 region had a shorter chronic phase than those with a
breakpoint in the C1 region (Fig. 5A, Table 1). Further, in
o " ~o " m • lio " io " m the C2 region the patients were heterogenous with expres-
DulTtlon of chronic phaee(monthe) sion of b2-a or b2-a and b3-a or b3-a (10 of 12 patients had
b2-a and b3-a or b3-a expression and two of 12 patients had
b2-a expression). Taking into consideration our results for
C ~v~,
N oIN .0~_~. both the DNA and RNA analyses in the same patients (Fig.
3), we believe that the occurrence of a shorter chronic phase
in patients with a breakpoint in the C2 region does not seem
|,, , to be due to the presence of exon 3 in the C2 region of the
bcr locus. The C2 region, which is about I kb in length, con-
Z tains bcr exon 3 and also splicing donor sites to express two
3N kinds of bcr-ABL mRNA [37]. The presence of DNA-specific
binding proteins in and around bcr exon 3 would also sup-

I" . .
| lii
Om',,Uonof ~
Figure 5 Cumulative survival curves showing the duration of
chronic phase according to breakpoint regions. A) Comparison
among the four groups C1, C2, D, and El; B) comparison between
the two groups, C1 and C2 + D + El; C) comparison between the
two groups, C1 + C2 and D + El. Mantel Cox's test was used for
statistical analysis. A statistically significant differencewas observed
among C1, C2, D, and E1 regions (P = 0.028). Numbers of patients
in regions C1, C2, D, and E1 were 13, 29, 40, and 6, respectively.
restriction enzyme fragment length polymorphism (RFLP)
and comigration of rearranged bands with germ line frag-
ments. RFLP involving BglIl, BamHI, and EcoRI restriction
enzyme sites in the bcr have been reported by several authors
. .
!
. o . . .
port the idea that transcriptional regulatory factors in the bcr
locus might affect the expression of bcr-ABL mRNA [39] and
that this might be associated with the duration of the chronic
phase. The carboxy-terminal BCR protein domains encoded
I rim N IN
phame(monU,~,) by the 3' site of the BCR gene might also involve the bcr re-
gion, which has a high homology with p85 proteins bind-
ing to the SH2 region of c-ABL protein [40] and also GTPase-
activating proteins for RAS-related GTP-binding protein,
p21 rac [41]. This suggests that the 3' site of the BCR region
might play an important role in the activation and regula-
tion of c-ABL-or RAS-related oncogenes and that it may also
influence the clinical course of CML.
There have been affirmatory [6-13] and contradictory
reports [14-20] concerning the relationship between the
breakpoint sites at the 3' region and the greater duration of
chronic phase in such patients than in patients with the
breakpoint at the 5' site of the bcr. Jaubert et al. [17]and others
[14-16, 18-20] found no significant difference in the dura-
tion of the chronic phase in patients who had breakpoints
at the 5' and 3' sites of the bcr locus. This different finding
M-BCR Breakpoint Site i n CML 45

Table 1 Hazard ratio for prognostic factors estimated by proportional hazard model °
Standardized Influence for
regression Hazard the duration of
Parameters b coefficient ratio Probability chronic phase d
Breakpoint region (I)c C2 vs C1 2.5421 3.7393 p K 0.05 Negative
D vs C1 1.1623 1.6638 n.s.
E1 vs C1 1.0834 2.2614 n.s.
Breakpoint region (II)c C1 vs D - 1.1623 0.6010
C2 vs D 2.2024 2.2474 p ( 0.05 Negative
E vs D 0.4633 1.3592
Sex Male vs Female 1.2251 1.5090 n.s.
Age (years) age 1> 60 vs age ~ 60 1.1559 1.5844 n.s.
WBC (/mm3) WBC /> 100,000 vs WBC K 100,000 1.4510 1.5667 n.s.
Platelet counts PLT ~ 150,000 vs 150,000 x< PLT K 800,000 3.3879 4.8319 p ( 0.001 Negative
(/mma) PLT >/800,00 vs 150,000 ~< PLT K 800,000 - 1.0650 0.5991 n.s.
Basophil (%) basophil >/ 7.0% vs basophil ~( 7.0% 1.2883 1.5452 n.s.
n.s.: not significant
a The five patients in blastic phase at the time of diagnosiswere excluded in this analysis.Numberof patients in regions C1, C2, D, and E1 were 13, 29,
40, and 6, respectively.
b Patients correspondingto the conditions shown in the right hand of 'VS' were regarded as a baseline population.
c With respect to the evaluation of breakpoint regions, either C1 or D was independentlyregarded as a baseline population.
d Breakpoint at C2 and platelet count less than 150,000/mma were significantlycorrelated with the shorter duration of chronic phase.

might be due to the presence of a large n u m b e r of censored
cases, misidentification of the breakpoint region, and varia-
tions in the division of the bcr locus as per the cleavage sites.
Most authors divided intron 3 into two regions, separated
by the HindIII cleavage site between exons 3 a n d 4, as 5' and
3' bcr regions. Using this division, we were unable to detect
any statistically significant prognostic difference between the
two groups [C1 + C2 and D + El). Moreover, w h e n we
divided the bcr locus into two regions separated by the BamHI
cleavage site just before exon 3 (C1 and C2 + D + El), in
order to elucidate the difference i n biological significance
between bcr exons 2 and 3, there was no statistically signifi-
cant difference between these two groups. However, we can-
not deny the possibility that no prognostic difference was
detected due to the small n u m b e r of patients whose break-
points were w i t h i n the C1 region. In our series, 61.4 % of the
patients (54 of 88) i n c l u d e d for statistical analyses had de-
veloped blastic phase. Therefore, this study showed more
definite duration of the chronic phase in each patient than
that shown in other reports.
Seven of the 100 CML patients had been exposed to high
doses of atomic bomb radiation in Hiroshima. Although a
significant increase of CML has been observed among
Hiroshima atomic bomb survivors [42], those patients with
atomic bomb radiation-induced CML did not show any spe-
cific breakpoint region in the bcr [43]. Thus, the presence
of patients exposed to atomic bomb radiation in our series
did not seem to influence our results.
Because all CML patients generally develop blastic phase,
it is considered that the "re-rearrangement" or altered expres-
sion of the bcr-ABL gene itself, and also that of other genes
closely associated with the function of the bcr-ABL gene,
might be involved in the development of blastic crisis [44].
The occurrence of the deletion at the 5' bcr site or of break-
points outside bcr could indicate that the "re-rearrangement"
at this bcr locus is not a rare p h e n o m e n o n . One patient with
the 5' bcr deletion showed l y m p h o i d phenotype in the blas-
tic phase (data not shown). Other evidence showing a possi-
ble relationship between 5' bcr deletion and lymphoid pheno-
type has been observed in a CML patient with blast crisis
[45] and in a patient with Ph positive acute lymphocytic
leukemia with T-cell phenotype [46].
The correlation between the breakpoint site in the bcr lo-
cus and the duration of the chronic phase suggests an im-
portant role for the bcr-ABL gene in the pathogenesis of CML.
Daley et al. [47] and others [48-50] have clearly demonstrated
that the bcr-ABL gene or p210 protein behaves as an oncogene
i n hematopoietic stem cells, i n d u c i n g granulocytic or lym-
phocytic neoplasms i n mice. Further studies of the function
of the bcr-ABL chimeric gene and p210 protein are essential
for the u n d e r s t a n d i n g of the pathogenesis of h u m a n CML.
Our multi-faceted study of the bcr-ABL gene in 100 patients
has demonstrated the possible influence of breakpoint sites
on the progression of CML.
We wish to thank Dr. Ahmed M. Mansoor for his critical reading
and correction of this manuscript, Dr. S. Mizutani, Dr. K. Nakamura,
and Mr. H. Tazawa for their kind help in PCR analysis, and Dr. S.
Hirosawa for kindly providing the 5' bcr probes. This study was sup-
ported, in part, by a Grant-in-Aid for Scientific Research from the
Ministry of Education, Science and Culture, Japan (#01480534) and
Grant from Sankyo Foundation of Life Science (1990 and 1991).
REFERENCES
1. Groffen J, Stephenson JR, Heisterkamp N, de Klein A, Bartram
CR, Grosveld G (1984): Philadelphia chromosomal breakpoints
are clustered within a limited region, bcr, on chromosome 22.
Cell 36:93-99.
2. Heisterkamp N, Stephenson JR, Groffen J, Hansen PF, de Klein
46
A, Bertram CR, Grosveld G (1983): Localization of the c-abl on-
cogene adjacent to a translocation break point in chronic my-
elocytic leukemia. Nature 306:239-242.
3. Stem K, Heisterkamp N, Grosveld G, de Klein A, Verma RS, Cole-
man M, Dosik H, Groffen J (1985): Evidence of a new chimeric
bcr/abl mRNA in patients with chronic myelocytic leukemia and
the Philadelphia chromosome. N Engl J Med 313:1429-1433.
4. Gale RP, Canaani E (1984): An 8-kilobase abl RNA transcript
in chronic myelogenous leukemia. Proc Natl Acad Sci USA
81: 5648-5652.
5. Ben-Neriah Y, Daley GO~ Mes-Masson AM, WiRe ON, Baltimore
D (1986): The chronic myelogenous leukemia-specific P210 pro-
tein is the product of the bcr/abl hybrid gene. Science 233:
212-214.
6. Schaefer-Rego K, Dudek H, Popenoe D, Arlin Z, Meats JG, Bank
A, Leibowitz D (1987): CML patients in blast crisis have break-
points localized to a specific region of the BCR. Blood 70:
448-455.
7. Mills KI, Mackenzie ED, Birnie GD (1988): The site of the break-
points within the bcr is a prognostic factor in Philadelphia-
positive CML patients. Blood 72:1237-1241.
8. Eisenberg A, Silver R, Soper L, Arlin Z, Coleman M, Bernhardt
B, Benn P (1988): The location of breakpoints with the break-
point cluster region (bcr) of chromosome 22 in chronic myeloid
leukemia. Leukemia 2:642-647.
9. Mills KI, Hynds SA, Burkitt AK, Mackenzie ED, Birnie GD
(1989): Further evidence that the site of breakpoint in the major
breakpoint cluster region (M-bcr) may be a prognostic indica-
tor. Leukemia 3:837-840.
10. Dreazen O, Barman M, Gale RP (1988): Molecular abnormali-
ties ofber and c-abl in chronic myelogenous leukemia associated
with a long chronic phase. Blood 71:797-799.
11. Shtalirid M, Talpaz M, Kurzrock R, Kantarjian H, Trnjillo J, Gut-
terman J, Yoffe G, Blick M (1988): Analysis of breakpoints within
the bcr gene and their correlation with the clinical course of
Philadelphia positive chronic myelogenous leukemia. Blood
72:485-490.
12. Grossman A, Silver RT, Arlin Z, Coleman M, Camposano E, Gas-
con P, Benn PA (1989): Fine mapping of chromosome 22 break-
points within the breakpoint cluster region (bcr) implies a role
for bcr exon 3 in determining disease duration in chronic my-
eloid leukemia. Am J Hum Genet 45:729-738.
13. Johansson B, Mertens F, Fioretos T, Heim S, Kristoffersson V,
Mandahl N, Bartram CR, Mitelman F (1990): Remarkably long
survival of a patient with Phi-positive chronic myeloid leuke-
mia and 5' bcr rearrangement. Leukemia 4:448-449.
14. Ogawa H, Suhiyama H, Soma T, Masaoka T, Kishimoto S (1989):
No correlations between locations ofbcr breakpoints and clini-
cal states in Phi-positive CML patients. Leukemia 3:492-496.
15. Prezepiorka D (1988): Breakpoint zone of bcr in chronic my-
elogenous leukemia does not correlate with disease phase or
prognosis. Cancer Genet Cytogenet 36:117-122.
16. Dyck JA, Bosco JJ (1982): Clinical stage of chronic gmnulocytic
leukemia and BCR breakpoint location in South-East Asian pa-
tients. Br J Haematol 72:64-67.
17. Jaubert J, Martiat P, Dowding C, Ifran N, Goldman JM (1990):
The position of the M-BCR breakpoint does not predict the du-
ration of chronic phase or survival in chronic myeloid leuke-
mia. Br J Haematol. 74:30-35.
18. Morris SW, Daniel L, Ahmed CMI, Elias A, Lebowitz P (1990):
Relationship of bcr breakpoint to chronic phase duration, and
blast crisis lineage in chronic myelogenous leukemia patients
presenting in early chronic phase. Blood 75:2035-2041.
19. Teffer A, Bren GD, Wagner KV, Schaid DJ, Ash RC, Thipodeau
SN (1990): The location of the Philadelphia chromosomal break-
point site and prognosis in chronic granulocytic leukemia.
Leukemia 4:839-842.
K. T a n a k a et al.
20. Tien HE Wang CH, Chen YC, Shen MC, Wu HS, Lee FY, Chu-
ang SM, Liu CH (1990): Chromosome and bcr rearrangement
in chronic myelogenous leukemia and their correlation with
clinical states and prognosis of the disease. Br J Haematol
75:469-475.
21. Mills KI, Benn P, Birnie GD (1991): Does the breakpoint within
the major breakpoint cluster region (M-bcr) influence the du-
ration of the chronic phase in chronic myeloid leukemia? An
analytical comparison of current literature. Blood 78:1155-1161.
22. Kamada N, Tanaka K (1983): Cytogenetic studies of hematolog-
ical disorders in atomic bomb survivors. In: Sandberg AA, ed.
Radiation-Induced Chromosome Damage in Man. Alan R. Liss,
New York, pp. 455-474.
23. ISCN (1985): An International System for Human Cytogenetic
Nomenclature, in Harnden DG, Klinger HP, eds. Published in
collaboration with Cytogenet Cell Genet, Karger, Basel.
24. Southern EM (1975): Detection of specific sequences among DNA
fragments separated by gel electrophoresis. J Mol Bio198:503-517.
25. Hirosawa S, Aoki N, Matsushima H, Shibuya M (1988): Unde-
tectable bcr-abl rearrangements in some CML patients are due
to a deletion mutation in the bcr gene. Am J Haemato128:33-36.
26. Chirgwin JM, Przyblyla AE, Mac Donald RJ, Rutter WJ (1979):
Isolation of biologically active ribonucleic acid from sources
enriched in ribonuclease. Biochemistry 18:5294-5299.
27. Nakamura K, Miyashita T, Ozaki M, Zhong WK, Iwai M,
Nakazawa S, Okamum J, Takayama J, Ohira M, Kamada N, Tanaka
K, Tamura K, Kobayashi N, Mizutani S (1991): Molecular studies
of chronic myelogenous leukemia by polymerase chain reac-
tion. Cancer 68:2426-2430.
28. Kawasaki ES, Clark SS, Coyne MY, Smith SD, Champliu R, Witte
ON, McCormick FP (1988): Diagnosis of chronic myeloid and
acute lymphocytic leukemias by detection of leukemia-specific
mRNA sequences amplified in vitro. Proc Natl Acad Sci USA
85:5698-5702.
29. Roth MS, Antin JH, Bingham EL, Ginsburg D (1989): Detection
of Philadelphia chromosome-positive cells by the polymerase
chain reaction following bone marrow transplant for chronic
myelogenous leukemia. Blood 74:882-885.
30. Chambers JM, Cleaveland WS, Kleiner B, Turkey PA (1983):
Graphic methods for data analysis. Duxbury Press, Boston, pp.
57-64.
31. Cox DR (1972): Regression models and life tables. J Royal Stat
Soc Series B34, pp. 187-220.
32. Kleinbaum DQ, Kupper LL (1978): In: Applied Regression Anal-
ysis and Other Multivariate Methods. Duxbury Press, Boston.
33. Dixon WJ, Brown MB (1983): Statistical Software, 1983 revised
printing. University of California Press, Los Angeles, CA.
34. Tukey JW (1977): Some thoughts on clinical trials, especially
problems of multiplicity. Science 198:679.
35. Grossman A, Mathew A, O'Connell MP, Tiso P, Distenfeld A,
Benn P (1990): Multiple restriction enzyme digests are required
to rule out polymorphism in the molecular diagnosis of chronic
myeloid leukemia. Leukemia 4:63-64.
36. Shtivelman E, Lifshitz B, Gale RP, Roe BA, Canaani E (1986):
Altemtive splicing of RNAs transcribed from the human abl gene
and from the bcr-abl fused gene. Cell 47:277-284.
37. Shtivelman E, Gale RP, Dreazen O, Berrebi A, Zaizov R,
Kubonishi I, Miyoshi I, Canaani EC (1987): bcr-abl RNA in pa-
tients with chronic myelogenous leukemia. Blood 69:971-973.
38. Mills KI, Sproul AM, Leibowitz D, Burner AK (1991): Mapping
of breakpoints, and relationship to BCR-ABL RNA expression,
in Philadelphia-chromosome-positive chronic myaloid leukae-
mia patients with a breakpoint around exon 14(b3) of the BCR
gene. Leukemia 5:937-941.
39. Leibowitz D, Young K, Gregory C (1991): Sequence-specific DNA-
binding proteins within the Mbcr on the Phl chromosome. Genes
Chrom Cancer 3:308-312.
M-BCR B r e a k p o i n t Site in CML
40. Pendergast AM, Muller AJ, Havlik MH, Maru Y, Witte ON (1991):
BCR sequences essential for transformation by the BCR-ABL on-
cogene bind to the ABL SH2 regulatory domain in a non-
phosphotyrosine-dependent manner. Cell 66:161-171.
41. Diekmann D, Brill S, Garrett MD, Totty N, Hsuan J, Monfries
C, Hall C, Lim L, Hall A (1991): Bcr encodes a GTPase-activating
protein for p21 rac. Nature 351:400-402.
42. Ichimaru M (1986): Cancers in atomic bomb survivors. In: Mono-
graph on Cancer Research, No. 32, Japanese Cancer Associa-
tion, p. 120.
43. Tanaka K, Takechi M, Hong J, Shigeta C, Oguma N, Kamada
N, Takimoto Y, Kuramoto A, Dolly H, Kyo T (1989): 9;22 trans-
location and bcr rearrangements in chronic myelocytic leuke-
mia patients among atomic bomb survivors. J Rad Res 30:
352-358.
44. Morgan GJ, Hernandez A, Chan LC, Hughes T, Martiat P,
Weidemann LM (1990): The role of alternative splicing patterns
of BCR/ABL transcripts in the generation of the blast crisis of
chronic myeloid leukaemia. Br J Hematol 76:33-44.
47
45. Bartram CR, Janssen JWG, Becher R, de Klein A, Grosveld G
(1986): Persistence of chronic myelocytic leukemia despite de-
letion of rearranged bcr/c-abl sequences in blast crisis. J Exp
Med 164:1389-1396.
46. Takechi M, Ohnishi A, Tanaka K, Kimura N, Kamada N (1990):
Acute lymphocytic leukemia with Philadelphia chromosome
and rearrangements of bcr gene and of TCR-8 gene. Leukemia
Res 14:885-893.
47. Daley GO~ van Etten RV, Baltimore D (1990): Induction of chronic
myelogenous leukemia in mice by the p210 b~rabl gene of the
Philadelphia chromosome. Science 247:824-830.
48. Heisterkamp N, Jenster G, yen Hoeve J, Zovich D, Pattengale PK,
Groffen J (1990): Acute leukemia in bcr/abl tmnsgenic mice. Na-
ture 344:251-253.
49. Elefanty AG, Hariharan IK, Cory S (1990): bcr-abl, the hallmark
of chronic myeloid leukaemia in man, induces multiple hae-
mopoietic neoplasms in mice. EMBO J 9:1069-1078.
50. Kelliher MA, McLaughlin J, Witte ON, Rosenberg N (1990): In-
duction of a chronic myelogenous leukemia-like syndrome in
mice with v-abl and BCR/ABL. Proc Natl Acad Sci USA
87:6649-6653.