Myeloproliferative Disorders original paper

haematologica 2001; 86:252-259

Real-time quantification of different http://www.haematologica.it/2001_03/0252.htm
types of bcr-abl transcript in chronic
myeloid leukemia

MARILINA AMABILE, BARBARA GIANNINI, NICOLETTA TESTONI,

VITTORIO MONTEFUSCO, GIANANTONIO ROSTI, CLAUDIA ZARDINI,
CAROLINA TERRAGNA, SILVIA BUONAMICI, EMANUELA OTTAVIANI,
SIMONA SOVERINI, MAURO FIACCHINI, SIMONA BASSI, ANTONIO
DE VIVO, ELENA TRABACCHI, GIUSEPPE SAGLIO,° FABRIZIO PANE,#
MICHELE BACCARANI,* SANTE TURA, GIOVANNI MARTINELLI
Institute of Hematology and Medical Oncology “Seràgnoli”, Uni-
versity of Bologna, *Institute of Hematology, University of Udine; Correspondence: Giovanni Martinelli, M. D., Institute of Hematology and
Medical Oncology “Seràgnoli”, University of Bologna, via Massarenti 9,
°Department of Medical Science and Human Oncology, Universi- 40138 Bologna, Italy.
ty of Turin; #Department of Biochemistry & Medical Biotechnolo- Phone: international +39-051-6364075 - Fax: international +39-051-
gies, Federico II University, Naples, Italy 6364037 - E-mail: [email protected]

n
M
Background and Objectives. The most common translo- olecular analysis of chromosomal transloca-
tions associated with hematopoietic tumors

tio
cation in chronic myeloid leukemia (CML) t(9;22)
(q34;q22) produces the BCR/ABL fusion gene. We set has provided important tools for the diagnosis
up and evaluated a rapid and reliable real-time reverse- and monitoring of patients’ response to therapy. A con-

da
transcription-polymerase chain reaction (RT-PCR) sistent translocation, t(9;22), is present in the cells of
approach using TaqMan technology for detection and most chronic myeloid leukemia (CML) patients.1 Recip-
un
quantification of bcr-abl transcripts in CML patients at rocal translocation of the two genes involved – ABL and
diagnosis and during therapy. BCR on chromosomes 9 and 22, respectively – gives rise
Fo
Design and Methods. A pair of primers and probe com- to a fusion gene which produces the neoplastic BCR-
plementary to ABL exon 2 were designed, enabling detec- ABL fusion protein.2 In 99% of CML patients, two types
tion of the most frequent bcr-abl transcripts, and also of of bcr-abl junction (namely, “b2-a2” and “b3-a2”) can
the normal ABL-Ia transcript as an internal control. Con-
ti

be detected by reverse transcription (RT) polymerase

ditions were established to amplify less than 1-10 target chain reaction (PCR), due to the exclusion (b2-a2) or
or

molecules/reaction and detect one CML cell in 106 cells

from healthy donors. To determine the utility of the assay, inclusion (b3-a2) of 75 base pairs (bp) from BCR exon
St

we quantified the bcr-abl/ABL-Ia ratio in 59 bone mar- 14 (or b3 of a major-breakpoint cluster region, M-BCR).3
row samples (45 samples with evidence of different Ph+ RT-PCR detection of the bcr-abl transcript can be used
chromosome percentages and 14 samples in complete for diagnosis of CML and monitoring of the Ph clone.4
ta

cytogenetic remission) from 48 CML patients, 34 of them It has allowed more precise definitions of disease sub-
at diagnosis and 14 in clinical remission (CR). In 14 cas- sets, and provided potentially valuable prognostic
ra

es, this ratio was compared with results obtained by a information for the management of patients.4,5 Quali-
competitive-quantitative RT-PCR/capillary electrophore-
r

tative RT-PCR has already been widely used to reveal

sis method from contemporary specimens.
Fe

the presence of detectable levels of minimal residual

Results. By real-time RT-PCR, the median value of bcr- disease (MRD) in patients in complete cytogenetic
abl/ABL-Ia ratio at diagnosis was 15.334 (range 3.3- remission (CR) after α-interferon therapy (α-IFN).4-6
©

28.81) and fell to 0.9 (range 0.003-26.1) in CR. The However, the recent introduction of ‘real-time’ quanti-
median value of bcr-abl/ABL-Ia ratio at cytogenetic tative RT-PCR7 should eventually allow actual quantifi-
remission was 0.7 (range 0.003-2.83). The real-time bcr-
abl/ABL-Ia ratios correlated with those obtained by com- cation of the target gene to be routinely performed at
petitive RT-PCR (p <0.0001) and the percentage of Ph+ diagnosis and during follow-up. In real-time RT-PCR,
metaphases (p <0.0001). The high sensitivity and speci- the 5'-3' nuclease activity of the Taq polymerase cleaves
ficity of the real-time RT–PCR procedure was confirmed an internal fluorogenic probe specific for the target
in all 14 patients with minimal residual disease. sequence, causing the emission of a fluorescent signal
Interpretation and Conclusions. We conclude that this that can be detected during amplification. Several
real-time RT-PCR procedure is a reliable and sensitive groups have used real-time RT-PCR with TaqMan tech-
method of monitoring CML patients after therapy, and nology to quantify MRD with leukemia-specific chro-
that the bcr-abl/ABL-Ia ratio correlates strongly with mosome aberrations.8 Complete suppression of the Ph+
cytogenetic analysis. clone has been observed in the majority of allotrans-
©2001, Ferrata Storti Foundation planted patients9 and a few of those treated with α-
IFN4,10-11 These considerations suggest that routine
Key words: chronic myelogenous leukemia, real-time RT- quantitative analysis using real-time RT-PCR could be
PCR, bcr-abl transcript, minimal residual disease, quanti- of clinical value in CML patients submitted to these
tative PCR treatments.

haematologica vol. 86(3):March 2001

 Real-time quantification of bcr-abl transcripts in CML 253

Table 1. Qualitative molecular characterization of chronic myeloid leukemia patients.

CML patients studied/samples CML patients studied Male/female Type e1/a2 Type e13/a2 Type e14/a2
at diagnosis/after BMT bcr-abl bcr-abl bcr-abl

48/59 34/14 26M/22F 0 (0%) 20 (41%) 28 (59%)

In the present study, we have developed bcr-abl quan- where.17 Total cellular RNA was extracted as previously
tification by real-time RT-PCR using the ABI PRISM described,18 and 1µg of total RNA was used in the RT-
7700 (Perkin Elmer), a new technique which allows sim- reaction.13
ple and rapid quantification of a target sequence dur-
Qualitative RT-PCR analysis
ing the extension phase of PCR amplifications. A fluo-
Qualitative RT-PCR was routinely performed at diag-
rogenic probe labeled with both a reporter-dye at the 5' nosis and during follow-up, as previously described.13,15,17
end and a quencher-dye at the 3' end hybridizes to the To assess the quality and quantity of the amplifiable
target sequence on the second exon of the ABL gene. RNA isolated from samples, RT-PCR of ABL control gene
The exonuclease activity of the Taq DNA polymerase transcripts was performed as previously described.13,15,17

n
cleaves the probe and releases the reporter-dye, result- The following stringent criteria for negativity were

tio
ing in an increase in the fluorescence signal. This allows applied: no amplification of the bcr-abl transcript in
identification and quantification of specific RT-PCR three independent assays at a sensitivity of 1:105 (see
products as the reaction proceeds.12 In an unknown

da
below) always accompanied by successful amplification
sample, the absolute copy numbers of the bcr-abl tar- of the ABL transcript. To determine the level of sensi-
get sequence and of a control gene, such as ABL, can be un
tivity of breakpoint sequence amplification, experiments
calculated at the end of the reaction using a calibration using bcr-abl positive RNA (types e13/a2 and e14/a2)
curve prepared from a set of bcr-abl RNA standards. The were conducted by serially diluting bcr-abl positive RNA
Fo
results are then expressed as a bcr-abl/ABL ratio. Quan- in HL60 cell line RNA, as reported elsewhere;13,15,17 TOM1
titative data can be rapidly produced with a very wide cell line was used as an e1/a2 positive control. Positive
dynamic detection range of over five orders of magni- and negative controls were performed in all assays. Pos-
ti

tude. We set up probes for bcr-abl and tested the itive controls were bcr-abl positive RNA extracted from
or

method on bone marrow (BM) samples previously a BM sample from a CML patient. Negative controls
obtained from a group of CML patients, some of whom consisted in reactions with either RNA from a bcr-abl
St

had already been studied by an established competitive negative CML patient or HL60 cell line RNA. Precau-
RT-PCR method.13,14 In the latter cases, we compared tions taken to avoid contamination included use of a
ta

the reliability and sensitivity of the two methods. specifically designed UV-flow cabinet and PCR-desig-
nated pipettes with filter tips.15,17 All tests were done
ra

Design and Methods twice to confirm the results.

Patients and therapy
r

Competitive RT-PCR and capillary

Fe

A total of 59 BM samples (45 samples with cytoge- electrophoresis (CE) assays

netic evidence of disease and 14 samples from patients Competitive RT-PCR and CE assays were performed as
in complete cytogenetic remission) (Table 1) from 48
©

described previously.13
CML patients (26 males, 22 females), thirty-four at diag-
nosis and 14 in clinical remission (CR) after allotrans- Real-time quantitative RT-PCR
plantation were analyzed. Informed consent was always Real-time quantitative RT-PCR was retrospectively
obtained, as required by the Declaration of Helsinki. At performed on stored samples. The principles and proce-
diagnosis, all patients were Ph+. In 14 cases, this ratio dure of real-time RT-PCR quantification using the Taq-
was compared with results obtained by a competitive- Man probe (or a similar approach)19 have already been
quantitative RT-PCR/capillary electrophoresis method described.20
from contemporary specimens. Primers and probe for real-time RT-PCR of
Cytogenetic analysis bcr-abl
Cytogenetic analysis was routinely performed accord- In order to define an amplicon no longer than 200 bp,
ing to a standard technique.15 At least 20 mitoses were the TaqMan probe (BOP ABL probe) and reverse primer
analyzed for each sample.16 (BOR ABL primer) were located in exon 2 of the ABL
gene for all types of transcript (Figure 1 and Table 2). The
Sample and RNA isolation forward primers were located on exon 1 (BOF E1A2
Mononuclear cells from BM (3-5 mL) samples were primer), on exon 13 (BOF B2A2 primer) and exon 14
obtained by Ficoll-Hypaque density gradient centrifu- (BOF B3A2 primer) of the BCR gene for types e1/a2,
gation and stored at –80°C in GITC as reported else- e13/a2 and e14/a2 respectively. Probes were labeled by

haematologica vol. 86(3):March 2001

254 M. Amabile et al.

Table 2. Oligonucleotides used in the real-time RT-PCR

assay for quantification of BCR-ABL transcripts.

Primers and Sequences (5’ - 3’) Mapping

detection probes

BOF E1A2 primer CGCAAGACCGGGCAGAT BCR exon 1

BOF B2A2 primer GCATTCCGCTGACCATCAAT BCR exon 13
BOF B3A2 primer TCCACTCAGCCACTGGATTTAA BCR exon 14
BOF ABL primer TCCTCCAGCTGTTATCTGGAAGA ABL exon Ia
BOR ABL primer TGGGTCCAGCGAGAAGGTT ABL exon 2
BOP ABL probe FAM-CCAGTAGCATCTGACTTTGAGCCTCAGGG-TAMRA ABL exon 2

Table shows primers and probes used for bcr-abl and for ABL-Ia real-time RT-PCR
quantification. FAM, 6-carboxyfluorescein; TAMRA, 6-carboxy-tetramethylrhodamine.

300/300, 300/900, 900/50, 900/300 and 900/900. RT-

PCR reactions were set in MicroAmp optical 96-well

n
reaction plates closed with MicroAmp optical caps
(Perkin Elmer/Applied Biosystem). After 2 minutes at

tio
50°C to allow UNG to destroy potential contaminant
RT-PCR products, and 10 minutes at 95°C to denature

da
UNG and activate AmpliTaq Gold, the amplification was
Figure 1. Figure shows a schematic representation (not in carried out by 50 cycles at 95°C for 15 seconds and 65°C
un
scale) of the set of primers and probes used for real-time RT- for 60 seconds in the ABI/Prism 7700 Sequence Detec-
PCR assay for amplification of either e1/a2 and e13/a2 or tor System (ABI/Perkin Elmer, Foster City, CA, USA).
e14/a2 bcr-abl transcripts. Open boxes and numbers rep-
Fo
resent BCR exons. Dotted boxes and numbers represent Construction of “standard curves” for
ABL exons. Arrows indicate primers used for RT-PCR with BCR/ABL fusion gene and ABL control
their orientations. Filled black rectangles represent the Taq-
Man probe. Numbers below the boxes represent nucleotides gene
ti

at exon boundaries. Sequences of primers and probes as in We cloned the RT-PCR products obtained by amplifi-
or

Table 2 are also named under the boxes. cation with the primers listed in Table 2, derived at diag-
nosis from TOM1 cell line, from patients #1 and 2 cor-
St

responding to bcr-abl types e1/a2, e13/a2 and e14/a2,

respectively, and the ABL-Ia amplification product. The
ta

four products were cloned into the pCR II-TOPO vector

a 5’ FAM reporter and 3’ TAMRA quencher. One set of using a TOPO TA cloning kit (Invitrogen, Carlsbad, CA,
ra

primers and probe was also defined for the ABL control USA). We called the resulting plasmids pTA-e1/a2, pTA-
gene (exon Ia) (BOF ABL primer in Table 2).2 Primers and e13/a2 and pTA-e14/a2, pTA-a1/a2, respectively. In order
r

to construct the standard curve for the quantification of

Fe

TaqMan probes were all designed using Primer Express

software (Perkin Elmer, Foster City, CA, USA). each type of bcr-abl and ABL-Ia transcript, serial dilu-
tions of the corresponding plasmid were used. For each
©

PCR conditions for real-time RT-PCR of assay, 10-fold dilutions, starting at 104 plasmid copies,21
bcr-abl were analyzed in triplicate using Sequence Detector Sys-
Reaction mixtures of 25 µL contained 12.5 µL of Taq- tem software V1.6 (Perkin Elmer, Applied Biosystem). A
Man buffer A with the ROX dye as the passive reference, standard curve was established by plotting the CT and
5 mM MgCl2, 200 µM dATP, dCTP, dGTP, 400 µM dUTP, the known copy number on a logarithmic scale.
1.25 U AmpliTaq Gold DNA polymerase, 0.5 U AmpErase
uracil N-glycosylase (UNG), 300 nM forward and reverse Statistical analysis
primers, 200 nM specific TaqMan probe and 6 µL of Comparison between qualitative and real-time RT-
plasmid or cDNA (diluted 1:3). All the reagents were PCR positivity was performed as previously reported.3
from Perkin Elmer/Applied Biosystem. All analyses were carried,out using the SPSS software
All real-time RT-PCR experiments were performed at package (SPSS Inc., Chicago, IL, USA).22
least in triplicate. Before determining the sensitivity of
the target, the real-time RT-PCR set up was optimized. Results
In particular, the amounts of forward and reverse primer Cytogenetic analysis
producing the highest ∆Rn and lowest CT were deter- Cytogenetic results are summarized in Table 3 (a, b,
mined. In the primer-matrix experiment, nine combina- and c). A total of 59 BM analyses were performed. At
tions of 50, 300 and 900 mM for each primer were test- diagnosis, all 34 patients showed Ph+ chromosome
ed in triplicate: i.e. 50/50, 50/300, 50/900, 300/50, (Table 3a). Furthermore, all the 14 CML patients who

haematologica vol. 86(3):March 2001

 Real-time quantification of bcr-abl transcripts in CML 255

Table 3 (a), (b), (c). Molecular and cytogenetic evaluation of clinical samples of CML patients.

CML
Cases studied Phase of Disease Ph+% Quantitative real time Qualitative RT-PCR
RT-PCR analysis for analysis for bcr-abl
bcr-abl/ABL-Ia ratios

(a) Cases from 1 to 34 Chronic phase 100 median 15.334 + all (positive)
and diagnosis (range 3.36-28.81)

CML
Cases studied Phase of Disease Ph+% bcr-abl/ABL-Ia ratios Qualitative RT-PCR
analysis for bcr-abl

(b) Cases from 35 to 41 allotransplantation median 0.013

(UPN of patients transplanted (months from BMT) (range 0.003-0.7720)
in chronic phase)

164 +13 0 0.0120 - (negative)

194 +35 0 0.0700 -
249 +2 0 0.3180 -
262 +6 0 0.0010 -
274 +15 0 0.0170 -

n
292 +3 0 0.7720 -

tio
+6 0 0.0003 -
301 +6 0 0.0008 -

da
CML
Cases studied Phase of disease Ph+ % Quantitative real-time Qualitative RT-PCR
un RT-PCR analysis for analysis for bcr-abl
bcr-abl/ABL-Ia ratios
Fo
(c) Cases from 42 to 48 allo-transplantation median 2.5
(UPN of patients transplanted (months from BMT) (range 0.04-20.16)
in accelerated or blastic phase)
ti

141 +62 0 2.8340 + (positive)

or

+64 9.3 1.0760 +

+72 0 0.1040 -
St

+77 0 2.5000 -
190 +15 3.9 2.5910 +
+25 8.5 15.9690 +
ta

+36 16.6 17.4040 +

+41 2 8.9610 +
ra

+51 4 1.8650 +
+56 28 21.4310 +
+61 58 26.1130 +
r

199 +2 40.5 20.1610 +

Fe

214 +2 3.3 0.7720 +

267 +3 0 0.0880 -
+4 0 0.2050 -
©

285 1 0 0.04 -
288 +3 25 20.014 +

This table shows clinical and molecular follow-up of three groups of CML patients: at diagnosis (a); in allotransplanted patients in chronic phase (b); in allotransplanted patients
in accelerated or blastic phase (c). UPN = unidentified progressive number of transplanted CML patients. Bcr-abl/Ia ratios are expressed as described in Design and Methods;
qualitative PCR analysis was considered positive (+) or negative (-) as described in ref. #15.

were studied after BMT had been Ph+ before transplan- Qualitative RT-PCR analysis
tation. Eleven samples from 5 allotransplanted patients Assays of all 59 samples from the patients were stud-
(all transplanted in blastic phase: UPN141, UPN190, ied. All 48 patients were bcr-abl positive by conven-
UPN199, UPN214, and UPN288) showed different tional PCR at diagnosis. Twenty out of 48 (41%) patients
degrees of Ph positivity. All the 8 samples from seven expressed e13/a2 transcript,13 and 28 (58%) displayed
patients allotransplanted in chronic phase (UPN164, type e14/a2; no patient expressed e1/a2. All samples
UPN194, UPN249, UPN262, UPN274, UPN292 and that turned out to be ABL-Ia negative were excluded
UPN301 in Table 3b) were in karyotypic remission (KR). from subsequent analyses.23 Log sensitivities of the tran-

haematologica vol. 86(3):March 2001

256 M. Amabile et al.

script primer sets were 10–6 for all bcr-abl types, as

reported.3,24-27 Qualitative RT-PCR analysis for bcr-abl
on samples is reported in Table 3 a, b, and c.
Application of real-time quantitative
RT-PCR for bcr-abl analysis
For all types of bcr-abl transcript studied involving
ABL exon 2 (e1/a2, e13/a2 and e14/a2), it was possible
to develop primer/probe combinations that permitted
amplification and real-time RT-PCR analysis. For all
types of transcript, the TaqMan probe and forward
primer were located in exon 2 of the ABL-Ia gene. The
reverse primers were located in exons 1, 13 and 14 of
the BCR gene for transcripts e1/a2, e13/a2 and e14/a2,
respectively (Figure 1 and Table 2). The number of tar-
get molecules of transcript in each sample was
expressed as a percentage ratio between bcr-abl and
ABL-Ia in 6 µL of cDNA. Figure 2. Bcr-abl/ABL-Ia ratios, as evaluated by real-time
PCR, of samples from Ph+ CML patients at diagnosis [group

n
Sensitivity of real-time PCR (a), Table 3] or during clinical remission and karyotypic

tio
In repeated tests, we reliably amplified ten e1/a2, remission after allogeneic bone marrow transplantation
e13/a2 or e14/a2 bcr-abl and ten ABL-Ia transcripts per [group (b), and some patients of group (c) Table 3].

da
reaction. To calculate the maximum sensitivity, serial
dilutions of 10, 100, 1,000, and 10,000 myeloid PB cells
taken from e13/a2 and e14/a2 bcr-abl positive CML un
patients at diagnosis or from the TOM1 cell line were (inter-assay comparison, day-to-day variation with new
processed and analyzed in 10×107 leukocytes from a mixtures of reagents) resulted in a CV of r=0.01, the
Fo
healthy donor. We routinely achieved a sensitivity of cycle threshold crossing points having a CV of r=0.01.
10–6 (i.e. we could detect one Ph+ cell in 106 normal
white blood cells). The same approach was performed Correlation between real-time
ti

with a dilution of TOM1 (e1/a2 bcr-abl positive cell line), quantification of bcr-abl/ABL ratio
or

of BV173 (e13/a2 bcr-abl positive cell line) and K562 and karyotypic status
(e14/a2 bcr-abl positive cell line) cells in HL60 (bcr-abl Figure 2 summarizes the results of 59 BM samples ret-
St

negative cell line) cells. Reproducible sensitivities of 10-7 rospectively studied by real-time RT-PCR. All had posi-
were reached for all cell lines. tive ABL quantification, with amounts of transcript rang-
ing from 38 to 93,544 (median 4,217). The median bcr-
ta

Reliability of real-time RT-PCR

Quantitative analysis of bcr-abl in 20 identical sam- abl/ABL-Ia ratio at diagnosis (34 patients, 36 samples)
ra

ples of 500 molecules of plasmids pTA-e13/a2 (similar was 15.334 (range 3.361-28.810) (Table 3a). At clinical
results also being obtained with pTA-e1/a2 and with remission, 7 patients were also in cytogenetic remission
r

(UPN164, UPN194, UPN249, UPN262, UPN274, UPN292,

Fe

pTA-e14/a2) in one run (intra-assay comparison) result-

ed in a coefficient of variation (CV, calculated for the UPN301; 8 BM samples) and 7 patients were mostly with
determined concentrations) of 0.18; the respective cycle minimal residual disease by cytogenetic analysis
©

threshold crossing points resulted in a CV of r=0.02. (UPN141, UPN190, UPN199, UPN214, UPN267, UPN285,
Analysis of 20 identical samples in 20 runs on 20 days UPN288; 21 BM samples) (Table 3b and c). The median

Table 4 (a), (b), (c). Molecular and cytogenetic evaluation of clinical samples of CML patients.

CML
Cases studied Phase of disease Ph+% Quantitative competitive RT-PCR Quantitative real-time RT-PCR Quantitative real-time RT-PCR Qualitative RT-PCR
analysis for bcr-abl analysis for bcr-abl analysis for bcr-abl/ABL-Ia ratios analysis for bcr-abl

(a) Cases from 1 to 34 Chronic phase 1 100 135,418 157,690 median 15,334 + all (positive)
and diagnosis (range 178,278-48,938) (range 4,898- 1,778,278) (range 3.36-28.81)

This table shows comparison between quantitative competitive and real-time quantification RT-PCR methods for bcr-abl: All samples from CML patients were studied at diagno-
sis (group a). Either quantitative competitive (as described in ref. #15) or real-time RT-PCR analyses for bcr-abl were expressed as bcr-abl transcript for 1mg of total RNA. Bcr-
abl/Ia ratios are expressed as described in Design and Methods; qualitative PCR analysis was considered positive (+) or negative (-).

haematologica vol. 86(3):March 2001

 Real-time quantification of bcr-abl transcripts in CML 257

bcr-abl/ABL-Ia ratio at clinical remission on 25 samples (e19a2).48 A possible limitation of our system could be
was 0.9 (range 0.003-26.1) showing a decrease of the an inability to recognize and quantify some very rare
transcript from diagnosis to CR. Fourteen samples from bcr-abl transcripts that lack ABL exon 2 (b2a3 or
ten patients (UPN164, UPN194, UPN249, UPN262, b3a3).24,25,49
UPN274, UPN292, UPN301, UPN141, UPN267 and The sensitivity of our single step real-time RT-PCR
UPN285) (Table 3b and c) showed complete karyotypic technique turned out to be almost as high as standard
remission: the median value of bcr-abl/ABL-Ia ratio was nested RT-PCR, and significantly higher than qualitative
0.7 (range 0.003-2.83) (Figure 2). A highly significant single step RT-PCR. To standardize bcr-abl mRNA levels
correlation was seen between the proportion of Ph+ with respect to variability in RNA and cDNA quality, we
metaphases determined by cytogenetics and the bcr- used ABL-Ia transcripts as an internal control. Thus,
abl/ABL ratios determined by the Taqman (p < 0.01). amplification of ABL-Ia cDNA sequences between exons
Ia and a2 was done to produce appropriate control
Real-time RT-PCR is more sensitive than
sequences expressed at similar levels as the target genes
qualitative RT-PCR
and to avoid non-specific cDNA amplification. The
Real-time RT-PCR revealed the presence of bcr-abl tran-
advantage of using ABL-Ia as a control gene is that it
scripts in all 14 patients in CR who turned out to have
can be quantified by the same probe and reverse primer
MRD, whereas only 5 (UPN141, UPN190, UPN199,
as bcr-abl. However, we also tested the variability of
UPN214 and UPN288) were found to be positive at qual-
the ABL quantification by comparing ABL-Ia with β2-

n
itative RT-PCR (Table 4): this suggests that real-time
microglobulin and GAPDH transcript levels in samples of
RT–PCR is highly specific, and indicates that it is signifi-

tio
different qualities and with different levels of bcr-abl
cantly more sensitive than qualitative RT-PCR (p =0.001).
(data not shown). The levels of these three housekeep-

da
ing genes correlated significantly, and similar results
Discussion
were obtained with bcr-abl/ABL, bcr-abl/β2 microglob-
In CML, cytogenetic analysis is still the standard tech-
ulin and bcr-abl/GAPDH ratios with respect to percent-
nique for assessing the proportion of malignant BM cells
un
ages of Ph+ cells at cytogenetic analysis. These findings
after therapy and defining patients’ response to treat-
suggest that any of these three genes may be used as
Fo
ment. A drawback of this technique is that BM samples
an internal standard. However, both β2-microglobulin
are required in order to obtain metaphases. We15, 28-29
and GAPDH level of expression turned out to be 5 to 10
and others30 have previously demonstrated the advan-
times higher than that of ABL-Ia, and we are currently
ti

tages of quantitative RT-PCR techniques for monitoring

investigating whether these differences could be rele-
or

CML patients during or after treatment. Results from

vant in patients who show clinical and cytogenetic
competitive RT-PCR correlate well with cytogenetic
responses.
response in patients treated with α-IFN31,32 or with allo-
St

We tested the reliability of our system by comparing

transplantation.28,29 Variable numbers of bcr-abl tran-
it with an established method of quantification, name-
scripts persist in CML patients who achieve CR with α-
ta

ly competitive RT-PCR followed by CE.28 The two meth-

IFN,29 and levels of detectable MRD may vary by as much
ods gave virtually identical results. The correlation with
as two29 or even four orders of magnitude.29,33 After allo-
ra

cytogenetic analysis was also very good. In the future,

transplantation, rising or persistently high levels of bcr-
the possibility of using peripheral blood samples for real-
r

abl mRNA can be detected prior to cytogenetic or hema-

time RT-PCR27 instead of the BM samples required for
Fe

tologic relapse.29,34-38
cytogenetic analysis could enable molecular monitoring
Quantitative data from competitive RT-PCR have been
to be done on a regular basis with a minimum of patient
used to initiate donor lymphocyte transfusions for treat-
©

discomfort.
ment after relapse, and to monitor response to thera-
We conclude that the real-time RT-PCR technique
py.39,40 Real-time RT-PCR allows direct measurement of
described herein for quantification of bcr-abl is a rapid,
the amount of RT-PCR product during the amplification
sensitive and reliable method for monitoring CML
process and provides a far more accurate indication of
patients before and after therapy. The method offers the
the initial number of targets. Following other early
opportunity to standardize the assay and to develop rig-
reports of real-time RT-PCR for the detection and quan-
orous standards and controls. It could provide an effec-
tification of bcr-abl transcripts using the TaqMan sys-
tive and convenient substitute for cytogenetic monitor-
tem with a double fluorescence labeled probe,41-44 for
ing. It is likely that real-time PCR will enable the quan-
the present study we set up a novel primer/probe com-
titative analysis of MRD to become more widely avail-
bination to provide a similar system bearing some par-
able. This could guide therapeutic decision-making in
ticular characteristics. We tested our technique on 65
CML patients treated either with allotransplantation or
BM samples taken from 48 CML patients. The use of a
α-IFN.
probe matching ABL exon 2 sequences in combination
with appropriate primers allows the detection of the Contributions and Acknowledgments
two most common bcr-abl transcripts,2,13 but could also MA was the principal investigator: she designed the
be effectively applied in other rarer types of bcr-abl study. BG set up PCR procedures and drafted the paper. NT
transcripts such as e1a2,27 e6a2,45 e8-int-a2,46,47 or c3a2 was responsible for cytogenetic and molecular analyses.

haematologica vol. 86(3):March 2001

258 M. Amabile et al.

VM, GR, CZ, CT, SB, EO, SB, AdV, ET, GS, FP, MB were 5. Saglio G, Pane F, Martinelli G, Guerrasio A. BCR/ABL tran-
responsible for clinical management of patients. ST gave scripts and leukemia phenotype: an unsolved puzzle. Leuk
the final approval for submission. GM was responsible for Lymphoma 1997; 26:281-6.
6. Chronic myeloid leukemia, BCR/ABL transcript, response
ethical approval of the program, for funding and direct to α-interferon and survival. The Italian Cooperative
supervision. The order of authorship reflects the contri- Study Group on chronic myeloid leukemia. Leukemia
bution given to the study. 1995; 9:1648-52.
The authors are extremely grateful to the other mem- 7. Holland PM, Abramson RD, Watson R, Gelfand DH. Detec-
tion of specific polymerase chain reaction product by uti-
bers of the “RQ-PCR European Network in MRD” for lizing the 5’--3’ exonuclease activity of Thermus aquati-
divulging their invaluable technical know-how on real- cus DNA polymerase. Proc Natl Acad Sci USA 1991; 88:
time quantification, and for ongoing exchanges of per- 7276-80.
sonal experiences. We thank Robin MT Cooke for editing. 8. Eder M, Battmer K, Kafert S, Stucki A, Ganser A, Herten-
stein B. Monitoring of BCR-ABL expression using real-
Funding time RT-PCR in CML after bone marrow or peripheral
This study was supported by the Associazione Italiana blood stem cell transplantation. Leukemia 1999; 13:
per la Ricerca sul Cancro (A.I.R.C.), by MURST 40% (Sante 1383-9.
9. Monitoring treatment and survival in chronic myeloid
Tura) (AML), by A.I.L., by the Italian 98.00526.CT04 C.N.R. leukemia. Italian Cooperative Study Group on Chronic
target projects, by “30 Ore per la Vita” A.I.L. grants, by Myeloid Leukemia and Italian Group for Bone Marrow
CML Cofin 99 (M. Fiacchini, and G Saglio) funds. Valuable Transplantation. J Clin Oncol 1999; 17:1858-68.
experience was gained during the meetings of the “RQ- 10. Long-term follow-up of the Italian trial of interferon-α

n
versus conventional chemotherapy in chronic myeloid
PCR European Network in MRD Detection in Leukemia”

tio
leukemia. The Italian Cooperative Study Group on Chron-
(http://194.214.97.12/Scripts/defaultIE.asp). ic Myeloid Leukemia. Blood 1998; 92:1541-8.
11. Interferon α-2a as compared with conventional
Disclosures

da
chemotherapy for the treatment of chronic myeloid
Conflict of interest: none. leukemia. The Italian Cooperative Study Group on Chron-
Redundant publications: no substantial overlapping ic Myeloid Leukemia. N Engl J Med 1994; 330:820-5.
un
with previous papers. 12. Barbany G, Hagberg A, Olsson-Stromberg U, Simonsson
B, Syvanen AC, Landegren U. Manifold-assisted reverse
Manuscript processing transcription-PCR with real-time detection for measure-
Fo

This manuscript was peer-reviewed by two external- ment of the BCR-ABL fusion transcript in chronic myeloid
leukemia patients. Clin Chem 2000; 46:913-20.
referees and by Prof. Mario Cazzola, who acted as an 13. Martinelli G, Testoni N, Montefusco V, et al. Detection of
Associate Editor. The final decision to accept the paper
ti

bcr-abl transcript in chronic myelogenous leukemia

was taken jointly by the Editors. Manuscript received patients by reverse-transcription-polymerase chain reac-
or

November 2, 2000; accepted January 31, 2001. tion and capillary electrophoresis. Haematologica 1998;
83:593-601.
St

14. Martinelli G, Montefusco V, Amabile M, et al. Quantita-

Potential implications for clinical practice tive evaluation of BCR-ABL amount of transcript post
Real-time RT-PCR - such as the TaqMan assay set up mobilization with G-CSF of peripheral blood stem cells
ta

from chronic myeloid leukemia patients in cytogenetic

by us - for quantification of bcr-abl is a rapid, sen- response. Leuk Lymphoma 2000; 39:113-20.
sitive and reliable method for monitoring CML
ra

15. Testoni N, Martinelli G, Farabegoli P, et al. A new method

patients before and after therapy (allotransplanta- of “in-cell reverse transcriptase-polymerase chain reac-
tion, α-IFN etc.), and should provide a major advance tion” for the detection of BCR-ABL transcript in chron-
r
Fe

on cytogenetics in routine practice. Use of peripher- ic myeloid leukemia patients. Blood 1996; 87:3822-7.
16. Russo D, Marin L, Bertone A, Tiribelli M, Testoni N, Mar-
al blood samples for real-time RT-PCR would spare tinelli G. Pilot study of combined therapy with interfer-
patients the discomfort of the bone marrow biopsies
©

on-α, arabinosyl cytosine and all-trans retinoic acid in

necessary for cytogenetic analysis, and would facil- patients with chronic myeloid leukemia in the chronic
itate more regular and frequent monitoring. phase. Haematologica 1999; 84:185-7.
17. Frassoni F, Martinelli G, Saglio G, et al. Bcr/abl chimeric
transcript in patients in remission after marrow trans-
plantation for chronic myeloid leukemia: higher fre-
References quency of detection and slower clearance in patients
grafted in advanced disease as compared to patients
1. Sawyers CL. Chronic myeloid leukemia. N Engl J Med grafted in chronic phase. Bone Marrow Transplant 1995;
1999; 340:1330-40. 16:595-601.
2. Chissoe SL, Bodenteich A, Wang YF, et al. Sequence and 18. Martinelli G, Ottaviani E, Testoni N, Visani G, Pagliani G,
analysis of the human ABL gene, the BCR gene, and Tura S. Molecular analysis of granulocytic sarcoma: a
regions involved in the Philadelphia chromosomal single center experience. Haematologica 1999; 84:380-
translocation. Genomics 1995; 27:67-82. 2.
3. Martinelli G, Saglio G, Baccarani M, et al. The use of 19. Hochhaus A, Weisser A, La Rosee P, et al. Detection and
quantitative RT-PCR for bcr-abl in the clinical manage- quantification of residual disease in chronic myeloge-
ment of CML. Haematologica 1998; 83:18-25. nous leukemia. Leukemia 2000; 14:998-1005.
4. Meloni G, Russo D, Baccarani M, et al. A prospective study 20. Marcucci G, Livak KJ, Bi W, Strout MP, Bloomfield CD,
of α-interferon and autologous bone marrow transplan- Caligiuri MA. Detection of minimal residual disease in
tation in chronic myeloid leukemia. The Italian Cooper- patients with AML1/ETO-associated acute myeloid
ative Study Group on Chronic Myeloid Leukaemia. leukemia using a novel quantitative reverse transcrip-
Haematologica 1999; 84:707-15. tion polymerase chain reaction assay. Leukemia 1998;

haematologica vol. 86(3):March 2001

 Real-time quantification of bcr-abl transcripts in CML 259

12:1482-9. Bone Marrow Transplant 1996; 18:1147-52.

21. Cross NC, Feng L, Chase A, Bungey J, Hughes TP, Goldman 36. Lin F, van Rhee F, Goldman JM, Cross NC. Kinetics of
JM. Competitive polymerase chain reaction to estimate increasing BCR-ABL transcript numbers in chronic
the number of BCR-ABL transcripts in chronic myeloid myeloid leukemia patients who relapse after bone mar-
leukemia patients after bone marrow transplantation. row transplantation. Blood 1996; 87:4473-8.
Blood 1993; 82:1929-36. 37. Gaiger A, Lion T, Kalhs P, et al. Frequent detection of
22. Kaplan EL, Meyer P. Non parametric estimation for BCR-ABL specific mRNA in patients with chronic myeloid
incomplete observation. J Am Stat Assoc 1958; 53:457. leukemia (CML) following allogeneic and syngeneic bone
23. van Dongen JJ, Macintyre EA, Gabert JA, at al. Standard- marrow transplantation (BMT). Leukemia 1993; 7: 1766-
ized RT-PCR analysis of fusion gene transcripts from 72.
chromosome aberrations in acute leukemia for detection 38. Lion T, Henn T, Gaiger A, Kalhs P, Gadner H. Early detec-
of minimal residual disease. Report of the BIOMED-1 tion of relapse after bone marrow transplantation in
Concerted Action: investigation of minimal residual dis- patients with chronic myelogenous leukaemia. Lancet
ease in acute leukemia. Leukemia 1999; 13:1901-28. 1993; 341:275-6.
24. Martinelli G, Amabile M, Terragna C, et al. Concomitant 39. van Rhee F, Lin F, Cullis JO, et al. Relapse of chronic
expression of the rare E1/A3 and B2/A3 types of BCR/ABL myeloid leukemia after allogeneic bone marrow trans-
transcript in a chronic myeloid leukemia (CML) patient. plant: the case for giving donor leukocyte transfusions
Leukemia 1999; 13:1463-4. before the onset of hematologic relapse. Blood 1994; 83:
25. Amabile M, Martinelli G, Terragna C, Montefusco V, 3377-83.
Tabilio A, Tura S. An atypical (b3/a3) junction of the 40. Raanani P, Dazzi F, Sohal J, et al. The rate and kinetics of
bcr/abl gene lacking abl exon a2 in a patient with chron- molecular response to donor leucocyte transfusions in
ic myeloid leukemia. Haematologica 1999; 84:573-5.

n
chronic myeloid leukaemia patients treated for relapse
26. Visani G, Martinelli G, Piccaluga P, et al. α-interferon after allogeneic bone marrow transplantation. Br J

tio
improves survival and remission duration in P-190BCR- Haematol 1997; 99:945-50.
ABL positive adult acute lymphoblastic leukemia. 41. Gibson UE, Heid CA, Williams PM. A novel method for real
Leukemia 2000; 14:22-7.

da
time quantitative RT-PCR. Genome Res 1996; 6:995-
27. Saglio G, Pane F, Gottardi E, et al. Consistent amounts of 1001.
acute leukemia-associated P190BCR/ABL transcripts are 42. Heid CA, Stevens J, Livak KJ, Williams PM. Real time
expressed by chronic myelogenous leukemia patients at
diagnosis. Blood 1996; 87:1075-80.
un
quantitative PCR. Genome Res 1996; 6:986-94.
43. Mensink E, van de Locht A, Schattenberg A, et al. Quan-
28. Martinelli G, Montefusco V, Testoni N, et al. Clinical val-
ue of quantitative long-term assessment of bcr-abl titation of minimal residual disease in Philadelphia chro-
Fo

chimeric transcript in chronic myelogenous leukemia mosome positive chronic myeloid leukaemia patients
patients after allogeneic bone marrow transplantation. using real-time quantitative RT-PCR. Br J Haematol
Haematologica 2000; 85:653-8. 1998; 102:768-74.
44. Preudhomme C, Révillion F, Merlat A, et al. Detection of
ti

29. Martinelli G, Testoni N, Amabile M, et al. Quantification

BCR-ABL transcripts in chronic myeloid leukemia (CML)
or

of BCR-ABL transcripts in CML patients in cytogenetic

remission after interferon-α-based therapy. Bone Mar- using a 'real time' quantitative RT-PCR assay. Leukemia
row Transplant 2000; 25:729-36. 1999; 13:957-64.
St

30. Cross NC. Assessing residual leukaemia. Baillieres Clin 45. Hochhaus A, Reiter A, Skladny H, et al. A novel BCR-ABL
Haematol 1997; 10:389-403. fusion gene (e6a2) in a patient with Philadelphia chro-
31. Goldman JM, Kaeda JS, Cross NC, Hochhaus A, Hehlmann mosome-negative chronic myelogenous leukemia. Blood
ta

R. Clinical decision making in chronic myeloid leukemia 1996; 88:2236-40.

based on polymerase chain reaction analysis of minimal 46. Martinelli G, Terragna C, Amabile M, et al. Alu and
ra

residual disease. Blood 1999; 94:1484-6. translisin recognition site sequences flank translocation
32. Hochhaus A, Lin F, Reiter A, et al. Quantification of resid- sites in a novel type of chimeric BCR-ABL transcript and
r

ual disease in chronic myelogenous leukemia patients on suggest a possible general mechanism for BCR-ABL
Fe

interferon-α therapy by competitive polymerase chain breakpoints. Haematologica 2000; 85:40-6.

reaction. Blood 1996; 87:1549-55. 47. Martinelli G , Terragna C, Amabile M, et al. Translisin
33. Hochhaus A, Lin F, Reiter A, et al. Variable numbers of recognition site sequences flank translocation break-
©

BCR-ABL transcripts persist in CML patients who achi- points in a Philadelphia chromosome positive chronic
eve complete cytogenetic remission with interferon-α. Br myeloid leukemia patient expressing a novel type of
J Haematol 1995; 91:126-31. chimeric BCR-ABL transcript (E8-INT-A2). Leukemia
34. Martinelli G, Terragna C, Lemoli RM, et al. Clinical and 1999; 13:1635-7.
molecular follow-up by amplification of the CDR-III 48. Pane F, Frigeri F, Sindona M, et al. Neutrophilic-chronic
region in IgH region in multiple myeloma patients after myeloid leukemia: a distinct disease with a specific mark-
autologous transplantation of hematopoietic CD34+ stem er. Blood 1996; 88:2410-4.
cells. Haematologica 1999; 84:397-404. 49. van der Plas DC, Soekarman D, van Gent AM, Grosveld G,
35. Lin F, Kirkland MA, van Rhee FV, et al. Molecular analy- Hagemeijer A. bcr-abl mRNA lacking abl exon a2 detect-
sis of transient cytogenetic relapse after allogeneic bone ed by polymerase chain reaction in a chronic myeloge-
marrow transplantation for chronic myeloid leukaemia. neous leukemia patient. Leukemia 1991; 5:457-61.

haematologica vol. 86(3):March 2001