1. No category

From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
1991 77: 331-339
Use of polymerase chain reactions to monitor minimal residual
disease in acute lymphoblastic leukemia patients
S Yokota, TE Hansen-Hagge, WD Ludwig, A Reiter, A Raghavachar, E Kleihauer and CR Bartram
Updated information and services can be found at:
http://www.bloodjournal.org/content/77/2/331.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
Use of Polymerase Chain Reactions to Monitor Minimal Residual Disease in
Acute Ly m ph ob 1a s tic Leu ke mi a Patients
By Shouhei Yokota, Thomas E. Hansen-Hagge, Wolf-Dieter Ludwig, Alfred Reiter, Anand Raghavachar,
Enno Kleihauer, and Claus R. Bartram
T-cell receptor (TCR) 6 gene rearrangements are observed in
more than 80% of acute lymphoblastic leukemia (ALL) patients. Moreover, a preferential usage of specific genetic
elements has been shown in different ALL subtypes: VS,DJ6,
rearrangements predominante in T-ALL, while most B-precursor ALLs show a recombination of VS, t o D6,. Recently we
have proposed a strategy for the detection of minimal
residual disease (MRD) based on the isolation of clonospecific probes following the in vitro amplification of V6,DJ6,
junctions by polymerase chain reaction (PCR) and now have
adapted this method t o the preparation of specific V6,D6,
fragments. In the present study, clonospecific probes were
generated from 11 T-ALL and 16 CALLpatients (21children, 6
adults). The sensitivity of these 27 probes in detecting
residual leukemia cells varied between lo-‘ t o lo-‘ as determined by semiquantitative evaluation of dilution experiments. PCR analysis of 55 bone marrow (BM) and peripheral
blood (PB) samples obtained from the 27 patients during
complete clinical remission showed the following results: (1)
Evidence for MRD was obtained in the BM of all patients
(eight of eight) investigated 2 t o 6 months after remission
induction and also in 6 of 11 cases on maintenance therapy 7
t o 19 months after diagnosis. (2)In contrast, all patients but
one (10 of 11) analyzed 6 t o 41 months after the termination
of treatment lacked apparent evidence for leukemia DNA;
residual cells
the exception was a girl exhibiting IO-‘ t o
in her PB 5.5 years after diagnosis. (3)Longitudinal analysis in
nine patients disclosed marked individual differences in the
intervals between achievement of clinical remission and
complete eradication of the leukemia cell clone. (4)Differences in the duration of MRD were not associated with
distinct clinical-hematologic features. (5) Detection of residual disease by PCR proceeded clinical relapse in t w o cases.
0 1991by The American Society of Hematology.
T
children, ALL/NHL-BFM (Berlin Frankfurt Miinster) 1983 and
1986, and adults, ALUAUL-BMFT (Bundesministerium fur Forschung und Technologie) February 1984 to March 1987, for the
following reasons: (1) Leukemia cells showed a TCRG gene
recombination accessible to PCR analyses. (2) Sufficient cryopreserved and/or fresh cell material was available of presentation and
remission samples. (3) Clinical follow-up was known. The design of
the studies has been described el~ewhere,’~.’~
comprising a total
duration of treatment of 104 weeks in the BFM trials (remission
induction [weeks 0 to 111, reinduction [weeks 23 to 291, maintenance therapy [weeks 13 to 21 and 31 to 1041) or 130 weeks
according to the BMFT protocols (remission induction [weeks 0 to
81, consolidation [weeks 20 to 261, maintenance therapy [weeks 10
to 18 and 20 to 1301). ALL had been diagnosed in all patients by
local and central reviews of cytologic and cytochemical features.
Immunologic markers were analyzed by one of us (W-D.L., Berlin)
by a terminal deoxynucleotidyl transferase (TdT) assay and a broad
panel of murine monoclonal antibodies (MoAbs) following criteria
described else~here.’~*’’
Based on these phenotypic analyses,
cryopreserved cell samples of the patients were sent for Southern
blot analyses to Ulm. Determination of the patients’ immunophenotype and genotype was performed with informed consent.
HE QUANTITY AND KINETIC behavior of minimal
residual disease (MRD) in leukemia patients following successful remission induction therapy poses a major
problem of clinical oncology. Better knowledge of these
issues would certainly be helpful in terms of evaluating the
efficacy of treatment regimen, monitoring individual responses of patients, early detection of impending relapses,
and quantification of remaining leukemic cells in autologous bone marrow (BM) grafts before transplantation. The
possibilities and limitations of methods for the detection of
minimal residual leukemia depend on their specificity,
sensitivity, and reproducibility. Apart from many methods
with a detection limit comparable with cytomorphologic
techniques (1 to 5 malignant cells between 100 normal
cells), double-color immunofluorescence analysis has
emerged as a reliable tool for the identification of as few as
neoplastic cells in acute leukemias characterized by
phenotypic features that are extremely rare or absent on
normal hematopoietic
More recently, the application of techniques based on the enzymatic amplification of
DNA target sequences by polymerase chain reactions
(PCRS)~
has attracted much attention, because these strategies allow the detection of minimal residual leukemia cells
at a
to
level. An approach developed in our
laboratory4 takes its advantage from the observation that
the majority of ALL cases shows a T-cell receptor (TCR) 6
gene rearrangement and is characterized by a preferential
use of specific genetic elements depending on the immunologic subtype. Thus, V6, and J6, segments are frequently
rearranged in T-cell acute lymphoblastic leukemia (TALL) while V6, is predominantly used in B-cell precursor
ALL.”13 In this study we have amplified and isolated the
TCRG junctional regions of 27 ALLs and used them as
clonospecific probes in PCR studies of the patients’ remission samples.
PATIENTS AND METHODS
Patients. Cases included in the present study were selected
from patients enrolled in the German multicenter ALL studies for
Blood, Vol77, No 2 (January 151,1991: pp331-339
From the Section of Molecular Biology, Department of Pediam’csII,
Department of Internal Medicine III, University of Ulm; Department
of Hematology and Oncology, Minikum Steglitz, Berlin; and Department of Pediahics f l Hannover Medical School, Germany.
Submitted July 23,1990; accepted September 18, 1990.
Supported by grants from the Deutsche Forschungsgemeinschafi,
Deutsche hjebshilfe and Forderkreis fur tumor- und leukdmiekranke
Kinder Ulm. S.Y. is a recipient of a fellowship from the Alexander-voni
Humboldt-Stifiung.
Address reprint requests to Claus R. Bartram, MD, Section of
Molecular Biology, Department of Pediatrics Il, Prittwitzstrasse 43,
D- 7900 Ulm, Germany.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C.section I734 solely to
indicate this fact.
0 1991 by The American Society of Hematology.
0006-4971/91/7702-0OO7$3.00I0
33 1
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
332
YOKOTA ET AL
Clinical and laboratory data of the 27 patients are presented in
precautions: (1) The PCR processor was kept in a separate room
Table 1. Three of the 21 children with ALL did not qualify for the
away from the laboratory where DNA preparation was performed.
BFM-trial, but were treated according to the BFM protocol and
(2) No amplified samples were allowed to be brought back into the
room where cell collection, DNA preparation, and Southern blot
therefore also analyzed in the present study. The reasons for
analysis was performed. (3) All reagents used in the PCR (includexclusion were inadequate induction treatment (cases 2 and 16)
ing the oligomers) were prepared, aliquoted, and stored in an area
and leukemia relapse (4 years after initial diagnosis) in case 8.
that was free of PCR-amplified products. (4) Positive-displacement
The estimated risk for treatment failure is determined in the
pipettes were used. (5) At least one negative control was run for
BFM-trials by two parameters, the risk factor (RF) and the initial
each experiment and all samples were analyzed on at least two
prednisone response. The R F is calculated from the number of
different occasions. In addition, one should keep in mind that
leukemic blasts in the peripheral blood (PB) and the enlargement
analyses using clonospecific probes are less prone to false-positive
of liver and spleen below the costal
Patients with R F
results than PCR analyses identifying similar or identical amplificaless than 1.2, R F 1.2 to 1.7, and R F 2 1.7 are considered to be at
tion products in different patients.
standard, medium, and high risk for treatment failure, respectively.
Isolation and hybridization of clonospecificprobes. Clonospecific
Second, initial response to monotherapy with prednisone for 7 days
probes were prepared and hybridized to amplified DNA samples
emerged as a powerful prognostic parameter from the BFMaccording to a previously published protoco1,4 except for the
trials." More than 1,000 blasts/",
in PB at day 8 indicate a poor
following modifications. During the first round of PCR, 1 kg of
prognosis as compared with patients with less than 1,OOO leukemic
leukemic cell DNA was amplified with primers 1and 6 in cases with
blasts. The ALL/AUL-BMFT trials distinguish two risk groups
V6,DJ6, recombinations and with oligomers 7 and 6 in patients
(high, low). Patients exhibiting at least one of the following
exhibiting V6,D6, rearrangements (Table 2). In the second round
parameters are considered to be at high risk white blood cells
of PCR, 1% (1 WL of the primary PCR product was further
(WBC) greater than 30,000/mm3 in PB, age over 35 years, acute
amplified for 35 cycles by using primers 3 and 4 (V6,DJ6,) or 9 and
undifferentiated leukemia (AUL) immunophenotype, or failure to
10 (V6,D6,) (Table 1). In cases with V6,DJ6, recombination the
achieve complete remission (CR) within 4 weeks of induction
amplified
DNA was digested with Fokl (New England Biolabs,
therapy.M
Beverly, MA) to exclude universal J6, sequences from the clonospeSouthem blot analysis. High molecular weight DNA was precific probes.' DNA samples were separated on a 3% agarose gel
pared from cryopreserved or fresh mononuclear BM cells. After
(Ultrapure, BRL, Gaithersburg, MD) and DNA fragments containwashing in phosphate-buffered saline (PBS), cells were resusing the specific (NDN) sequences were bound to diethyl aminopended in TNE (50 mmol/L Tris-HCI, pH 7.5; 100 m m o m NaCI; 5
ethyl (DEAE) membranes (Schleicher & Schuell, Dassel, Germmol/L EDTA). Proteinase K and sodium dodecyl sulfate (SDS)
many) and purified as described?" In cases characterized by
were added to a final concentration of 10 &mL and 1% (wthol),
V6,DJ6, or V6,D6, recombinations on both alleles resulting in two
respectively. Cells were incubated overnight at 55°C. After adding
different amplification products, both fragments were isolated and
sodium perchlorate (0.5 mol/L final concentration), DNA was
used together as a single probe. Following denaturation at 95°C for
precipitated in 1 vol isopropanol and washed in 70% ethanol.
3 minutes, clonospecific probes (20 ng) were labeled for 45 minutes
DNA, 10 pg, was digested with BamHI, BglII, and Hind111
using DNA polymerase/Klenow fragment (PharmaciaLKB, Upp(Pharmacia, Uppsala, Sweden), electrophoresed on a 0.7% agasala, Sweden). The labeling reaction was primed by 200 pmol of
rose gel, and transferred onto nylon membranes (Nytran, Schleieither a V6,-specific hexanucleotide (5'-CTCTTG-3') identical to
cher & Schuell, Dassel, Germany). Following hybridization, the
the 3' end of oligomer 3, or a VG,-specific primer (5'-GTGCCT-3'),
filters were washed and exposed to XAR-5 film (Kodak, Rochester,
derived from the 3' end of V6, germline ~equences.2~
NY) using intensifying screens for 24 to 72 hours at -70°C. To
For the detection of residual leukemic cells in remission specishow TCRS rearrangements, we used the J6S16 probe? kindly
mens of the patients or in dilution series, DNA samples were
provided by Drs T. Boehm and T.H. Rabbitts (Cambridge, UK).
amplified by a two-step PCR procedure using consecutively amAll cases were reanalyzed by V6,- and V6,-specific probes generplimers 1/6 and 2/5 in cases characterized by V6,DJ6, recombinaated by PCR-directed amplification of V6, and V6, sequences in
tions or primers 7/6 and 8/11 in CALL patients initially showing a
our laboratory4 (see below).
rearranged V6,D6, allele. After the second PCR rounds, amplified
PCR. PCR was essentially performed as described by Saiki et
DNA fractions (20 ng) were spotted onto nylon filters and fixed by
al? A 100-pL reaction mixture contained 1 to 5 pg of genomic
UV illumination. Hybridization to clonospecific probes and washDNA, 30 pmol of each 5' and 3' oligonucleotide primer, 200
ing conditions, including a final high-stringency step in 0.05 X SSC
fimol/L dATP, dCTP, dGTP, and dlTP, 10 mmol/L Tris-HC1 pH
(sodium ch1oride:sodium citrate) at 68°C were performed as
8.3, 50 mmolL KCI, 3.0 mmol/L MgCI,, and 0.001% gelatine
described?
(wt/vol). Synthetic oligonucleotides were prepared using the solidphase triester method" according to published sequen~es.5'~~~~~"
RESULTS
Sequences of the primers are listed in Table 2, and their position in
T-ALL and CALL exhibit distinct TCRG rearrangements.
recombined VG,DJS, and V6,D6, segments is shown in Fig 1. Two
The backbone of the present PCR study is an immunogenorounds of amplification using nested primers were performed
according to a strategy published elsewhere?
type analysis performed prospectively in more than 500
The reaction mixture was first incubated at 92°C for 12 minutes
patients enrolled in the German multicenter ALL trials for
to denature double-stranded DNA followed by 2 minutes at 56°C
children and adults. Details of molecular genetic and
to anneal primer and template. Primer extension was started by the
immunologic data and their correlation to clinical features
addition of 1 U Taq polymerase (Amplitaq, Cetus, Norwalk, CT)
will be described elsewhere after completion of the respecand allowed to proceed for 90 seconds at 72°C. Subsequently
tive trials. In the present investigation we included 27
denaturing, annealing, and extension steps were performed at 92°C
leukemias with a rearrangement of the TCRG locus accessifor 30 seconds, 56°C for 60 seconds, and 72°C for 90 seconds,
ble to PCR analysis. Because of the limited repertoire of
respectively, for 35 cycles in an automatic PCR processor (Bio
germline elements, specific types of TCRG recombinations
Med, Theres, Germany). To minimize contamination problems
can be identified in hematopoietic neoplasias by Southern
caused by the carry-over of amplified DNA, we took the following
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
333
DETECTION OF MINIMAL RESIDUAL ALL BY PCR
Table 1. Clinical and LaboratoryData of 21 Children and 6 Adults With ALL
Initial Diagnosis
Patient
No.
Sex
Age
(y:mo)
1
M
2
Detection Limit
of Clonospecific
Cell
Probe
Source
lmmunophenotype
WBC
(mm3)
Risk
Group*
10;O
T-ALL
65.800
M.G
10-4
M
3;5
T-ALL
48.200
NQ
10-8
3
4
M
M
6;7
5;8
T-ALL
T-ALL
73.000
116.000
H, G
M,P
10-8
10-5
5
M
6;lO
T-ALL
10.400
S.G
10-4/10-5
6
7
9;6
6;l
6;l
Pre T-ALL
T-ALL
cALL
61.800
32.300
42.800
H. P
M,G
8
F
F
F
NQ
10-5
10-5
10-5
9
M
3;11
cALL
60.300
M.G
10-4/10-5
10
M
2;9
cALL
21.400
S,G
10-4
11
12
M
M
2;3
6;2
cALL
cALL
3.900
6.800
S,G
S, G
10-4
10-5
13
M
2;9
cALL
199,000
M,G
10-5
PB
BM
BM
PB
BM
BM
PB
BM
PB
PB
PB
BM
PB
BM
PB
PB
BM
BM
BM
BM
PB
BM
BM
PB
BM
BM
BM
21
F
2;3
cALL
19.100
S,G
10-5
PB
PB
BM
PB
BM
PB
BM
PB
PB
BM
BM
BM
PB
BM
PB
22
M
21;2
cALL
13.000
H
10-5/10-6
BM
23
M
24;4
cALL
4.200
L
10-4
24
25
M
M
26;6
34;O
T-ALL
T-ALL
24.600
28.000
L
L
10-4
10-5
26
27
M
M
19;7
22;O
T-ALL
T-ALL
10.600
35.300
L
H
10-5
10-5
BM
PB
BM
BM
BM
PB
BM
PB
BM
BM
PB
14
M
5;9
cALL
23.700
M.G
10-5
15
16
17
18
19
20
F
F
5;11
4;3
2;l
7;11
3;11
5;11
cALL
cALL
cALL
cALL
cALL
cALL
32.000
76.500
12.600
115.000
27.400
5.300
S,G
10-4
10-3/10-4
10-4
10-5
10-4
10-4
M
F
F
M
NO
S,G
H,G
S, G
S, G
PB
Remission Sample
Mo after Diagnosis
(termination of therapy)
4
5
11
4
4
4
5
5
12
9
12
12
20
24
10
10
5
MRD
by PCR
Clinical
Follow-up
10-3
CCR (11 mo +)
CCR (13 mo +)
CCR (12 mo +)
1. CR (8 mo), relapse
2. CR (4 mo +)
CCR (25 mo +)
1. CR (14 mo), relapse, deatht
CCR (22 mo +)
2. CR (11 mo +)S
8
6
12
15
15
14
17
17
2
13
27 (2)
40 (15)
40 (15)
7
12
45 (21)
45 (21)
44 (20)
52 (28)
53 (29)
54 (30)
55 (32)
59 (35)
59 (35)
65 (41)
CCR (13 mo +)
CCR (22 mo +)
CCR (17 mo +)
CCR (21 mo +)
CCR (44 mo +)
CCR (49 mo +)
CCR (47 mo +)
CCR (54 mo +)
CCR (54 mo +)
CCR (55mo +)
CCR (57 mo +)
CCR (60 mo +)
CCR (68 mo +)
21 (post BMT) 10-5/10-6 1. CR (70 mo), relapse, allog.
21 (post BMT) 10-4/10-5
BMT in 2. CR (23 mo t)
12
CCR (24 mo +)
Neg
20
Neg
20
Neg
6
10-3’10-4
CCR (17 mo +)
9
10-3
CCR (25 mo +)
19
Neg
10-5
19
24
Neg
24
Neg
Ne9
CCR (34 mo +)
31 (6)
Neg
CCR (37 mo +)
34 (10)
Abbreviations: L, low; H, high; S, standard; M, medium; H, high RF; G,good initial response to prednisone; P, poor initial response to prednisone;
NO, not qualified as protocol patient; BMT, BM transplantation; Neg, negative.
*According to criteria of the ALUAUL-BMFT trial or ALL-BFM study.
tAfter 2 weeks from therapy-resistant leukemia.
SCR lasted 48 months, probe derived from relapse sample.
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
YOKOTA ET AL
334
Table 2. Oligomers Used for PCRs
V6,- and V6,-specific sequences (not shown). Thus the V6,
probe
identified a rearranged 9.6-kb Hind111 fragment in
5' GTGTGTAmGTGGCCTTCA3'
the T-ALLs (and a germline configuration in CALL), while
5' ACTCAAGCCCAGTCATCAGT3'
5' GCAAAGTACTTTTGTGCTCTTG3'
the V6, probe hybridized in all respective CALLsamples to
5' GGGTTCCTTTTCCAA~GATGAG 3'
the same rearranged 7.2-kb HindIII, 18.5-kb BamHI, and
5' GAGTTACTTACTTGGTTCCAC
3'
9.5-kb BglII fragments, previously shown by JSS16. Based
5' AAATGCTAGCTATITCACCCA3'
on these Southern blot analyses we proceeded to the
5' TCATCCATCTCTCTCTCTTC3'
isolation of clonospecific probes by PCR and succeeded in
5' GAGTCATGTCAGCCATTGAG3'
all T-ALL cases and 16 of 18 CALLpatients (Table 1).
5' GCACCATCAGAGAGAGATGA3'
Clonospecific probes from T-ALLs with V6,DJ6, recombi5' UGTAGCACTGTGCGTATCC3'
nation.
The PCR strategy for the isolation of clonospecific
5' AGGGAAATGGCACTTTTGCC3'
probes recently introduced by us was initially applied to
See Fig 1 for position of oligomers; in Oligomer 4 a G (*) was
leukemias exhibiting V6,DJ6, recombinations: According
substituted for wild-type A, thus creating an artificial Fokl cleavage site
to this protocol two rounds of PCR using a set of nested
(underlined).
primers (1/6 and 3/4) are performed to amplify DNA
fragments of approximately 120 bp containing the NDNblot analysis. Accordingly, a rearrangement of V6, to D6,
functional regions, which specifically characterize the resequences is indicated after hybridization to the J6S16
arranged TCRG loci (Fig 1). This approach was successfully
probe by a pattern of aberrant 7.2-kb HindIII, 18.5-kb
applied to all 11 T-ALLs of the present study (Fig 3A).
BamHI, and 9.5-kb Bgl I1 fragments, while a 9.6-kb HindIII
Cases 5, 7, and 25 have also been included in our previous
fragment suggests a V6,DJ6, recombinati~n.~-~~'~,'~,~,~
Moreinvestigation (patients Si, Ha, and Ma, re~pectively).~
All
over, striking differences between immunologically defined
probes gave specific signals when hybridized to the leukesubgroups in the usage of TCRG elements have recently
mia cell DNA of the patient from which they were derived.
been established. Along this line we showed in 324 children
None of the 11probes showed cross-hybridization to DNA
entering the BFM-studies a TCRG rearrangement in 96%
obtained from leukemia cells of the other T-ALL patients
(58 of 60) of T-ALL and 81% (162 of 210) of CALL patients,
(not shown) or from healthy individuals (Fig 4A), conrespectively, and further specified the immunogenotype by
firming our previous experience: Because we wanted to use
Southern blot analysis as VS,DJG, rearrangement in 25% of
these probes as detectors for MRD it became essential to
the T-ALL and as V6,D6, recombination in 57% of the 162
determine their respective sensitivities. Therefore, we diCALL patients (C.R. Bartram, unpublished results). We
luted genomic DNA of each leukemia into DNA obtained
tend to use the term V6,D6, rather than V6,(D)D6,
from PB cells of healthy controls at lo-' to lo-', amplified
because all 14 CALLSinvestigated thus far in our laboratory
the DNA samples by using consecutively primers 1/6 and
by PCR-directed sequence analysis (including cases 9 and
2/5 (Fig I), and hybridized the different fractions to the
10 of this report) lacked convincing evidence for the
clonospecific probes. At least two independent dilution and
participation of D6, or D6, elements at the TCRG junction?'"
amplification series using different control DNAs were
For this study we selected 11 T-ALL (7 children, 4
performed for each patient. Assuming that one human cell
adults) and 18 CALLcases (16 children, 2 adults) charactercontains approximately 10 pg of DNA and 1 to 10 pg of
ized by a putative V6,DJ6, or V6,D6, rearrangement
DNA are used for amplification, the detection limit of this
according to the criteria discussed above. Figure 2 shows
PCR approach is theoretically
to
diluting leukethe characteristic pattern of aberrant HindIII fragments
mic DNA lo6 and amplifying 1 kg of DNA should yield
observed in 14 patients. Moreover, the interpretations
results according to the Poisson distribution, ie, statistically
derived from Southern blot analysis using the J6S16 probe
1of 10 samples will be PCR-positive. Representative results
were supported by rehybridizations of the same filters to
of these analyses are shown in Fig 4A. In fact, the detection
limit of the clonospecific probes derived from the 11T-ALL
patients varied between
to
as determined by
semiquantitative analysis (Table 1, Fig 4A). In the majority
of cases, leukemia DNA could be detected when representT l o D oJ61 -recombination
ing as little as 0.001% of total DNA, in good agreement
with our previous r e s ~ l t s . ~
L
'61
J61
.............
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Adaptation of the PCR strategy to V6@6, rearrangements.
+
+
-t
t t
t
Because V6,D6, recombinations apparently predominate in
1
2
3
4 5
6
CALL (the most frequent ALL subtype of children and
D6,-recombination
adults), we tried to modify our PCR method accordingly.
L
D63
J61
'62
Based on published TCRG sequences we designed a proto.
.
.
.
.
.
.
.
.
.
.
.
.
.............
/Ha---col outlined in Fig 1. Again, two rounds of PCR were
t t
e
+
+
+8
10 11
6
7
9
included. A first step using oligomers 1 and 6 amplifies a
rather
large fragment of approximately 1.6 kb. Because the
Fig 1. Partial diagrams of wild-type and rearranged TCRG loci.
published sequences of the area spanning between D6, and
Position of amplimers used for PCR analyses are shown by arrows
J6, contain some discrepancies, we preferred to use oligo(see Table 2).
t 2 0
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
335
DETECTION OF MINIMAL RESIDUAL ALL BY PCR
1
2
3
4
5
6
7
15 17
18
20 2l
22
23 C
kb
~ 9 . 6
Fig 2. Southem blot analysis
of the TCRG locus in seven 1-ALL
(1 to 7) and seven CALL (15 to 23)
patients (see Table 1) and a
healthy individual (C). Following
hybridization t o TCRG probe
JGS16, Hindlll-digestedDNA samples (10 pg) exhibit 6.4-kb germline fragments (G) and rearranged
fragments of 9.6 kb or 7.2 kb
indicating V&,DJ&, and VG,D6, recombinations, respectively.
c7.2
-6.4
a-
mer 6 (having served as reliable amplimer in V6,DJ6,
studies) as outer primer. Following this primary enrichment
for recombined TCRG molecules, the second PCR round
using oligomers 9 and 10 focuses on the region flanking and
including the clonospecificjunctional region and amplifies
DNA fragments of approximately 80 to 100 bp (Fig 3B).
These clonospecific fragments are isolated and labeled with
a hexanucleotide primer identical with V6, sequences
immediately upstream of the N-region. Following this
procedure we were able to prepare clonospecific probes
from 16 ALLS. The successful amplification of the V6,D6,
junctions also supported the conclusion drawn from the
initial Southern blot analyses of respective leukemias. In
two other CALL we did not obtain a distinct DNA
fragment after the second round of PCR but rather a smear
of 80- to 120-bp sized amplification products. Both leukemias were therefore excluded from further analyses.
Taking into account the possible limitation of junctional
diversity within V6,D6, segments it became crucial to
determine the specificity and sensitivity of the probes. To
this end DNA samples obtained from the CALLSand PB
cells of 30 healthy probands were amplified by using
consecutivelyprimers 716 and 8/11 (Fig 1). After the second
PCR round the amplification products representing either
monoclonal (leukemia) or polyclonal (PB control) V6,D6,
junctions had an approximate size of 300 bp. Each of the 16
clonospecific probes was then hybridized to the amplified
DNA samples obtained from at least six other CALLSand
three healthy individuals (not shown). The results of these
analyses matched remarkably well with the data obtained
for VG,DJG,-specific probes: Thus, none of the individual
V6,D6, probes showed any cross-hybridization with DNA
samples of the controls or other CALLStested. Representative results of the corresponding mixing experiments are
shown in Fig 4B. The resolution power of clonospecific
V6,D6, probes again varies from case to case and tends to
be slightly less sensitivethan V6,DV6, sequences. However,
it should be taken into account that the detection limits
A
1-118
72
-i
-
15
B
17
18 20
21
22
23
7
bp
-118
- 72
agarose gel and visualized by ethidium bromide
staining. Ha8 Ill-digested6x174 DNA was included as
size marker. (A) In T-ALLs DNA fragments of approximately 100 to 120 bp are visible (lane a). Two
fragments observed in patients 2, 3, and 6 correspond to V6,DJG, recombinations on both alleles (see
Fig 2).Foki-digested aliquots of the samples (lane b)
exhibit fragments of 30 bp (arrow) containing universal J6, sequences and one or two fragments comprislngtheclono (allele)-specificV6,DJG, junctions (shown
bv dots). (6) DNA fragments of 80 to 100 bp containing the V8,D8, junctions are visible in the CALL
samples. Patient 22 is characterizedby V&,DG, recombinations on both alleles (see Fig 2) resulting in two
amplification products.
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
336
@
YOKOTA ET AL
D
101
-2
-3
-4 -5
-6
c
-?
PB
4
BM
4
2
PB
BM
5
12
4
5
@...
pSB BM
12
.e.-
PB PB BM
12
m
24
0
0
I
BM
9
PB BM
PB BM
19
24
19
24
0
25
13
e
D 16' -2 -3 -4 -5
O O O . ' O
C
-6 -7
BM BM PB BM
2
13
27
40
PB
40
a
e...*@
I
I
PB BM PB BM
7
12
45
45
14
PBRM
21
22
PB
65
21 a
b
\
I
I
a
21
Fig 4. Detection of minimal
residual disease In four T-ALL (A)
and four CALLpatients (E). DNAs
of leukemic cells at diagnosis (D)
were diluted Into PB cell DNA of
16 healthy individuals (C) at 10.'
to lo-'. B M or PB DNA samples
obtained during the patients'
complete clinical-hematologic r e
mission were also Included(numbers indicate months after diagnosis, see Table 1). After amplification, the corresponding DNA
fractions (20 ng) were spotted
onto nylon filters and hybridized
to the clonospecific probes. In
addition, 100-ng samples of patient 21 were also run on a 1.5%
agarose gel, Southern blotted,
and hybridized (B, lane 21b). Res u b of two independent dllution and amplification series are
shown for every patient.
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
DETECTION OF MINIMAL RESIDUAL ALL BY PCR
given in Table 1 represent conservative estimations based
on the general use of very stringent washing conditions
(final step in 0.05 x SSC at 68°C). In some leukemias less
stringent washes could reduce nonspecific background
signals to a negligible level. This reduction was associated
with an approximately 10-fold increase in the detection
level. In any case, it was undoubtedly possible to identify in
all patients leukemia DNA representing 0.01% of total
DNA.
Detection of residual leukemia in remission samples. We
next used the clonospecific probes to analyze BM or PB
samples obtained from the 27 ALL patients during complete clinical and hematologic remission. Southern blot
analyses on all these samples failed to detect the TCRG
gene rearrangement initially characterizing the leukemias
and also showed a germline configuration of TCRP and JgH
loci. Following amplification of the VG,DJG,- or VG,DG,junctional region by the two-step PCR strategy described
above, DNA fractions were either directly spotted onto
nylon filters or run on an agarose gel, Southern blotted, and
subsequently hybridized to the specific probe. The readout
of both detection systems matched perfectly (Fig 4B). We
always included a dilution series in the experiments to
estimate the detection limit of the clonospecific probes
under the actual hybridization and washing conditions.
Moreover, at least two independent analyses were performed for each sample. Representative results are shown
for eight patients (Fig 4A and B) and the data of all cases
are summarized in Table 1.
One interesting result of this study is the observation that
BM samples of virtually all cases analyzed during the phase
of consolidation therapy following remission induction (ie,
2 to 6 months after diagnosis) exhibited residual leukemic
cells. In these eight patients (nos. 1 through 4,8, 9, 13, and
24) the level of residual cells varied between
to
(Table 1). The findings obtained in 11 patients studied
during maintenance therapy might be even more remarkable. BM samples of six cases (nos. 5, 6, 9, 10, 12, and 25)
still showed remaining blasts 7 to 19 months after diagnosis
at frequencies of
to
However, a different pattern
emerged from remission samples of 11 patients being off
therapy for 6 to 41 months (median 24 months). In all but
one of the 11 cases, BM (seven cases) or PB samples (four
cases) were apparently free from leukemia cells. The
exception was a girl (case 21) exhibiting
to
residual
cells 3.5 years after the termination of therapy. This result
was confirmed in two independent blood samples. Unfortunately, a BM sample was not available for comparative
analysis. In all instances, where both PB and BM specimen
of a patient were available for PCR analysis, PB samples
contained significantly less residual leukemia cells, if at all
(Table 1).
At least two consecutive remission samples could be
evaluated in nine patients (nos. 1, 4, 5, 8, 9, 13, 14, 23, and
25). These serial analyses disclosed marked individual
differences in the intervals between achievement of clinical
remission and eradication of residual disease as determined
by PCR criteria (Table 1). Cases in point are patients 8 and
25. In the former patient, leukemia cells became undetectable within 8 months after diagnosis. However, in patient 25
337
it took more than twice as long (> 19 months) before PCR
analysis showed the complete elimination of the malignant
cell clone (Fig 3B).
In this respect it appears noteworthy that the dynamic
disparities observed in the eradication of minimal residual
disease did not necessarily correlate with known risk factors
(Table 1). Thus, in some patients at standard risk, leukemia
cells persisted more than 1 year after achieving continuing
CR (CCR) (nos. 5, 10, 12, 21, and 25), while patients at an
elevated risk (nos. 8 and 14) became PCR negative in less
than 8 months. Two of the children with T-ALL experienced relapses within the observation period of this study
(nos. 4 and 6). In both patients PCR analyses had shown
residual leukemia approximately 4 months before clinical
manifestation.
DISCUSSION
We have analyzed 55 BM and PB samples from 27 ALL
patients considered to be in CR according to clinical and
laboratory criteria. This study represents to the best of our
knowledge the largest series of ALL cases investigated for
the presence of MRD by PCR technology thus far. The
investigation became feasible through a successful adaptation of our recently introduced PCR strategy based on the
amplification of clonospecific TCRG junctions to the analysis of YG,DG, recombinations in CALL. Along this line one
might envisage significant difficulties posed by the limited
junctional diversity of VG,DG3 segments as compared with
VG,DJG, junctions in T-ALL. In fact, sequence analysis of
VG,DG, regions from CALL patients recently showed complete 5' boundaries of most DG, elements and the absence
of DG, and DG, sequences; moreover, an identical VG2DG,
junction was observed in a cALL and a PB cell clone from a
healthy individ~al.~'"Taking into account these possible
shortcomings, it is remarkable that we did not encounter
diagnostic problems in any of the cases included in the
present study. Virtually all 16 cALL probes specifically
hybridized to the patients' leukemia DNA and could readily
detect
neoplastic cells. However, these promising
results by no means exclude possible pitfalls in future
studies using this approach.
In this context it appears to be appropriate to address
some other possible limitations of the PCR approach.
Theoretically, any subpopulation of leukemic cells undergoing further recombination would escape detection by PCR.
The frequency of secondary alteration at rearranged TCRG
loci during the clinical course of ALL is currently unknown.
Yet, clonal variations at rearranged Ig heavy chain loci have
Moreover,
been observed in up to 30% of ALL
PCR could amplify DNA sequences of leukemic cells (or
even cell debris) persisting in BM niches for a considerable
time, although these cells may have lost their proliferative
capacity. This possible shortcoming of PCR should be
particularly considered in samples obtained during the first
weeks after starting treatment.
A main result of our investigation is the demonstration of
residual leukemia cells in all cases analyzed up to 6 months
after successful remission induction m d frequently also in
specimens of patients being in CCR for more than 1 year.
This molecular genetic evidence for long-term persistence
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
YOKOTA ET AL
338
of neoplastic cell populations in the majority of ALL
patients might be regarded as the biologic basis of the
well-established clinical requirement for maintenance therapy. The complementing observation that long-term survivors (beyond 2 years of diagnosis) in general did not exhibit
residual leukemia makes us confident to believe that PCR
analysis is not inappropriately sensitive in addressing the
problem of MRD, but rather constitutes a valuable tool for
the identification of clinically relevant leukemia cell populations.
The present study illustrates the limited value of a single
PCR analysis. More relevant is the actual proliferative
capacity of residual leukemic cells as shown by longitudinal
study. Thus, a steady, albeit prolonged, decrease of neoplastic cells, as in case 25, may predict a favorable course. The
opposite development, ie, a continuous increase of residual
blasts to the point of clinical manifestation, was not
documented in one of our patients (only single remission
samples were available from cases 4 and 6). In this context it
will be of particular interest to follow-up the future course
of patient 21 exhibiting residual leukemia more than 3 years
after termination of therapy. The presenting features of this
leukemia did not show an increased risk; moreover, the girl
responded without any complication to chemotherapy.
Despite the fact that the patient has still not passed the
period at risk for relapse, the current probability of such an
event is well below 1%.Finally, one must also consider the
focal nature of residual disease, which might interfere with
the detection of minimal residual leukemia if based on the
analysis of a single BM specimen.”
Although the patients recruited for the present study
were all treated according to the German multicenter ALL
trials, the presenting features of the leukemias differed not
only with respect to age and immunophenotype but also as
to initial blast count and other risk factors (Table 1). This
heterogeneity must be kept in mind when interpreting the
clinical significance of the PCR data. Interestingly enough,
persistence of minimal residual leukemia for more than 1
year after diagnosis does not appear to be associated with
distinct clinical or laboratory characteristics. Monitoring of
leukemia patients by PCR analysis might therefore identify
compohents of the individual response to chemotherapy
that are not realized by currently applied parameters.
Along this line one could speculate that PCR analysis will
offer a unique tool for the evaluation of a patient’s actual
demand for maintenance therapy and ultimately might
define a rationale for a case-adapted treatment modification. However, we would like to emphasize at the same time
that the currently available PCR data are far too sparse to
justify any (far-reaching) recommendation for clinical settings. It certainly remains a challenge for future studies,
including many more patients, to define more precisely the
clinical significance, if any, of PCR analyses. In fact, it is still
an open question whether the complete eradication of
leukemic cells is really an imperative prerequisite to guarantee long-term disease-free survival and cure. Prospective
analyses of the remission status using the PCR strategies
discussed in this report have, therefore, been initiated in
the German multicenter ALL trials. Fortunately, additional
methods for MRD detection are available. With respect to
ALL these include double-color immunofluorescence’.‘and
PCR studies on TCRy” and IgH genes’* or the BCR-ABL
oncogene$ All techniques bear relevant limitations and
specific advantages. We envisage that the combined use of
several approaches will be required to investigate most
patients and to balance interpretations derived from individual methods.
After submission of this manuscript, Yamada et a1
reported on PCR analyses of BM samples from eight
B-lineage ALL patients demonstrating minimal residual
disease up to 18 months after diagn0sis.3~
ACKNOWLEDGMENT
This report is dedicated to Prof H. Heimpel on the occasion of
his 60th birthday. We thank Profs H. Heimpel, D. Hoelzer, B.
Kubanek, H. Riehm, and E. Thiel for continuous support, and
gratefully acknowledge the fruitful cooperation with the participants of the German Multicenter ALL Trials for Children and
Adults, namely Drs Bliitters-Sawatzki, Bode, Breu, Dopfer,
Gaedicke, Gerein, Gnekow, Graf, Gussetis, Havers, Jacobi, Kohne,
Mertens, Niethammer, Ritter, Sternschulte, Stollmann, Suttorp,
Treuner, Vokl, Wehinger, and Wolf. We thank U. Mehr, C. Tell,
and A. Wunderlich for expert technical assistance as well as H.
Barro and A. Jacobs for editing the manuscript.
REFERENCES
1. Van Dongen JJM, Hooijkaas H, Adriaansen HJ, Hahlen K,
Van Zanen GE: Detection of minimal residual disease acute
lymphoblastic leukemia by immunological marker analysis:
Possibilities and limitations, in Hagenbeek A, Liiwenberg B (eds):
Minimal Residual Disease in Acute Leukemia. Dordrecht, The
Netherlands, Martinus Nijhoff, 1986, p 113
2. Campana D, Coustan-Smith E, Janossy G: The immunologic
detection of residual disease in acute leukemia. Blood 76:163,1990
3. Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn
GT, Mullis KB, Erlich H A Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science
239:487,1988
4. Hansen-Hagge TE, Yokota S, Bartram CR: Detection of
minimal residual disease in acute lymphoblastic leukemia by in
vitro amplification of rearranged T-cell receptor 6 chain sequences.
Blood 741762,1989
5. Loh EY, Cwirla S, Serafini AT, Phillips JH, Lanier L L
Human T-cell receptor 6 chain: Genomic organization, diversity,
and expression in populations of cells. Proc Natl Acad Sci USA
85:9714,1988
6. Takihara Y, Reimann J, Michalopoulos E, Ciccone E, Moretta L, Mak TW:Diversity and structure of human T cell receptor 6
chain genes in peripheral blood ?/&-bearing T lymphocytes. J Exp
Med 169:393,1989
7. Hata S, Satyanarayana K, Devlin P, Band H, McLean J,
Strominger JL, Brenner MB, Krangel MS: Extensive junctional
diversity of rearranged human T cell receptor 6 genes. Science
240:1541,1988
8. Hara J, Benedict SH, Champagne E, Takihara Y, Mak TW,
Minden M: T cell receptor 6 gene rearrangements in acute
lymphoblastic leukemia. J Clin Invest 82:1974,1988
9. Griesinger F, Greenberg JM, Kersey JH: T cell receptor
From www.bloodjournal.org by guest on October 15, 2014. For personal use only.
DETECTION OF MINIMAL RESIDUAL ALL BY PCR
gamma and delta rearrangements in hematologic malignancies.
Relationship to lymphoid differentiation. J Clin Invest 84506,1989
10. Dyer MJS: T-cell receptor 6/01rearrangements in lymphoid
neoplasms. Blood 74:1073,1989
11. Asou N, Hattori T, Matsuoka M, Kawano F, Takatsuki K:
Rearrangements of T-cell antigen receptor 6 chain gene in hematological neoplasms. Blood 74:2707,1989
12. Loiseau P, Guglielmi P, Le Paslier D, MacIntyre E, Gessain
A, Bories JC, Flandrin G, Chen Z, Sigaux F Rearrangements of
the T cell receptor 6 gene in T acute lymphoblastic leukemia cells
are distinct from those occurring in B lineage acute lymphoblastic
leukemia and preferentially involve one V6 gene segment. J
Immunol142:3305,1989
13. Biondi A, di Celle PF, Rossi V, Casorati G, Matullo G,
Giudici G, Foa R, Migone N: High prevalence of T-cell receptor
V8 2-(D)-D6 3 or DS,,-8 3 rearrangements in B-precursor acute
lymphoblastic leukemias. Blood 75:1834,1990
14. Riehm H, Gadner H, Henze G, Kornhuber B, Lampert F,
Niethammer D, Reiter A, Schellong G: Results and significance of
six randomized trials in four consecutive ALL-BFM studies, in
Biichner T, Schellong G, Hiddemann W, Ritter J (eds): Acute
Leukemias 11. Prognostic Factors and Treatment Strategies. New
York, NY, Springer-Verlag. 1990, p 439
15. Hoelzer D, Thiel E, Liiffler H, Bodenstein H, Plaumann L,
Biichner T, Urbanitz D, Koch P, Heimpel H, Engelhardt R, Muller
U, Wendt FC, Sodomann H, Riihl H, Herrmann F, Kaboth W,
Dietzfelbinger H, Pralle H, Lunscken C, Hellriegel KP, Spors S,
Nowrousian RM, Fischer J, Fiille H, Mitrou PS, Pfreundschuh M,
Gorg C, Emmerich B, Queisser W, Meyer P, Labedzki L, Essers U,
Konig H, Mainzer K, Herrmann R, Messerer D, Zwingers T
Intensified therapy in acute lymphoblastic and acute undifferentiated leukemia in adults. Blood 64:38,1984
16. Ludwig WD, Bartram CR, Ritter J, Raghavachar A, Hiddemann W, Heil G, Harbott J, Seibt-Jung H, Teichmann JV, Riehm
H: Ambiguous phenotypes and genotypes in 16 children with acute
leukemia as characterized by multiparameter analysis. Blood
71:1518,1988
17. Ludwig WD, Bartram CR, Harbott J, Koller U, Haas OA,
Hansen-Hagge T, Heil G, Seibt-Jung H, Teichmann JV, Ritter J,
Knapp W, Gadner H, Thiel E, Riehm H: Phenotypic and genotypic
heterogeneity in infant acute leukemia. I. Acute lymphoblastic
leukemia. Leukemia 3:431,1989
18. Riehm H, Feickert HJ, Schrappe M, Henze G, Schellong G:
Therapy results in five ALL-BFM studies since 1970: Implications
of risk factors for prognosis, in Biichner T, Schellong G, Hiddemann W, Urbanitz D, Ritter J (eds): Acute Leukemias, New York,
NY,Springer-Verlag, 1387, p 139
19. Riehm H, Reiter A, Schrappe M, Berthold F, Dopfer R,
Gerein V, Ludwig R, Ritter J, Stollmann B, Henze G: Die
Corticosteroid-abhangige Dezimierung der Leukamiezellzahl im
Blut als Prognosefaktor bei der akuten lymphoblastischen Leukamie im Kindesalter (Therapiestudie ALL-BFM 83). Klin Padiatr
199:151,1987
20. Hoelzer D, Thiel E, Gffler H, Biichner T, Ganser A, Heil G,
Koch P, Freund M, Diedrich H, Ruhl H, Maschmeyer G, Lipp T,
Nowrousian MR, Burkert M, Gerecke D, Pralle H, Muller U,
Lunscken C, Fiille H, Ho AD, Kiichler R, Busch FW,Schneider W,
Gorg C, Emmerich B, Braumann D, Vaupel HA, von Paleske A,
Bartels H, Neiss A, Messerer D: Prognostic factors in a multicenter
339
study for treatment of acute lymphoblastic leukemia in adults.
Blood 71:123,1988
21. Boehm T, Buluwela L, Williams D, White L, Rabbitts TH: A
cluster of chromosome llp13 translocations found via distinct D-D
and D-D-J rearrangements of the human T cell receptor 8 chain
gene. EMBO J 7:2011,1988
22. Seliger H, Ballas K, Herold A, Kotschi U, Lyons J, Eisenbeiss F, Sinah ND, Talwar G P New preparative methods in
oligonucleotide chemistry and their application to gene synthesis.
Chem Scripta 26:569,1986
23. Hockett RD, de Villartay JP, Pollock K, Poplack DG, Cohen
DI, Korsmeyer SJ: Human T-cell antigen receptor (TCR) 6-chain
locus and elements responsible for its deletion are within the TCR
a-chain locus. Proc Natl Acad Sci USA 859694,1988
24. Dariavach P, Lefranc MP: First genomic sequence of the
human T-cell receptor 62 gene (TRDV2). Nucleic Acids Res
174880,1989
25. Takihara Y, Tkachuk D, Michalopoulos E, Champagne E,
Reimann J, Minden M, Mak TW:Sequence and organization of
the diversity, joining, and constant region genes of the human
T-cell &chain locus. Proc Natl Acad Sci USA 85:6079,1988
26. Danner DB: Recovery of DNA fragments from gels by
transfer to DEAE-paper in an electrophoresis chamber. Anal
Biochem 125:139,1982
27. Studencki AB, Wallace RB: Allele-specific hybridization
using oligonucleotide probes of very high specific activity; Discrimination of the human PA-and p-globin genes. DNA 3:7,1984
27a. Yokota S, Hansen-Hagge TE, Bartram CR: T-cell receptor
8 gene recombination in common acute lymphoblastic leukemia:
Preferential usage of V6, and frequent involvement of the Jcx
cluster. Blood 1991 (in press)
28. Raghavachar A, Thiel E, Bartram CR: Analyses of phenotype and genotype in acute lymphoblastic leukemias at first
presentation and in relapse. Blood 701079,1988
29. Bird J, Galili N, Link M, Stites D, Sklar J Continuing
rearrangement but absence of somatic hypermutation in immunoglobulin genes of human B cell precursor leukemia. J Exp Med
168:229,1988
30. Martens ACM, Schultz FW, Hagenbeek A: Nonhomogenous distribution of leukemia in the bone marrow during minimal
residual disease. Blood 70:1073, 1987
31. d’Auriol L, MacIntyre E, Galibert F, Sigaw F In vitro
amplification of T cell y gene rearrangements: A new tool for the
assessment of minimal residual disease in acute lymphoblastic
leukemias. Leukemia 3:155,1989
32. Yamada M, Hudson S, Tournay 0, Bittenbender S, Shane
SS, Lange B, Tsujimoto Y, Caton AJ, Rovera G: Detection of
minimal disease in hematopoietic malignancies of the B-cell
lineage by using third-complementarity-determining region (CDR111) specific probes. Proc Natl Acad Sci USA 865123,1989
33. Kawasaki ES, Clark SS, Coyne MY, Smith SD, Champlin R,
Witte ON, McCormick FP 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,
1988
34. Yamada M, Wasserman R, Lange B, Reichard BA, Womer
RB, Rovera G: Minimal residual disease in childhood B-lineage
lymphoblastic leukemia. Persistence of leukemic cells during the
first 18 months of treatment. N Engl J Med 323:488,1990