British Journal of Haematology, 2001, 114, 349±357
Cytogenetic responses in high-risk myelodysplastic syndrome
following low-dose treatment with the DNA methylation
inhibitor 5-aza-2
0
-deoxycytidine
Michael LuÈbbert,
1
Pierre Wijermans,
2
Regina Kunzmann,
1
Gregor Verhoef,
3
Andre Bosly,
4
Christophe Ravoet,
5
Marc Andre
6
and Augustin Ferrant
7 1
Department of Medicine,
Division of Haematology/Oncology, University of Freiburg Medical Centre, Freiburg, Germany,
2
Department of
Haematology, Leyenburg Hospital, The Hague, The Netherlands,
3
Gasthuisberg Hospital, Leuven,
4
Mont-Godinne UCL
University Hospital, Yvoir,
5
Jolimont Hospital, Haine Saint Paul,
6
Centre Hospitalier Notre-Dame et Reine Fabiola,
Charleroi, and
7
Cliniques Universitaires Saint-Luc, Brussels, Belgium
Received 4 January 2001; accepted for publication 6 April 2001
Summary. Decitabine (5-aza-2
0
-deoxycytidine) acts as a
powerful demethylating agent in vitro. Clinically, low-dose
decitabine ameliorates cytopenias including induction of
trilineage responses in <50% of patients with high-risk
myelodysplastic syndrome (MDS). We examined the inci-
dence and kinetics of cytogenetic responses to decitabine in
these patients. Of 115 successfully karyotyped patients, 61
(53%) had clonal chromosomal abnormalities prior to
treatment. Major cytogenetic responses were observed in
19 patients (31% of those with abnormal cytogenetics, 17%
of all patients by intention-to-treat) after a median of three
courses (range, 2±6) until best cytogenetic response.
Progressive decrease of the abnormal clone over time was
also determined using fluorescence in situ hybridization
(FISH) analysis in two patients. Median duration of
cytogenetic responses was 7´5 months (range, 3±15).
Analysis of response by the International Prognostic Scoring
System (IPSS) cytogenetic risk groups revealed three out of
five cytogenetic responses (60%) in the IPSS `low-risk'
group, 6 out of 30 with `intermediate risk' (20%) and 10
out of 26 in the `high-risk' group (38%). Median survival in
these cytogenetic subgroups was 30, 8 and 13 months
respectively. The relative risk of death in patients achieving a
major cytogenetic response was 0´38 (95% confidence
interval 0´17±0´88) compared with patients in whom the
cytogenetically abnormal clone persisted (P ˆ 0´0213). In
conclusion, repeated courses of low-dose decitabine induce
cytogenetic remissions in a substantial number of elderly
MDS patients with pre-existing chromosomal abnormalites;
these are associated with improved survival compared with
patients in whom the cytogenetically abnormal clone
persists. Patients with `high-risk' chromosomal abnormal-
ities may particularly benefit from this treatment.
Keywords: DNA methyltransferase, decitabine, 5-azacyti-
dine, monosomy 7, remission.
The myelodysplastic syndromes (MDS) are clonal disorders
originating from an early haematopoietic progenitor cell
(Koeffler, 1996). They are most frequent in elderly patients,
with an incidence of .20 in 100 000/year in patients aged
70 years and older (Aul et al, 1992). Clinical hallmarks of
MDS are cytopenia, variable expansion of `preleukaemic'
clonal myeloblasts, and an increased risk of developing
acute myeloid leukaemia (AML) (Bennett et al, 1982). MDS
with .20% bone marrow blast excess and AML have been
proposed to constitute a pathobiological continuum rather
than distinct disorders (Ghaddar et al, 1994; Harris et al,
1999; Rossi et al 2000), and treatment options are similar
in younger patients.
However, most MDS patients, owing to age and co-
morbidity, are not eligible for intensive chemotherapy. In
these patients, best supportive care constitutes the only
generally accepted treatment standard. Combinations of
haematopoietic growth factors are effective in ameliorating
anaemia and neutropenia (Negrin et al, 1993; Ganser et al,
1996; HellstroÈm-Lindberg et al, 1998). Low-dose chemo-
therapy with cytarabine does not improve survival (Miller
et al, 1992) and differentiation-inducing agents (retinoids,
1,25 dihydroxy vitamin D3 analogues) have not resulted in
q 2001 Blackwell Science Ltd 349
Correspondence: Michael LuÈbbert, MD PhD, Department of
Medicine, Division Haematology/Oncology, University of Freiburg
Medical Centre, Hugstetter Str. 55, D-79106 Freiburg, Germany.
E-mail: luebbert@mm11.ukl.uni-freiburg.de

a proven clinical benefit thus far (Koeffler et al, 1988;
Motomura et al, 1991).
Decitabine (5-aza-2
0
-deoxycytidine) and its ribose analo-
gue 5-azacytidine are powerful inhibitors of DNA methyla-
tion and have been developed for clinical use based on their
strong antileukaemic activity in vitro and in animal models
(Pinto & Zagonel, 1993; LuÈbbert, 2000). The antiprolifera-
tive effect of decitabine may result from DNA strand breaks
(JuÈ ttermann et al, 1994), or from demethylation of genes
involved in cellular proliferation and differentiation that are
silenced by promoter hypermethylation (Baylin et al, 1998).
This was shown in cell lines and animals models (Bender
et al, 1998). Even at a total dose that is much lower than
that used in AML (Pinto et al, 1989; Schwartsmann et al,
1997; Willemze et al, 1997) and chronic myeloid leukaemia
(CML) in blast crisis (Kantarjian et al, 1997), decitabine
induced haematological responses including complete
remissions in <50% of MDS patients (Wijermans et al,
1997, 2000). The aims of the present study were to
determine (i) the ability of low-dose decitabine to suppress
the abnormal cytogenetic clone in patients with high-risk
MDS and chromosomal abnormalities, (ii) whether the type
of cytogenetic abnormality had an effect upon the
cytogenetic response to decitabine, and (iii) whether patients
achieving major cytogenetic responses had better survival
than those in whom the cytogenetic abnormality persisted
despite treatment.
PATIENTS AND METHODS
Patients. Between January 1991 and December 1999, a
total of 124 patients with MDS [refractory anaemia with
excess blasts (RAEB), RAEB in transformation (RAEB-t),
chronic myelomonocytic leukaemia (CMMoL), and transfu-
sion-dependent refractory anaemia (RA) and RA with
ringed sideroblasts (RARS)] were studied in three clinical
protocols of low-dose decitabine (Pharmachemie BV, Haar-
lem, The Netherlands) approved by local Ethics Committees.
The first protocol (phase I/II study) used continuous
infusion (c.i.) at two starting dose levels, subsequent
protocols had a single dose level with infusions of 4 h
duration, three times daily. Starting doses were 50 mg/m
2
/d
c.i. for 3 d (total dose 150 mg/m
2
, 15 patients) and 40 mg/
m
2
/d c.i. for 3 d (total dose 120 mg/m
2
, eight patients) in
the phase I/II study. Fifteen milligrams/m
2
three times daily
for 3 d (total dose 135 mg/m
2
) were administered in both
the phase II study (66 patients) and the subsequent
`compassionate-use' protocol (35 patients). Inclusion cri-
teria in all three protocols were: patients 18 years and older
(no upper age limit) with MDS of subtype RAEB, RAEB-t,
CMMoL, or transfusion-dependent RA and RARS. Thus,
younger patients in whom AML-type induction treatment
was not feasible could also be treated. In the phase I/II
study, recruited patients with progressive MDS who had
.30% bone marrow blasts at the start of treatment could
also be included. The compassionate-use programme
allowed previous treatment with decitabine. Courses were
repeated every 6±8 weeks for up to seven courses (c.i.
schedule) and every 6 weeks up to six courses (t.i.d.
schedule) respectively. No dose reduction was allowed.
Clinical results of both studies have been published (Wijer-
mans et al, 1997, 2000).
Cytogenetic analyses. Bone marrow samples were obtained
prior to treatment and karyotype analyses were performed
on bone marrow cells from 122 out of 124 patients (98%)
after short-term cultivation (24±72 h). If possible, 20±30
metaphases were analysed. Chromosome aberrations were
designated according to the recommendations of the
International System for Human Cytogenetics Nomencla-
ture (Mitelman, 1991). An abnormal clone was defined by
the presence of at least two cells with the same structural
rearrangement or extra chromosome, or at least three cells
with the same missing chromosome. Cytogenetic responses
were defined according to those used in CML treatment with
interferon-alpha (Talpaz et al, 1991): complete response, 0%
abnormal metaphases; partial response, 1±35% abnormal
metaphases; minor response, 36±95% abnormal meta-
phases; no response: 96±100% abnormal metaphases.
Complete and partial responses were scored as major
response. When karyotyping did not yield sufficient mitoses,
this was considered as no cytogenetic response.
Fluorescence in situ hybridization (FISH) analyses. Loss of
chromosome 7 was detected by FISH analysis using a
centromere-specific probe (D7Z) and a probe for band q31 of
chromosome 7 (Vysis, Stuttgart, Germany). The interstitial
deletion of the long arm of chromosome 5 was detected
using a probe for epidermal growth receptor 1 (EGR-1)
located on the long arm of chromosome 5, band q31 (Vysis).
Hybridizations were performed using standard methods. The
cut-off for background signals of both probes was 1%.
Statistical analysis. The effect of cytogenetic response upon
survival from start of treatment was analysed using the
Mantel±Byar test which stratifies to either group at the time
of best cytogenetic response (Mantel & Byar, 1974) and
plotted according to Simon and Makuch (1984). Occur-
rence of cytogenetic responses was analysed as a time-
dependent covariate in a Cox regression model. Survival
probabilities in the different cytogenetic risk groups were
estimated using log-rank statistics (Kaplan & Meier, 1958).
Evaluation was as of January 1, 2000. Cytogenetic response
and survival analyses were performed on all 61 patients
with clonal chromosomal abnormalities (intention-to-treat),
whereas the 54 patients without the informative, clonal
cytogenetic markers were excluded.
RESULTS
Chromosomal abnormalities of MDS patients treated with
decitabine
Of 124 patients with high-risk MDS (Table I) treated at
seven study centres, 115 were successfully karyotyped prior
to treatment with low-dose decitabine (Table II). Treatment
schedules were similar with respect to total dose, timing of
administration and responses. As previously reported,
<50% of haematological responses including complete
responses were observed (Wijermans et al, 1997, 2000),
with late myelotoxicity as the major toxicity, but allowing
for outpatient management after administration of the drug.
350 M. LuÈbbert et al
q 2001 Blackwell Science Ltd, British Journal of Haematology 114: 349±357