ORIGINAL ARTICLE
Predictors of outcome following myeloablative allo-SCT
for therapy-related myelodysplastic syndrome and AML
TJ Nevill
1,2,3
, DE Hogge
1,2,3
, CL Toze
1,2,3
, SH Nantel
1,2,3
, MM Power
1,2,3
, YR AbouMourad
1,2,3
,
KW Song
1,2,3
, JC Lavoie
1,2,3
, DL Forrest
1,2,3
, MJ Barnett
1,2,3
, JD Shepherd
1,2,3
, JY Nitta
2
, S Wong
2
,
HJ Sutherland
1,2,3
and CA Smith
1,2,3
1
The Leukaemia Bone Marrow Transplantation Program of British Columbia, Division of Hematology, Vancouver
General Hospital, Vancouver, British Columbia, Canada;
2
BC Cancer Agency, Vancouver, British Columbia, Canada and
3
Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada
Administration of alkylating agents (Alk), topoisomerase
II inhibitors (Topo II) and radiotherapy (RT) can result in
therapy-related myelodysplastic syndrome or acute mye-
logenous leukaemia (t-MDS/t-AML), the optimal treat-
ment for which is allo-SCT. A retrospective review was
performed of 24 patients who underwent related- or
unrelated-donor SCT for t-MDS/t-AML at our institu-
tion. Eight patients remain alive and in continuous
remission (median follow-up 54 months (range, 12–161))
with estimated 5-year EFSbeing 30% (95% confidence
intervals 16–58%). Corresponding actuarial risks
of relapse and non-relapse mortality (NRM) are 39%
(19–60%) and 30% (13–50%), respectively. EFSwas
40% in Alk/RT-related t-MDS/t-AML and 11% in Topo
II-related t-MDS/t-AML (P¼ 0.05), with an increased
risk of relapse in the latter (56 vs 29%, respectively
(P¼ 0.05)). In multivariate analysis, development of acute
GVHD (P¼ 0.009) and Topo II-related t-MDS/t-AML
(P¼ 0.018) were associated with inferior EFS. Patients
with acute GVHD had an increased risk of NRM
(P¼ 0.03) whereas risk of relapse was higher for patients
of advanced age (P¼ 0.046) and for patients who under-
went bone marrow (vs blood) SCT (P¼ 0.032). Allo-SCT
can result in long-term survival for individuals with
t-MDS/t-AML although outcome in Topo II-related
t-MDS/t-AML patients remains suboptimal.
Bone Marrow Transplantation (2008) 42, 659–666;
doi:10.1038/bmt.2008.226; published online 4 August 2008
Keywords: allogeneic; transplantation; therapy-related;
myelodysplasia; acute myelogenous leukaemia
Introduction
Patients who have previously received chemotherapy and/
or radiation treatment are predisposed to develop therapy-
related myelodysplastic syndrome or AML (t-MDS/
t-AML).
1–3
The incidence of these disorders is rising
because of improved primary malignancy outcomes, most
notably in childhood ALL and non-Hodgkin’s lymphoma.
4
In contrast to de novoMDS, BM cytogenetic abnormalities
are demonstrable in over 90% of cases of t-MDS/t-AML.
1,2
Moreover, it has now been recognized that there are two
broad categories of t-MDS/t-AML, one that occurs after
exposure to alkylating agents (Alk) or radiation (RT) and is
typically associated with abnormalities of chromosomes
5 and/or 7, and a second disorder, following treatment with
topoisomerase II inhibitors (Topo II), which is associated
with a shorter latency period, balanced chromosomal
translocations and a more favourable initial response to
chemotherapy.
5–7
CR is achieved in a minority of patients with t-MDS/t-
AML after conventional chemotherapy is given and, even
when documented, is typically brief.
8
As a result, long-term
survival is unlikely without the use of high-dose therapy
and SCT. Auto-SCT has been used in selected patients with
t-MDS/t-AML who enter CR with conventional che-
motherapy with modest success.
9,10
However, allo-SCT is
generally recommended for these individuals, provided a
suitably matched related or unrelated donor can be
identified. Initial results of allo-SCT in this patient cohort,
including our own institution’s experience, were disap-
pointing and suggested that outcomes were inferior to that
observed in de novo (primary) MDS/sAML.
11–13
It is
noteworthy that these early SCT reports frequently
grouped t-MDS/t-AML patients with those that had
AML evolving from primary MDS (‘secondary AML’ or
sAML), making interpretation of outcomes more difficult.
To accurately determine the success rate with allo-SCT for
t-MDS/t-AML and to better define risk factors for
outcome in this population, a review was performed
of patients who underwent related- and unrelated-donor
SCT for this diagnosis over an 18-year period at our
institution.
Received 15 January 2008; revised 13 May 2008; accepted 16 May 2008;
published online 4 August 2008
Correspondence: Dr TJ Nevill, Division of Hematology, Gordon &
Leslie Diamond Centre, 10th Floor-Room 10149, 2775 Laurel Street,
Vancouver, British Columbia, Canada V5Z 1M9.
E-mail: tnevill@bccancer.bc.ca
Bone Marrow Transplantation (2008) 42, 659–666
& 2008 Macmillan Publishers Limited All rights reserved 0268-3369/08 $32.00
www.nature.com/bmt

Patients and methods
Patient characteristics
A retrospective computerized data base review was
performed to identify patients referred to the Leukaemia/
Bone Marrow Transplantation Program of British Columbia
with a diagnosis of t-MDS/t-AML. Between March 1989
and January 2007, 122 patients with t-MDS/t-AML were
evaluated at Vancouver General Hospital. The majority
(74 patients, 61%) were not considered as candidates for
SCT based upon advanced age or medical comorbidities.
Six patients with good-risk karyotypes received standard
chemotherapy only, six patients underwent auto-SCT and
one patient had a reduced-intensity conditioning allo-SCT.
Eight patients died of disease progression before a donor
could be identified and three patients left the province to
receive treatment. Twenty-four patients (20%) actually
proceeded to myeloablative allo-SCT, with the character-
istics of this cohort shown in Table 1. Some of the patients
in the cohort have previously been reported in meeting
abstracts or published manuscripts.
10,11,14
All subjects
provided informed consent and all research studies were
approved by the University and Institutional Review
Boards. Bone marrow histopathology was centrally
reviewed at Vancouver General Hospital and diagnoses
were based upon published WHO criteria.
6
Seven patients
had t-MDS and IPSS scores were intermediate-2 in five
patients, intermediate-1 in one patient and high-risk in one
patient. The remaining 17 patients had t-AML. Of the 21
patients who had received prior chemotherapy, three had
received an Alk without any history of exposure to Topo II
and two had received an epipodophyllotoxin without any
history of exposure to an Alk or radiotherapy; one patient
had been given only azathioprine for Crohn’s disease. The
remaining 15 patients had received both an Alk and a Topo
II inhibitor. As a result, Topo-II-related t-MDS/t-AML
was diagnosed based on specific established clinical,
morphologic and cytogenetic criteria.
5
All nine of the
patients with Topo II-related tAML had at least two of the
following: lack of a preceding MDS phase (eight patients),
typical karyotypic changes (t(9;11) in four cases; t(8;16)
or t(8;21) in one case each), monoblastic morphology
(six patients), a short latency period (less than 3 years)
and/or a history of having been treated on dose-intensive
epipodophyllotoxin-containing Pediatric Oncology Group
protocols.
15
As per institutional policy, conventional
cytoreductive chemotherapy (high-dose cytosine arabino-
side and daunorubicin or mitoxantrone/etoposide) was
administered only when required for hematologic stabiliza-
tion to allow for tissue typing of potential donors.
16
Of the
11 patients who required conventional chemotherapy, all
achieved an initial CR. However, three patients had
evidence of recurrent t-MDS/t-AML on bone marrow
exams performed just before starting the SCT preparative
regimen. Thirteen patients did not require standard AML
induction chemotherapy before commencing transplanta-
tion conditioning although, two patients had previously
received immunosuppressive treatment with anti-thymo-
cyte globulin and CYA without benefit. Four patients in
the study group previously underwent auto-SCT for a
separate primary malignancy.
Cytogenetics
Cytogenetic analyses were performed on direct and/or
unstimulated cultured marrow specimens at four reference
laboratories and reviewed at Vancouver General Hospital.
All 24 patients in this report had successful bone marrow
karyotyping before SCT. Patients were divided into
prognostic cytogenetic subgroups according to accepted
MRC criteria for AML.
7
In brief, favourable karyotypes
were those patients with t(15;17), t(8;21), or inv(16)/
t(16;16), whether or not other abnormalities were also
present. Adverse karyotypes were those with monosomy 5
or 7, del(5q), abnormalities of 3q or a complex karyotype
(Xfive abnormalities), in the absence of the aforementioned
favourable translocations. All patients who did not have
favourable or adverse karyotypes were classified as inter-
mediate-risk. All seven patients with therapy-induced MDS
were categorized in the identical cytogenetic subgroup that
each would have been, had the IPSS cytogenetic grouping
17
been used (six patients poor-risk and one patient inter-
mediate-risk).
Table 1 Patient characteristics
Median age, years (range) 45.5 (15–55)
Gender, M/F 7/17
WHO Dx
t-AML
Alk/RT-related 8
Topo II-related 9
t-MDS
RA 1
RARS 1
RCMD 3
RAEB-1 1
RAEB-2 1
Prior malignancy
Lymphoma 8
Breast Ca 6
Acute leukaemia 3
Germ cell tumour 2
Other
a
5
Prior therapy
CT
a
11
RT 3
CT+RT 10
Median latency [CT/RT-Dx], mos (range) 47.5 (13–324)
MRC cytogenetic risk group
Favourable 1
Intermediate 12
Normal 3
Other 9
Adverse 11
Abbreviations: Alk/RT¼ alkylating agent/radiation; Ca¼ carcinoma;
CT¼ chemotherapy; Dx¼ diagnosis; F¼ female; M¼male;
mos¼months; MRC¼Medical Research Council; RA¼ refractory
anemia; RARS¼RA with ringed sideroblasts; RCMD¼ refractory
cytopenias with multilineage dysplasia; RAEB¼RA with excess blasts;
t-MDS¼ therapy-related myelodysplastic syndrome; Topo
II¼ topoisomerase II inhibitor; WHO¼World Health Organization.
a
Includes one patient who received azathioprine therapy for Crohn’s
disease.
SCT for tMDS/AML
TJ Nevill et al
660
Bone Marrow Transplantation