VOLUME 25 䡠 NUMBER 14 䡠 MAY 10 2007

JOURNAL OF CLINICAL ONCOLOGY R E V I E W A R T I C L E

Acute Myeloid Leukemia and Myelodysplastic Syndromes in

From the Department of Leukemia, The
University of Texas M.D. Anderson
Cancer Center, Houston, TX.
Submitted December 5, 2006; accepted
February 8, 2007.
Author’s disclosures of potential con-
flicts of interest and author contribu-
tions are found at the end of this
article.
Address reprint requests to Elihu Estey,
MD, Department of Leukemia, Unit
428, The University of Texas M.D.
Anderson Cancer Center, 1515
Holcombe Blvd, Box 61, Houston, TX
77030; e-mail: ehestey@mdanderson.org.
© 2007 by American Society of Clinical
Oncology
0732-183X/07/2514-1908/$20.00
Older Patients
Elihu Estey
A B S T R A C T
The median age of patients with acute myeloid leukemia (AML) is 65 to 70 years. The majority of
older patients with AML probably do not receive specific treatment, and those who receive
standard regimens have a median survival time of less than 1 year. This suggests that, in general,
older patients should receive investigational therapy; however, factors other than age influence
survival after administration of standard treatment and need to be accounted for when making
treatment recommendations. In some cases where investigational therapy is unavailable, palliative
care may be the best option. Like AML, myelodysplastic syndrome (MDS) is a disease of the
elderly. It is divided into higher and lower risk groups. The natural history of high-risk MDS (eg,
⬎ 10% marrow blasts) bears more resemblance to that of AML than to that of an indolent
disorder; accordingly, similar therapeutic considerations apply. The more benign natural history of
lower risk MDS leads to consideration of reduction in transfusion needs and improvement in
quality of life as primary goals of therapy. Lenalidomide, azacitidine, and decitabine, each recently
approved by the US Food and Drug Administration, are useful in achieving these objectives.
J Clin Oncol 25:1908-1915. © 2007 by American Society of Clinical Oncology

DOI: 10.1200/JCO.2006.10.2731
ACUTE MYELOID LEUKEMIA
Any discussion of acute myeloid leukemia (AML) in
a geriatric population should begin with recognition
of the difficulty in accurately defining such a popu-
lation. In particular, it seems that each additional
year of age above age 5 to 10 years is associated with
a worse outcome.1 Thus, although clinical trials of-
ten consider patients aged 60 years and older to be
elderly, there is less difference in outcome between,
for example, a 58-year-old patient and a 61-year-old
patient than between the same 61-year-old patient
and a 69-year-old patient, although the 58-year-old
patient would be considered young, and both the
61-year-old and 68-year-old patients would be con-
sidered elderly. However, for practical purposes, age
ⱖ 60 years is a commonly accepted criterion for
elderly AML (with 60 years being 5 to 10 years less
than the median age of all AML patents). Impor-
tantly, the effect of age is not constant throughout a
patient’s course of treatment. Rather, approximately
1 year after treatment begins, the effect of age on
prognosis disappears; this reflects the dispropor-
tionate influence of age on early death.2
Clinical trials typically include as eligible all
older patients except those with a poor performance
status, abnormal organ function, or active infection.
These exclusions suggest that factors (covariates)
other than age affect outcome. Such outcome is a
sum of the probability of treatment-induced death
(TID) and the probability of therapeutic resistance.
Resistance means either a transient remission or no
remission, with the inability to maintain a remission
incompatible with survival. The distinction between
TID and resistance is arbitrary, with overlap be-
tween the two categories, but often, deaths occur-
ring within the first 6 weeks after beginning therapy
are considered TID, with deaths occurring later as-
cribed to resistance; using this definition, resistance
is the cause of death in 75% of older patients. Of
particular note, at any given age, the risk of TID
depends on, among other factors, performance sta-
tus, bilirubin and creatinine (even within the normal
range), the presence of infection, albumin, and
beta2-microglobulin.3-5 The major predictor of re-
sistance is not age but leukemia-cell cytogenetics,6
with a normal karyotype and the very rare cases with
inv(16) or t (8;21) associated with the best outcome,
abnormalities of chromosomes 5 and/or 7 (⫺5/⫺7)
associated with the worst outcome (except when,
rarely, present as single abnormalities7), and other
karyotypes associated with an intermediate out-
come. Furthermore, at any given age, an abnormal
blood count for at least 1 month before diagnosis of
AML (antecedent hematologic disorder [AHD]) or
AML arising after chemotherapy for another dis-
ease (secondary AML) confers a worse prognosis,
usually independent of cytogenetics.8 Leith et al9
found that presence of a protein (MDR1, also
called P-glycoprotein) that promoted extrusion

1908
Information downloaded from jco.ascopubs.org and provided by at Christian Medical College-Vellore on August 9, 2011 from
Copyright © 2007 American Society of Clinical Oncology. All rights reserved.
220.225.126.138
 AML and MDS in the Elderly

of anthracyclines commonly used to treat AML was associated with
a lower complete response (CR) rate, but not relapse-free survival
rate, in older patients. In contrast, internal tandem duplications
(ITD) of the FLT3 gene (but seemingly not mutations that do not
affect the gene’s length) decrease relapse-free survival time but not
CR rate; such ITDs are most commonly seem in the normal cyto-
genetic group.10 However, although older age is associated with
poorer performance status, ⫺5/⫺7, an AHD, secondary AML, and
MDR1 (but not FLT3 ITDs), the deleterious effect of advancing age
generally remains after accounting for these covariates.1,3-5 None-
theless, it is important to consider covariates other than age when
deciding whether to recommend standard anti-AML therapy, investi-
gational anti-AML therapy, or no specific anti-AML therapy (ie, sup-
portive care only) for an older patient with untreated AML. This
decision is at the heart of the management of AML in the elderly.
General Inadequacy of Standard Anti-AML Therapy
Two types of regimens may be considered standard. The first,
known as 3⫹7, consists of 3 days of an anthracycline (usually dauno-
rubicin or idarubicin) and 7 days of cytarabine (ara-C). Table 1 lists
series reporting outcome with 3⫹7 in older patients. Despite varia-
tions in postremission therapy in these studies and possible differences
in therapy after relapse, the table indicates that the average older
patient with untreated AML can expect to live 8 to 12 months if
administered 3⫹7 as initial therapy.11-15 The second standard regi-
men omits anthracyclines and reduces the dose of ara-C (low-dose
ara-C [LDAC]). In one of the only randomized comparisons of 3⫹7
and LDAC, which involved 87 patients aged more than 65 years, Tilly
et al16 found median survival times of 9 months with LDAC and 13
months with 3⫹7; however, patients administered LDAC spent less
time in hospital (median time, 28 v 34 days, respectively) and required
fewer transfusions (median, seven v 10 transfusions, respectively).
Reported survival data likely overestimate the effectiveness of
standard therapy in older patients. This overestimation results from
the tendency to treat only very fit older patients.17,18 Indeed, most
elderly patients probably do not receive anti-AML therapy, leading to
the contention that that there is no standard therapy for AML of the
elderly. The characteristics of untreated patients, as a group, are often
not known in adequate detail. Hence, the natural history of AML
becomes difficult to specify, particularly compared with outcome with
standard anti-AML therapy. In one of the few randomized studies
to address the issue, the European Organisation for Research and
Treatment of Cancer assigned patients aged more than 65 years to
immediate 3⫹7 or to observation/supportive care, with use of chem-
otherapy (hydroxyurea or LDAC) if blood counts worsened or symp-
toms developed.19 Median survival time was 21 weeks in the 3⫹7 arm
and 10 weeks in the observation arm, and the number of days spent in
hospital was essentially identical. However, the degree to which the
enrolled patients represented AML in the elderly is in some doubt
because the enrolled patients had a performance status of less than 3
and relatively normal organ function, and 50% of the patients ran-
domly assigned to observation had to begin therapy within 1 month,
suggesting that they were on the verge of progression when randomly
assigned. In contrast, it is commonly observed that AML in older
patients can have an indolent course for months if not longer. How-
ever, regardless of their relative merits, 3⫹7, LDAC, and supportive
care only are not appealing approaches for the majority of older
patients with AML.
Most Elderly Patients Are Candidates for
Investigational Therapies
The data in Table 1 indicate that a principal reason to recom-
mend investigational therapy for older patients with AML is the un-
satisfactory outcome with standard therapy. Although scientific
promise is a frequently cited reason for recommending an investiga-
tional therapy, history suggests that the promise of many therapies is
not borne out by clinical trials. Furthermore, investigational therapies
can prove worse than standard therapies. (An example is a trial of the
MDR1 antagonist PSC-833 in older patients with untreated AML.20)
If this was not the case, random assignment between standard and
investigational therapies would be problematic. Nonetheless, I believe
that, in the majority of older patients with AML, the results with newer
therapy cannot be much worse than those obtained with standard or
no therapy. The view that elderly patients with AML are typically
candidates for investigational therapy is widespread. For example,
the National Comprehensive Cancer Network,21 a consortium of

Table 1. Outcomes in Older Patients Administered Anthracycline Plus Cytarabine for AML
Median Probability
Age of Survival of Survival Induction
No. of Patients Time at 2 Years CR Rate Death Rate
Study Group or Institution Patients (years) (months) (%) (%) (%) Comment
Rowe et al11 ECOG 234 ⱖ 56 7-8 ⬇ 20 41 19 Results same with daunorubicin,
idarubicin, or mitoxantrone and ⫾
GM-CSF priming
Goldstone et al12 NCRI (formerly MRC) 1,314 ⱖ 56 ⬇ 12 ⬇ 25 62 16 Results same with 1 or 4 courses
after CR and ⫾ G-CSF starting 8
days after end of induction
Anderson et al13 SWOG 161 ⱖ 56 9 19 43 15 Survival worse with mitoxantrone ⫹
etoposide
Van der Holt et al14 HOVON 211 ⱖ 60 ⬇ 10 ⬇ 25 48 15 Results same ⫾ PSC-833
Estey et al15 M.D. Anderson Cancer 31 ⱖ 65 ⬇ 12 ⬇ 20 48 Not given Cytarabine at 1.5 g/m2 daily ⫻ 3;
Center survival worse with single-agent
gemtuzumab

Abbreviations: AML, acute myeloid leukemia; CR, complete response; ECOG, Eastern Cooperative Oncology Group; GM-CSF, granulocyte-macrophage
colony-stimulating factor; NCRI, National Cancer Research Institute; MRC, Medical Research Council; G-CSF, granulocyte colony-stimulating factor; SWOG,
Southwest Oncology Group; HOVON, Stichting Haemato-Oncologie voor Volwassenen Nederland.

www.jco.org 1909
Information downloaded from jco.ascopubs.org and provided by at Christian Medical College-Vellore on August 9, 2011 from
Copyright © 2007 American Society of Clinical Oncology. All rights reserved.
220.225.126.138
 Elihu Estey

prominent American cancer centers, explicitly cites clinical trial as the

1.0
preferred option in patients with untreated AML who are aged 60 Total Died
years and olders. Better* 115 77
0.8 + Worse 533 465
Thinking in terms of covariates other than age allows identifica-

Survival Probablity
tion of groups of older patients for whom standard therapy (3⫹7 or ++ +
+
LDAC) might be reasonable. From the preceding discussion, one such 0.6 +
+ +
+
group might be patients with the following characteristics: aged 60 to ++ +++ +++++
69 years, performance status less than 2, normal bilirubin and creati- 0.4 ++
+
nine, neither an AHD nor secondary AML, and no pretreatment +++ ++++ + + ++++ + + +
+++
++ ++
infection. Such patients have had a median survival time of 14 months +++ + +
0.2 ++
after administration of various ara-C– containing regimens at M.D. ++ + ++++ ++
+ +++ + +++ +
Anderson Cancer Center over the past decade, with a 38% actuarial +++
survival rate at 2 years, a 5% TID rate at 6 weeks, and a 70% CR rate
(Table 2). Because investigational therapy can be worse than standard 0 1 2 3 4 5

therapy, some patients in this group might prefer standard therapy.
The same might be said, although to a lesser extent, of patients aged 70
to 79 years but with an otherwise favorable configuration of perfor-
mance status, bilirubin/creatinine/infection, and AHD/secondary
AML (Table 2; the number of such patients aged 80 years and older is
too small to warrant comment). However, patients less than age 80
years with otherwise favorable features composed only 17% of the
patients seen during this time. The outcome of the remaining 83% of
patients after ara-C– containing therapy is more problematic (Fig 1).22
For example, those patients aged ⱖ 80 years only had an 11-week
median survival time, with 39% dead within 6 weeks of beginning
therapy (Table 2), which is similar to reports from an earlier period.
The dominant effect of performance status is seen in the 21-month
median survival time and 24% 6-week mortality rate in otherwise
favorable patients who have a performance status more than 1. Pa-
tients with an AHD or secondary AML as their only unfavorable
characteristic still had a median survival time of only 35 weeks, despite
only a 9% 6-week TID rate. All of these patients are candidates for
investigational therapy. However, if such therapy is unavailable, then
the preferred option in those patients aged ⱖ 80 years or with a
performance status more than 1 would seem to be supportive care
only given their high early death rates; in contrast, 3⫹7 or LDAC
might be more plausible in patients with only an AHD or second-
ary AML. As the number of unfavorable covariates increases, so
should the reluctance to give standard therapy rather than investi-
gational therapy.22,23
Any discussion of choice of therapy must refer to the observa-
tions of Sekeres et al24 that 74% of older patients estimated that their
chances of cure with 3⫹7 were at least 50%; in contrast, 85% of their
physicians estimated this chance to be less than 10%. Although the
Time (years)
Fig 1. Survival in two groups of older patients with untreated acute myeloid
leukemia (log-rank test, P ⬍ .0001). As described in the text, patients in the better
group are more plausible candidates for standard therapy (eg, 3 days of an
anthracycline plus 7 days of cytarabine or low-dose cytarabine) than patients in
the worse group. (ⴱ) Age ⬍ 88 years, performance status ⬍ 2, no antecedent
hematologic disorder or secondary acute myeloid leukemia, bilirubin/creatinine
⬍ 2, and no infection or fever of unknown origin.
most plausible cause of this discrepancy is patients’ natural tendency
to believe what they want to believe, there may also be gaps in com-
munication between physicians and patients.
Role of Cytogenetics
Optimally, cytogenetics would inform initial management of
untreated AML. For example, considering the most favorable patients
in Table 2 (row 1), median survival time was 90 weeks in patients with
a normal karyotype or, more rarely, an inv(16) or t (8;21) karyotype,
with 46% alive at 2 years, versus a median survival time of 40 weeks in
patients with other karyotypes, with 20% alive at 2 years (Fig 2).
Accordingly, the former might be less receptive than the latter to
investigational therapy. However, results of cytogenetic tests are often
unavailable for at least 1 week, necessitating that physicians decide
whether it is riskier to delay initiation of therapy or to administer
potentially toxic standard therapy that is unlikely to be effective
given the patient’s cytogenetics (or, in the case of a normal karyo-
type, to administer investigational therapy when standard therapy
might be reasonably successful). The Eastern Cooperative Oncol-
ogy Group reported that the 113 patients aged ⱖ 55 years in whom
logistical reasons necessitated a 3- to 5-day delay in beginning

Table 2. Outcome According to Indicated Covariates After Cytarabine-Containing Therapy for AML at M.D. Anderson Cancer Center 1996 to 2006
Median % of Probability
Survival Patients of Survival
Age Performance AHD or Bilirubin/Creatinine No. of Time Dead By CR Rate at 2 Years
(years) Status Secondary AML (mg/dL) Infection/FUO Patients (weeks) Day 42 (%) (%)
60-69 0-1 No ⬍2 No 64 58 5 70 38
70-79 0-1 No ⬍2 No 51 46 2 59 28
ⱖ 80 Any Yes or no Any Yes or no 46 11 39 39 11
60-79 ⬎1 No ⬍2 No 21 21 24 67 5
60-69 0-1 Either or both ⬍2 No 127 35 9 55 18

Abbreviations: AML, acute myeloid leukemia; AHD, antecedent hematologic disorder; FUO, fever of unknown origin; CR, complete response.

1910 JOURNAL OF CLINICAL ONCOLOGY

Information downloaded from jco.ascopubs.org and provided by at Christian Medical College-Vellore on August 9, 2011 from
Copyright © 2007 American Society of Clinical Oncology. All rights reserved.
220.225.126.138

1.0
Cytogenetics
Normal, Inv(16), t(8;21)
0.8 + Other
Survival Probablity
0.6 + +
++
+ ++ + + + +
0.4 ++
+ +
0.2
+
0 1 2 3
Time (years)
Fig 2. Influence of cytogenetics on survival in patients in the better group
described in Figure 1 who were aged 60 to 69 years (log-rank test, P ⫽ .02).
induction therapy had a CR rate of 38% compared with a CR rate of
50% in the 235 similarly aged patients who began the same thera-
pies without such delay.11 M.D. Anderson Cancer Center data
(unpublished data) suggest that a 1-week delay in beginning ther-
apy in older patients with WBC counts less than 50,000/␮L is
associated with decreases in median survival time ranging from 4
to 28 weeks depending on other covariates.
If it does not enter into initial management decisions, cytogenetic
information should be used in planning postremission therapy. In
particular, patients with an abnormal karyotype would generally be
poorer candidates for ara-C or anthracycline plus ara-C than patients
with a normal karyotype.
Investigational Approaches
The first three agents listed in Table 3 (tipifarnib,25 lestaurtinib,26
and decitabine27) are examples of targeted therapy, although, partic-
ularly with tipifarnib and less so with lestaurtinib and decitabine, there
is no direct correlation between response and the pretreatment status
of, or the effect of the drug on, the presumed target. As seen in Table 3,
response rates seem higher when the new agent is not a targeted
therapy but bears more resemblance to a traditional cytotoxic agent
(eg, clofarabine28 or VNP40101M [Cloretazine; Vion Pharmaceuti-
cals, New Haven, CT]29). Accordingly, it is generally assumed that
many targeted agents will eventually be combined with cytotoxic
agents, as has been done with gemtuzumab ozogamicin (GO)30 and
oblimersen sodium (Marcucci et al, submitted for publication; Table
3. In younger patients with favorable or intermediate karyotypes,
gemtuzumab ozogamicin plus chemotherapy has been reported to
lengthen relapse-free survival time32; whether the same is true in older
patients is unknown.) Although not exhaustive, Table 3 makes several
other relevant points. First, there is an increasing tendency to report
responses (eg, CR with incomplete platelet recovery) less demanding
than CR. Although it is known that achievement of CR prolongs
survival, the effect of these lesser responses on survival remains to be
established.33 Second, although patients are more interested in the
comparative activity of different drugs than in the activity of a given
drug, particularly with respect to survival, most new drug studies are
single-arm studies. The National Cancer Research Institute (formerly
the Medical Research Council) in the United Kingdom is addressing
www.jco.org
Information downloaded from jco.ascopubs.org and provided by at Christian Medical College-Vellore on August 9, 2011 from
Copyright © 2007 American Society of Clinical Oncology. All rights reserved.
220.225.126.138
AML and MDS in the Elderly
this issue by conducting relatively small randomized trials intended to
Total Died select which of several agents to investigate in larger trials.34,35 Third,
37 21 new drugs are typically studied in patients with active disease. Given
27 21 average CR rates less than 50% in older patients, this approach is
necessary. However, because the volume of disease is less in CR than at
diagnosis, new therapies might prove to be more effective in prolong-
ing than in inducing remissions. Furthermore, based on well-known
relationships between probability of response to previous new drugs
(eg, administered at relapse) and beneficial effect of a prior therapy
(eg, as measured by duration of initial CR),36 it is plausible that a new
drug’s benefit will be more readily demonstrable in patients in CR
than in untreated patients. Finally, in none of the studies in Table 3 is
it noted that the patients were consecutive patients, raising the possi-
4 5 bility of selection bias, similar to that described for minitransplanta-
tion (see next paragraph).
Numerous studies comparing allogeneic stem-cell transplanta-
tion (SCT) with standard chemotherapy in patients less than age 60
years have found that the likely greater anti-AML effect of SCT is
counterbalanced by greater TID, resulting on average in equivalent
survival.37-41 Current thinking assigns more of a role to an immuno-
logically mediated graft-versus-leukemia effect than to high-dose
chemotherapy as the principal mediator of the anti-AML effect of
SCT.41 If this is the case, it becomes feasible to use reduced-intensity
conditioning (RIC) SCT (minitransplantation) to reduce TID. Estey
et al42 described a systematic attempt to perform RIC-SCT in first CR
in all M.D. Anderson patients aged ⱖ 50 years with an abnormal
karyotype and a sibling or matched unrelated donor. Matching for
known prognostic factors suggested that the14 patients who received
transplantation had longer survival time than the 83 patients who did
not receive transplantation, but these 14 patients represented only 5%
of all treated patients aged more than 50 years with abnormal karyo-
types. These results question the general applicability of RIC-SCT and
suggest that it should be performed before patients enter first CR to
increase the number of patients who might be candidates given the
low CR rates in older patients. For example, older patients might
initially receive a new agent. While waiting to see this agent’s effect, a
donor search might be completed, as might various logistical arrange-
ments. RIC-SCT would be performed once response to the new agent
is known.
Cytotoxic T lymphocytes (CTLs) are proposed mediators of
graft-versus-leukemia.43 Molldrem et al44 have identified PR1, an
HLA-A2– binding peptide derived from proteinase-3, as a myeloid
leukemia–associated antigen for CTL, finding PR1-specific high-
avidity CTLs in 11 of 12 patients with CML responding to interferon
compared with none of seven patients without a response. They sug-
gested that failure of myeloid leukemia patients to spontaneously
develop an immune response to their disease resulted from overex-
pression of proteinase-3, leading to apoptosis of PR1-specific high-
avidity CTLs.45 Clinical responses have followed PR1 vaccination of
patients with AML and have occurred only in patients in whom there
was an immune response, which was defined as ⱖ two-fold increase in
the number of PR1-specific CTLs; in contrast, there was no relation-
ship between response and the number of CTLs directed against the
control antigen PP65.43,46 Immune response was more common in
patients in remission when beginning vaccination, leading to a soon to
be initiated trial comparing PR1 vaccination with no further therapy
in patients who are HLA-A2 positive and have completed three
courses of postremission therapy.
1911
 Elihu Estey

Table 3. Investigational Therapies in Patients Age ⱖ 60 Years With Untreated AML

Response
Median Age ⬍ CR
Type of Drug Example Study No. of Patients (years) CR Rate (%) Rate (%) Effect on Survival
25
Farnesyltransferase Tipifarnib Lancet et al 158 74 14 9 Median survival time,
inhibitor 5.3 months
26
FLT3 inhibitor Lestaurtinib Knapper et al 27 73 0 11 Not given
(CEP701)
Hypomethylating agent Decitabine (plus Lubbert et al27 29 72 14 17 Median survival time,
all-trans-retinoic 7.5 months
acid
Nucleoside analog Clofarabine Burnett et al28 66 71; included 29 20 Median survival time,
only patients 5 months
aged ⱖ 65
Alkylating agent VNP40101M Karp et al29 45 (no AHD or 72 49 — 22% survival at 1
(Cloretazine) secondary year
AML)
Anti-CD33 antibody Gemtuzumab De Angelo et al30 21 67 43 — 48% of patients alive
attached to toxin ozogamicin with median
(⫹ ara-C) follow-up time of 7
months
Bcl-2 antagonist: Oblimersen sodium Marcucci et al 503 — 48 in patients who — None; 1-year survival
enhances apoptosis (added or not (submitted for received rate: 36% with
added to 3 ⫹ 7) publication) oblimersen oblimersen
sodium; 52 in sodium, 40%
patients who did without oblimersen
not receive sodium
oblimersen
sodium
MDR1 antagonist
PSC-833 Added or not Baer et al20 120 70 39 in patients — Median survival time
added to 3 ⫹ 7 who received of 2 months with
plus etoposide PSC-833; 46 in PSC-833 and 7
patients who months without
did not receive PSC-833; study was
PSC-833 stopped because of
excess early deaths
with PSC-833
Zosuquidar Added or not Cripe et al31 442 67 — — None; median
added to 3 ⫹ 7 survival time of 7.7
with zosuquidar
and 9.4 months
without zosuquidar

Abbreviations: AML, acute myeloid leukemia; CR, complete response; AHD, antecedent hematologic disorder; AML, acute myeloid leukemia; ara-C, cytarabine;
3 ⫹ 7, 3 days of an anthracycline and 7 days of cytarabine.

Clinical Practice Issues

Several practices that may have little effect on mortality but
considerable effect on patients’ lives should be mentioned. First is
hospitalization; given the more serious nature of hospital-acquired
infection compared with community-acquired infection, the use of
routine hospitalization after completion of a course of therapy should
be questioned. The second practice involves masks and avoidance of
crowds. The advice to avoid crowds seems inconsistent with observa-
tions that bacteria and fungi, rather than viruses, are the typical causes
of infection in AML.47,48 The bacteria are invariably residents of the
patients’ own skin, mouth, or intestines. The fungi are similarly found
on the skin or are airborne. However, because of the organism’s size, it
is unclear whether masks will restrict entry of Aspergillus into the nose.
The third practice involves consumption of fresh fruits and vegetables.
There is conflicting data as to whether avoidance of these foods lessens
the risk of infection. An ongoing M.D. Anderson Cancer Center trial
suggests that patients who are encouraged to eat fresh fruits and
vegetables have similar infection rates as patients randomly assigned to
not eat them (unpublished data).
MYELODYSPLASTIC SYNDROME
Like AML, myelodysplastic syndrome (MDS) is a disease that has an
onset usually after age 60 years. Patients are usually classified as high or
low risk.
High-Risk MDS
High-risk MDS generally implies International Prognostic Scor-
ing System (IPSS) scores of intermediate-2 or high.49 All patients with
10% to 20% marrow blasts will meet these criteria, as will patients with
two to three cytopenias and either 5% to 10% blasts and intermediate
cytogenetics or less than 5% blasts and unfavorable cytogenetics, as
defined by the IPSS.49 Life expectancy without treatment is typically
less than 1.5 to 2 years. Thus, the natural history of high-risk MDS
bears more resemblance to that of AML than to that of an indolent
smoldering disease, which is how MDS, at times, is regarded. Further-
more, patients with AML and patients with high-risk MDS have sim-
ilar outcomes when treated identically, after accounting for the older
age of MDS patients and their tendency to have ⫺5/⫺7 and longer

1912 JOURNAL OF CLINICAL ONCOLOGY

Information downloaded from jco.ascopubs.org and provided by at Christian Medical College-Vellore on August 9, 2011 from
Copyright © 2007 American Society of Clinical Oncology. All rights reserved.
220.225.126.138
 AML and MDS in the Elderly
durations of abnormal blood counts.50 Indeed, the boundary between
high-risk MDS and AML is malleable; previously defined by a marrow
blast count of 30%, AML is now diagnosed when more than 20%
blasts are seen.51 Although there are undoubtedly biologic differ-
ences between high-risk MDS and AML, there is also likely to be
biologic overlap, given the somewhat arbitrary distinction between
the two entities.52
Given the natural history of high-risk MDS, the primary of goal
of therapy is to lengthen survival.53 The US Food and Drug Adminis-
tration has recently approved lenalidomide, decitabine, and azaciti-
dine for treatment of MDS. However, too few patients with high-risk
MDS have received lenalidomide to draw conclusions.54-56 Approval
of decitabine in MDS was based on a randomized trial comparing
supportive care plus decitabine (15 mg/m2 every 8 hours for 3 days
intravenously repeated every 6 weeks) with supportive care alone;
approximately 70% of patients had high-risk MDS, with 25% of these
patients having 20% to 30% blasts.57 Decitabine produced a response
rate (CR ⫹ partial response ⫹ hematologic improvement, as defined
by List et al54) of 30%, reduced RBC transfusion need, and improved
quality of life. An analogous trial randomly assigning patients to either
supportive care or supportive care plus azacitidine (75 mg/m2 subcu-
taneously on days 1 through 7) prompted approval of azacitidine.58,59
Unlike the decitabine trial, crossover was permitted for patients in the
supportive care group who remained transfusion dependent, devel-
oped worsening cytopenias, or had increased blasts. Forty-six percent
of the patients for whom IPSS scores were known had high-risk MDS,
and substantial numbers of these patients had AML, using the 20%
blast criterion. The CR ⫹ partial response ⫹ hematologic improve-
ment rate was 47% in the azacitidine arm compared with 17% in
the supportive care arm. Both decitabine and azacitidine produced
statistically significant delays in time to AML or death (eg, median
time, 12.0 months for decitabine v 6.8 months for supportive care
in high-risk patients). But neither drug resulted in a more than
5-month delay in median time to death; in the decitabine trial, the
median time to death was 14.0 months for decitabine compared
with 14.9 months for supportive care, with data not broken down
by risk group; and in the azacitidine trial, the median time to death
in patients who would probably currently be considered as high
risk was 18 months in the azacitidine arm compared with 13
months in the supportive care arm. Although the ability to cross
over to azacitidine decreased the difference in survival between
supportive care and azacitidine, such a crossover would of course
be done in clinical practice. The inability of azacitidine and decit-
abine to have more of an effect on survival may reflect the low CR
rates each drug produced (9% for decitabine and 10% for azaciti-
dine). In general, I believe time to AML or death is considerably less
important than time to death; after all, patients presumably care
little about whether they die with or without AML, and the defini-
tion of AML is itself arbitrary.
Given the data, I believe that decitabine, azacitidine (as adminis-
tered in the randomized trials), and lenalidomide are not satisfactory
options for patients with high-risk MDS. Accordingly, like older pa-
tients with AML, older patients with high-risk MDS are primarily
candidates for investigational treatments.
Besides the investigational treatments noted earlier and in Table
3, there are other investigational therapies to consider. The first is
low-dose decitabine. Two confounding factors in any analysis of the
effects of azacitidine and decitabine are the number of courses admin-
www.jco.org
Information downloaded from jco.ascopubs.org and provided by at Christian Medical College-Vellore on August 9, 2011 from
Copyright © 2007 American Society of Clinical Oncology. All rights reserved.
220.225.126.138
istered and the interval between courses. Thus, in 50% of patients, at
least three courses are needed to observe response to azacitidine58,59;
25% of responses or best responses were seen between course 5 and
course 17. A similar phenomenon occurs with decitabine.57 Use of
lower doses (eg, 20 mg/m2 daily for 5 days) facilitates administration
of more courses at more frequent intervals (eg, every 4 to 5 weeks).
Using such an approach, Kantarjian et al60 reported a CR rate of 34%
in 95 MDS patients, two thirds of whom were high-risk patients. In
contrast, the CR rate was 9% in the randomized (higher dose, fewer
and less frequent courses) trial.57 A similar approach might be taken
with azacitidine.
Another investigational therapy is hypomethylating agents plus
histone deacetylase inhibitors. The effect of azacitidine and decitabine
is believed to be mediated by removal of methyl groups from various
tumor suppressor genes, with hypomethylation restoring gene expres-
sion.61 Through a different mechanism, histone deacetylase inhibitors
are also postulated to reactivate transcription of tumor suppressor
genes.62 The possibility of synergy between these two classes of
epigenetic-acting drugs has led to their combination (eg, azacitidine
plus valproic acid, decitabine plus suberoylanilide hydroxamic acid)
in ongoing clinical trials. Given the overlap between high-risk MDS
and AML, low-dose decitabine and combination epigenetic therapy
are also plausibly effective in AML.
Lower Risk MDS
Lower risk MDS includes patients with IPSS scores of
intermediate-1 or low.48 The natural history is quite variable, but
median survival times usually range from 2 to 10 years. This more
benign natural history leads to consideration of reduction in RBC
transfusion need and concomitant improvement in quality of life as
primary goals.53 Table 4 presents an approach to management of
lower risk MDS in transfusion-dependent patients. Lenalidomide is
particularly effective in patients with deletion of the long arm of
chromosome 5 (5q–).55 Sixty-seven percent of 148 patients with this
abnormality became transfusion independent within a median of 4.6
weeks (range, 1 to 49 weeks) of beginning lenalidomide 10 mg daily for
21 days every 4 weeks, with independence maintained for at least 1
year in 62% of patients. There was no relationship between the com-
plexity of the 5q karyotype and the likelihood of response, but multi-
variate analysis indicated that response was less likely in patients with
thrombocytopenia or who required more than 4 units of RBCs in the
8 weeks before study entry.
Preliminary data suggest that lenalidomide may also be effec-
tive (but less so) in patients with a normal karyotype.56 However,
Table 4. Treatment of Low-Risk MDS
Candidate for
EPO Level Immunosuppressive
Cytogenetics (mU/mL) Therapy Treatment
5q- Any NA Lenalidomide
Not 5q- ⬍ 500 NA EPO ⫾ G-CSF
Not 5q- ⬎ 500 Yesⴱ ATG
Not 5q- ⬎ 500 Noⴱ Azacitidine, decitabine
Abbreviations: MDS, myelodysplastic syndrome; EPO, erythropoietin; NA,
not applicable; G-CSF, granulocyte colony-stimulating factor; ATG, antithymo-
cyte globulin.
ⴱ
See text.
1913
 Elihu Estey

at present, I recommend erythropoietin (EPO) or darbepoetin

(⫾ granulocyte colony-stimulating factor) for patients without 5q–
with a serum EPO level of less than 500 mU/mL, primarily because
these drugs have fewer adverse effects than lenalidomide. Such pa-
tients have a 25% to 75% response rate, with the specific rate inversely
related to the transfusion rate (eg, ⬍ 2 v ⱖ 2 units/month).63
Some patients with MDS respond to immunosuppressive ther-
apy (eg, antithymocyte globulin [ATG]), reflecting the autoimmune
nature of MDS; indeed, there is overlap between aplastic anemia and
MDS. ATG-responsive patients tend to be positive for HLA-DR15 and
have an age in years and a duration of RBC transfusion dependence in
months summing to less than 72.64
Patients who are not candidates for lenalidomide, EPO, or
ATG are candidates for azacitidine, decitabine, or investigational
therapies. One example of the latter is the PR1 vaccine discussed in
the section on AML.43
SUMMARY
Most patients with untreated AML or high-risk MDS should receive
investigational therapy. Covariates other than age affect outcome and
should be used in deciding whether specific patients receive investiga-
tional or standard therapy or supportive care only in cases where
investigational therapy is not feasible. There are several problems with
the approaches used to conduct trials of new agents. Supportive care
practices for older patients with AML or high-risk MDS should be
questioned. Lenalidomide, azacitidine, and decitabine are important
advances for patients with lower risk MDS.
AUTHOR’S DISCLOSURES OF POTENTIAL CONFLICTS
OF INTEREST
The author indicated no potential conflicts of interest.

REFERENCES
1. Estey EH, Thall PF, Cortes JE, et al: Compar-
ison of idarubicin ⫹ ara-C-, fludarabine ⫹ ara-C-, and
topotecan ⫹ ara-C-based regimens in treatment of
newly-diagnosed acute myeloid leukemia, refractory
anemia with excess blasts in transformation, or
refractory anemia with excess blasts. Blood 98:
3575-3583, 2001
2. Thall P, Estey E: Graphical methods for
evaluating covariate effects in the Cox model, in
Crowley J (ed): Handbook of Statistics in Clinical
Oncology. New York, NY, Marcel-Dekker, 2001,
pp 411-433
3. Estey E, Smith T, Keating M, et al: Prediction
of survival during induction therapy in patients with
newly diagnosed AML. Leukemia 3:257-263, 1989
4. Appelbaum F, Gundacker H, Head D, et al:
Aging and acute myeloid leukemia. Blood 107:3481-
3485, 2006
5. Tsimberidou A-M, Kantarjian H, O’Brien S, et
al: Prognostic significance of beta-2 microglobulin
levels in acute myeloid leukemia. Blood 108:240a,
2006 (abstr 802)
6. Grimwade D, Walker H, Harrison G, et al: The
predictive value of hierarchical cytogenetic classifi-
cation in older adults with acute myeloid leukemia:
Analysis of 1065 patients entered into the United
Kingdom Medical Research Council AML 11 trial.
Blood 98:1312-1320, 2001
7. Estey E, Pierce S, Keating M: Identification of
a group of AML/MDS patients with a relatively
favorable prognosis who have chromosome 5
and/or 7 abnormalities. Haematologica 85:246-249,
2000
8. Goldstone AH, Burnett AK, Avivi I, et al:
Secondary AML has a worse outcome than de novo
AML even taking into account cytogenetics and age:
AML 10, 11, 12 MRC trials. Blood 100:88a, 2002
(abstr)
9. Leith C, Kopecky K, Godwin J, et al: Acute
myeloid leukemia in the elderly: Assessment of
multidrug resistance (MDR1) and cytogenetics dis-
tinguishes biologic subgroups with remarkably dis-
tinct responses to standard chemotherapy—A
Southwest Oncology Group study. Blood 89:3323-
3329, 1997
10. Schnittger S, Schoch C, Dugas M, et al:
Analysis of FLT3 length mutations in 1003 patients
with acute myeloid leukemia: Correlation to cytoge-
netics, FAB subtype, and prognosis in the AMLCG
study and usefulness as a marker for the detection
of minimal residual disease. Blood 100:59-66, 2002
11. Rowe JM, Neuberg D, Friedenberg W, et al: A
phase 3 study of three induction regimens and of
priming with GM-CSF in older adults with AML: A
trial by the Eastern Cooperative Oncology Group.
Blood 103:479-485, 2004
12. Goldstone AH, Burnett AK, Wheatley K, et al:
Attempts to improve treatment outcomes in AML in
older patients: The results of the United Kingdom
Medical Research Council AML11 trial. Blood 98:
1302-1311, 2001
13. Anderson JE, Kopecky KJ, Willman CL, et al:
Outcome after induction chemotherapy for older
patients with AML is not improved with mitox-
antrone and etoposide compared to daunorubicin
and cytarabine: A Southwest Oncology Group
Study. Blood 100:3869-3876, 2002
14. Van der Holt B, Lowenberg B, Burnett AK, et
al: The value of the MDR1 reversal agent PSC-833 in
addition to daunorubicin and cytarabine in the treat-
ment of elderly patients with previously untreated
AML in relation to MDR1 status at diagnosis. Blood
106:2646-2654, 2005
15. Estey EH, Thall PF, Giles FG, et al: Gemtu-
zumab ozogamicin with or without interleukin 11 in
patients 65 years of age or older with untreated
AML or high-risk MDS: Comparison with idarubicin
plus high-dose continuous infusion cytosine arabino-
side. Blood 99:4343-4349, 2002
16. Tilly H, Castaigne S, Bordessoule D, et al: Low
dose cytarabine versus intensive chemotherapy in
the treatment of acute nonlymphocytic leukemia in
the elderly. J Clin Oncol 8:272-279, 1990
17. Mengis C, Aebi S, Tobler A, et al: Assessment
of differences in patient populations selected for or
excluded from participation in clinical phase III acute
myelogenous leukemia trials. J Clin Oncol 21:3933-
3939, 2003
18. Hutchins LF, Unger JM, Crowley JJ, et al:
Underrepresentation of patients 65 years of age or
older in cancer-treatment trials. N Engl J Med 341:
2061-2067, 1999
19. Lowenberg B, Zittoun R, Kerkhofs H, et al: On
the value of intensive remission induction therapy in
elderly patients of 65⫹ years with AML: A random-
ized phase III study of the European Organization for
Research and Treatment of Cancer Group. J Clin
Oncol 7:1268-1274, 1989
20. Baer MR, George SL, Dodge RK, et al: Phase
3 study of the multidrug resistance modulator PSC-
833 in previously untreated patients 60 years of age
and older with acute myeloid leukemia: Cancer and
Leukemia Group B Study 9720. Blood 100:1224-
1232, 2002
21. National Comprehensive Cancer Network
AML Panel: AML clinical practice guidelines. J Natl
Comprehensive Cancer Network 1:520-539, 2003
22. Estey E: General approach to, and perspec-
tives on clinical research in, older patients with
newly-diagnosed acute myeloid leukemia. Semin
Hematol 43:89-95, 2006
23. Estey E: How I treat older patients with AML.
Blood 96:1670-1673, 2000
24. Sekeres MA, Stone RM, Zahrieh D, et al:
Decision-making and quality of life in older adults
with AML or advanced MDS. Leukemia 18:809-816,
2004
25. Lancet J, Gojo I, Gotlib J, et al: A phase 2
study of the farnesyl transferase inhibitor tipifarnib
in poor-risk and elderly patients with previously
untreated acute myelogenous leukemia. Blood 109:
1387-1394, 2007
26. Knapper S, Burnett A, Littlewood T, et al: A
phase 2 trial of the FLT3 inhibitor lestaurtinib
(CEP701) as first-line treatment for older patients
with acute myeloid leukemia not considered fit for
intensive chemotherapy. Blood 108:3262-3270,
2006
27. Lubbert M, Ruter B, Schmid M, et al: Contin-
ued low-dose decitabine is an active first-line treat-
ment of older AML patients: First results of a
multicenter phase II study. Blood 106:527a, 2005
(abstr 1852)
28. Burnett A, Baccarani M, Johnson P, et al: A
phase 2 study of clofarabine monotherapy in pa-
tients aged 65 years or older with acute myeloid
leukemia for whom standard intensive chemothera-
py is not considered suitable. Blood 108:130a, 2006
(abstr)
29. Karp J, Rizzieri D, Vey N, et al: Cloretazine is
an effective induction therapy in elderly patients
with poor-risk de novo AML. J Clin Oncol 24:340s,
2006 (suppl, abstr 6512)
30. De Angelo D, Stone RM, Durant S, et al:
Gemtuzumab ozogamicin (Mylotarg) in combination
with induction chemotherapy for the treatment of
patients with de novo acute myeloid leukemia: Two
age-specific phase 2 trials. Blood 102:100a, 2003
(abstr)
31. Cripe L, Li X, Litzow M, et al: A randomized
placebo-controlled, double-blind trial of the MDR
modulator zosuquidar during conventional induction
and post-remission therapy for patients ⬎ 60 years

1914 JOURNAL OF CLINICAL ONCOLOGY

Information downloaded from jco.ascopubs.org and provided by at Christian Medical College-Vellore on August 9, 2011 from
Copyright © 2007 American Society of Clinical Oncology. All rights reserved.
220.225.126.138

of age with newly-diagnosed acute myeloid leuke-
mia or high-risk myelodysplastic syndrome. Blood
108:129, 2006 (abstr 423)
32. Burnett A, Kell W, Goldstone A, et al: The
addition of gemtuzumab ozogamicin to induction
chemotherapy for AML improves disease-free sur-
vival without extra toxicity: Preliminary results of
1115 patients in the MRC AML15 trial. Blood 108:
8a, 2006 (abstr 13)
33. Estey E, Garcia-Manero G, Giles F, et al:
Clinical relevance of CRp in untreated AML. Blood
106, 2005 (abstr 541)
34. Estey E: Clinical trials in AML of the elderly:
Should we change our methodology? Leukemia
18:1772-1774, 2004
35. Estey E, Thall P: New designs for phase 2
clinical trials. Blood 102:442-448, 2003
36. Breems D, Van Putten W, Huijgens P, et al:
Prognostic index for adult patients with acute my-
eloid leukemia in first relapse. J Clin Oncol 23:1969-
1978, 2005
37. Cassileth PA, Harrington DP, Appelbaum FR,
et al: Chemotherapy compared with autologous or
allogeneic bone marrow transplantation in the man-
agement of acute myeloid leukemia in first remis-
sion. N Engl J Med 339:1649-1656, 1998
38. Zittoun RA, Mandelli F, Willemze R, et al:
Autologous or allogeneic bone marrow transplanta-
tion compared with intensive chemotherapy in
AML. N Engl J Med 332:217-223, 1995
39. Burnett AK, Wheatley K, Goldstone AH, et al:
The value of allogeneic bone marrow transplant in
patients with acute myeloid leukaemia at differing
risk of relapse: Results of the UK MRC AML 10 trial.
Br J Haematol 118:385-400, 2002
40. Jourdan E, Boiron J, Dastugue N, et al: Early
allogeneic stem-cell transplantation for young adults
with acute myeloblastic leukemia in first complete
remission: An intent-to-treat long-term analysis of
the BGMT experience. J Clin Oncol 23:7676-7684,
2005
41. Kolb HJ, Schattenberg A, Goldman JM, et al:
Graft-versus-leukemia effect of donor lymphocyte
transfusions in marrow grafted patients: European
Group for Blood and Marrow Transplantation Work-
ing Party Chronic Leukemia. Blood 86:2041-2050,
1995
www.jco.org
Information downloaded from jco.ascopubs.org and provided by at Christian Medical College-Vellore on August 9, 2011 from
Copyright © 2007 American Society of Clinical Oncology. All rights reserved.
AML and MDS in the Elderly
42. Estey E, de Lima M, Tibes R, et al: Prospec-
tive feasibility analysis of reduced intensity condi-
tioning regimens for hematopoietic stem cell
transplantation (HSCT) in elderly patients with acute
myeloid leukemia and high-risk myelodysplastic syn-
drome. Blood 109:1395-1400, 2007
43. Molldrem JJ: Vaccination for leukemia. Biol
Blood Marrow Transplant 12:13-18, 2006 (suppl 1)
44. Molldrem JJ, Lee PP, Wang C, et al: Evidence
that specific T lymphocytes may participate in the
elimination of chronic myelogenous leukemia. Nat
Med 6:1018-1023, 2000
45. Molldrem JJ, Lee PP, Kant S, et al: Chronic
myelogenous leukemia shapes host immunity by
selective elimination of high-avidity leukemia-
specific T cells. J Clin Invest 111:639-647, 2003
46. Qazilbash M, Wieder E, Rios R, et al: Vaccina-
tion with the PR1 leukemia-associated antigen can
induce complete remission in patients with myeloid
leukemia. Blood 104, 2004 (abstr 259)
47. Anderlini P, Luna M, Kantarjian HM, et al:
Causes of initial remission induction failure in pa-
tients with acute myeloid leukemia and myelodys-
plastic syndromes. Leukemia 10:600-608, 1996
48. Mattiuzzi G, Alvarado G, Alvarez R, et al:
Causes of death in patients with acute myeloid
leukemia and myelodysplastic syndromes undergo-
ing induction chemotherapy. Blood 102:878a, 2003
(abstr)
49. Greenberg P, Cox C, LeBeau M, et al: Inter-
national scoring system for evaluating prognosis in
myelodysplastic syndromes. Blood 89:2079-2088,
1997
50. Estey E, Thall P, Beran M, et al: Effect of
diagnosis (refractory anemia with excess blasts,
refractory anemia with excess blasts in transforma-
tion, or acute myeloid leukemia [AML]) on outcome
of AML-type chemotherapy. Blood 90:2969-2977,
1997
51. Vardiman JW, Harris NL, Brunning RD: The
World Health Organization (WHO) classification of
the myeloid neoplasms. Blood 100:2292-2302, 2002
52. Albitar M, Manshouri T, Shen Y, et al: Myelo-
dysplastic syndrome is not merely “preleukemia.”
Blood 100:791-798, 2002
53. Cheson BD, Greenberg PL, Bennett JM, et al:
Clinical application and proposal for modification of
■ ■ ■
220.225.126.138
the International Working Group (IWG) response
criteria in myelodysplasia. Blood 108:419-425, 2006
54. List A, Kurtin S, Roe DJ, et al: Efficacy of
lenalidomide in myelodysplastic syndromes. N Engl
J Med 352:549-557, 2005
55. List A, Dewald G, Bennett J, et al: Lenalido-
mide in the myelodysplastic syndrome with chromo-
some 5q deletion. N Engl J Med 355:1456-1465,
2006
56. Raza A, Reeves JE, Feldman EJ, et al: Long
term clinical benefit of lenalidomide treatment in
patients with myelodysplastic syndrome without del
5q cytogenetic abnormalities. Blood 108:78a, 2006
(abstr)
57. Kantarjian H, Issa JP, Rosenfeld CS, et al:
Decitabine improves patient outcomes in myelodys-
plastic syndromes. Cancer 106:1794-1803, 2006
58. Silverman LR, Demakos EP, Peterson BL, et
al: Randomized controlled trial of azacitidine in pa-
tients with the myelodysplastic syndrome: A study
of the Cancer and Leukemia Group B. J Clin Oncol
20:2429-2440, 2002
59. Silverman LR, McKenzie DR, Peterson BL, et
al: Further analysis of trials with azacitidine in pa-
tients with myelodysplastic syndrome: Studies
8421, 8921, and 9221 by the Cancer and Leukemia
Group B. J Clin Oncol 24:3895-3903, 2006
60. Kantarjian H, Oki Y, Garcia-Manero G, et al:
Results of a randomized study of 3 schedules of
low-dose decitabine in higher risk myelodysplastic
syndrome and chronic myelomonocytic leukemia.
Blood 109:52-57, 2007
61. Egger G, Liang G, Aparicio A, et al: Epigenet-
ics in human disease and prospects for epigenetic
therapy. Nature 429:457-463, 2004
62. Bolden JE, Peart MJ, Johnstone RW: Antican-
cer activities of histone deacetylase inhibitors. Nat
Rev Drug Discov 5:769-784, 2006
63. Hellstrom-Lindberg E, Gulbrandsen N, Lindberg
G, et al: A validated decision model for treating the
anaemia of myelodysplastic syndromes with erythro-
poietin ⫹ granulocyte colony-stimulating factor: Sig-
nificant effects on quality of life. Br J Haematol
120:1037-1046, 2003
64. Saunthararajah Y, Nakamura R, Wesley R, et
al: A simple method to predict response to immu-
nosuppressive therapy in patients with myelodys-
plastic syndrome. Blood 102:3025-3027, 2003
1915