Diagnosis and Management of Anemia in Patients With the Myelodysplastic
Syndrome
John M. Bennett, MD, and Peter A. Kouides, MD
While not appropriate for all patients with the myelodysplastic syndrome,
recombinant erythropoietin (EPO) is a possible alternative to red blood cell
transfusion. Specific factors such as the presence of cytopenias, the bone marrow
blast percentage, and cytogenetic findings determine which patients are good candidates
for treatment with EPO.
Diagnosis of the Myelodysplastic Syndromes
Myelodysplastic syndrome (MDS) is composed of a group of diseases
that for decades have been challenging to diagnose and manage.1,2 MDS is a
clonal disease of the bone marrow. The pathologic manifestation of morphologic
abnormalities (dysplasia) include ringed sideroblasts, megaloblastic changes,
pseudo-Pelger cells, and micromega and multinucleated megakaryocytes. The clinical
spectrum ranges from a mild anemia to pancytopenia and acute leukemic evolution.
Pathogenesis of Myelodysplastic Syndrome
A major advance in the understanding of the pathogenesis of MDS has
been the recognition of accelerated programmed cell death (apoptosis) in this disease.3,4
The demonstration of both increased apoptosis by in situ end-labeling of DNA and increased
cell proliferation by bromodeoxyuridine in vivo labeling distinguishes these disorders
from acute myeloid leukemia (AML). To some extent, these observations reconcile the
seemingly opposing findings of marrow hypercellularity seen in the majority of bone
marrows of patients with MDS and the presentation of pancytopenia. This may be related to
the levels of the apoptosis-related oncogene products such as c-myc, which enhances
programmed cell death, and bcl-2, which has the opposite effect.5 Other
cytokines identified as playing a role include tumor necrosis factor alpha, transforming
growth factor beta, and interleukin-1 beta-converting enzyme. Potential therapies to block
the common lipid signaling pathways shared by these cytokines include the use of
pentoxifylline, lisophylline, and ciprofloxacin.
Presenting Symptoms
Most patients with MDS do not have specific abnormalities on
physical examination. The exception is the proliferative phase of chronic myelomonocytic
leukemia. Twenty percent of patients present with gum hypertrophy and splenomegaly or
hepatomegaly. The most common presentation in a patient over the age of 65 years is the
discovery of anemia or pancytopenia without any other explanation. The incidence of MDS is
similar to that of AML (ie, approximately five cases per 100,000 and rising to 30 cases
per 100,000 in patients over age 70 years).6
Before proceeding with a bone marrow aspirate and biopsy in a
patient suspected of having MDS, the possibility of the following conditions should be
considered: vitamin B12 or folate deficiency, exposure to heavy metals, recent cytotoxic
chemotherapy (including patients receiving methotrexate and other immunosuppressive agents
for autoimmune disorders), chronic liver diseases, and human immunodeficiency virus (HIV)
infection (Table 1). Once these conditions have been excluded, the bone marrow examination
becomes of value. Both an aspiration and a biopsy are recommended (the latter to assess
the cellularity, the presence of fibrosis, and islands of immature cells). The presence of
significant dysplasia in any one of the major three cell lines (erythroid, granulocytic,
or megakaryocytic), usually defined as 10% of cells affected, confirms the diagnosis.7
To further classify the type of MDS, an accurate blast percentage must be determined (eg,
less than 5%; 5% to 10%; 11% to 20%; 21% to 30%). An iron stain is necessary to subdivide
the under 5% group into greater or less than 15% abnormal sideroblasts. Prognosis is
inversely related to the percentage of blasts and has been confirmed in many studies.8
For example, patients with refractory anemia with ringed sideroblasts have a median
survival of five years, whereas patients with refractory anemia with excess blasts in
transformation have a median survival of less than one year.
Table 1. -- Diagnostic
Evaluation for MDS |
| Establish serum folate, vitamin B12, ferritin, erythropoietin, serum
Fe/total iron-binding capacity |
| |
| Determine complete blood count, reticulocyte count, red blood cell
indices |
| |
| Rule out renal disorder, thyroid dysfunction, occult bleeding |
| |
| Evaluate bone marrow and biopsy (percentage of blasts, degree of
dysplasia, iron stain for iron stores, and percentage of abnormal sideroblasts) |
Bone marrow cytogenetics are obtained at the time of the marrow
procedure. Several abnormal karyotypes have been described in as high as 60% of primary
MDS and as high as 85% in secondary MDS caused by treatment or known toxic exposure.
Patients with abnormalities including 5q– or 20q– have a good prognosis, whereas
monosomy 7 or complex abnormalities carry a worse prognosis.9
Over the years, many features of MDS have been studied in an attempt
to better clarify the overall survival of patients with MDS. These features include
degrees of dysplasia, the degree of cytopenia, lactic dehydrogenase levels, bone marrow
blast percentage, presence of abnormal localization of immature precursors on the bone
marrow biopsy, marrow immunophenotype (presence of CD34, for example), and marrow
cytogenetics. An International Prognostic Scoring System (IPSS)8 was recently
published that used only three variables: cytopenias, percentage of marrow blasts, and
marrow cytogenetics. More than 800 patients with primary MDS made up this large data base.
Four risk groups were identified with median survivals: (1) low risk, 5.7 years, (2)
intermediate-1 risk, 3.5 years, (3) intermediate-2 risk, 1.2 years, and (4) high risk, 0.4
years. An additional stratification for age less than or greater than 60 years provides an
even better separation for the first two groups, doubling the median survival for the
younger cohort in the low and intermediate-1 groups. It is hoped that this new type of
risk assessment will permit a better comparison of therapies by different study groups.
Management of Anemia in Myelodysplastic Syndromes
Until the availability of recombinant EPO, patients were arbitrarily
given transfusions upon becoming symptomatic. After determining that B12 and folate levels
were normal, alternative interventions (eg, pyridoxine or prednisone) were attempted, but
very few patients responded. Transfusions would start at an approximate hemoglobin level
of 9 g/dL and then be administered at a rate of two units every three to four weeks. Since
each unit of blood yields 250 mg of elemental iron and the body has no way to excrete
excess iron, patients who were given transfusions developed an increased risk of secondary
hemosiderosis and possibly hemochromatosis and its associated phenomena.
Recombinant Erythropoietin
Recombinant erythropoietin (EPO) has provided an attractive
alternative to red cell transfusion for a wide range of anemic patients but not all (Table
2). A recent publication8 has provided a new prognostic index for patients with
MDS. Utilizing the presence of cytopenias, bone marrow blast percentage, and cytogenetic
findings, four distinctive groups were identified as defined previously. Patients with an
IPSS score in the high-risk and intermediate-2 risk groups have a short median survival of
one year or less, and they usually are candidates for intensive chemotherapy rather than
EPO. Patients with IPSS scores in the low-risk to intermediate-1 risk groups are excellent
candidates for EPO.
Table 2. -- Guidelines
for Treating MDS With Erythropoietin |
| Hematocrit <30 OR hemoglobin <10.0 g/dL |
| Transfusion requirement |
| Endogenous EPO level <500 I.U./dL |
Studies with EPO have been numerous and varied. In more than 40
published papers on the use of EPO from the mid-1990s to the present, endogenous EPO
levels prior to the administration of exogenous EPO were widely divergent, and in some
studies, endogenous EPO levels were not measured. Studies included patients who had
transfusions as well as those who never received a transfusion, and EPO was used either
alone or with other growth factors -- most often G-CSF and GM-CSF. Most trials included
patients with ringed sideroblasts/refractory anemia and a few cases of refractory anemia
with excess blasts.
EPO dosages also ranged widely, from a low of 80 U/kg to as high
1,000 to 10,000 U/kg. Schedules have been intravenous or subcutaneous and daily to three
times a week. The duration of treatment has ranged from as little as 12 weeks to longer
than a year.
The numbers in each of these individual studies are small, and some
demonstrated remarkable success while others showed trivial or no success. One of the most
notable studies with EPO was published by Rose and colleagues.10 It employed a
compassionate use protocol approved by the Food and Drug Administration and recombinant
EPO provided by Ortho Biotech, Inc. Patients had to be significantly anemic; those who
were likely to move on quickly to AML were excluded. EPO levels in patients accepted into
the study were not to be above 500, although some subjects had higher levels. Other
criteria were reasonably normal creatinine and correction of overt hemolysis and iron
deficiency. More than 80% of patients had fewer than 5% blasts, providing an opportunity
to observe the effects of EPO over a longer period of time. Ten percent of patients
demonstrated a rise of at least 6 hematocrit points, and an additional 18% had a 50%
decrease in transfusion requirements. The median serum EPO levels of these patients was
134; 86% of the responding cases had EPO levels of 100 or less.
Data from Hellström-Lindberg11 suggest that patients who
have ringed sideroblasts have a response rate to EPO of less than 10%. Using a
meta-analysis approach, responses were seen in 21% of patients with other types of MDS,
with somewhat higher responses when patients received no transfusions or only occasional
transfusions and had low EPO serum levels (under 200 U). A more recent study12
by this group, including American investigators, has combined EPO with G-CSF. The dosages
of EPO ranged between 5,000 to 10,000 U per day SC and G-CSF at 1.0 µg/kg per day SC to
double the pretreatment granulocyte count. Thirty-six percent of patients showed a
response, and no difference occurred among morphologic subgroups, but the highest
responses were again noted in patients who had the lowest EPO levels (<100 U) and fewer
than two transfusions per month (74% response rate). The investigators noted that it may
take up to two to three months before responses occur. The duration of the responses is at
least one year, with much variability in maintenance regimens. Verification of these
results is being undertaken by the Eastern Cooperative Oncology Group, but it will be
several years before more information becomes available.
Quality-of-Life Studies
The ECOG trial is also studying issues involving quality of life.
Its trial design will enable investigators to discern quality of life with improvement of
hemoglobin levels vs quality of life with or without transfusions. Results will reveal
whether simply raising the hematocrit will materially affect other parameters of general
well being. Overall, fewer transfusions are safer and more comfortable for the patient,
and those receiving transfusions are still at risk for complications such as hepatitis,
HIV, viral infection, transfusion reactions, longer hospital stays, and iron overload.
Conclusions
Patients with MDS clearly have an abnormal clone. This situation
stands in contrast with the majority of solid tumor patients who are currently receiving
EPO in whom the bone marrow stem cell is not primarily affected. Whether there is a
residual normal clone in MDS patients or whether we are taking advantage of erythroid
precursors that retain some sensitivity to EPO is currently unknown. Today, the entire
therapeutic armamentarium is designed either to modulate the clone with the use of
cytokines and hormones or to destroy the clone in patients who are suitable for
therapeutic assault with an allogeneic transplant or with intensive chemotherapy.
References
1. Kouides PA, Bennett JM. Myelodysplastic syndromes. In: Abeloff MD, Armitage JO,
Lichter AS, et al, eds. Clinical Oncology. New York, NY: Churchill Livingstone;
1995:1977-1997.
2. Hofmann WK, Ottmann WK, Ganser A, et al.. Myelodysplastic syndromes: clinical
features. Semin Hematol. 1996;33:177-185.
3. Raza A, Gezer S, Mundle S, et al. Apoptosis in bone marrow biopsy samples involving
stromal and hematopoietic cells in 50 patients with myelodysplastic syndromes. Blood.
1995;86:268-276.
4. Yoshida Y, Anzai N, Kawabata H. Apoptosis in myelodysplasia: a paradox or paradigm. Leuk
Res. 1995;19:887-891.
5. Rajapaksa R, Ginzton N, Rott LS, et al. Altered oncoprotein expression and apoptosis
in myelodysplastic syndrome marrow cells. Blood. 1996;88:4275-4287.
6. Aul C, Gatterman N, Schneuder W. Age-related incidence and other epidemiological
aspects of myelodysplastic syndromes. Br J Haematol. 1992;82:358-367.
7. Bennett JM, Catovsky D, Daniel MT, et al. Proposals for the classification of the
myelodysplastic syndromes. Br J Haematol. 1982; 51:189-199.
8. Greenberg P, Cox C, LeBeau MM, et al. International scoring system for evaluating
prognosis in myelodysplastic syndromes. Blood. 1997;89:2079-2088.
9. Mufti GL. A guide to risk assessment in the primary myelodysplastic syndrome. Hematol
Oncol Clin North Am. 1992;6:587-606.
10. Rose EH, Abels RI, Nelson RA, et al. The use of r-HUEpo in the treatment of anaemia
related to myelodysplasia (MDS). Br J Haematol. 1995;89:831-837.
11. Hellström-Lindberg E. Efficacy of erythropoietin in the myelodysplastic syndromes;
an analysis of 205 patients in 17 studies. Br J Haematol. 1995;89:67-71.
12. Hellström-Lindberg E, Negrin R, Stein R, et al. Erythroid response to treatment
with G-CSF plus erythropoietin for the anemia of patients with myelodysplastic syndromes:
proposal for a predictive model. Br J Haematol. 1997;99:344-351.
DR DALTON
Have studies on red cell survival been performed, both before and
after receiving EPO? The question comes up because EPO increases production, but some in
vitro studies also show that EPO may actually be an antiapoptotic hormone. This may be
important in this particular disease.
DR ZUCKERMAN
It would not be an issue of apoptosis in the mature, circulating red
cells. The apoptosis issue would come in the erythroid precursors in the marrow, and that
is what EPO appears to do: reduce the apoptosis. Of course, once you have a nonnucleated
erythroid cell, apoptosis does not come into play.
DR BENNETT
If you are referring to chromium-labeled studies on red cell
survival, I am not aware that any studies have been done. As far as alterations in
apoptosis, the answer is yes. Eva Hellström has shown that before and after the
administration of EPO and G-CSF, there is a marked decrease in apoptosis. Therefore, I
believe that there is some evidence that one is having an impact on the balance between c-myc
and bcl-2, and other factors. Peter Greenberg and his group at Stanford have a
paper that studies bcl-2 evolution. As the disease evolves, with increasing
percentages of blast proliferation and a decrease in apoptosis, the status of bcl-2
increases, as measured on a paraffin model.
DR CRAWFORD
I know of a couple of mechanisms by which you might see that effect
of G-CSF synergy with EPO. The effects of GM-CSF on early stem cells are well known. In
addition, however, G-CSF has an anti-inflammatory effect of reducing TNF-alpha and IL-1.
DR BENNETT
In addition to looking at an effect of erythropoiesis, which would
be measured by looking at stool and urinary protoporphyrins, it would be interesting to
know if a study has been done on the surviving non- nucleated red cells in the peripheral
blood to show whether they have a shortened red cell survival. For example, do they have a
shortened survival in pernicious anemia or folic acid deficiency?
DR ZUCKERMAN
It would also be interesting to know whether, during the course of
erythroid maturation, EPO produces hardier red cell that might survive longer. I am
unaware of any studies in any EPO trials for any disease in which that has been addressed.
DR BENNETT
That could be studied quickly in MDS patients. Data on red cell
survival could be determined prior to the introduction of EPO; then, following treatment,
determine if there is a normal red cell survival or a high-normal red cell survival of
longer than 28 days.
DR ZUCKERMAN
The cleanest study might actually be in the renal failure patients,
where there is simply an EPO deficiency disorder and little else impacting on red cell
survival.
DR CRAWFORD
Regarding desferal, do you supplement iron because these people are
not mobilizing it, or do you give them desferal and then EPO? Are these questions being
addressed in either direction?
DR BENNETT
The ECOG trial gives no recommendations for supplementing iron. The
statement is that patients cannot be iron-deficient, based on ferritin levels and iron
stains in the bone marrow, as well as serum iron and total iron-binding capacity (TIBC).
Anecdotally, we saw a dramatic phenomenon with EPO in a 40-year-old woman with chronic
lymphocytic leukemia. She completed a trial with fludarabine and reversed her
neutrophil/lymphocyte ratio perfectly. She became profoundly anemia as part of this
process. Previously, she had hematocrits in the mid-30s and went down to hematocrit as low
as 24 with no evidence of blood loss. We started recombinant EPO, and she became
iron-deficient, with a mean corpuscular volume that plummeted to approximately 65. A serum
iron at that point was 5, although it previously had been normal. She had not responded to
EPO. She had not become more anemic; her hematocrit remained around 22. We began iron and
kept her on EPO, and her hematocrit at six weeks later was 40. She recovered 5 to 6
hematocrit points every 10 days, while one would expect about 3 hematocrit points every 10
days. We stopped the EPO and continued only with iron. This is the first time I have seen
surreptitious iron deficiency being provoked by the administration of EPO. I believe it is
critical to be sure that a patient is not borderline iron-deficient; if so, iron should be
given. However, as a standard recommendation for patients receiving EPO, we do not
normally recommend iron therapy.
DR LEE
We used EPO without ferrous sulfate when we gave chemoradiation
therapy to non-small-cell lung cancer patients. Our idea was that if the patient was not
iron-deficient, there was normal storage in the bone marrow. However, the mobilization and
utilization of iron may not be normal in patients with cancer. We found that without
ferrous sulfate supplement, our patients with lung cancer developed anemia despite the
therapy with EPO, but when ferrous sulfate was added, only one developed anemia. Even in
the patient population with MDS who have much iron storage in the bone marrow, iron
supplements will be required for the erythropoiesis.
DR BENNETT
There is no reliable information about giving supplemental iron to
patients with malignancies, exclusive of MDS, whose iron stores you expect to be
reasonably normal. Whether the recommendation should be to do a quick survey of iron
stores is, I suppose, open-ended. Many patients with Hodgkin’s disease, for example,
present with a picture that mimics iron deficiency. They can have low serum iron but
usually low-to-normal TIBCs, and sometimes low-normal or elevated ferritin levels. Whether
they will respond to oral iron is unclear; traditionally, they have not, but they might
with EPO administration.
Regarding a leukemia-triggering phenomenon, I believe there is no
evidence to support a potential role for EPO in that. The evidence for other growth
factors is marginal at best. In the few randomized trials in which patients were
administered growth factor vs no growth factor, the evolution of AML was identical. Some
early reports of growth factor that triggered AML were published, but those investigators
agree that the patients were at high risk for evolution to AML and probably were in the
process of developing AML when the growth factor was prescribed.
In patients with MDS who are receiving growth
factors, we see changes in the character of the early precursors, ie, more promyelocytes
and more late myeloblasts. Thus, you can become alarmed as to whether they actually are
evolving. However, if you leave them alone (or stop the growth factor, if you are really
concerned), the appearance reverses. There is little evidence that these growth factors
will trigger an AML evolution in MDS.
From the Department of Pathology and Laboratory Medicine, University of Rochester
Cancer Center, Rochester, NY.
Address reprint requests to John M. Bennett, MD, Department of
Medicine, University of Rochester School of Medicine & Dentistry, Rochester, NY 14620.
Dr. Bennett is a member of Speakers’ Bureau for Ortho Biotech,
Inc.
Back to Cancer Control Journal Supplement Volume 5 Number 2