New drugs in the treatment of myelodysplastic syndromes: are they changing the role of transfusion support?

Introduction

Myelodysplastic syndromes (MDS) are a group of clonal disorders of haematopoietic stem cells characterised by ineffective haematopoiesis¹ and a tendency to evolve into acute myeloid leukaemia (AML)² . Anaemia is often the sole cytopenia at diagnosis and patients, therefore, present with symptoms due to low haemoglobin levels. Neutropenia and thrombocytopenia generally follow later during the course of the disease, but may be present at the beginning, sometimes as the initial laboratory finding³. So far only bone marrow transplantation can cure MDS, but this can be offered only to a minority of patients, because the advanced mean age of MDS patients makes the risk of the procedure unacceptable.

As patients with MDS may have different life expectancies, a great deal of effort has been made in order to distinguish the different entities within MDS subtypes. The French-American-British (FAB) classification⁴ was used for a long time, but MDS are currently classified according to the WHO classification which takes into account the number of dysplastic cytopenias, morphology of cells and the presence of blasts and their type⁵ (Table I). In 1997 Greenberg et al.⁶ proposed a risk (of evolution toward acute myeloid leukemia) classification of MDS [the International Prognostic Scoring System (IPSS)] which identifies four classes [low, intermediate-1 (Int-1), intermediate-2 (Int-2), and high risk] (Table II), according to a score based on the number of cytopenias (neutrophils < 1800/mL, Hb <10 g/dL, platelets < 100,000/mL), chromosome abnormalities and number of blasts (Table II). The time free from leukaemia is clearly different between the four classes–in particular it decreases significantly and progressively from high-risk to Int-2 patients–thus enabling the relevant treatment to be chosen accordingly (Figure 1). Patients in the low and Int-1 risk classes may take advantage of support therapy mainly aiming at correcting anaemia, while more aggressive treatments are justified for patients in the Int-2 and high-risk classes. Recently modifications of these criteria based on transfusion requirements and cytogenetics and blast patterns have been suggested⁷.

Disease	Blood findings	Bone marrow findings
Refractory anaemia	Anaemia No or rare blasts < 1x109/L monocytes	Erythroid dysplasia only <10% dysplastic granulocytes or megakaryocytes < 5% blasts < 15% ringed sieroblasts
Refractory anaemia with ringed sideroblasts (RARS)	Anaemia No blasts	Erythroid dysplasia only <10% dysplastic granulocytes ≥ 15% ringed sieroblasts < 5% blasts
Refractory cytopenia with multilineage dysplasia (RCMD)	Cytopenias (bicytopenia or pancytopenia) No or rare blasts No Auer rods < 1x109/L monocytes	Dysplasia in ≥ 10% of cells in two or more myeloid cell lines < 5% blasts in marrow No Auer rods < 15% ringed sieroblasts
Refractory cytopenia with multilineage dysplasia and ringed sieroblasts (RCMD-RS)	Cytopenias (bicytopenia or pancytopenia) No or rare blasts No Auer rods < 1x109/L monocytes	Dysplasia in ≥ 10% of cells in two or more myeloid cell lines ≥ 15% ringed sieroblasts < 5% blasts No Auer rods
Refractory anaemia with excess blasts–1 (RAEB-1)	Cytopenias < 5% blasts No Auer rods < 1x109/L monocytes	Unlineage or multilineage dysplasia 5%–9% blasts No Auer rods
Refractory anaemia with excess blasts–2 (RAEB-2)	Cytopenias 5%–9% blasts Auer rods ± < 1x109/L monocytes	Unilineage granulocyte or megakaryocyte dysplasia 10%–19% blasts Auer rods ±
Myelodysplastic syndrome, unclassified (MDS-U)	Cytopenias No or rare blasts No Auer rods	Unilineage granulocyte or megakaryocyte dysplasia < 5% blasts No Auer rods
MDS associated with isolated del (5q)	Anaemia < 5% blasts Platelets normal or increased	Normal to increased megakaryocytes with hypolobulated nuclei < 5% blasts No Auer rods Isolated del (5q)

Table IIA.

Myelodysplastic syndromes: International Prognostic Scoring System (IPSS) (Greenberg et al. 1997)

Prognostic variable	0	0.5	1.0	1.5	2.0
BM blast %	<5	5–10	-	11–20	21–30
Karyotype*	Good	Intermediate	-	Poor	-
Cytopenias	0/1	2/3	-	-	-

^*Good: normal, Y, del (5q), del (20q); Poor: complex (≥ 3 abnormalities) or chromosome 7 abnormalities; Intermediate: other abnormalities

Table IIB.

Survival and risk of AML evolution by International Prognostic Scoring System (IPSS) score

IPSS Risk Group
	Low	int-1	Int-2	High
Score	0	0.5–1.0	1.5–2.0	≥ 2.5
Lifetime AML evolution	19%	30%	33%	45%
Median years to AML	9.4	3.3	1.1	0.2
Median survival (years)	5.7	3.5	1.2	0.4

^*Good: normal, Y, del (5q), del (20q); Poor: complex (≥ 3 abnormalities) or chromosome 7 abnormalities; Intermediate: other abnormalities

Figure 1.

Survival and freedom from acute myeloid leukemia in patients with MDS divided according to their IPSS-defined risk group

Finally, criteria have been published (and recently reviewed) permitting a more correct evaluation of the response to treatment in MDS and other clonal myeloid disorders, which is particularly useful when data from clinical trials are to be evaluated^8,9 .

Together with a complete haematological evaluation of each patient, clinical aspects including age and co-morbidities¹⁰ must be taken into account in order to make the most appropriate therapeutic choice.

Transfusion support is still the mainstay of the treatment of MDS, and almost all patients will probably receive red blood cell transfusions, the sole therapeutic option in many cases¹¹. However, in the last decade growth factors, new drugs and old drugs whose mechanisms of action have been partially elucidated have been proposed for the purpose of correcting anaemia and reducing transfusion support, and perhaps even postponing evolution to leukaemia in high-risk patients. This review focuses on the approaches to treatment of MDS patients that have recently proven to modify, at least temporarily, the course of the disease. Induction chemotherapy and bone marrow transplantation are not discussed.

Erythropoiesis-stimulating agents (ESA)

Until the advent of ESA no robust studies had been devised to evaluate the ability of a drug to correct haemoglobin values in MDS. One meta-analysis¹², two open studies^13,14 and one randomised trial¹⁵ demonstrated that recombinant erythropoietin alpha (rHuEpo-a) was effective in improving haemoglobin levels and reducing transfusion needs in patients with low-risk MDS, while response was poor in patients with excess of blasts. The lower overall response rate reported in the meta-analysis by Hellstrom-Lindberg et al.¹² compared to the rates in the other studies^13–15 (see Table III) probably reflects different schedules and dosing. For example, 150 U/Kg rHuEpo-a were used daily in the Italian co-operative study (the only randomised study vs placebo), while the same dosage was used every other day by Terpos et al.¹⁴ and others authors whose studies were considered in the meta-analysis. Moreover different schedules were equally effective, such as a biweekly administration of rHuEpo 40000 U followed by a reduction to 40000 once a week¹⁶. In this study the authors also reported that quality of life, evaluated using validated tests, improved as a consequence of the increased haemoglobin levels.

Table III.

Results of trials of rHuEpo in myelodysplastic syndrome

	Hellstrom- Lindberg 1995 (meta-analysis)	Rose et al. 1996 (open)	Italian Cooperative Group 1998 (randomised)	Terpos et al. *2002 (open)
Overall response (%)	16.1	28	36.8	45.1
RA	21.8	39	50	48.3
RARS	7.5	17.5	37.5	58.4
RAEB	22.7	12.5	16.7	33.8
RAEB II				13
Not transfused	44.4			60
Predictors for response:
sEpo (mU/mL)	< 200	< 100	< 200	< 150 (81% 26 weeks)

^*Response higher in patients with favourable cytogenetics. Prolonged administration increased the probability of response. The median duration of response was 68 weeks

Besides proving that rHuEpo-a can be an effective treatment of anaemia of MDS (increasing haemoglobin level and/or decreasing transfusion support), the studies mentioned above demonstrated that the treatment was most effective if candidate patients presented with: a) refractory anaemia (RA)-RA with ringed sideroblasts (RARS) (FAB classification) or low-risk–Int 1 (IPSS classification) MDS; b) no or reduced transfusion support before treatment; c) a baseline endogenous erythropoietin level = 200 mU/mL.

The combination of rHuEpo with granulocyte colony-stimulating factor (G-CSF) or granulocyte-monocyte colony-stimulating factor (GM-CSF) has been claimed to be more effective than rHuEpo alone^17–21. However response rates were not clearly better than with rHuEpo alone, although in one study it was observed that withdrawal of G-CSF led to loss of response in 50% of patients. More recently Balleari et al.²² reported response rates of 73%, 100% and 83.3 % in patients with, respectively, RA, RARS and refractory cytopenia with multilineage dysplasia (RCMD) (WHO classification) and confirmed that the combination therapy may also be effective in patients with endogenous erythropoietin levels up to 500 mU/mL, as previously suggested by Hellstrom-Lindberg et al.²¹.

Darbepoetin-a, an ESA whose half-life is longer than that of rHuEpo, was also tested in MDS patients. Overall response rates from 40% to 75%, evaluated according 2001 International Working Group (IWG) criteria, were reported in phase II studies using 150-300 ìg once or week or 500 once every 3 weeks with some response observed in patients pre-treated with rHuEpo^23–27. In 2006 Mannone et al.²⁸ reported an erythroid response in 44/62 patients (71%–34 major and 10 minor) after 12 weeks, including 8/13 previous non-responders to conventional erythropoietin. These data were confirmed by a French co-operative group²⁹ in a larger trial in which the 2000 and 2006 IWG response criteria were compared (Table II). For the first time this study demonstrated that darbepoetin also has a favourable impact on the survival of MDS patients compared to the untreated cohort of patients used to design the IPSS risk classification. It was also confirmed that less than 10% blasts, low and Int-1 IPSS class, red blood cell transfusion independence, serum erythropoietin level < 200 UI/L, and shorter interval between diagnosis and treatment (with IWG 2006 criteria only) were predictors of response, which was not affected by the karyotype pattern.

ESA have been demonstrated to be not only an effective but also a safe treatment in MDS patients. In particular, a recent meta-analysis by Ross et al.³⁰ showed that this treatment does not seem to induce a higher frequency of AML. rHuEpo-a and darbepoetin appear to be equally effective in improving anaemia in low-risk MDS patients, although some patients unresponsive to rHuEpo may respond to darbepoetin²⁸; a head-to head study comparing rHuEpo and darbepoetin in the treatment of low-risk MDS is, however, lacking. Finally it must be remembered that although ESA are widely used in MDS, this is still an off-label use, as it has not yet been approved by regulatory authorities.

Lenalidomide

Immunomodulatory and anti-angiogenic agents have been proposed for the treatment of MDS, but their mechanisms of action still remain unclear^31–33. Thalidomide, at doses of 100–400 mg/day, has been used in some clinical trials with mixed results regarding efficacy in increasing haemoglobin levels in MDS patients, but side effects are not negligible^34–38.

Lenalidomide is a 4-amino-glutarimide analogue of thalidomide with potent immunomodulatory and anti-angiogenic activities. List et al.³⁹ reported that lenalidomide (Revlimid) had activity (transfusion independence, >50% reduction of transfusion support, increase in haemoglobin level) in a significant proportion of patients (95/215, 44%) with low-risk MDS in a phase II study. Moreover, they also demonstrated that the response was particularly impressive in patients with 5q- MDS, as 112/148 (76%) such patients had an erythroid response⁴⁰. Furthermore, lenalidomide induced a complete cytogenetic response in 38/85 and a partial response in 24/85 patients, suggesting that the drug is capable of inhibiting the dysplastic clone. Bone marrow toxicity was, however, significant as neutropenia occurred in 65% of the patients and thrombocytopenia in 74%, requiring interruption of treatment in 58% of patients. These studies led to the Food and Drug Administration (FDA) approval of lenalidomide for the treatment of transfusion-dependent patients with low- or Int-1-risk MDS with a chromosome 5q deletion with or without additional cytogenetic abnormalities. Very recently Raza et al.⁴¹ confirmed, in a phase II study, that lenalidomide is also effective in increasing haemoglobin levels and reducing transfusion requirements in non 5q-MDS patients, although its efficacy is limited to a smaller proportion of patients (43% overall rate of haematological improvement).

The use of lenalidomide in MDS patients other than those in the low or Int-1 risk classes seems to be disappointing. Results from a study by Burcheri et al.⁴² showed that this drug was effective in Int-2- or high-risk MDS patients with 5q deletion in the absence of complex cytogenetics and thrombocytopenia, but severe bone marrow toxicity and other complications, including sepsis due to neutropenia (sometimes fatal) and bleeding, occurred in 28/29 evaluable patients (out of 49 enrolled).

A direct comparison between ESA and lenalidomide in low-risk MDS patients is lacking. Both treatments have been proven to be able to improve haemoglobin levels and reduce transfusion requirements. A practical approach could be to use ESA in patients with a normal karyotype or karyotypic abnormalities other than 5q- who present with clinical and laboratory characteristics that make them suitable for this treatment. In 5q- patients lenalidomide should be used as front-line treatment although, while already approved by the FDA, it still awaits approval from European Regulatory Authorities.

Epigenetic therapy

Tumour suppressor genes can undergo abnormal gene silencing through DNA hypermethylation of promoter regions. Re-expressor strategies can take advantage from inhibitors of DNA methyltransferases and inhibitors of histone deacetylases (HDAC). DNA methyltransferase inhibitors are substances with the capacity to reverse DNA methylation, leading to reactivation of abnormally silenced genes and potential elimination of neoplastic clones. Two DNA methyltransferase inhibitors, 5-azacytidine (5-AZA) and 5-aza-20-deoxycytidine, were initially developed for clinical use at cytotoxic levels and with schedules similar to those for cytarabine, whose clinical usage was limited by intense haematological and non-haematological toxicity. In a phase II study, Silverman and the Cancer and Leukemia Group B (CALGB)⁴³ demonstrated that intravenous low-dose 5-AZA (Vidaza) (7 days, total dose of 525 mg/m², repeated every 4 weeks) was active in patients with high-risk MDS, with an overall response rate of 49%, [12% complete remissions] and a median response duration of 14.7 months. Comparable results were obtained in 68 elderly MDS patients in whom the drug was administered by subcutaneous injection⁴⁴. Moreover, the activity of subcutaneous 5-AZA was evaluated in a phase III study in which patients were randomized against best supportive care with the possibility of ‘‘cross-over’’ from best supportive care in the case of progressive disease⁴⁵. In this study the overall response rate was 60% with 7% complete remissions and 16% partial remissions, a median response duration of 15 months, and a median time to response of about four cycles. It is also important to emphasise that these findings were associated with: a) a delay of in the progression to leukaemia (median: 21 months with 5-AZA versus 13 months with supportive care); b) improved survival; c) improved quality of life in the arm treated with 5-AZA⁴⁶.

Decitabine (5-aza-20-deoxycytidine, Dacogen) has also been tested in elderly patients by Wijermans and co-workers in a multicentre phase II study⁴⁷ in which it was able to elicit a response in about 50% of patients, particularly in those presenting with high-risk features, including karyotype, using a 3-day continuous infusion with total doses of 125–225 mg/ m² repeated every 6 weeks. Similarly to 5-AZA, decitabine induced cytogenetic remissions⁴⁸ . In a phase III study carried out by Kantarjian et al.⁴⁹ a total of 170 patients were randomised: 89 patients received 5-aza-20-deoxycytidine in a schedule of 15 mg/m² over 72 h for a total of nine doses, repeated every 6 weeks, while 81 patients received best supportive care. The overall response rate was 30% (9% complete remissions, 8% partial remissions, 13% haematopoietic improvement with a median response duration of 10.3 months compared to 7% haematopoietic improvement in the best supportive care/control group. It was also observed that the time to progression to AML or death was longer in the 5-aza-20-deoxycytidine arm compared to in the arm receiving best supportive care: this difference was statistically significant, especially for IPSS high-risk patients. Results of an European Organisation for Research and Treatment of Cancer (EORTC)/German MDS Study Group phase III study randomising patients with Int-2 or high-risk MDS according to the IPSS are pending.

As regards side effects and toxicity, emesis and skin rashes were more frequently associated with 5-AZA, and myelosuppression was common with both drugs. However these untoward effects were found to be manageable and both treatments appear suitable for elderly patients.

The FDA approved the use of 5-AZA for treatment of all subtypes of MDS and 5-aza-2'-deoxycytidine for patients with Int-1 or high-risk MDS according to the IPSS. The European Medicines Agency (EMEA) have not, so far, approved this usage, and results from a large confirmatory study being carried out in Europe are awaited.

Inhibitors of HDAC have also been tested in MDS. The anticonvulsant drug valproic acid showed some efficacy in MDS patients, particularly in those with an excess of blasts^50,51. The combination of hypomethylating agents and HDAC inhibitors may offer further advantages in the treatment of patients with MDS^52,53.

Conclusions

MDS still await effective treatments capable of producing durable responses. Nevertheless ESA, immunomodulatory and epigenetic drugs offer substantial possibilities of haematological improvement, including the correction of anaemia with the reduction or abolition of transfusion support, and improved quality of life. An adequate evaluation of each single patient, based on WHO and IPSS classifications, and erythropoietin levels and transfusion needs, is the key to making treatment choices. Besides the drugs mentioned previously, alternatives such as arsenic trioxide and amifostine^54–56 can be employed usefully in special subsets of patients and immunosuppression with antithymocyte globulin or cyclosporine A in those presenting with hypoplastic MDS⁵⁷. In a recent study Lim et al. confirmed that antithymocyte globulin is effective in patients with hypocellular bone marrow and low or Int-1 (IPSS) risk MDS, while patients in Int2 or high IPPS risk groups had a poorer survival⁵⁸. Younger age and HLA-DR phenotype were also factors associated with haematological response in a population of 129 patients with MDS, studied by Sloand and coworkers⁵⁹. Thirty-nine of 129 patients (30%) receiving immunosuppressive therapy responded either completely (12/39), including the achievement of transfusion independence, or partially, the highest rate of response being observed in patients who received combination therapy (antithymocyte globulin plus cyclosporine A). Responses were less frequent when antithymocyte globulin (20 out of 48) or cyclosporine A (only 1/13 patients) was used as a single agent. Nevertheless red cell transfusion therapy still plays, and will do so in the foreseeable future, an important role in the management of patients, in particular those with Int-2 or high-risk MDS, whose disease does not respond to or is likely to relapse after the remission obtained with new targeted therapies or induction chemotherapy.

Much of the effort has been devoted to the correction of anemia, but neutropenia also contributes to morbidity and mortality in MDS. However, available data do not recommend a preventive use of GM-CSF or G-CSF in order to raise neutrophil counts and reduce the risk of infections¹¹ . Thrombocytopenia is generally present in advanced MDS, and often requires platelet transfusion to control severe bleeding episodes, but growth factors stimulating platelet production are still not commercially available. Very promising results have, however, recently been reported by Kantarjian et al. who showed, in a phase I/II study, that the peptide antibody AMG 531, an agonist of thrombopoietin receptor, is able to induce a sustained increase in platelet counts and is well-tolerated in patients with low-risk MDS⁶⁰ . These results, if confirmed, will open new perspectives in the treatment of MDS which may take advantage of the combination of drugs acting on different bone marrow cell targets.