SEARCH

SEARCH BY CITATION

Keywords:

  • epoetin;
  • myelodysplastic syndromes;
  • granuylocyte–colony-stimulating factor;
  • granulocyte macrophage–colony-stimulating factor;
  • anemia

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. Conflict of Interest Disclosures
  8. References

BACKGROUND:

Epoetin α (EPO) continues to be the initial treatment of choice for most anemic patients with myelodysplastic syndromes (MDS). Over the years, different therapeutic strategies have been adopted to optimize the clinical benefits of EPO in this setting.

METHODS:

In the current meta-analysis of published literature, erythroid response (ER) rates with EPO as a single agent versus its combination with granulocyte–colony-stimulating factor (G-CSF) or granulocyte-macrophage–colony-stimulating factor (GM-CSF) were compared.

RESULTS:

The assessment indicated that the ER rates were comparable between the 2 EPO-based therapeutic strategies. Furthermore, EPO monotherapy at a higher dose of 60,000 to 80,000 U weekly produced significantly higher ER rates (64.5%) compared with the standard oncology dose of 30,000 to 40,000 U weekly either as a single agent (49%; P < .001) or in combination with G-CSF/GM-CSF (50.6% P = .007). In addition, when transfusion-dependent patients were assessed separately, both EPO monotherapy and its combination with G-CSF/GM-CSF produced comparable and appreciable levels of transfusion independence (28.8% and 24.8%, respectively).

CONCLUSIONS:

In the current meta-analysis, higher doses of EPO demonstrated better ER rates compared with EPO at standard doses alone or in combination with G-CSF/GM-CSF. Furthermore, the authors concluded that prospective clinical studies are warranted to evaluate the use of higher doses of EPO in anemic patients with MDS. Cancer 2009. © 2009 American Cancer Society.

Refractory anemia in myelodysplastic syndromes (MDS) is a well established clinical hallmark observed in the vast majority of patients. Despite the availability of 3 therapeutic agents approved for the treatment of MDS by the United States Food and Drug Administration (FDA) (lenalidomide, 5-azacytidine and decitabine), for the majority of patients, erythropoietin continues to be the initial choice of treatment.1 Recombinant human erythropoietin or epoetin α (EPO) was the first and had remained the most widely used erythropoiesis-stimulating agent (ESA) in MDS since the early 1990s. In a previous meta-analysis, we demonstrated that the erythroid response (ER) rates in more recent studies have improved significantly, primarily with the use of standardized response evaluation criteria from the International Working Group (IWG) in MDS.2 In addition, the appropriate selection of patients with low erythrocyte (RBC) transfusion requirement and low endogenous serum erythropoietin (sEPO) levels also have enhanced the ER rates.2, 3

Historically, although the clinical benefit of EPO therapy always was appreciated in MDS, the ER rates observed with EPO monotherapy were modest (approximately 15%-20%).4 In an attempt to improve the response to EPO therapy, white blood cell growth factors, such as granulocyte– or granulocyte-macrophage–colony-stimulating factors (G-/GM-CSF), have been combined with EPO. The initial rationale for this combination was that cytopenia of granulocytic and/or megakaryocytic lineage often accompany anemia in patients with MDS. However, subsequently, a direct effect of G-CSF was demonstrated in suppressing apoptotic death and promoting the survival of early hematopoietic progenitors derived from the bone marrow of patients with MDS.5

Because of the paucity of direct comparative studies, it is difficult to discern the true clinical benefit of combination therapy over EPO alone. Moreover, it has been demonstrated that a higher dose of EPO as a single agent (60,000-80,000 U per week) also improved ER rates compared with a standard weekly oncology dose of 30,000 to 40,000 U.3, 6 With the availability of a patient selection model and standardized IWG response evaluation criteria, it would be interesting to study which therapeutic strategy yields a better response to EPO. In the current meta-analysis, we attempted to draw a comparative profile between EPO monotherapy versus EPO in combination with G-/GM-CSF. In addition, a separate assessment of ER rates in transfusion-dependent patients was conducted.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. Conflict of Interest Disclosures
  8. References

Literature Search

We performed a systematic search of the medical literature for studies evaluating ER rate of various treatments in patients with MDS during a period of 1990 to 2007 using the PubMed database. The search was conducted using the following combination of keywords:

  • “Myelodysplastic or myelodysplasia or MDS”; AND

  • “Treatment or therapy or monotherapy or polytherapy or agent”; AND

  • “Erythroid or response”; AND

  • “G-CSF or GM-CSF” or “erythropoietin or epoetin” or “thalidomide or IL-3 or interleukin-3 or retinoic acid or tretinoin or pentoxifylline or ciprofloxacin or steroid or antithymocyte globulin or lenalidomide or 5-azacytidine or decitabine or (cytarabine or cytosine arabinoside and low dose) or etanercept or infliximab or amifostine or arsenic trioxide.”

The search cutoff date was September 30, 2007. An additional search for abstracts from the American Society of Clinical Oncology (ASCO) and the American Society of Hematology (ASH) proceedings also was performed.

We conducted this wider review of the literature with several study objectives in mind. For the current research, we identified the studies that evaluated epoetin α as a single agent or in combination with G-/GM-CSF. The results of the search initially were analyzed in title and abstract format. Reference sections from the selected publications were checked for additional articles or abstracts. Items were selected based on an initial search for a full-text analysis to determine the eligibility of the article for our analysis.

Criteria for Inclusion of Studies for Analysis

We included all studies and abstracts that evaluated the effectiveness of epoetin α as a monotherapy or in combination with G-CSF or GM-CSF for the treatment of anemia in MDS. The only additional criterion applied was the availability of the ER rate to assess treatment effectiveness in patients who were naive to erythropoietic treatment. Studies that were published in languages other than English were included in this review if the relevant data for the analysis were available from the abstract.

Data Extraction

Two reviewers independently assessed the quality of the data collected. Each reviewer evaluated the relevant data from eligible studies and entered the information electronically into an Excel data collection form with prespecified fields. If relevant data were reported only graphically, then values were estimated by measuring the charts. Quality control was done by comparing the 2 independent datasets, and any differences noted were reconciled by a third party by referring to the original sources. Data on the study design and outcomes were collected for each included study. These included the number of patients enrolled, the study design (eg, definition of ER), baseline characteristics of patients, dosing regimen, and the ER rate.

Analytical Approach

Only those studies that used IWG 2000 response evaluation criteria or IWG-like criteria were stratified by treatment group as standard-dose EPO (30,000-40,000 U per week) alone or in combination with G-/GM-CSF or as single-agent EPO at high dose (60,000-80,000 U per week). For the first objective, the 3 groups were compared for overall ER rate and for major ERs. Our second objective was to assess the ER rate specifically in transfusion-dependent patients with MDS comparing EPO monotherapy versus EPO in combination with G-/GM-CSF.

The ER rate was the outcome measure for our meta-analyses. According to IWG response criteria, an ER is defined as either major (ie, an increase ≥2 g/dL in the hemoglobin [Hb] level from baseline and/or complete transfusion independence) or minor (ie, an increase from ≥1 to <2 g/dL in the Hb level from baseline and/or a reduction ≥50% in transfusion requirements) response that is sustainable for 2 months.7

Univariate baseline statistics were generated for all studies. Frequency counts and percentages were used to summarize categorical variables, whereas means and ranges were used to describe continuous variables. Statistical comparisons between groups were conducted using chi-square tests for categorical variables and 2-sided Student t tests for continuous variables.

Meta-Analysis

Pooled estimates of ER rates and 95% confidence intervals (CIs) were calculated for each group using fixed-effects or random-effects meta-analysis methods, depending on the heterogeneity test.8 Under the random-effects model, it is assumed that, in addition to variability within studies, there also is variability among studies. This interstudy variability was assessed using the I2 statistic and was tested by the heterogeneity test, in which a low P value implies that the study results are heterogeneous. In such cases, a random-effects model is more appropriate than a fixed-effects model.

A 2-sided α error of .05 was used to determine statistical significance. All statistical analyses were performed using SAS software (release 9.1 or newer; SAS Institute, Inc., Cary, NC) or Intercooled Stata 9.0 software.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. Conflict of Interest Disclosures
  8. References

In total, 15 studies with 741 evaluable patients were included in the current meta-analysis, as shown in Figure 1.6, 9-22 Twelve of these 15 studies (6 EPO monotherapy and 6 EPO plus G-/GM-CSF) were assessed further for the benefits of EPO in transfusion-dependent disease.6, 9-13, 17-22 In total, 307 evaluable patients were categorized as transfusion dependent in these 12 studies, representing 46.3% (307 of 663 patients) of the overall evaluable population. The studies were stratified for analysis as standard-dose EPO (30,000-40,000 U per week) as a single agent (n = 5 studies; 393 evaluable patients), standard-dose EPO in combination with G-/GM-CSF (n = 6 studies; 152 evaluable patients), or high-dose EPO (60,000-80,000 U per week) as a single agent (n = 4 studies; 196 evaluable patients). Table 1 provides baseline characteristics of these study populations. The average initial weekly dose of EPO in these 3 categories was 32,445 U per week, 34,213 U per week, and 78,740 U per week, respectively. The proportion of patients with refractory anemia (RA) or refractory anemia with ringed sideroblasts (RARS) patients was significantly greater in those who received high-dose EPO compared with those who received standard-dose EPO or standard-dose EPO plus G-/GM-CSF (84% vs 69% or 75%, respectively; P ≤ .05). The mean endogenous serum EPO (sEPO) levels were comparable between the patients who received standard-dose EPO plus G-/GM-CSF and the patients who received high-dose EPO (P = .4097), and sEPO levels were significantly lower in both groups compared with the patients who received standard-dose EPO (P ≤ .05). In addition, it is worth noting that the proportion of transfusion-dependent patients was greater among the patients who received the combination of standard-dose EPO plus G-/GM-CSF compared with those who received standard-dose or high-dose EPO monotherapy (76% [P < .001] vs 35% or 39%, respectively).

thumbnail image

Figure 1. Study disposition. ASCO indicates American Society of Clinical Oncology; ASH, American Society of Hematology; MDS, myelodysplastic syndrome; IWGc, International Working Group criteria; G-/GM-CSF, granuylocyte–/granulocyte macrophage–colony-stimulating factor.

Download figure to PowerPoint

Table 1. Baseline Characteristics of Patients Treated With Different Therapeutic Strategies Using Epoetin α
CharacteristicStd-Dose EPOStd-Dose EPO+G-/GM-CSFHigh-Dose EPO
  • Std indicates standard; EPO, epoetin α; G-/GM-CSF, granulocyte–/granulocyte macrophage–colony-stimulating factor; RA, refractory anemia; RARS. refractory anemia with ringed sideroblasts; Hb, hemoglobin; sEPO, serum endogenous erythropoietin.

  • *

    The distribution in the corresponding group was significantly different at P ≤ .05 compared with the standard-dose EPO group.

  • The distribution in the high-dose EPO group was significantly different at P ≤ .05 compared with the standard-dose EPO+G-/GM-CSF group.

Starting EPO dose, U/wk30,000-40,00030,000-40,00060,000-80,000
No. of studies564
No. of enrolled patients406181213
No. of evaluable patients393152196
RA/RARS (range), %69 (53-100)75 (47-81)84 (68-100)*
Women, % (range)46 (38-75)43 (25-58)51 (27-64)
Transfusion-dependent patients (range), %35 (25-83)76 (37-100)*39 (18-61)
Mean age (range), y71.2 (62-74)69.2 (62-73)70.5 (65-74)
Mean baseline Hb (range), g/dL8.7 (7.6-10.7)8.5 (8.2-8.8)8.6 (8.2- 8.8)
Mean sEPO (range), mU/mL403.8 (300-418)167.7 (49-354)*70.1 (44-129)*
Initial EPO wkly dose (range), U32,445 (30,000-40,000)34,213 (13,000-49,000)78,740 (74,000-80,000)*

Erythroid Response to Different Strategies of Epoetin α Treatment

All studies that were included in the current meta-analysis evaluated ER rates using IWG 2000 or IWG-like criteria.7 According to these criteria, a major ER was defined as an increase ≥2 g/dL in Hb and/or complete transfusion independence, whereas a minor ER was defined as an increase from ≥1 to <2 g/dL in Hb and/or a reduction ≥50% in the transfusion requirement compared with the pretreatment level or requirement. A response had to be sustainable for 2 months.

The pooled overall ER rates were comparable when the 2 standard EPO strategies were assessed. Figure 2 shows that the ER rate with standard-dose EPO monotherapy was 49% (95% CI, 44%-54%), whereas the rate with a combination of standard-dose EPO plus G-/GM-CSF was 50.6% (95% CI, 43%-58%; P = .731). In contrast, the high-dose EPO studies revealed a significantly higher ER rate of 64.5% (95% CI, 58%-71%) compared with the rate for standard-dose EPO monotherapy (P < .001) or standard-dose EPO plus G-/GM-CSF (P = .007). In parallel, the major ER rates also were highest in high-dose EPO studies among the 3 therapeutic strategies studied (P < .01 vs both standard-dose EPO alone and standard-dose EPO plus G-/GM-CSF). In all studies taken together, a univariate metaregression analysis revealed that greater proportions of patients who had RA or RARS (P = .001) and lower sEPO levels (P < .001) were associated with significantly higher ER rates.

thumbnail image

Figure 2. Comparison of the pooled estimates of erythroid response (ER) with different therapeutic strategies using epoetin α (EPO). An asterisk indicates that major ER rates were available from 5 of 6 studies. Std., standard; G-/GM-CSF, granuylocyte–/granulocyte macrophage–colony-stimulating factor.

Download figure to PowerPoint

Table 2 shows that the ER rates were compared between EPO monotherapy versus EPO plus G-/GM-CSF in different French-American-British (FAB) histologic categories. Because of the limited availability of data on individual FAB groups in different studies, the standard-dose and high-dose EPO monotherapy groups were analyzed as a single comparator against combination therapy. With both therapeutic strategies, as expected, patients who had RA or RARS had higher ER rates compared with the other FAB groups taken together. Within the individual FAB groups tested, however, the overall ER rates were comparable between patients treated with EPO monotherapy versus EPO plus G-/GM-CSF.

Table 2. Crude Percentage of Erythroid Response by French-American-British Subtypes
VariableEPO Monotherapy: High and Std DoseStd-Dose EPO+G-/GM-CSFP
  • EPO indicates epoetin α; Std, standard; G-/GM-CSF, granulocyte–/granulocyte macrophage–colony-stimulating factor; RA, refractory anemia; RARS. refractory anemia with ringed sideroblasts; RAEB, refractory anemia with excess blasts; RAEB-t, refractory anemia with excess blasts in transformation; CMML. chronic myelomonocytic Leukemia.

  • *

    The N represents the total number of evaluable patients in each French-American-British subtype category.

All patients   
 No. of studies65 
 No. of patients401141 
 Erythroid response, n/N (%)*   
  RA89/179 (49.7)29/56 (51.8).7874
  RARS53/97 (54.6)24/48 (50).5983
  RAEB/RAEB-t/CMML32/125 (25.6)14/37 (37.8).1470
Transfusion-dependent patients   
 No. of studies43 
 No. of patients5156 
 Erythroid response, n/N (%)*   
  RA14/25 (56)9/23 (39.1).2425
  RARS3/8 (37.5)7/20 (35).9007
  RAEB/RAEB-t/CMML6/18 (33.3)1/13 (7.7).0920

Erythroid Response in Transfusion-Dependent Patients

It is well established that the clinical benefit of EPO is more pronounced in anemic patients with MDS who do not require regular RBC transfusion. However, transfusion independence with EPO therapy in regularly transfused patients also has been noted previously. Therefore, in the current meta-analysis, we also assessed the clinical benefit of adding G-/GM-CSF to EPO specifically in transfusion-dependent patients. The data on transfusion dependent patients were derived from 6 EPO monotherapy studies6, 9-13 (4 standard-dose EPO studies and 2 high-dose EPO studies) and from all 6 studies that used the combination therapy.17-22 Because of the limited availability of data, the standard-dose and high-dose EPO monotherapy groups were analyzed as a single comparator against the combination therapy.

There were 192 evaluable transfusion-dependent patients in 6 studies that used EPO monotherapy and 115 evaluable patients in 6 studies that used EPO plus G-/GM-CSF. The mean pretherapy RBC transfusion requirement per month in these patients in the 2 study groups was comparable (2.37 U vs 2.33 U, respectively). In addition, the mean (±standard deviation) sEPO levels were comparable between the 2 groups (EPO monotherapy, 374.3 ± 72.2 mU/mL; EPO plus G-/GM-CSF, 288 ± 172.1 mU/mL). The initial mean(±standard deviation) weekly dose of EPO was 49,326 ± 22,752 U in the monotherapy group and 34,789 ± 12,083 U in the combination therapy group.

Figure 3 shows that there was no significant difference in the overall ER rates elicited by EPO as a single agent or in combination with G-/GM-CSF (41.6% vs 36.4%, respectively; P = .643). Moreover, in both groups, similar proportions of patients achieved major ER or complete transfusion independence (28.8% vs 24.8%, respectively; P = .705). In addition, when the ER rates were assessed according to FAB categories in transfusion-dependent patients, no significant differences were noted between the 2 treatment strategies (Table 2).

thumbnail image

Figure 3. Comparison of erythroid response (ER) rates with epoetin α (EPO) monotherapy versus EPO plus granuylocyte–/granulocyte macrophage–colony-stimulating factor (G-/GM-CSF) in transfusion-dependent patients.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. Conflict of Interest Disclosures
  8. References

For well over 15 years, erythropoietin has been used therapeutically in patients with MDS to alleviate the major symptom of these disorders, namely RA. Several attempts have been made in the past to improve ER rates and optimize the clinical benefit of this therapeutic agent in MDS. Currently, it is used as a single agent at a standard oncology dose of 30,000 to 40,000 U per week; or as a single agent at higher doses of 60,000 to 80,000 U per week; or in combination with white blood cell growth factors, G-CSF, or GM-CSF. To our knowledge, with the exception of 1 study on epoetin β,23 no information is available on how these strategies of using EPO therapy compare in terms of the ER rate in patients with MDS. Our current meta-analysis demonstrates clearly that both the overall ER rate and the major ER rate were comparable between standard-dose EPO monotherapy and standard-dose EPO plus G-/GM-CSF. Conversely, increasing the dose of EPO to 60,000 to 80,000 U per week as a single agent appeared to yield a higher ER rate than either of the standard-dose EPO strategies.

The results of the current meta-analysis support a recent meta-analysis by French investigators, who compared patient-level data that indicated no increase in the ER rate when an ESA was combined with G-CSF.24 Furthermore, in the French study, an ESA was used at a higher dose (60,000 U per week EPO or epoetin β or 300 μg per week darbepoetin α), the overall ER rate reported using IWG 2000 criteria was 62%, and the major ER rate was 40%. In parallel, the high-dose EPO group in our analysis had an overall ER rate of 64.5% and a 44.9% major response rate, corroborating the observations reported in the French study.

Although currently there does not appear to be a prospective trial directly comparing the combination of EPO plus G-CSF with the combination of EPO plus GM-CSF, the previous meta-analysis of independent clinical data on the 2 combinations underscored the higher efficacy of G-CSF over GM-CSF in every histologic subgroup of MDS.25 In the current analysis, in the combination therapy group, there were 3 studies each that used EPO plus G-CSF (96 patients) and EPO plus GM-CSF (56 patients). Although it was shy of statistical significance, with the G-CSF combination, a higher ER rate (55.2%) was observed, similar to the report by Kasper et al, compared with the GM-CSF combination (ER rate; 42.8% P = .142).

It is well established that requiring a transfusion is a poor prognostic factor for response to EPO therapy. According to the ER prediction model proposed by Hellstrom-Lindberg and coworkers,26 if the transfusion requirement is >2 U of RBC per month, then the ER rate may be affected significantly. The studies that were included in our meta-analysis had a sizeable proportion of patients who reportedly were transfusion dependent. Although the percentage of transfusion-dependent patients appeared to be higher in the EPO plus G-/GM-CSF group, the mean pretherapy RBC transfusion requirement per month was comparable both in the EPO monotherapy groups (2.37 U per month) and in the EPO plus G-/GM-CSF group (2.33 U per month). Unfortunately, the pretransfusion Hg levels were not reported in the reports that were included in the current meta-analysis. Thus, no comparison could be drawn on the trigger for transfusion initiation between the groups. However, in both groups, the mean sEPO levels were <500 mU/mL. Thus, according to the response prediction model described above, with a transfusion requirement >2 U per month and an sEPO level <500 mU/mL, the expected ER rate would be approximately 23%.26 Indeed, our assessment produced ER rates of 42% and 36.4% with monotherapy and combination therapy, respectively, and approximately 25% of patients achieving complete transfusion independence in both groups. This observation is supported by the earlier patient-level retrospective analysis by the Nordic group, which reported a 29% transfusion independence rate with EPO plus G-CSF.27 In addition, when only the transfusion-dependent patients were considered, other agents (including decitabine, arsenic trioxide, and thalidomide) appeared to yield comparable or lower levels of transfusion independence.28-30 Moreover, lenalidomide in patients without chromosome 5q deletion yielded comparable level of transfusion independence (approximately 25%).31 However, in patients with 5q− abnormalities, lenalidomide appeared to yield a significantly higher proportion of transfusion independence (>65%) and, hence, is an approved therapy for this category of low-risk patients with 5q deletion.32 In our analysis, lower sEPO levels, at least in part must have overcome the adverse effect of transfusion dependency and, in turn, contributed to the appreciable responses observed in transfusion-dependent patients with both EPO-based strategies studied. In general, the lower sEPO levels in our meta-analysis were associated with higher ER rates. Otherwise, with sEPO levels >500 mU/mL and regular transfusion dependency, the response prediction model would estimate an ER rate <10% on EPO with or without G-CSF.26

Along with a low RBC transfusion requirement and low sEPO levels, the probability of achieving a response to EPO therapies appears to be greater if the bone marrow blast count is <10%.24 Thus, patients primarily in low-risk and intermediate-1–risk categories according to the International Prognostic Scoring System appear to be appropriate candidates for EPO therapy. Unfortunately, because of the lack of adequate data, it was not possible to analyze this parameter in our current literature meta-analysis. Alternatively, however, according to the FAB diagnostic classification, the RA or RARS category, which generally includes patients with less advanced disease compared with other FAB categories, was associated positively with the ER rate.24

Safety parameters were not included in the current meta-analysis. However, 2 publications on the long-term follow-up of Nordic and French patients who received ESAs with or without G-CSF have reported no adverse effect on leukemic progression compared with historic controls who received supportive care only.24, 27 In addition, whereas the Nordic study also reported no adverse effect on the overall survival of patients who received an ESA with or without G-CSF, the French study demonstrated a significantly reduced hazard ratio for treated patients and especially a benefit in patients who responded to an ESA with or without G-CSF. Similarly, another recent meta-analysis specifically in low-risk to intermediate-1–risk patients reported no adverse effect of hematopoietic growth factor treatment on the rate of either leukemic transformation or survival.33 In light of several deleterious effects of prolonged RBC transfusions in patients with MDS, as reviewed previously,34, 35 the objectives of treatment in patients who have low-risk to intermediate-1– risk disease are to delay transfusion dependence or avoid transfusions in those who require them regularly. On the basis of the current meta-analysis, EPO-based therapies appear to fulfill these therapeutic objectives.

It must be acknowledged that, as listed in Table 1, a relatively greater proportion of transfusion-dependent patients in the EPO plus G-/GM-CSF group and patients with RA or RARS who had the lowest sEPO levels in the high-dose EPO group may have influenced the ER rates in the current meta-analysis. Furthermore, this study, similar to meta-analyses in general, has some additional limitations. First, only EPO studies were considered for the current analysis. Studies that used either epoetin β (an ESA that is not available in the United States) or darbepoetin α were not included because of the very small number of available reports with these agents. Second, 1 study15 from peer-reviewed abstracts/posters was included in the current analysis. A differential distribution of such reports in different study groups may introduce an information bias. Third, although meta-analysis has the ability to improve the power of small or inconclusive studies, it cannot improve the quality of reporting of the original studies. Fourth, the lack of primary source data from the original studies is also another limitation to this analysis, because an ecologic bias may influence results.36 In this analysis, the original studies were weighted based on standard meta-analysis methods,8 which tend to favor studies with a larger sample size. However, the quality of the original studies (eg, randomized vs observational design) was not considered. Furthermore, the total number of studies on EPO monotherapy or its combination with G-/GM-CSF in MDS was relatively small, and data regarding reported durations of follow-up were limited, suggesting that these results may be considered hypothesis-generating. Finally, the studies reported in this meta-analysis included very little data regarding the long-term safety of using ESAs in the setting of MDS. Further prospective, randomized, comparative trials are necessary to compare the validity of the current results and to evaluate long-term safety.

Among all patients with newly diagnosed MDS, as many as 60% to 70% may have low-risk to intermediate-1–risk disease and are suitable for EPO-based therapy.37 Two previous patient-level meta-analyses in independent demographic populations and our current literature meta-analysis, which included results from diverse demographic populations, demonstrated that the efficacy of EPO is well established.24, 27 The patient-level meta-analyses also demonstrated that the rates of leukemic transformation and overall survival in these patients are not affected by EPO-based treatment24, 27; in fact, there may be a survival advantage in responding patients.24 Our meta-analysis also indicated that a higher dose of EPO alone actually may produce a better response compared with the combination of a standard dose with G-/GM-CSF in anemic patients with MDS. This possibility needs to be tested in a prospective trial in the future.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. Conflict of Interest Disclosures
  8. References

We thank Marie-Helene Lafeuille and Dominick Latremouille-Viau of Analysis Group, Inc. along with Behin Yektashenas of Ortho McNeil Janssen Scientific Affairs, LLC for their assistance in this project.

Conflict of Interest Disclosures

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. Conflict of Interest Disclosures
  8. References

Financial support for this study was provided by Centocor Ortho Biotech Services, LLC.

Patrick Lefebvre, Francis Vekeman, and Mei Sheng Duh have received research grants from Centocor Ortho Biotech Services, LLC.

Suneel Mundle, Ruchi Rastogi, and Victor Moyo are employees of Centocor Ortho Biotech Services, LLC.

References

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. Conflict of Interest Disclosures
  8. References