Background

Hellström-Lindberg score (HLS-1997) is a useful tool to predict the probability of response to erythropoietin (EPO) treatment in anemic MDS patients. HLS response criteria was not assessed according to the IWG criteria published by Cheson et al. in 2006. To date, international guidelines suggest a front-line therapy with high doses of EPO (80000 U/week) in MDS anemic patients with Hb <10 g/dL and endogenous EPO levels <500 mU/mL. In our previous report, we validated the HLS in a single center cohort according to the IWG criteria of response (Molteni et al. Int J Hematol 2013). Of note, observing the response rate of the ''good'' responders group (score 3 or 4), a significant difference between patients with a score 3 vs 4 was detected (p= 0.004), identifying two new categories ("good" and "very good") with different probability of response. The high likelihood of responding with patients with a "very good" score was such as to suggest that even standard therapy doses (40000 U/week) would be effective for this group.

Aim

To validate our previous report on a new cohort of MDS patients treated with erythropoietin.

Patients and methods

Hellström-Lindberg score variables considered were: erythropoietin serum levels in pre-treatment (<100, 100-500 or >500 mU/mL) and RBC transfusion requirements (≥2 or <2 RBC per month). We decided not to consider the patients with "poor" HLS (EPO >500 mU/mL and ≥2 RBC per month) on the basis of our previous report and on the evidence of the very modest efficacy of EPO in these cases. Therefore, we applied a "revised HLS" (figure 1), that distinguishes three prognostic groups: "very good" response (score 4), "good" (score 3) and "intermediate" response (score -1 to +1). With this premise, we retrospectively evaluated 57 "low" or "intermediate-1" IPSS risk MDS patients treated with a standard dose of EPO between 2005 and 2017. According to our score: 15 of them were "intermediate", 12 "good" and 30 "very good". Response to EPO therapy was based on the revised IWG criteria (2006). Statistical analysis was conducted using Fisher's exact test and ordinal logistic regression followed by Wald's test. Then the quality of the response with the different scores was compared using the Bonferrion - Adjusted for multiple comparison test.

Results

Among the 57 patients 14 were transfusion dependent before treatment. Initial EPO dose was 40000 U/week. Dose was modulated in relationship to response from a maximum of 40000 U every 5 days to a minimum of 40000 U every two weeks. Response rate observed was 76% (64% in the previous series described) in "good" + "very good" group and 33% (33% in the previous series) in "intermediate" group. Scoring "very good" patients had a higher response rate (80%) than scoring good patients (67%). In the previous experience the results were 79% and 33% respectively. The difference in response rate between "very good" and "intermediate" scoring was statistically significant (p=0.009). No difference was instead noticed between "good" and "very good" scoring (p=0.187) and between "good" and "intermediate" scoring (p=0.384). Furthermore, scoring "very good" patients showed a 700% (it was 667% in the previous report) increase in odds of response compared to scoring "intermediate" patients (p=0.004), meanwhile a not significant difference in odds of response between scoring "good" and scoring "intermediate" was found, just as in the previous described series.

Conclusion:

These data confirm our previous report which describes a very high probability of responding to standard EPO dose (40000 U/week) in the "very good" group and suggests a "standard" starting dose in these patients. "High" initial doses (80000 U/week) have to be recommended for the "intermediate" group. Considering the "good" group, the difference between the two analysis suggests that it would be appropriate to expand the sample in order to assess whether these patients may benefit from high doses of therapy or may also be candidate, in the first instance, for the standard doses.

Figure
Figure.
View largeDownload slide
Disclosures

No relevant conflicts of interest to declare.

Topics:
anemia, conflict of interest, disclosure, erythrocyte transfusion, erythropoietin, guidelines, logistic regression, mauritius, oman, prostatic hypertrophy risk score

Author notes

*

Asterisk with author names denotes non-ASH members.

© 2017 by The American Society of Hematology
2017
Sign in via your Institution