Early Versus Late Onset of ESA in Lower Risk Anemic MDS Patients: Results of the GFM Randomized Phase III EPO-Pretar Trial

BACKGROUND

Most lower risk MDS patients (pts) with anemia currently receive frontline ESA in Europe, but when ESA should be started in those pts (i.e., before or when RBC transfusion dependence (TD) occurs) is not clearly established. In a retrospective study, we found that early onset of treatment with an ESA, within 6 months of diagnosis, in non-TD anemic lower-risk MDS, significantly delayed the time to RBC TD and therefore possibly slowed disease evolution (Park et al , Leuk Res 2010 PMID: 20580086).

METHODS

In this open-label, randomized, multicenter, phase III EPO-PRETAR trial (NCT03223961), we prospectively compared, in non-RBC TD lower-risk MDS with anemia, the time to RBC TD between early and delayed onset of ESA.

Lower-risk MDS patients (pts) (IPSS-R<4) with baseline Hb between 9 and 10.5g/dL were randomized to receive EPO Alfa 60000 UI/week for at least 12 weeks at inclusion (early onset arm) or at the Hb threshold prospectively chosen for RBC transfusions for each pt (based on age, comorbidities, but which should be < 9g/dL) (late onset arm). TD was defined by 2 transfusion episodes during an 8-week interval after inclusion (early onset arm), or after at least 12 weeks of ESA (late onset arm). Baseline Cytokines, DNA -Seq and RNA-seq were performed in pts having received ESA.

RESULTS

Between February 2018 and Dec 2022, 121 pts were screened of which 84 fulfilled inclusion criteria, and 37 were screen failures. 43 were assigned to the early onset and 41 to the late onset arm. 40 pts started ESA in the early onset arm (3 withdrew consent), and 23 (56%) in the late onset arm had started ESA by 8 th July 2024, closing date of the study, after a median interval of 10.4 mo.

Median age was 84 years, with 54% males. 75% of the pts had cardio-vascular comorbidities, 46% had sideroblastic and 54% non-sideroblastic lower risk MDS. Median time between diagnosis and inclusion was 4.1 mo (range 2.1-11). IPSS-R was very low or low in 88%. Median baseline Hb level was 10.2 g/dL (range 8.8-10.5), and median sEPO level 37 IU/L (range 5-335).

Median follow-up was 34 months. Of the 84 included pts, 14 (32%) and 18 (44%) became RBC TD after inclusion in the early and late onset arms, respectively (p=0.28). Of the 63 pts who received at least 12 weeks of ESA, HI-E (Hematologic improvement-erythroid, IWG 2018 criteria) was 79.5% in the early onset arm and 54% in the late onset arm ( p =0.03). HI-E duration was longer in the early onset than in the late onset arm (median 30.6 months and 12.7 months respectively) (p=0.02). However, median time to RBC TD was similar (36.4 mo, 95% CI [28,3-NA] in the early arm, 36.4 months ,95% CI [22.4-44,1] in the late arm)) (p= 0.3).

13 and 10 pts progressed to higher risk MDS or AML in the early and late arms respectively (9.5 % in both arms for AML). Median PFS from inclusion was 60.6 mo in the early onset and not reached in the late onset arm (p=0.4). Median OS from inclusion was 49.6 and 55 months, respectively (p=0.3).

No significant differences in changes in the every-12 weeks QOL FACT-An and EQ5D questionnaires were seen between the 2 arms. The number of cardio-vascular events was also similar.

In the whole series of pts who received ESA, HI-E was associated with lower sEPO level (88% if < 50IU/L and 46% if> 50IU/L) (p=0.0003), lower IPSS-M (median IPSS-M: -0.603 in non- responders (NR) and -1.07 in responders (R )) (p=0.007), SF3B1 mutation: 81.8% HI-E in SF3B1 pos and 56.7% in SF3B1neg) (p=0.01). Time to TD was longer in IPSS-M lower risk (42.8 mo vs 28.3 mo) (p=0.002). TGFbeta-1 and IL6 plasma levels were lower in responders. Differentially expressed genes analysis (log2FC > I1I, BH-adjusted P-value <0.05) at baseline showed the activation of adaptive and innate immunity pathways (T cell activation, IL2/IL4 signaling and IFN alfa/IFN beta signalling) in responders. The expression level of a set of genes including erythropoiesis markers (ALAS2, ERFE, GDF15, BCL2L1), allowed the hierarchical clustering of non-responders (NR) vs responders and, within the NR pts, the segregation of TD vs non-TD pts.

CONCLUSIONS:

Early introduction of ESA in lower risk MDS patients with anemia increased HI-E and its duration but not the time to RBC TD from inclusion, and had no effect on PFS, AML transformation, quality of life and survival. IPSS-M also predicted HI-E and duration of response to ESA.