Efficacy of Imetelstat in Achieving Red Blood Cell Transfusion Independence (RBC-TI) across Different Risk Subgroups in Patients with Lower-Risk Myelodysplastic Syndromes (LR-MDS) Relapsed/Refractory (R/R) to Erythropoiesis-Stimulating Agents (ESAs) in IMerge Phase 3 Study

BACKGROUND: IMerge (NCT02598661) is a phase 2/3 global study of imetelstat, a first-in-class telomerase inhibitor, for patients with RBC transfusion-dependent non-del(5q) LR-MDS R/R to or ineligible for ESAs, a population with unmet needs. Phase 3 results showed that the rate of ≥8-week, ≥24-week and 1-year RBC-TI was higher with imetelstat than placebo (Zeidan et al. ASCO 2023. Abstr 7004; Platzbecker et al. EHA 2023. Abstr S165).

METHODS: To evaluate clinical efficacy of imetelstat across different International Prognostic Scoring System (IPSS), revised IPSS (IPSS-R), IPSS-R cytogenetic, or IPSS-molecular (IPSS-M) risk categories, cytogenetic analysis was performed centrally on bone marrow aspirates by karyotyping, and mutation profile was analyzed using peripheral blood samples by next-generation sequencing on a panel of 36 genes commonly mutated in MDS. Baseline IPSS, IPSS-R, IPSS-R cytogenetic, or IPSS-M risk categories were derived for patients in the phase 3 portion of IMerge based on specific classification criteria, and rates of RBC-TI ≥8-week, ≥24-week, ≥1-year were assessed for each risk subgroup.

RESULTS: In 105 IPSS-low patients, 95 (90.5%) remained IPSS-M very low/low/moderate low, and 10 patients (9.5%) upstaged to IPSS-M moderate high and high. In 50 IPSS-intermediate-1 patients, 39 (78%) remained IPSS-M very low/low/moderate low, and 11 patients (22%) upstaged to IPSS-M moderate high/high. Among 129 IPSS-R low/very low patients, 118 (91.5%) remained IPSS-M very low/low/moderate low, and 11 (8.5%) patients upstaged to IPSS-moderate high/high. In 25 IPSS-R–intermediate patients, 16 (64%) remained IPSS-M low/moderate low, and 9 patients (36%) upstaged to IPSS-M moderate high/high/very high. One IPSS-R high-risk patient remained IPSS-M high.

Overall, significantly higher rates of ≥8-week, ≥24-week, and ≥1-year RBC-TI were achieved in patients receiving imetelstat (39.8%, 28%, and 13.6%) vs those receiving placebo (15%, 3.3%, and 1.7%). Further subgroup analysis demonstrated that imetelstat consistently had higher TI response rates than placebo across different risk subgroups, irrespective of classification system.

For IPSS risk groups, the TI rates of ≥8-week, ≥24-week, and ≥1-year with imetelstat vs placebo were 40.0% vs 20.5% (P = 0.034), 28.8% vs 5.1% (P = 0.003), and 12.5% vs 2.6% (P = 0.082), respectively, in patients with low risk and 39.5% vs 4.8% (P = 0.004), 26.3% vs 0% (P = 0.009), and 15.8% vs 0% (P = 0.048), respectively, in patients with intermediate-1 risk.

For IPSS-R risk groups, the TI rates of ≥8-week, ≥24-week, and ≥1-year with imetelstat vs placebo were 42.5% vs 19.6%, 29.9% vs 4.3%, and 11.5% vs 2.2%, respectively, in the low-risk subgroup. The TI rates of ≥8-week, ≥24-week, ≥1-year TI with imetelstat were 35%, 25%, and 20%, respectively, in the intermediate-risk subgroup, whereas placebo treatment had no response. The number of patients was too low in both arms of the very low or high IPSS-R subgroups to assess difference in TI response.

For IPSS-R cytogenetic risk groups, the TI rates of ≥8-week, ≥24-week, and ≥1-year with imetelstat vs placebo were 37.1% vs 17.0%, 24.7% vs 4.3%, and 10.1% vs 2.1%, respectively, in very good/good risk groups and 54.5% vs 11.1%, 40.9% vs 0%, and 22.7% vs 0%, respectively, in the intermediate-risk group.

For IPSS-M risk groups, the TI rates of ≥8-week, ≥24-week, and ≥1-year with imetelstat vs placebo were 47.8% vs 21.2%, 34.8% vs 3%, and 14.5% vs 0%, respectively, in patients with very low/low risk and 20.7% vs 6.3%, 10.3% vs 0%, and 6.9% vs 0%, respectively, in patients with moderate low/moderate high risk. In patients with high/very high risk, ≥8-week TI rates were 40% vs 0% with imetelstat vs placebo, with no ≥24-week or ≥1-year TI observed in either arm. All results are summarized in the Table.

CONCLUSIONS: Improvement in RBC-TI rates was observed in patients treated with imetelstat vs placebo across different risk subgroups as defined by IPSS, IPSS-R, IPSS-R cytogenetic, or IPSS-M risk profiles. Notably, placebo had not achieved durable (≥24-week and ≥1-year) TI response in the higher-risk groups irrespective of the risk classification assessment model used, while TI response rates with imetelstat in higher-risk subgroups with poor prognosis were similar to TI response rates in lower-risk subgroups of heavily transfused R/R ESA LR-MDS, indicating that clinical efficacy of imetelstat is independent of risk categories.