Enasidenib or ivosidenib in combination with intensive chemotherapy in patients with newly diagnosed AML

Isocitrate dehydrogenase (IDH)1 and 2 gene mutations are frequently seen in patients with acute myeloid leukemia (AML). These IDH1 and IDH2 mutations can be targeted using ivosidenib or enasidenib, respectively, which act as small molecule inhibitors. Overall response rates of ~40% have been achieved in phase I trials of either ivosidenib or enasidenib used alone; therefore, investigators decided to assess the benefit of combining these treatments with chemotherapy.

Eytan Stein and colleagues published an article in Blood on the results of a phase I trial (NCT02632708) of enasidenib or ivosidenib combined with intensive induction and consolidation chemotherapy in patients with newly diagnosed AML.¹

Study design

In this phase I trial, four treatment arms were assessed as shown in Figure 1. Initially, six patients were enrolled in each arm of the trial. If two or fewer patients experienced a dose limiting toxicity (DLT), six more patients would be enrolled on that arm. Out of this cohort of 12, if less than of equal to three patients experienced a DLT this dose was taken forward for evaluation. If these conditions were not met, a lower dose was administered.

Figure 1. Study design for the initial phase of testing*

SID, once a day.
*Data from Stein et al.¹

If patients achieved at least a partial response by the end of induction they were given consolidation therapy as follows:

• up to four cycles of intermediate- or high-dose cytarabine or one cycle of mitoxantrone/etoposide;
• plus ivosidenib or enasidenib continuously once daily.

Maintenance therapy was given to patients in remission at the end of consolidation, consisting of:

• ivosidenib or enasidenib continuously once daily until relapse/unacceptable toxicity or allogeneic stem cell transplant (allo-SCT).

The eligibility criteria included:

• patients aged ≥18 years diagnosed with AML (de novo or secondary);
• Eastern Cooperative Oncology Group score of 0–2; and
• IDH1 or IDH2 mutation by laboratory testing.

Patients were ineligible if:

• they had received prior chemotherapy for AML (except hydroxyurea).

The primary endpoint of this trial was the safety and tolerability of the combinations tested.

Results

In the ivosidenib + chemotherapy arm:

• 15% received two cycles of induction;
• 58% went on to consolidation therapy; and
• 32% received ivosidenib maintenance.

With respect to treatment continuation, 78% of the ivosidenib cohort discontinued (mostly to receive allo-SCT; 28 patients).

In the enasidenib + chemotherapy arm:

• 24% received two cycles of induction;
• 49% went on to consolidation therapy; and
• 26% received enasidenib maintenance.

Regarding treatment continuation, 87% of the enasidenib cohort discontinued (mostly to receive allo-SCT; 43 patients).

Patient baseline characteristics are shown in Table 1. The median age was around 63 years in both cohorts and the majority of patients were diagnosed with de novo AML. In the enasidenib arm, almost 50% of patients (with secondary AML) had received a hypomethylating agent, compared with only 22% in the ivosidenib arm.

Table 1. Baseline patient characteristics*

Efficacy

Complete response (CR)/CR with incomplete neutrophil recovery (CRi)/CR with incomplete platelet (CRp) rates were similar between treatment arms (ivosidenib, 77%; enasidenib, 74%), with patients with de novo AML achieving higher rates than patients diagnosed with secondary AML (Table 2). In the ivosidenib cohort, 68% of patients achieved CR, compared with 55% in the enasidenib arm.

Table 2. Overall responses*

Median overall survival was not reached for the ivosidenib arm after a median follow-up of 9.3 months (range, 0.4–32.1 months).

In the enasidenib arm, median overall survival was 25.6 months (95% confidence interval, 25.5 months–not calculable) after a median follow-up of 14.5 months (range, 0.5−31.8 months).

IDH-1/2-clearance was assessed in the patients with best response of CR/CRi/CRp. For the ivosidenib group, 39% had IDH1 mutant clearance from the bone marrow, as assessed by polymerase chain reaction. In the enasidenib group, 23% had IDH2 mutant clearance.

Measurable residual disease (MRD) was assessed by multiparameter flow cytometry in patients with best response of CR/CRi/CRp:

• 80% of the ivosidenib arm were MRD negative; and
• 63% of the enasidenib arm were MRD negative.

The most frequent baseline mutations were:

• DNMT3A (41%), NPM1 (34%), ASXL1 (20%), and BCOR (14%) in the ivosidenib arm; and
• DNMT3A (39%), SRSF2 (25%), ASXL1 (23%), and RUNX1 (20%) in the enasidenib arm.

ASXL1, NRAS, U2AF1, and TP53 were associated with a lack of CR/CRi/CRp only in the enasidenib arm.

Safety

Only one DLT was recorded during the initial safety evaluation. A 64-year-old man with de novo AML developed persistent Grade 4 thrombocytopenia on Day 42 of Cycle 1.

Both treatment arms exhibited a similar median time to recovery of absolute neutrophil count and platelet count of 28 days.

The most common non-hematologic treatment emergent adverse events (TEAEs) were diarrhea, nausea, and rashes (Table 3). The majority of TEAEs were Grade <3. The most frequent Grade ≥3 TEAEs were increased blood bilirubin (16.1% in the enasidenib arm) and hypophosphatemia (16.7% in the ivosidenib arm).

In the ivosidenib arm, the 30-day mortality rate was 5%, increasing to 10% at 60-days. Results were similar in the enasidenib arm, at 5% and 9% for the 30- and 60-day mortality rates, respectively.

During this trial, 13 patients in the ivosidenib arm died and 31 deaths occurred in the enasidenib cohort. None of the deaths were considered to be related to treatment.

Table 3. Non-hematologic TEAEs experienced in >30% of patients in any treatment arm*

In the ivosidenib arm, two patients developed IDH-differentiation syndrome (DS). The same was true in the enasidenib arm, and all four cases occurred during induction. IDH-DS was resolved in three of the patients and they went on to achieve CR. The remaining patient died as a result of a lung infection and had ongoing IDH-DS at the time of death.

QT prolongation in the ivosidenib cohort was reported in 26.7% and 10% experienced Grade ≥3 QT prolongation. In the enasidenib group, three patients reported QT prolongation and in one case it was Grade ≥3.

Mitoxantrone/etoposide was only used in two patients, therefore the safety of this consolidation treatment could not be assessed.

Conclusion

This phase I study demonstrated that the IDH inhibitors in combination with intensive induction and consolidation chemotherapy were well tolerated in both arms. CR/CRi/CRp rates were >70% in both the ivosidenib and enasidenib arms. A further phase III trial is being conducted to assess the value of ivosidenib or enasidenib and chemotherapy for newly diagnosed patients with IDH-mutant AML.