FDA Approval Summary: Ivosidenib in Combination with Azacitidine for Treatment of Patients with Newly Diagnosed Acute Myeloid Leukemia with an IDH1 Mutation

Abstract

On May 25^th, 2022, FDA approved a supplemental application for ivosidenib (Tibsovo; Servier) extending the indication in patients with newly diagnosed IDH1-mutated acute myeloid leukemia (AML) in older adults or those with comorbidities to include the combination with azacitidine. The efficacy of ivosidenib in combination with azacitidine was evaluated in Study AG120-C-009, a phase 3, multicenter, double-blind, randomized (1:1), controlled study of ivosidenib or matched placebo in combination with azacitidine in adults with previously untreated AML with an IDH1 mutation who were 75 years or older or had comorbidities that precluded use of intensive induction chemotherapy. Efficacy was established based on improved event-free survival (EFS) and overall survival (OS) on the ivosidenib + azacitidine arm (HR 0.35, 95% CI 0.17, 0.72, p= 0.0038 and HR 0.44, 95% CI 0.27, 0.73, p=0.0010), respectively. Furthermore, the rate and duration of complete remission (CR) were improved with ivosidenib versus placebo (CR 47% versus 15%, 2-sided p<0.0001; median duration of CR not estimable [NE] [95% CI 13.0, NE] months versus 11.2 [95% CI 3.2, NE] months). The safety profile of ivosidenib in combination with azacitidine was consistent with that of ivosidenib monotherapy, with important adverse reactions including differentiation syndrome (15%) and QT interval prolongation (20%).

Introduction

Acute myeloid leukemia (AML) can occur in all ages, but predominantly occurs in older adults with a median age of diagnosis at 68 years.¹ There have been several recent advances in novel therapeutics for treatment of AML within the last decade, specifically for patients that are “unfit” to receive intensive induction chemotherapy. This includes approval of the combination of venetoclax with a hypomethyating (HMA) or low-dose cytarabine (LDAC), glasdegib in combination with LDAC, gemtuzumab ozogamicin for CD33 positive AML, and the targeted therapy ivosidenib for patients harboring an IDH1 mutation.^2,3

Currently, no universal consensus exists regarding the exact definition of “unfit” for intensive chemotherapy. Ferrara et al, listed nine (9) operational criteria to define unfitness to receive intensive chemotherapy, which included age ≥75 years, severe organ impairment, and low performance status.⁴ More recently, the American Society of Hematology (ASH) released guidelines on treatment of AML in older adults.⁵ In these guidelines, experts state that more research is needed to better determine “fitness” for intensive chemotherapy, as the expert panel could not clearly define a patient population “unfit” for intensive chemotherapy. However, since the approval of venetoclax in combination with hypomethylating (HMA) therapy for AML, there is a precedent for using age ≥75 years or specific comorbidities (baseline Eastern Cooperative Oncology Group [ECOG] Performance Status [PS] >2, severe cardiac, severe pulmonary, hepatic impairment with bilirubin >1.5x the ULN, or creatinine clearance <45mL/min) to identify an indicated population for which intensive induction chemotherapy is precluded.⁶

Ivosidenib, a small molecule inhibitor of isocitrate dehydrogenase 1 (IDH1), was first approved as monotherapy for adult patients with relapsed/refractory (R/R) AML in 2018 and then was subsequently approved as monotherapy for adult patients with newly diagnosed AML who were 75 years or older or who had comorbidities that precluded use of intensive chemotherapy in 2019.⁷ In this review, we provide a summary of the FDA review of the supplemental marketing application of ivosidenib in combination with azacitidine for treatment of adult patients with newly diagnosed AML who are 75 years or older or have comorbidities that preclude use of intensive induction chemotherapy with a susceptible IDH1 mutation as detected by an FDA-approved test.

Clinical Pharmacology

The FDA reviewed subject-level data collected from patients with newly diagnosed AML with an IDH1 mutation as well as the population pharmacokinetic (PK) and exposure-response analyses in studies AG120-C-009 (NCT03173248) and AG221-AML-005 (NCT02677922).

The selection of the 500 mg ivosidenib dose once daily (QD) for the combination therapy with azacitidine in patients with newly diagnosed acute myeloid leukemia was based on previous ivosidenib monotherapy data, which reported overlapping PK exposures between 500 and 800 mg QD doses, and the maximized 2-HG inhibition in blood and bone marrow at 500 mg QD in patients with IDH1 mutated hematologic malignancies.⁸ The addition of azacitidine to the combination therapy did not affect the PK exposures (steady state maximum concentration and area under the curve) of ivosidenib, suggesting that there is no additional risk of QTc-prolongation for this combination. Within the exposure ranges of ivosidenib at 500 mg QD, there appeared to be a flat exposure-response relationship for efficacy of the combination with azacitidine. A positive trend for all grade aspartate aminotransferase (AST) elevation with increasing ivosidenib exposure was observed while a flat exposure-response relationship was generally observed for other relevant safety endpoints. However, exposure-response analyses for safety and efficacy should be interpreted cautiously due to the limited sample size and narrow range of dose levels evaluated in combination with azacitidine.

Assessment of Efficacy

Clinical Trial Overview

Study AG120-C-009 (NCT03173248) was a phase 3, multicenter, double-blind, randomized (1:1), controlled study of ivosidenib or matched placebo in combination with azacitidine in adult patients with previously untreated IDH1-mutated AML. Randomization was stratified by de novo status and geographic region. Adults must have been 75 years or older or had at least one comorbidity that precluded the use of intensive induction chemotherapy: baseline ECOG performance status of 2, severe cardiac or pulmonary disease, hepatic impairment with bilirubin > 1.5 times the upper limit of normal, creatinine clearance < 45 mL/min, or another comorbidity. The IDH1 mutation was confirmed centrally using the FDA-approved Abbot RealTime IDH1 Assay.

The primary efficacy endpoint was event-free survival (EFS), which was defined as the time from randomization until treatment failure (TF; failure to achieve complete remission [CR] by week 24), relapse from remission, or death from any cause, whichever occurred first. Patients who experienced TF were considered to have had an EFS event on day 1 of randomization. Key secondary endpoints included complete remission (CR) rate, CR + CR with partial hematological recovery rate (CRh), and overall survival (OS). The original design for study AG120-C-009 utilized a sample size of 392 patients and a primary endpoint of OS. A protocol amendment dated 09 January 2020 changed the primary endpoint to EFS and the sample size to 200 patients. However, after review of safety and efficacy by an independent data monitoring committee (IDMC) in May 2021, the trial was halted after enrollment of 148 patients (including 2 patients randomized after the data cutoff date). This was due to clinically significant observed differences in deaths between the two treatment arms. The p-values for the primary and key secondary endpoints were assessed against boundaries constructed with the O’Brien Fleming alpha spending function using the Lan DeMets method to account for the unplanned analysis by the IDMC.

Treatment Plan

Patients received either ivosidenib 500 mg orally QD on days 1–28 or matched placebo in combination with azacitidine 75 mg/m² intravenous (IV) or subcutaneous (SC) on days 1–7 of each 28-day cycle, for a minimum of 6 cycles. Treatment continued until disease progression, unacceptable toxicity, or undergoing hematopoietic stem cell transplantation (HSCT).

Disposition and Demographics

The demographics of the efficacy population are shown in Table 1. Overall, all patients enrolled met the age or comorbidities criteria expected for the proposed population. Most patients met criteria based on age alone, with the second most common criterion being ECOG performance status =2. At the time of the primary efficacy analysis, the median follow-up was approximately 15 months on both treatment arms. A total of 27 patients (38%) on the ivosidenib + azacitidine arm and 12 (16%) on the placebo + azacitidine arm remained on therapy at the time of data cut-off. The primary reasons for treatment discontinuation included adverse events (28% ivosidenib + azacitidine vs 27% placebo + azacitidine), disease progression (14% vs 19%), and patient withdrawal (7% vs. 14%). The median duration of treatment was longer in the ivosidenib + azacitidine arm compared to the placebo + azacitidine arm (6 [range 0–33] months vs. 2.7 [range 0–19] months), with 32% (23 patients) on the ivosidenib + azacitidine arm receiving treatment for >12 months compared to 10% (7 patients) on the placebo + azacitidine arm. Two patients in the ivosidenib + azacitidine arm and one patient on the placebo + azacitidine arm proceeded to allogeneic HSCT.

Table 1:

Characteristics of randomized patients with newly diagnosed IDH1-mutated AML in Study AG120-C-009

Variable		Ivosidenib + Azacitidine (N=72)	Placebo + Azacitidine (N=74)
Age	Median (Range) ≥75 (%)	76 (58–84) 39 (54.2)	76 (45–94) 43 (58.1%)
Sex	Male Female	42 (58.3) 30 (41.7)	38 (51.4) 36 (48.6%
Race	Asian White Black Other Not Reported	15 (20.8) 12 (16.7) 0 1 (1.4) 44 (61.1)	19 (25.7) 24 (16.4) 2 (2.7) 1 (1.4) 40 (54.1)
Nature of AML	De Novo Secondary History of MDS MPN Related Therapy-Related	54 (75.0) 18 (25.0) 10 (13.9) 4 (6) 2 (3)	53 (71.6) 21 (28.4) 12 (16.2) 8 (11) 1 (1)
ECOG Performance Status	0 1 2	14 (19.4) 32 (44.4) 26 (36.1)	10 (13.5) 40 (54.1) 24 (32.4)
IDH1 Mutation Type	R132C R132H R132G R132L R132S	45 (62.5) 14 (19.4) 6 (8.3) 3 (4.2) 2 (2.8)	51 (68.9) 12 (16.2) 4 (5.4) 0 6 (8.1)
Cytogenetics	Favorable Intermediate Poor Other Missing	3 (4.2) 48 (66.7) 16 (22.2) 3 (4) 2 (3)	7 (9.5) 44 (59.5) 20 (27.0) 1 (1) 2 (3)
Transfusion Dependent at Baseline	Yes	39 (54)	40 (54)
Baseline Severe Cardiac Disease	Yes	11 (15.3)	16 (21.6)
Baseline Severe Pulmonary Disorder	Yes	5 (6.9)	8 (10.8)
Baseline CrCl < 45 mL/min	Yes	5 (6.9)	10 (13.5)
Baseline Bilirubin > 1.5x ULN	Yes	1 (1.4)	1 (1.4)
Other*	Yes	19 (26.4)	24 (32.4)

^*Other included, “other comorbidity”, “investigator judgement”, or both.

Abbreviations: CrCl, creatinine clearance; ECOG, Eastern Cooperative Oncology Group; MPN, myeloproliferative neoplasm; ULN, upper limit of normal.

Source: U.S. Food and Drug Administration.

Efficacy Results

Efficacy was established based on EFS, OS, and the rate and duration of CR. Efficacy results are shown in Figures 1 and 2 and Table 2. The study showed a significant improvement in EFS, with significantly less events (treatment failure, relapse, and death), for the ivosidenib + azacitidine arm compared to the placebo + azacitidine arm (HR 0.35, 95% CI 0.17, 0.72, p = 0.0038). Due to the high incidence of treatment failure on both arms, median EFS was short at 0.03 months (95% CI, 0.03, 4.90) for ivosidenib + azacitidine and 0.03 months (95% CI, count not be estimated) for placebo + azacitidine. The improvement in EFS on the ivosidenib + azacitidine arm was generally consistent across most of the prespecified demographic and disease subgroups. However, patients with the R132H mutation had HR for EFS of 1.24 (95% CI 0.24, 6.36) (Supplementary Fig. S1). Firm conclusions based on this subset analysis are limited based on the small sample size (n=14 ivosidenib + azacitidine, n= 12 placebo + azacitidine).

Figure 1:

Event-Free Survival. Kaplan-Meier plot of event-free survival for patients with newly diagnosed IDH1-mutated AML on Study AG120-C-009. (Source: U.S. Food and Drug Administration.)

Figure 2:

Overall Survival. Kaplan-Meier plot of overall survival for patients with newly diagnosed IDH1-mutated AML on Study AG120-C-009. (Source: U.S. Food and Drug Administration.)

Table 2:

Efficacy results in patients with newly diagnosed IDH1-mutated AML on Study AG120-C-009

Endpoint	Ivosidenib + Azacitidine N=72	Placebo + Azacitidine N=74
EFS, events (%) Treatment Failure Relapse Death	47 (65) 43 (60) 3 (4) 1 (1)	62 (84) 59 (80) 2 (3) 1 (1)
Hazard Ratio (95% CI)	0.35 (0.17, 0.72)
p-value	0.0038
Median OS (95% CI)	24.0 (11.3, 34.1)	7.9 (4.1, 11.3)
Hazard Ratio (95% CI)	0.44 (0.27, 0.73)
p-value	0.0010
CR, n (%)	34 (47)	11 (15)
95% CI	(35,59)	(8, 25)
p-value	<0.0001
Median Duration of CR (95% CI), months	NE (13.0, NE)	11.2 (3.2, NE)
CR + CRh, n (%)	37 (51)	13 (18)
95% CI	(39,63)	(10,28)
p-value	<0.0001
Median Duration of CR + CRh (95% CI), months	NE (13.0, NE)	9.2 (5.8, NE)

Source: U.S. Food and Drug Administration.

There was also a significant improvement in OS; patients that received ivosidenib + azacitidine had a significantly longer median OS compared to patients that received placebo + azacitidine (24.0 months vs. 7.9 months; HR 0.44 [95% CI 0.27, 0.73] p=0.0010). CR was also significantly improved on the ivosidenib + azacitidine arm, with median duration of CR being not estimable (Table 2). The median time to first CR for ivosidenib + azacitidine was 4 months (range, 1.7 to 11.9 months). Regarding efficacy across IDH1 mutation subtypes, a survival benefit of ivosidenib + azacitidine combination therapy was seen in all subtypes except the R132S mutation (Supplementary Figure 2). This should be interpreted with caution as the sample size for patients with the R132S was limited (N=2 on the ivosidenib + azacitidine arm vs. N=6 on the placebo+ azacitidine arm).

Assessment of Safety

Safety Events

The summary of the most frequent (>10%) adverse reactions (ARs) by arm is listed in Table 3. Common non-hematological laboratory abnormalities more frequent in the ivosidenib + azacitidine arm included hyperglycemia, hypophosphatemia, elevated AST, hypomagnesemia, elevated alkaline phosphatase, and hyperkalemia (Supplementary Table S2).

Table 3.

Treatment-emergent AEs in the safety population^a

	Ivosidenib + Azacitidine N=71		Placebo + Azacitidine N=73
Adverse Eventb	All Grades n (%)	Grade ≥3 n (%)	All Grades n (%)	Grade ≥3 n (%)
Nausea	30 (42)	2 (3)	28 (38)	3 (4)
Vomiting	29 (41)	0	20 (27)	1 (1)
Arthralgia	21 (30)	3 (4)	6 (8)	1 (1)
Electrocardiogram QT prolonged	14 (20)	7 (10)	5 (7)	2 (3)
Dyspnea	14 (20)	2 (3)	11 (15)	4 (5)
Insomnia	13 (18)	1 (1)	9 (12)	0
Differentiation Syndrome	11 (15)	7 (10)	6 (8)	6 (8)
Hematoma	11 (15)	0	3 (4)	0
Leukocytosis	9 (13)	0	1 (1)	0
Hypertension	9 (13)	3 (4)	6 (8)	4 (5)
Headache	8 (11)	0	2 (3)	0

^aN = 144 adults with newly diagnosed IDH1-mutated AML treated with ivosidenib + azacitidine or placebo + azacitidine.

^bIncludes grouped terms. See Supplementary Table S1 for further information.

Source: U.S. Food and Drug Administration.

Patients in ivosidenib + azacitidine arm had a lower 30-day (7% vs. 12%), and 60-day mortality (13% vs. 25%), compared to patients on the placebo + azacitidine arm. There were three deaths due to ARs in the ivosidenib + azacitidine arm: two cases of differentiation syndrome and one case of cerebral ischemia.

Treatment with ivosidenib + azacitidine was interrupted in 56% of patients (49% on placebo + azacitidine), dose reduced in 14% (7% on placebo + azacitidine) and discontinued in 31% (30% in the placebo + azacitidine arm) due to adverse events (AEs). AEs leading to dose interruption in > 5% of patients on the ivosidenib + azacitidine arm were neutropenia, QTc prolongation, and thrombocytopenia. AEs leading to dose reductions included QTc prolongation , neutropenia, and thrombocytopenia. AEs leading to discontinuation in ≥ 2% of patients included differentiation syndrome and pulmonary embolism.

QTc prolongation is a known serious AR seen with ivosidenib therapy and was observed in 20% of patients in the ivosidenib + azacitidine arm. A total of 7% required treatment interruption, and 8% required a dose reduction, and one patient discontinued ivosidenib due to QTc prolongation. A total of 14% of patients on the ivosidenib + azacitidine arm had a QTc interval > 500 msec and 22% had an increase from baseline of > 60 msec (compared to 3% and 14% on the placebo arm, respectively). However, there were no cases of torsades de pointe or cardiac arrest.

Differentiation syndrome, a potentially fatal clinical syndrome characterized by shortness of breath, fever, weight gain, hypotension, acute kidney injury, and pulmonary infiltrates or pleuropericardial effusion,⁹ was seen in 15% of patients treated on the ivosidenib + azacitidine arm, with 10% being Grade 3 or higher (compared to 8% and 8% on the placebo arm, respectively). These rates were adjudicated by FDA using the comprehensive analysis of adverse events, laboratory, and vital sign data, as previously described.¹⁰ Of 11 patients with DS on the ivosidenib + azacitidine arm, 8 (73%) recovered following treatment or after dose interruption of ivosidenib.

Regulatory Insights

The efficacy of ivosidenib + azacitidine in adult patients who are ≥ 75 years old or have comorbidities that preclude use of intensive chemotherapy was established based on improved EFS, OS, and the rate and duration of CR compared to placebo + azacitidine on Study AG120-C-009. The FDA has previously accepted EFS, a reflection of durable CR and survival, as a clinical benefit for patients with newly diagnosed AML.¹¹ Thus, the significantly improved EFS, coupled with significant improvements in OS, and CR rate with demonstrated durability, all contributed to the substantial evidence of efficacy to support regular approval for the proposed indication.

While EFS has been used to evaluate efficacy in other trials of newly diagnosed AML, these trials were all in the intensive therapy setting where rates of complete remission are high. In Study AG120-C-009, the rates of treatment failure (defined as failure to achieve CR by week 24) were 60% and 80% in the treatment and control arms, respectively leading to a significant proportion of patients having an event on day 1 of randomization. Given the median time to CR for ivosidenib + azacitdine of 4 months, 24 weeks (roughly 6 months) was determined to be a reasonable timeframe for determination of treatment failure. Despite use of this extended “induction” period, only 7 EFS events across the two arms were relapse or death. In this case, the result of superior EFS is driven almost entirely by more patients reaching CR in the treatment arm. Thus, the EFS results do not robustly characterize long-term efficacy for this lower-intensive therapy regimen. It is unclear whether EFS will yield robust results as a primary endpoint in future trials of lower-intensity regimens.

It is important to emphasize that the definition of EFS on Study AG120-C-009 only took achievement of CR responses with full count recovery into consideration. CRh is an acceptable regulatory endpoint for drug approval in AML for relatively non-toxic and non-myelosuppressive agents, and as such, was used to support the monotherapy approvals of ivosidenib for patients with IDH1-mutated AML.¹² However, in the setting of combination with the myelosuppressive agent azacitidine, CRh was only considered supportive in nature.

Safety was assessed in all 144 patients that were enrolled and treated with ivosidenib + azacitidine or placebo + azacitidine on Study AG120-C-009. The safety profile of the combination of ivosidenib + azacitidine was like that reported with ivosidenib monotherapy, with comparable rates of differentiation syndrome, QTc prolongation, and leukocytosis. Fatal events of differentiation syndrome occurred in 2 patients (3%). The conduct of a randomized, controlled trial, with in-depth adjudication of DS cases per FDA’s algorithm elucidated the incidence of DS with ivosidenib + azacitidine at 15% compared to that of placebo + azacitidine at 8%.¹⁰ The incidence of DS with azacitidine was previously not well-described but is consistent with post-marketing reports of DS in patients with AML treated with hypomethylating agents.^13–16 The addition of ivosidenib clearly increased the incidence of DS, but the overall incidence remains about the same as ivosidenib monotherapy, including grade 3 or higher incidences of DS Clinical vigilance and immediate treatment for this adverse reaction must be exercised, and thus, DS continues to be described in a boxed warning in the United States Prescribing Information and in a Medication Guide for patients.

One major challenge of the pivotal Study AG120-C-009 is that the control of placebo + azacitidine would currently be considered sub-standard care for many patients in the United States, with the regular approval of venetoclax and azacitidine for patients with newly diagnosed AML who are ≥75 years or with comorbidities that preclude the use of intensive induction therapy in October 2020. However, we note that 1) Study AG120-C-009 was already well underway at the time of the regular approval of venetoclax for the treatment of patients with AML, 2) only two patients from the U.S. were enrolled in this study, both before the approval of the venetoclax combinations, 3) azacitidine monotherapy continues to be a recommended treatment option per NCCN guidelines,¹⁷ and 4) and FDA does not have a comparative efficacy requirement for regular approval. That being said, a similarly designed trial moving forward would not be considered applicable to the United States patient population. Future trials to investigate the comparative efficacy of ivosidenib + azacitidine compared to venetoclax + azacitidine and/or the combination of ivosidenib + azacitidine + venetoclax compared to each doublet therapy would be useful to better define the optimal use of ivosidenib in patients with newly diagnosed AML with mutated IDH1 who are unable to tolerate intensive induction chemotherapy.

Conclusions

In conclusion, treatment with ivosidenib + azacitidine improved EFS, OS, and durable CR rate compared to placebo + azacitidine. Notable adverse reactions included DS, QT interval prolongation, and leukocytosis. Overall, the submitted data indicated that the clinical benefit of ivosidenib 500 mg once daily in combination with azactidine outweighed the safety concerns for adult patients with newly diagnosed AML with a susceptible IDH1 mutation who are ≥ 75 years or who have comorbidities that preclude use of intensive induction chemotherapy.