Phase II Trial of MEK Inhibitor Binimetinib (MEK162) in RAS-mutant Acute Myeloid Leukemia

Abstract

Background:

Relapsed and refractory (R/R) acute myeloid leukemia (AML) continues to be a therapeutic challenge with poor outcomes. Dysregulation of the MAPK/ERK pathway frequently occurs in AML and myelodysplastic syndrome (MDS). Pre-clinical studies and early-phase trials have shown promise for MEK inhibition in AML. We sought to evaluate the safety and efficacy of MEK 1/2 inhibitor binimetinib in advanced myeloid malignancies.

Methods:

Nineteen patients with R/R AML and MDS, not candidates for intensive chemotherapy, or resistant/intolerant to standard treatment were enrolled on this phase II study of binimetinib dosed twice daily continuously in 28-day cycles.

Results:

The median age of the cohort was 64 years (range, 31–85). These patients had received a median of 3 prior lines of therapy (range, 1–6). The median bone marrow blast percentage was 49% (range, 2–94) and 14 patients had RAS mutations. Patients received a median of 2 cycles of binimetinib (range, 1–4) and were on treatment for a median duration of 1.2 months (range, 0.1–3.4). Sixteen patients (84%) received the 45 mg twice daily dose. The most common grade 3/4 treatment-emergent adverse events were hypokalemia (6%), hypotension (6%), lung infections (6%), and febrile neutropenia (6%). There were no treatment-related deaths. One of the 13 evaluable patients (8%) achieved a CRi lasting 2.1 months. The other 12 patients (92%) did not have a response. Six patients were inevaluable.

Conclusion:

Binimetinib had tolerable safety profile with minimal response in RAS-mutant AML. Future studies should focus on better patient selection and synergistic combination therapies involving MEK inhibition.

MICROABSTRACT

Dysregulation of the MAPK/ERK pathway frequently occurs in AML and MDS. Nineteen patients with advanced myeloid malignancies were treated with the MEK 1/2 inhibitor binimetinib. Minimal activity was noted with 1 out of 13 evaluable patients (8%) achieving a CRi, 12 patients (92%) did not respond, and 6 patients were inevaluable. Future studies need to evaluate synergistic combination therapies involving MEK inhibition.

INTRODUCTION

Patients with relapsed or refractory acute myeloid leukemia (R/R AML) have limited treatment options and poor response rates to traditional cytotoxic chemotherapy. Outcomes continue to be dismal with 2-year survival being <30% with intensive chemotherapy, and <5% with non-intensive strategies, particularly in older or unfit patients.¹ Hence, despite progress in the understanding of disease biology and the advent of newer therapies, R/R AML continues to be a formidable therapeutic challenge.

The mitogen-activated protein kinase/extracellular-signal regulated kinase (MAPK/ERK) pathway plays a complex role in cellular signal transduction, cell proliferation, and cell cycle regulation.² Mutations in pathway constituents and upstream and downstream regulators can disrupt cellular signaling and are among the most common genetic abnormalities in malignancy.^2–4 Mutations in MAPK/ERK pathway components have been frequently noted in AML and myelodysplastic syndrome (MDS).^3,5,6 Mutations in the RAS gene or its regulators that activate RAS signaling, upstream of MAP-ERK kinase (MEK), occur in over 30% of patients with AML and as high as 90% of patients with certain subtypes of AML.^7,8 However, despite being an attractive therapeutic target, MEK inhibition has had limited success in clinical trials due to complex feedback loops and alternative/redundant signaling involving the MAPK/ERK pathway.^4,9

Mutations in the RAS gene and the consequent protein product can lead to constitutive activation of MAPK/ERK signaling.^10,11 Preclinical studies have shown improvement in survival with MEK inhibition in mouse models of NRAS-mutant AML.¹² Two early phase clinical trials showed modest single agent activity with MEK inhibition in RAS mutant R/R AML using selumetinib, and trametinib.^13–15 Binimetinib (MEK162, ARRY 438162, Array BioPharma, Boulder Colorado) is a highly selective, potent, and orally bioavailable allosteric inhibitor of MEK 1/2 that is inhibits pERK and proliferation of BRAF mutant cancer cell lines.¹⁶ Multiple early phase clinical trials with single agent binimetinib and combinations have shown encouraging responses in various solid tumors and RAS-mutant cancers.^17–20 Hence, we sought to evaluate the tolerability and efficacy of binimetinib in R/R AML patients with RAS mutations. We selected the dose demonstrated to be safe and effective in solid tumor studies but decided to enroll some patients in the lead-in phase at a lower dose to ensure adequate tolerability.

PATIENTS AND METHODS

Patients

Adult patients with relapsed or refractory, primary or secondary AML or high-risk MDS were eligible. Patients with untreated AML were also eligible if they were older than 75 years, or older than 60 years with poor prognostic features e.g. secondary AML, unfavorable cytogenetics, or Eastern Cooperative Oncology Group (ECOG) performance status of 2. Although not required for eligibility, an emphasis was placed on recruiting patients with mutated RAS (NRAS or KRAS). Patients needed to have a performance status of ≤2, adequate cardiac function with ejection fraction ≥50% and were excluded if they had uncontrolled or significant cardiac abnormalities, poor organ function (creatinine >1.5 x upper limit of normal [ULN] or creatinine clearance of <30 mL/min, total bilirubin >1.5 xULN, alanine aminotransferase/aspartate aminotransferase >2.5 xULN), active CNS leukemia, uncontrolled infections, risk factor for or history of central serous retinopathy, retinal vein occlusion, myopathic disorder with elevated creatine kinase, diseases causing impairment of gastrointestinal function, or were pregnant.

Patients provided informed consent prior to enrollment. The study was approved by the Ethics Review Committee and the Institutional Review Board; and conformed with conducted in accordance Good Clinical Practice guidelines and the Declaration of Helsinki.

Study design and treatment

This was a single-center, open-label, non-randomized, phase II study (Study Sponsor: Array BioPharma, Boulder Colorado; Clinical Protocol 2013–0116; www.clinicaltrials.gov NCT02089230). The study enrolled patients between September 2014 and September 2017. The objectives of the study were to assess the safety, tolerability, and anti-leukemic activity of binimetinib in older patients with R/R AML, or newly diagnosed AML with poor prognosis, harboring NRAS and KRAS mutations, who were not candidates for intensive chemotherapy. Binimetinib was administered as a flat dose of 30–45 mg twice daily 12 hours apart. Three patients were treated at 30 mg twice daily in the lead-in phase and the subsequent patients were treated at 45 mg twice daily on days 1 to 28 of a 28-day cycle. Dose interruptions, reductions, and re-escalations were allowed for toxicities per predefined guidelines. Dose reduction below 30 mg twice daily was not allowed. Treatment was continued till disease progression, unacceptable toxicity, withdrawal of consent, noncompliance, pregnancy, or intercurrent illnesses requiring treatment interruptions of >21 days.

Study endpoints and response assessments

The primary safety endpoint was to identify the safety of binimetinib in patients with relapsed myeloid malignancies. The safety endpoints included adverse events (AE) and serious adverse events (SAE). Prespecified AEs of special interest included ocular events, retinal vein occlusion, skin toxicity, edema/anasarca, serum CK elevation, cardiac failure related events, and hepatic events. The primary efficacy endpoint was to determine overall response rate (ORR) including morphological complete response (CR) and morphological complete response with incomplete blood count recovery (CRi) after 1 or 2 cycles of therapy.

The common terminology criteria for adverse events (CTCAE) version 4.03 was used for grading of AEs and SAEs, which were reported per standard protocol. Responses were determined according to the modified International Working Group consensus criteria for AML.²¹

Patients were followed in clinic through frequent visits including physical examinations, complete blood counts, serum chemistries, and liver function tests. Bone marrow assessment was performed 28 days after initiation of therapy and included cytogenetic and molecular studies. Patients received ophthalmologic examinations and electrocardiograms after cycle 1. Patients continued therapy as long as they were deemed to be having clinical benefit.

Statistical analyses

Descriptive statistics were used to summarize categorical and continuous variables. Survival probabilities were estimated using the method of Kaplan and Meier.²² Survival was measured from start of therapy on trial to date of last follow-up or death from any cause at any time. GraphPad Prism 7.0 (GraphPad Software, La Jolla, CA) was used for statistical analyses. Safety analysis included all patients who had received at least 1 dose of binimetinib.

RESULTS

Subject characteristics

A total of 19 patients were enrolled on the trial. Baseline patient and disease characteristics are shown in Table 1. The median age of the cohort was 64 years (range, 31–85) and 12 (63%) patients were male. This was a heavily pre-treated population with a median of 3 prior lines of therapy (range, 1–6) and 5 patients (26%) had received prior allogeneic hematopoietic stem cell transplantation. Only 2 patients were enrolled after failing 1 prior line of therapy. Five patients (26%) did not have a RAS mutation.

Table 1.

Baseline demographics and disease characteristics

Patient characteristics (n=19)	Median [range], or n (%)
Age, median, years	64 [31–85]
Male / Female	12/7
ECOG Performance Status
0–1	14 (74)
2	5 (26)
Diagnosis
De novo AML	12 (64)
Secondary AML	5 (26)
Secondary MDS	1 (5)
CMML	1 (5)
White blood cell count, ×109/L	8.4 [0.2–34.6]
Hemoglobin, g/dL	9.1 [7.5–12.9]
Platelets, ×109/L	20 [5–117]
Peripheral blood blasts, %	27 [0–98]
Bone marrow blasts, %	49 [2–94]
No. of prior lines of therapy	3 [1–6]
Prior allogeneic hematopoietic stem cell transplantation	5 (26)
Cytogenetic risk group
Intermediate	12 (63)
Adverse	7 (37)
Favorable	0 (0)
RAS mutations
NRAS	10 (52)
None	5 (26)
KRAS	2 (11)
NRAS and KRAS	2 (11)

ECOG = Eastern Cooperative Oncology Group; AML = acute myeloid leukemia; MDS = myelodysplastic syndrome, CMML = chronic myelomonocytic leukemia

Subject disposition and treatment exposure

At a median follow-up of 1.8 months (range, 0.3–10.8), only 1 out of 19 patients was alive. Twelve patients (63%) were taken off study due to lack of response, 3 patients (16%) had early death (2 patients had pneumonia and 1 patient died in hospice), and 1 patient (5%) each was removed from study due to disease progression, toxicity, physician’s choice, and loss of follow-up. The cause of death was respiratory failure in 3 patients, pneumonia in 3 patients, intracranial hemorrhage in 2 patients, disease progression in 2 patients, febrile neutropenia in 1 patient, sepsis in 1 patient, fungal sinusitis in 1 patient, and unknown in 5 patients. Patients received a median of 2 cycles of binimetinib (range, 1–4) and were on treatment for a median duration of 1.2 months (range, 0.1–3.4). Sixteen patients (84%) received the 45 mg twice daily dose while 3 patients received the 30 mg twice daily dose. The trial was terminated due to slow accrual.

Toxicity

There were 324 treatment-emergent AEs (TEAE) regardless of causality in 16 patients. Out of these, 123 TEAEs were considered to be possibly or probably related to the study drug. There were 78 grade 3/4 AEs, and 8 grade 5 AEs including respiratory failure in 4 patients, disease progression in 2 patients, and febrile neutropenia, pneumonia, and intracranial hemorrhage in 1 patient each. None of the grade 5 events were related to the study drug. Table 2 shows TEAEs reported in at least 20% of patients, with at least one grade ≥3 event. The most common TEAEs were diarrhea (56%), hypokalemia (50%), hypotension (50%), hypoalbuminemia (50%), hypocalcemia (50%), and metabolic and nutritional disorders (not otherwise specified) (50%). The most common grade 3/4 TEAEs were hypokalemia (6%), hypotension (6%), lung infections (6%), and febrile neutropenia (6%). Seventy serious adverse events (SAE) were noted in 16 patients, out of which only 1 was considered to be related to study drug. The most common SAEs were febrile neutropenia (94%), hypotension (75%), and pneumonia (56%). No dose limiting toxicity was reported. There were no treatment-related deaths.

Table 2.

Treatment-emergent adverse events on binimetinib, regardless of causality, present in at least 20% of patients, with at least one grade ≥3 event

Treatment-emergent adverse events	Any grade	%	Grade 3/4	%
Diarrhea	9	56	0	0
Hypokalemia	8	50	6	38
Hypotension	8	50	6	38
Hypoalbuminemia	8	50	2	13
Hypocalcemia	8	50	2	13
Metabolism and nutrition disorders (NOS)a	8	50	0	0
Lung infection	7	44	6	38
Bilirubin elevation	7	44	3	19
Vomiting	7	44	2	13
Hypomagnesemia	7	44	1	6
Nausea	7	44	1	6
Eye disordersb	7	44	0	0
Gastrointestinal disorder (NOS)	7	44	0	0
Febrile neutropenia	6	38	6	38
Dyspnea	6	38	3	19
Fatigue	6	38	2	13
Edema	6	38	0	0
Hyperkalemia	5	31	1	6
Rash	5	31	1	6
Alkaline phosphatase elevation	5	31	0	0
Aspartate aminotransferase elevation	5	31	0	0
Cough	5	31	0	0
General disorders and administration site conditions	5	31	0	0
Hyponatremia	5	31	0	0
Sepsis	4	25	4	25
Hypernatremia	4	25	2	13
Abdominal pain	4	25	1	6
Hyperglycemia	4	25	1	6
Urinary tract infection	4	25	1	6
Cardiac disorders (NOS)	4	25	0	0
Chest pain	4	25	0	0
Epistaxis	4	25	0	0
International normalized ratio elevation	4	25	0	0
Tumor lysis syndrome	3	19	3	19
Hypoxia	3	19	2	13
Confusion	3	19	1	6
Infections (NOS)	2	13	2	13
Respiratory failure	2	13	2	13
Sinus tachycardia	2	13	2	13
Delirium	2	13	1	6
QTc prolongation	2	13	1	6
Sinusitis	2	13	1	6
Hematological toxicities
Neutropenia	2	13	2	13
Anemia	1	6	1	6
Leukocytosis	1	6	1	6

There were 9 grade 5 AEs including respiratory failure in 4 patients, disease progression in 2 patients, and 1 patient each with febrile neutropenia, pneumonia, and intracranial hemorrhage.

^a.NOS = not otherwise specified

^b.Eye disorders included retinal hemorrhage, exudative macular degeneration, and proptosis, all in one eye of 1 patient. Two other patients had retinal hemorrhages, 1 patient developed central serous retinopathy, and 1 patient had eye redness.

Efficacy

One out of the 13 evaluable patients (8%) achieved a CRi which lasted 2.1 months. This patient had a NRAS G12A mutation and was the longest survivor for a duration of 10.8 months and was 1 out of the 6 NRAS mutant patients (17%) evaluable for response. The other 12 patients (92%) did not have a response. Six patients were inevaluable due to early death in 3 patients, and early discontinuation due to toxicity, physician’s choice, and loss of follow-up in 1 patient each. No patients received allogeneic hematopoietic stem cell transplantation after receiving binimetinib. The median overall survival was 1.8 months, with 3-month survival of 42%, and 6-month survival of 12% (fig. 1).

Figure 1.

Overall survival of patients on binimetinib (n=19)

DISCUSSION

This study demonstrated that binimetinib (MEK162) at 45 mg twice daily dose was reasonably tolerated in patients with R/R AML, however anti-leukemic activity in this population was minimal. The trial was terminated after enrollment of 19 patients due to slow accrual. No new safety signals related to MEK inhibition were noted with binimetinib in this trial. The most common TEAE was diarrhea in 56% of patients while the most common grade 3/4 TEAEs were hypokalemia (6%), hypotension (6%), lung infections (6%), and febrile neutropenia (6%). Other known toxicities of MEK inhibitors in this population were eye disorders in 44% of patients, edema in 38% of patients and rash in 31% of patients. The 7 instances of eye disorders included retinal hemorrhage, exudative macular degeneration, and proptosis, all in the same eye in 1 patient. Two other patients had retinal hemorrhages, 1 patient developed central serous retinopathy, and 1 patient had eye redness. None of the patients had CK elevation. This finding differed from other binimetinib trials. One phase I study found CK elevation in 69% patients.¹⁹ However, this was likely because it was a dose escalation study going up to doses of 80 mg twice daily. Another study, using the same recommended phase 2 dose of 45 mg twice daily, as in our study, found CK elevation in 37% of patients.¹⁸ However, that study had more than 3 times larger number of patients, and much more longer duration of exposure compared to our study (3.3 months, range 0.6–8.7 months in that study compared to 1.2 months, range, 0.1–3.4 months in our study). Hence it may be possible that CK elevation may be related to duration of exposure. There were no treatment related deaths. The 45 mg twice daily dose and schedule used here has been established in other phase 1 trials of binimetinib in solid tumors.¹⁹

One out of the 13 evaluable patients (8%) who had a NRAS G12A mutation achieved a CRi lasting 2.1 months and survived for 10.8 months. Six patients (32%) were not evaluable and 12 patients (63%) did not have a response. The limited number of patients preclude drawing meaningful inferences from these results. Furthermore, this was a heavily pretreated population with a median of 3 prior lines of therapy and proliferative disease (median bone marrow blasts 49%) leading to the increased likelihood of there being multiple leukemic subclones in each patient which likely conferred resistance to single agent therapy. Furthermore, multiple alternate pathways and complex feedback loops within the MAPK/ERK pathway can enable leukemic clones to escape MEK inhibition.^4,9,23,24 Increased RasGRP1 and reduced p38 kinase activity have been implicated in MEK inhibitor resistance in AML cell lines with hyperactive RAS.²⁵ In KRAS mutant colorectal cancer cell lines, weak ERK1/2 signaling or strong PI3K signaling have been shown to cause intrinsic resistance to MEK1/2 inhibition.²⁶

Consistent with our findings, single agent binimetinib at doses of 45–60 mg twice daily has yielded modest response rates of 3–20% in solid tumors including biliary tract cancers, NRAS- and BRAF-mutant melanomas.^18,19,27 However, the differences in ORR to single agent binimetinib compared to other MEK 1/2 inhibitors like trametinib, selumetinib, and pimasertib in AML is not significant. In a phase I/II trial, trametinib, which is approved for combination therapy in melanoma, showed modest single agent activity in patients with R/R myeloid malignancies. It had an ORR of 20% in 50 patients with RAS mutant R/R AML and MDS, an ORR of 3% in a RAS wild-type cohort of 30 patients with myeloid malignancies and an ORR of 27% in 11 patients with chronic myelomonocytic leukemia with RAS mutations.¹³ Selumetinib showed modest single agent activity of 17% in 36 patients with R/R AML with FLT3 wild-type.¹⁵ Although caution is necessary when making cross-trial comparisons, the lower IC₅₀, longer half-life and better capacity to disrupt MEK phosphorylation can potentially explain higher ORR with trametinib compared to binimetinib.⁹ However, selumetinib and binimetinib have comparable pharmacodynamic and pharmacokinetic properties, and hence the reason for the difference in ORR remains unclear. In another phase I trial of a MEK 1/2 inhibitor pimasertib, we reported stable disease in 60% of the 15 patients with AML, but it did not result in any CR.²⁸ Tipifarnib is an oral farnesyl transferase inhibitor which disrupts RAS, PI3K/AKT signaling and was of interest in myeloid malignancies. However, phase 3 trials have failed to show any clinical benefit with tipifarnib in AML as a single agent, or in combination with low dose cytarabine.^29,30 A summary of all MEK inhibitors in hematological malignancies is presented in Table S1 (supplemental appendix).

Pre-clinical studies have also suggested potential role of binimetinib in other leukemias. One study showed MEK inhibitors including binimetinib severely impairs RAS-mutant pediatric acute lymphoblastic leukemia cells in vitro, and work synergistically with prednisolone.³¹ Another study showed single agent cytotoxic and cytostatic effect of binimetinib on cultured chronic lymphocytic leukemia cells and potentiates action of BH3 mimetics.³² However, biological understanding of the MAPK/ERK pathway, and clinical experience with MEK inhibitors thus far are not convincing to warrant further monotherapy trials of MEK inhibitors in AML. Instead, future pre-clinical and clinical studies should focus on synergistic combination therapies and better patient selection driven by predictive biomarkers. One phase I study combining selumetinib with azacytidine is currently enrolling patient with high-risk myeloid neoplasms and MDS (NCT03326310). Another phase Ib/II study is testing venetoclax in combination with cobimetinib in R/R AML (NCT02670044). Other attractive strategies include combining MEK inhibitors with statins, RAF, ERK, PI3K, and JAK targeted therapies.^23,33–35 Genetic mutations in the RAS pathway or those involved in activating MAPK signaling are known to be associated with primary resistance to mutant IDH2 inhibitor enasidenib.³⁶ Hence, combining MEK inhibitors with mutant IDH2 inhibitors may be an attractive strategy in these patients harboring mutations in IDH2 and RAS or other effectors of the MAPK/ERK pathway. Some combination therapies may offer better side effect profile, e.g., BRAF and MEK inhibition is better tolerated in metastatic melanoma than either monotherapy due to inhibition of paradoxical signaling pathways.⁹ MEK inhibitor combinations may be more attractive in advanced AML owing to the late acquisition of RAS mutations in AML.^4,13 The ratio of mutant and wild-type RAS could be a potential predictive biomarker for response to MEK inhibitors, if validated in future studies.³⁷ Constitutive MAPK phosphorylation has also been suggested as a predictive marker for sensitivity to MEK inhibition.³⁸

CONCLUSION

In summary, binimetinib had a tolerable safety profile and minimal single agent activity in AML with RAS mutation. Future studies need to focus on biomarker-driven patient selection for MEK inhibitor therapy and test synergistic combination of MEK inhibitors with other agents.

CLINICAL PRACTICE POINTS.

• Dysregulation of the MAPK/ERK pathway frequently occurs in myeloid malignancies.

• Pre-clinical studies and early-phase trials have shown promise of MEK inhibition in AML.

• In this study, the oral MEK 1/2 inhibitor binimetinib showed minimal activity in relapsed/refractory AML/MDS with CRi in 1 out of 13 evaluable patients (8%) and no response in 12 patients (92%), while 6 patients were inevaluable.

• Single agent binimetinib had acceptable toxicity with the most common grade 3/4 treatment-emergent adverse events being hypokalemia (6%), hypotension (6%), lung infections (6%), and febrile neutropenia (6%).

• Future studies need to focus on better patient selection and synergistic combination therapies involving MEK inhibition.