Leukemia (2020) 34:2903–2913
`https://doi.org/10.1038/s41375-020-0996-5
`
`ARTICLE
`
`Acute myeloid leukemia
`Safety and efficacy of BAY1436032 in IDH1-mutant AML:
`phase I study results
`
`4 ● Courtney DiNardo 5 ●
`1 ● Neil Palmisiano 2 ● Ioannis Mantzaris3 ● Alice Mims
`Michael Heuser
`Lewis R. Silverman6 ● Eunice S. Wang 7 ● Walter Fiedler8 ● Claudia Baldus9 ● Sebastian Schwind 10 ●
`Timothy Pardee 11 ● Alexander E. Perl12 ● Charles Cai13 ● Stefan Kaulfuss14 ● Eleni Lagkadinou14 ●
`Christine Rentzsch14 ● Markus Wagner14 ● Gary Wilkinson14 ● Bingyan Wu13 ● Michael Jeffers13 ● Isabelle Genvresse14 ●
`Alwin Krämer15
`
`Received: 23 June 2020 / Revised: 15 July 2020 / Accepted: 21 July 2020 / Published online: 30 July 2020
`© The Author(s) 2020. This article is published with open access
`
`Abstract
`The mutant IDH1 (mIDH1) inhibitor BAY1436032 demonstrated robust activity in preclinical AML models, supporting
`clinical evaluation. In the current dose-escalation study, BAY1436032 was orally administered to 27 mIDH1 AML subjects
`across 4 doses ranging from 300 to 1500 mg twice-daily. BAY1436032 exhibited a relatively short half-life and apparent
`non-linear pharmacokinetics after continuous dosing. Most subjects experienced only partial target inhibition as indicated by
`plasma R-2HG levels. BAY1436032 was safe and a maximum tolerated dose was not identified. The median treatment
`duration for all subjects was 3.0 months (0.49–8.5). The overall response rate was 15% (4/27; 1 CRp, 1 PR, 2 MLFS), with
`responding subjects experiencing a median treatment duration of 6.0 months (3.9–8.5) and robust R-2HG decreases. Thirty
`percent (8/27) achieved SD, with a median treatment duration of 5.5 months (3.1–7.0). Degree of R-2HG inhibition and
`clinical benefit did not correlate with dose. Although BAY1436032 was safe and modestly effective as monotherapy, the low
`overall response rate and incomplete target inhibition achieved at even the highest dose tested do not support further clinical
`development of this investigational agent in AML.
`
`1234567890();,:
`1234567890();,:
`
`Supplementary information The online version of this article (https://
`doi.org/10.1038/s41375-020-0996-5) contains supplementary
`material, which is available to authorized users.
`
`* Michael Heuser
`heuser.michael@mh-hannover.de
`
`1 Department of Hematology, Hemostasis, Oncology, and Stem Cell
`Transplantation, Hannover Medical School, Hannover, Germany
`2 Department of Medical Oncology, Sidney Kimmel Cancer Center,
`Thomas Jefferson University, Philadelphia, PA, USA
`3 Department of Medical Oncology, Montefiore Medical Center,
`Albert Einstein College of Medicine, Bronx, NY, USA
`4 Division of Hematology, Department of Medicine, The Ohio State
`University Comprehensive Cancer Center, Columbus, OH, USA
`5 Department of Leukemia, The University of Texas M.D. Anderson
`Cancer Center, Houston, TX, USA
`
`6
`
`Tisch Cancer Institute, Division of Hematology/Oncology,
`Icahn School of Medicine at Mount Sinai, New York, NY, USA
`7 Department of Medicine, Roswell Park Comprehensive Cancer
`Center, Buffalo, NY, USA
`8 Department of Hematology and Oncology, University Hospital
`Hamburg-Eppendorf, Hamburg, Germany
`
`9 Division of Hematology and Oncology, Charité University
`Hospital Berlin, Berlin, Germany
`10 Division of Hematology and Oncology, University Hospital
`Leipzig, Leipzig, Germany
`11 Section on Hematology and Oncology, Comprehensive Cancer
`Center of Wake Forest Baptist Health, Winston-Salem, NC, USA
`12 Division of Hematology-Oncology, Perelman School of Medicine,
`Abramson Cancer Center of the University of Pennsylvania,
`Philadelphia, PA, USA
`13 Pharmaceuticals Division, Bayer HealthCare Pharmaceuticals,
`Inc., Whippany, NJ, USA
`14 Pharmaceuticals Division, Bayer AG, Berlin, Germany
`15 Clinical Cooperation Unit Molecular Hematology/Oncology,
`German Cancer Research Center (DKFZ) and Department of
`Internal Medicine V, University of Heidelberg,
`Heidelberg, Germany
`
`Rigel Exhibit 1041
`Page 1 of 11
`
`

`

`2904
`
`Introduction
`
`Somatic hotspot mutations in isocitrate dehydrogenase 1
`(IDH1) have been identified in a variety of cancers, with a
`frequency of ~7% in acute myeloid leukemia (AML) [1–6].
`Tumor-associated IDH1 mutations (mIDH1) change the
`conserved arginine at codon 132 in the enzymatic active site
`to a variety of alternative amino acids (R132X), and in
`doing so confer a neomorphic activity to this enzyme.
`Whereas wild-type IDH1 (wtIDH1) catalyzes the conver-
`sion of isocitrate to α-ketoglutarate (α-KG), mIDH1 con-
`verts α-KG to R-2-hydroxyglutarate (R-2HG). Subjects
`with mIDH1 AML show elevated R-2HG levels, which
`inhibits α-KG-dependent enzymes, thereby leading to epi-
`genetic alterations and ultimately impaired hematopoietic
`differentiation [7–14].
`BAY1436032 is an oral small-molecule inhibitor of
`mIDH1 that is active in preclinical models of mIDH1 cancer
`[15–17]. Most mIDH1 inhibitors, including BAY1436032,
`reportedly interact with an allosteric site on the mutant
`enzyme, although an inhibitor which interacts directly with
`the active site was recently described [18]. Preclinical
`focusing on mIDH1 AML found that
`experiments
`BAY1436032 inhibits R-2HG production and colony
`growth in vitro, while promoting leukemic blast clearance,
`myeloid differentiation, and survival in animal models [16].
`Supported by these encouraging preclinical findings,
`BAY1436032 was evaluated in a phase I clinical study in
`subjects with mIDH1 AML (NCT03127735), the results of
`which are presented herein. Objectives of the study include
`determination of the maximum tolerated dose (MTD) and
`the recommended phase II dose (RP2D), and evaluation of
`the safety, tolerability, pharmacokinetics (PK), pharmaco-
`dynamics and clinical activity of BAY1436032.
`
`Materials and methods
`
`Study design
`
`This study was an open-label, nonrandomized, multicenter
`phase I trial. Subjects were screened at 13 hospital sites in 2
`countries (USA and Germany). The study was to consist of
`dose-escalation followed by dose-expansion. The MTD
`identified in dose-escalation was to be used in the dose-
`expansion. If MTD was not reached in dose-escalation, a
`dose for the expansion would be selected based on available
`PK,
`pharmacodynamic,
`safety,
`and
`efficacy
`data.
`BAY1436032 was administered twice-daily (BID) in con-
`tinuous 28-day cycles. In dose-escalation, up to 9 evaluable
`subjects could be enrolled per cohort, with a minimum of
`3 dose-limiting toxicity (DLT)-free evaluable subjects
`required prior
`to escalating to the next highest dose.
`
`M. Heuser et al.
`
`Bayesian dose-DLT modeling was performed to help guide
`dosing decisions [19]. MTD was defined as the highest dose
`of BAY1436032 that could be given such that ≤25% of
`subjects were predicted to experience a DLT.
`DLTs were differentially defined for nonhematopoietic
`versus hematopoietic toxicities. Nonhematopoietic toxi-
`cities of ≥grade 3 occurring during the first cycle of treat-
`ment were to be considered DLTs with the following
`exceptions: (1) Alopecia and nausea controlled by medical
`management; (2) Tumor lysis syndrome if successfully
`managed clinically and resolved within 7 days without any
`end-organ damage; (3) Differentiation syndrome (DS) if
`successfully managed clinically and resolved within 7 days
`without any end-organ damage; (4) Asymptomatic ≥grade 3
`electrolyte abnormalities not considered clinically sig-
`nificant by the investigator. Missing >20% of doses of study
`drug due to any drug-related toxicity, or delay in the start of
`cycle 2 by more than 14 days due to any drug-related
`toxicity, were also considered DLTs. For certain toxicities
`such as laboratory assessments without a clear clinical
`correlate, a discussion between the investigator and the
`sponsor determined whether the adverse event (AE) should
`be assessed as a DLT.
`thrombocytopenia of
`toxicities,
`For hematological
`≥grade 3 with clinically significant bleeding, or grade 4
`neutropenia persisting 42 days after the start of treatment in
`the absence of active AML, were considered DLTs.
`DLTs identified during the first cycle of treatment were
`used to guide dose-escalation decisions and to determine the
`MTD, and if safety issues appeared in subsequent cycles
`they were also to be considered. Hydroxyurea was per-
`mitted during the first cycle if white blood cell (WBC)
`20 × 109/L and was
`exceeded
`also
`permitted
`for
`treatment of DS.
`The study protocol was approved by the institutional
`review board of participating institutions and complied with
`the Declaration of Helsinki, current Good Clinical Practice
`guidelines, and local laws and regulations. Written informed
`consent was provided by all participants prior
`to the
`initiation of any study-specific procedure. Data were entered
`into clinical research forms by the study investigators and
`their staff. The study was sponsored by Bayer AG.
`
`Subjects
`
`Male and female subjects of ≥18 years of age with an
`Eastern Cooperative Oncology Group (ECOG) performance
`status of ≤2 and advanced AML were eligible. Information
`regarding ELN 2010 risk classification was collected during
`screening [20]. Subjects were required to harbor a missense
`mutation in IDH1-R132X based on local testing reported by
`study investigators, with sponsor review of test results prior
`to enrollment. IDH1 mutational status from a bone marrow
`
`Rigel Exhibit 1041
`Page 2 of 11
`
`

`

`Safety and efficacy of BAY1436032 in IDH1-mutant AML: phase I study results
`
`2905
`
`sample collected during screening was retrospectively
`evaluated at Foundation Medicine. Subjects were to be
`relapsed or refractory to at least 1 previous line of therapy,
`or intolerable to or unable to receive established therapies,
`and could have received any number and type of prior
`therapies prior
`to enrollment, except
`those targeting
`mIDH1. A cohort sample size of 3 to 9 DLT-evaluable
`subjects in dose-escalation was chosen based on experience
`and simulation results from adaptive Bayesian dose-DLT
`model. This number of subjects is anticipated to provide
`sufficient safety information to help guide dose escalation
`decisions in a reasonable time frame without exposing an
`excess number of subjects to potentially toxic or inactive
`doses of study drug. Dose-expansion was not conducted in
`this study.
`
`Study assessments
`
`The primary objectives of the study were to determine the
`safety,
`tolerability, MTD,
`and/or RP2D dose
`of
`BAY1436032 administered in a twice-daily dosing sche-
`dule in subjects with mIDH1 advanced AML. Secondary
`objectives were to evaluate PK and to assess pharmacody-
`namic effects and evidence of clinical efficacy.
`
`Safety
`
`Safety and tolerability were evaluated by analysis of
`adverse events, physical examinations, vital signs, ECOG
`performance status, and various laboratory assessments.
`For safety monitoring, subjects were scheduled for clinic
`visits every week for the initial three cycles of treatment,
`after which time visits could be reduced to every-other-
`week with investigator and sponsor agreement. Cardiac
`function was assessed with triplicate 12-lead electro-
`cardiograms (ECG) at screening, C1-D1, C1-D2, C1-
`D15, D1 of every subsequent cycle, and at treatment end.
`Severity of adverse events and toxicities were graded by
`investigators according to the National Cancer Institute
`Common Terminology Criteria for Adverse Events (NCI-
`CTCAE) version 4.03. AEs are presented by the Medical
`Dictionary for Regulatory Activities (MedDRA) v21.1.
`
`Efficacy
`
`Disease assessments and response evaluations from bone
`marrow aspirate or biopsy were scheduled at screening, C2-
`D1, C3-D1, D1 of every second cycle thereafter, and at
`treatment end (if not done on D1 of the last cycle). Per-
`ipheral blood was analyzed at each of these time points and
`at additional
`times between bone marrow assessments.
`Clinical efficacy was assessed by investigators using the
`modified 2003 International Working Group response
`
`criteria for AML [21] with some changes based on 2017
`European Leukemia Net recommendations [22]. Response
`categories included: complete remission (CR), morphologic
`CR with incomplete hematological recovery (CRh), mor-
`phologic CR with incomplete platelet recovery (CRp),
`morphologic leukemia-free state (MLFS), partial remission
`(PR), stable disease (SD), and progressive disease (PD).
`Overall response rate included CR, CRh, CRp, MLFS, and
`PR. Following study completion,
`investigators provided
`survival information.
`
`Pharmacokinetics
`
`Plasma samples were collected at various times on C1-D1,
`C1-D2, C1-D8, and C1-D15 and stored frozen for PK
`assessments. Collection times on these days are indicated on
`Supplementary Fig. S1 (C1-D8 PK results are not shown).
`The evening dose of BAY1436032 was withheld on C1-D1
`to facilitate assessment of the 24-h single-dose time point on
`C1-D2. Quantitative analysis of BAY1436032 (free acid) in
`plasma was performed as described in Supplementary
`Methods.
`
`Pharmacodynamics
`
`For quantification of R-2HG, plasma samples were col-
`lected at
`the following time points and stored frozen:
`screening, C1-D1 (pre-dose and post-dose), C1-D8 (pre-
`dose), C1-D15 (pre-dose and post-dose), C1-D22 (pre-
`dose), pre-dose on D1 and D15 of each cycle thereafter, and
`at
`treatment end. Plasma R-2HG concentrations were
`measured by Eurofins as described in Supplementary
`Methods.
`
`Retrospective mutational analysis
`
`Mutational analysis was performed on bone marrow aspi-
`rates or biopsies collected during screening and on a subset
`of samples collected during BAY1436032 treatment. Test-
`ing was performed by Foundation Medicine using the
`FoundationOne Heme panel which detects alterations in
`>400 tumor-associated genes via next-generation sequen-
`cing. Information provided by Foundation Medicine inclu-
`ded the allelic frequency and the likely pathogenic nature of
`each alteration identified.
`
`Results
`
`Subjects
`
`Thirty-three mIDH1 AML subjects signed informed con-
`sent, 27 received BAY1436032 treatment and 6 failed
`
`Rigel Exhibit 1041
`Page 3 of 11
`
`

`

`2906
`
`M. Heuser et al.
`
`Fig. 1 Subject disposition. The chart shows an overview of subjects who signed consent to participate in the study, their allocation into dose-
`escalation treatment cohorts, and reasons for their discontinuation from the study.
`
`screening for various reasons (e.g., presence of an uncon-
`trolled infection). The first subject started treatment on June
`28, 2017 and the last study visit was on December 5, 2018.
`All subjects were treated within dose-escalation, and a
`planned dose-expansion was not pursued. BAY1436032
`tablets were orally administered BID, with the evening dose
`withheld on the first day of treatment to facilitate PK ana-
`lysis. Administration was continuous and each treatment
`cycle was 28 days. Subjects were treated across 4 dosing
`cohorts: cohort 1 (300 mg BID; n = 7); cohort 2 (600 mg
`BID; n = 4); cohort 3 (1200 mg BID; n = 7) and cohort 4
`(1500 mg BID; n = 9) (Fig. 1). The dosing schedule and
`starting dose were selected based on preclinical PK mod-
`eling and safety data, and on the results of the ongoing first-
`in-human phase I trial of BAY1436032 in subjects with
`mIDH1 solid tumors
`(NCT02746081;
`[23]). Baseline
`demographics and disease characteristics are provided in
`Table 1. Subjects had received a median of 2 (0–8) prior
`systemic therapies for AML and 4 had received no prior
`systemic therapies. The prevalence of individual IDH1-
`R132X mutations across the 27 treated subjects based on
`investigator-reported information was as follows: R132C
`(n = 15), R132H (n = 5), R132G (n = 3), R132L and
`
`R132S (n = 2 each). Consistent with previous reports
`[4, 24], R132C and R132H were the most prevalent IDH1
`mutations identified.
`
`Pharmacokinetics and Pharmacodynamics
`
`PK analysis was performed on C1-D1 after single-dose oral
`administration and on C1-D15 following continuous BID
`dosing. Following single oral administration, BAY1436032
`plasma concentrations were detectable 30 min after admin-
`istration. Maximum plasma concentrations were observed
`~3 to 4 h after single-dose and continuous BID adminis-
`tration (Supplementary Fig. S1). In the evaluated dose
`range, BAY1436032 exposure after single-dose adminis-
`tration generally increased in a dose-proportional manner
`(1.6–2.0-fold); however, dose-proportionality was not
`apparent after continuous BID administration. Minimal
`accumulation was evident at C1-D15 and inter-subject
`variability was high in all cohorts for the main PK para-
`meters evaluated (Supplementary Fig. S1 and Table S1).
`To evaluate potential effects of the study drug on target
`inhibition, R-2HG levels were measured in plasma samples
`obtained at baseline and at various time points during
`
`Rigel Exhibit 1041
`Page 4 of 11
`
`

`

`Safety and efficacy of BAY1436032 in IDH1-mutant AML: phase I study results
`
`2907
`
`Table 1 Baseline demographic and disease characteristicsa.
`
`Cohort 1:
`300 mg
`BID
`n = 7
`
`Cohort 2:
`600 mg
`BID
`n = 4
`
`Cohort 3:
`1200 mg
`BID
`n = 7
`
`Cohort 4:
`1500 mg
`BID
`n = 9
`
`Total
`
`n = 27
`
`67 (37–86)
`
`72 (51–76)
`
`70 (42–83)
`
`68 (27–79)
`
`69 (27–86)
`
`4 (57)
`3 (43)
`
`5 (71)
`2 (29)
`
`1 (14)
`4 (57)
`2 (29)
`9 (1–32)
`
`0
`1 (14)
`4 (57)
`2 (29)
`
`3 (43)
`3 (43)
`1 (14)
`0
`0
`1 (0–4)
`
`2 (50)
`2 (50)
`
`3 (75)
`1 (25)
`
`1 (25)
`2 (50)
`1 (25)
`10 (2–19)
`
`0
`1 (25)
`2 (50)
`1 (25)
`
`2 (50)
`0
`2 (50)
`0
`0
`1 (1–3)
`
`3 (43)
`4 (57)
`
`4 (51)
`3 (43)
`
`1 (14)
`6 (86)
`0
`9 (1–25)
`
`1 (14)
`3 (43)
`2 (29)
`1 (14)
`
`5 (71)
`0
`0
`2 (29)
`0
`2 (0–4)
`
`3 (33)
`6 (67)
`
`7 (78)
`2 (22)
`
`1 (11)
`7 (78)
`1 (11)
`15 (5–44)
`
`2 (22)
`2 (22)
`5 (56)
`0
`
`5 (56)
`2 (22)
`0
`0
`2 (22)
`3 (1–8)
`
`12 (44)
`15 (56)
`
`19 (70)
`8 (30)
`
`4 (15)
`19 (70)
`4 (15)
`11 (1–44)
`
`3 (11)
`7 (26)
`13 (48)
`4 (15)
`
`15 (56)
`5 (19)
`3 (11)
`2 (7)
`2 (7)
`2 (0–8)
`
`5 (71)
`
`4 (100)
`
`5 (71)
`
`9 (100)
`
`23 (85)
`
`6 (86)
`1 (14)
`
`4 (100)
`0
`
`7 (100)
`0
`
`7 (78)
`2 (22)
`
`24 (89)
`3 (11)
`
`Age [median (range), in years]
`Sex, n (%)
`Male
`Female
`AML classification, n (%)
`De novo AML
`Secondary AML
`ECOG performance status, n (%)
`0
`1
`2
`Time from initial diagnosis to 1st dose of
`study drug [median (range), in months]b
`ELN risk classification, n (%)c
`Favorable
`Intermediate
`Adverse
`Missing
`mIDH1 identified, n (%)d
`R132C
`R132H
`R132G
`R132L
`R132S
`Number of prior systemic antileukemic
`lines of therapies for AML, [median
`(range)]
`Subjects having received at least 1 prior
`systemic antileukemic therapy for AML, n
`(%)
`Prior allogeneic transplantation
`No
`Yes
`
`BID twice-daily, ECOG Eastern Cooperative Oncology Group, ELN European LeukemiaNet, n number of subjects.
`aPercentages are calculated including missing values.
`bFor the calculation of time from initial diagnosis, only subjects with complete date information (year, month, day) are included.
`cELN classification as per 2010 recommendations [20].
`dInvestigator-reported mIDH1 results used for subject enrollment are shown. Retrospective evaluation of baseline leukemic samples at a central
`laboratory via next-generation sequencing confirmed investigator-reported results in all evaluable cases.
`
`treatment. Baseline R-2HG levels were highly variable,
`with a median concentration of 1755 ng/mL (78–14,749)
`(Supplementary Table S2). Of the 27 subjects treated, 26
`had baseline R-2HG levels above those seen in wtIDH1
`cancers (61 ng/mL) and above the 97th percentile upper
`reference limit found in healthy individuals (138 ng/mL)
`[25, 26].
`
`All subjects achieved a lowering of baseline R-2HG
`levels during BAY1436032 treatment, with a median
`maximal decrease across all subjects of 66% (16–99)
`(Supplementary Table S2). However, only 5/26 subjects
`with an elevated baseline R-2HG level experienced a
`reduction to a normal level of ≤138 ng/mL. Maximal R-
`2HG decreases did not show a clear relationship with
`
`Rigel Exhibit 1041
`Page 5 of 11
`
`

`

`2908
`
`M. Heuser et al.
`
`Table 2 BAY1436032-related treatment-emergent adverse events occurring in ≥5% of treated subjects.
`
`TEAE
`
`Sinus tachycardia
`Diarrhea
`Nausea
`
`Vomiting
`Fatigue
`
`Decreased appetite
`
`Hypomagnesaemia
`Differentiation syndromea
`
`Dyspnea
`
`CTCAE
`grade
`
`Grade 1
`Grade 2
`Grade 1
`Grade 2
`Grade 1
`Grade 1
`Grade 2
`Grade 3
`Grade 1
`Grade 2
`Grade 1
`Grade 2
`Grade 3
`Grade 1
`Grade 2
`
`Cohort 1:
`300 mg
`BID
`n = 7
`
`Cohort 2:
`600 mg
`BID
`n = 4
`
`Cohort 3:
`1200 mg
`BID
`n = 7
`
`Cohort 4:
`1500 mg
`BID
`n = 9
`
`Total
`
`n = 27 (%)
`
`1
`0
`1
`1
`2
`0
`1
`0
`1
`1
`1
`0
`0
`1
`1
`
`0
`0
`1
`0
`1
`1
`0
`0
`0
`0
`0
`0
`0
`0
`0
`
`0
`2
`1
`0
`0
`0
`1
`0
`0
`0
`0
`0
`1
`1
`0
`
`1
`0
`1
`0
`1
`0
`0
`1
`0
`0
`1
`1
`0
`0
`0
`
`2 (7)
`2 (7)
`4 (15)
`1 (4)
`4 (15)
`1 (4)
`2 (7)
`1 (4)
`1 (4)
`1 (4)
`2 (7)
`1 (4)
`1 (4)
`2 (7)
`1 (4)
`
`BID twice-daily, CTCAE common terminology criteria for adverse events, n number of subjects, TEAE treatment-emergent adverse events (as per
`MedDRA PT).
`aSymptoms were not reported, only the term differentiation syndrome (DS). Three additional subjects presented with symptoms associated with DS
`and are not listed in the table since the individual symptoms occurred in <5% of treated subjects.
`
`BAY1436032 dose, plasma exposure, or the specific IDH1-
`R132X mutation detected at baseline.
`
`Safety
`
`Seventeen of 27 treated subjects experienced at least 1
`treatment-emergent adverse event (TEAE) of any grade
`related to BAY1436032, with events occurring in ≥5% of
`treated subjects as listed in Table 2. Seven of 27 subjects
`least 1 TEAE of ≥grade 3 related to
`experienced at
`BAY1436032, with each of the following events occurring
`in a single subject: amylase increase, DS, fatigue, febrile
`neutropenia, hyponatremia,
`lung infiltration, peripheral
`edema, pneumonitis, and decreased WBC count (grade 3);
`anemia, ileus, decreased neutrophil count, decreased platelet
`count, and sepsis (grade 4). Twenty-four subjects experi-
`enced at least 1 TEAE ≥ grade 3 irrespective of relationship
`to study drug (Supplementary Table S3).
`DS was considered an AE of special interest (AESI).
`Five of 27 (19%) treated subjects (2 each in cohorts 1 and 4,
`1 in cohort 3) presented with DS, or symptoms associated
`with DS, which occurred between 8 and 76 days after the
`start of treatment. For treatment of DS symptoms, 4 subjects
`received steroids, 2 received hydroxyurea, and 1 had an
`interruption of study drug. All cases eventually resolved and
`2 of the subjects experienced a clinical response (1 CRp,
`1 MLFS).
`
`No MTD was identified. A single DLT was reported:
`grade 4 ileus in subject 21 (cohort 4) who had a prior his-
`tory of
`intestinal pseudo-obstruction. The ileus, which
`started on the 4th day of treatment, resolved following dose
`interruption and returned upon reintroduction of study drug.
`There were no clinically significant ECG such as pro-
`longation of corrected QT interval related to BAY1436032
`treatment.
`Six of 27 subjects died either during BAY1436032
`treatment or within 30 days of permanent treatment dis-
`continuation, and none of these deaths were attributed to
`study drug. Causes of death were reported as PD or AE, in
`3 subjects each. AEs associated with these deaths included
`sepsis,
`lung infection,
`and general physical health
`deterioration.
`
`Efficacy
`
`The median treatment duration for all 27 subjects was
`3.0 months (0.49–8.5) (Supplementary Table S2), and there
`was a trend towards increased neutrophil counts occurring
`during treatment (Supplementary Fig. S2).
`The overall response rate to BAY1436032, based on
`investigator assessment, was 15% (4/27): subject 16 (1200
`mg BID) attained CRp at C7; subject 5 (300 mg BID)
`attained MLFS at C3; subject 24 (1500 mg BID) attained
`MLFS at C2 and went on to receive a 2nd stem cell
`
`Rigel Exhibit 1041
`Page 6 of 11
`
`

`

`Safety and efficacy of BAY1436032 in IDH1-mutant AML: phase I study results
`
`2909
`
`transplant; and subject 13 (1200 mg BID) attained PR at C5
`(Supplementary Table S2). The median treatment duration
`of these four responding subjects was 6.0 months (3.9–8.5).
`Blast counts and plasma R-2HG levels determined at var-
`ious times during the treatment of
`these subjects are
`depicted in Fig. 2. In subject 13, R-2HG levels rapidly
`decreased to a normal level found in healthy individuals,
`whereas the reduction of R-2HG to a normal level took a
`longer time in subjects 5 and 16.
`Best response of SD and PD was achieved by 18/27 and
`5/27 subjects, respectively (Supplementary Table S2). Of
`the 18 subjects with SD, 8 remained stable for ≥2 con-
`secutive monthly assessments, stayed on treatment
`for
`≥3 months and are considered to have achieved relevant SD.
`This group of 8 subjects showed a median treatment dura-
`tion of 5.5 months (3.1–7.0). Some decrease in leukemic
`blast counts and plasma R-2HG levels was evident in most
`of these subjects, but R-2HG levels generally did not reach
`normal levels (Supplementary Table S2, Figs. S3 and S4).
`As of November 20, 2019, 26/27 subjects were reported
`as deceased. Survival data by dosing cohort, as well as for
`the 27 subjects as a single group, is presented in Fig. 3.
`Median overall survival for the group of 27 treated subjects
`was 6.6 months (95% CI 4.6–9.4). The small subject
`numbers in each dosing cohort (n = 4–9) preclude relevant
`comparisons among cohorts.
`
`Retrospective mutational analysis
`
`Following enrollment, retrospective mutational analysis of
`>400 tumor-associated genes was successfully performed
`on baseline leukemic samples obtained from 25/27 treated
`subjects and the mIDH1 results from this analysis were
`concordant with those obtained locally via investigator
`assessment (Supplementary Tables S2 and S4).
`Baseline leukemic samples from 24/25 of the subjects for
`whom retrospective mutational analysis was successfully
`performed harbored alterations in additional genes classified
`as known or likely pathogenic, with a median of 2.5 (0–4)
`co-occurring alterations detected per subject. Co-occurring
`alterations in the following genes were found in >10% of
`evaluated subjects: DNMT3A (n = 11), ASXL1 (n = 7),
`NRAS (n = 5), NPM1 (n = 5), FLT3 (n = 4; 2 with an
`internal
`tandem duplication (ITD) and 2 with a kinase
`domain point mutation), TET2 and U2AF1 (n = 3 each).
`Co-occurring mutations in tyrosine kinase pathway genes
`such as NRAS, FLT3, and PTPN11 were absent
`in
`responding subjects.
`A co-occurring IDH2-R140Q mutation was detected in
`subjects 16 and 25, with variant allele frequencies (VAF) of
`13% for mIDH1/26% for mIDH2 in subject 16 and 5% for
`mIDH1/20% for mIDH2 in subject 25. Subject 16 experi-
`enced a significant
`reduction in R-2HG levels during
`
`BAY1436032 treatment, achieved a clinical response (CRp)
`and remained on treatment for 8.5 months (Supplementary
`Table S2, Fig. 2). Subject 25, who had received the mIDH2
`inhibitor enasidenib for the 10 months just prior to enroll-
`ment, experienced a modest reduction in R-2HG levels
`during BAY1436032 treatment, achieved a best response of
`SD and remained on treatment for 6.2 months (Supple-
`mentary Table S2, Fig. S4).
`An analysis of baseline samples showed the median
`VAF of mIDH1 to be 28% (5–49)
`(Supplementary
`Tables S2 and S4). The VAF of mIDH1 in comparison to
`the VAF of co-occurring mutations suggested that mIDH1
`was present
`in the founding clone in 18/25 evaluated
`subjects. There was no apparent correlation between the
`mIDH1 VAF and the specific IDH1-R132X mutation
`identified, or clinical outcome. Maximal R-2HG decreases
`detected during BAY1436032 treatment also did not cor-
`relate with baseline mIDH1 VAF.
`In addition to the analysis on baseline samples, for a
`limited number of subjects a retrospective mutational ana-
`lysis was also performed on bone marrow samples collected
`during BAY1436032 treatment. Results from three subjects
`who had been on treatment for at least three full cycles of
`treatment and had data available from baseline, during
`treatment and end-of-treatment (EOT) are shown in Sup-
`plementary Fig. S5. Although the IDH1 mutation present at
`baseline was always detected in samples collected during
`treatment and at the EOT, changes in other genes were
`evident: for subject 12 the EOT sample showed mutations
`in WT1, BCOR, and FLT3 (ITD) and a PDGFRB-TPM3
`gene fusion that were undetectable at baseline or at C3-D1;
`for subject 14 the EOT sample showed a MLL duplication/
`rearrangement that was not detected at baseline or C3-D1;
`and for subject 18 a mutation in STAG2 that was detected at
`baseline and at C3-D1 was not detected in the EOT sample,
`whereas the converse was true for mutations in PIK3C2G
`and CREBBP. Potential resistance mechanisms such as
`isoform switching and the acquisition of secondary altera-
`tions in mIDH1 were evaluated and not detected.
`
`Discussion
`
`the mIDH1 inhibitor BAY1436032 was
`In this study,
`evaluated at 4 different dose levels in a total of 27 AML
`subjects who harbored a mutation which altered the residue
`at position R132 of IDH1 to any one of 5 different amino
`acids, each of which is known to generate the R-2HG
`oncometabolite and to be inhibited by BAY1436032
`[16, 17].
`BAY1436032 was found to be generally safe, with just a
`single DLT (ileus) identified among the 27 treated subjects.
`The only other gastrointestinal-related TEAEs ≥grade 3
`
`Rigel Exhibit 1041
`Page 7 of 11
`
`

`

`2910
`
`M. Heuser et al.
`
`Fig. 2 Per-subject blast percentages and R-2HG levels from sub-
`jects who experienced clinical responses during BAY1436032
`treatment. Data from the four subjects who experienced clinical
`responses (1 CRp, 2 MLFS, 1 PR) is shown. Blast percentages are
`shown on the left Y-axis, plasma R-2HG levels on the right Y-axis and
`
`treatment cycle number on the X-axis. Bone marrow blasts are illu-
`strated with blue circles and peripheral blood blasts/leukocytes with
`red triangles. R-2HG levels are represented by gray bars and the R-
`2HG level associated with healthy individuals (138 ng/mL) is shown
`as black squares connected by a dotted black line.
`
`Fig. 3 Subject survival after
`start of BAY1436032
`treatment. Overall survival is
`depicted according to dosing
`cohort (a) and for all 27 treated
`subjects (b). Censored subjects
`are indicated with an asterisk.
`
`were two episodes of grade 3 nausea in cohort 4, which
`were not attributed to study drug. Each ≥grade 3 TEAE
`attributed to BAY1436032 occurred in a single subject. The
`most frequent ≥grade 3 TEAEs associated with ivosidenib,
`which is approved for a subset of subjects with mIDH1
`AML [27, 28], are QT prolongation (8%) and IDH DS (4%)
`[29]. Five subjects treated with BAY1436032 presented
`with DS, or symptoms associated with DS (all ≤grade 3),
`and all cases eventually resolved. There were no episodes of
`≥grade 3 QT prolongation with BAY1436032. The favor-
`able safety profile of BAY1436032 is consistent with pre-
`clinical studies demonstrating that BAY1436032 is a highly
`
`specific inhibitor of mIDH1 which exhibits little activity
`against wtIDH1 or wtIDH2 [16, 17].
`Modest clinical benefit from single-agent BAY1436032
`was identified in this study as evidenced by an overall
`response rate of 15% and a median treatment duration of
`6.0 months among responders. Eight additional subjects
`who achieved SD for ≥2 consecutive monthly assessments
`and stayed on treatment for ≥3 months showed a median
`treatment duration of 5.5 months. Most of these eight sub-
`jects experienced decreases in both bone marrow and per-
`ipheral blood blast counts during treatment, consistent with
`some degree of clinical benefit from BAY1436032. Clinical
`
`Rigel Exhibit 1041
`Page 8 of 11
`
`

`

`Safety and efficacy of BAY1436032 in IDH1-mutant AML: phase I study results
`
`2911
`
`relationship to
`
`outcome did not exhibit an obvious
`BAY1436032 dose.
`Although BAY1436032 demonstrated some degree of R-
`2HG inhibition in all subjects, only 5/26 subjects with an
`elevated baseline R-2HG level experienced a reduction to a
`normal level during BAY1436032 treatment. In support of a
`potential relationship between the degree of target inhibition
`and clinical benefit, the four subjects who achieved clinical
`responses showed a median maximal inhibition of baseline
`R-2HG levels of 95% (75–99), compared with 58% (16–93)
`among the 23 other subjects. A correlation between the
`degree of R-2HG suppression and clinical response was not
`found for ivosidenib [29], although in that study R-2HG
`levels were strongly suppressed in the majority of subjects
`regardless of clinical response, thereby making it difficult to
`directly address this question.
`It is unknown why a subset of subjects treated with
`BAY1436032 exhibit robust R-2HG suppression coupled
`with clinical benefit, while others do not. A dose-response
`relationship was not apparent when the degree of maximal
`percent R-2HG inhibition or the extent of clinical benefit
`was compared between cohorts, and subjects who achieved
`clinical benefit did not exhibit plasma concentrations
`markedly higher than those who did not.
`The overall response rate of 15% for BAY1436032 is
`lower than the 33–42% reported for other mIDH1 inhibitors
`in relapsed and refractory AML [29–32], and the median
`overall survival (6.6 months; 95% CI 4.6–9.4) is somewhat
`lower than that reported for ivosidenib in similar populations
`(8.8 months; 95% CI 6.7–10.2) [29]. It is likely that the
`incomplete target inhibition generated by BAY1436032