Clinical pharmacology and tolerability of REC‐994, a redox‐cycling nitroxide compound, in randomized phase 1 dose‐finding studies

Abstract

Cerebral cavernous malformation (CCM) has variable clinical symptoms, including potentially fatal hemorrhagic stroke. Treatment options are very limited, presenting a large unmet need. REC‐994 (also known as tempol), identified as a potential treatment through an unbiased drug discovery platform, is hypothesized to treat CCMs through a reduction in superoxide, a reactive oxygen species. We investigated the safety, tolerability, and pharmacokinetic profile of REC‐994 in healthy volunteers. Single‐ and multiple‐ascending dose (SAD and MAD, respectively) studies were conducted in adult volunteers (ages 18–55). SAD study participants received an oral dose of REC‐994 or placebo. MAD study participants were randomized 3:1 to oral doses of REC‐994 or matching placebo, once daily for 10 days. Thirty‐two healthy volunteers participated in the SAD study and 52 in the MAD study. Systemic exposure increased in proportion to REC‐994 dose after single doses of 50–800 mg and after 10 days of dosing over the 16‐fold dose range of 50–800 mg. Median T _max and mean t _1/2 were independent of dose in both studies, and the solution formulation was more rapidly absorbed. REC‐994 was well tolerated. Treatment‐emergent adverse effects across both studies were mild and transient and resolved by the end of the study. REC‐994 has a favorable safety profile and was well tolerated in single and multiple doses up to 800 mg with no dose‐limiting adverse effects identified. Data support conducting a phase 2 clinical trial in patients with symptomatic CCM.

Cerebral cavernous malformation (CCM) pathogenesis involves elevated reactive oxygen species (ROS) levels. REC‐994 restores ROS balance, and in double‐blind, placebo‐controlled, trials in healthy volunteers, had low potential for off‐target adverse effects and pharmacokinetics suitable for phase 2 development.

Abbreviations

AE

adverse event

AUC

area under the plasma concentration versus time curve

AUC_0–24

AUC from time 0–24 h post‐dose

AUC_0–t and/or AUC_last

time 0 to time of last quantifiable concentration

AUC_inf

AUC from time 0 extrapolated to infinity

BMI

body mass index

C₂₄

trough plasma concentration observed at the end of dosing interval

CCM

cerebral cavernous malformation

CIs

confidence intervals

CL/F

apparent oral clearance

C _max

maximum observed plasma concentration

CV

coefficient of variation

ECG

electrocardiogram

HPLC

high‐performance liquid chromatography

LLOQ

lower limit of quantitation

LS

geometric least‐squares

MAD

multiple‐ascending dose

PK

pharmacokinetic

ROS

reactive oxygen species

SAD

single‐ascending dose

SD

standard deviation

t _1/2

terminal elimination half‐life

TEAE

treatment‐emergent adverse event

T _max

time to attain C _max

Vz/F

apparent volume of distribution

1. INTRODUCTION

Cerebral cavernous malformation (CCM) is a vascular endothelial disease characterized by vascular malformations in small vessels of the brain and spinal cord; while some cases are asymptomatic, others present with diverse clinical symptoms including headaches, seizures, neurologic deficits, and risk of fatal hemorrhagic stroke. ¹ Hemorrhage risk is significantly linked to the total number of CCM lesions (abnormal clusters of thin‐walled blood vessels), ¹ , ² and while life expectancy is consistent with the general population, morbidity‐free life expectancy declines substantially based on location and depth of lesions. ³ Approximately 20% of cases arise from inherited familial CCM syndrome and 80% arise sporadically, with overall prevalence estimated at 0.1%–0.8% of the general population ⁴ and 15/100 000 persons in the Orphanet database. ⁵ CCM is caused by inherited or acquired loss‐of‐function mutations in genes KRIT1 (CCM1), MGC4607 (CCM2), or PDCD10 (CCM3) in endothelial cells. ⁶ Proteins encoded by these genes are involved in redox homeostasis, and loss‐of‐function mutations activate pathways associated with downstream elevation in reactive oxygen species (ROS). ⁷ , ⁸ , ⁹ Molecular pathogenesis of CCM lesion formation has several explanations: a Knudson 2‐hit mechanism in which CCM develops following complete loss of 2 alleles on a CCM gene; haploinsufficiency, presence of a single functional copy of the CCM gene; or paradominant inheritance resulting in loss of heterozygosity and ultimate formation of mutant CCM genes. ¹⁰ Despite continued research, ¹¹ the only available treatment options for CCM are microsurgical resection and stereotactic radiosurgery ¹² ; thus, a large unmet need remains for pharmacologic treatments to ameliorate CCM symptoms. ¹¹

For over a decade, CCM research has focused on molecular mechanisms underlying the profound endothelial permeability and instability; however, no new therapeutics translated into the clinic. ¹¹ Ultimately, to bring new therapeutics to patients with CCM, an unbiased, machine learning–aided phenotypic drug discovery platform was developed to explore known, repurposed drugs to identify effective therapeutics for the disease. ¹³ By using an automated immunofluorescence and machine learning–based primary screen of structural phenotypes of CCM2‐deficient human endothelial cells, followed by a secondary screen of functional changes in endothelial monolayer permeability, a tertiary screen of dermal microvascular leak in mice with endothelial‐specific loss‐of‐function of CCM2, and a quaternary screen of CCM lesion burden in the same mouse model (Table 1), we successfully narrowed a field of 2100 repurposed drug candidates. ¹³ Through this platform, we then utilized animal models that recapitulate genotypes and phenotypes most commonly found in human CCM and adapted clinical imaging technology for these mouse models. Subsequently, we evaluated CCM lesions in a manner similar to analyses done in clinical settings to demonstrate that REC‐994 (a superoxide scavenger, also known as tempol) successfully reduced CCM lesions in mouse models of human CCM disease. ¹³ REC‐994 (4‐hydroxy‐2,2,6,6‐tetramethylpiperidine‐1‐oxyl) has been studied extensively preclinically for its beneficial effects on endothelial function and blood pressure, ¹⁴ , ¹⁵ , ¹⁶ traumatic brain injury, ¹⁷ and models of kidney and heart disease, ¹⁸ and as protection against acetaminophen‐induced liver injury. ¹⁹ However, to date, REC‐994 has been studied therapeutically in humans only for COVID‐19 (NCT04729595; trial terminated), to reduce severe treatment‐related mucositis in head/neck cancer patients (NCT03480971), and as a topical treatment to mitigate cancer radiation effects in the skin. ²⁰ , ²¹

TABLE 1.

REC‐994 targeted screening.

Screen level	Screen type	Model	CCM phenotype in model	Effect of REC‐994 in model
1°	Phenotypic In vitro	siRNA knockdown of CCM2 in human dermal microvascular endothelial cells → Quantify protein and profile cells with fluorescent microscopy	↓ Quantity/LoF of CCM protein	Reversal of LoF of CCM protein in terms of complex phenotypic signature in primary human endothelial cells
2°	Functional In vitro	CCM2‐deficient cells deficient in monolayer stability → Measure transendothelial resistance	↓ Transendothelial resistance	↑ Transendothelial resistance
3°	Functional Acute in vivo	CCM2‐deficient knockout mice → Measure leakiness of dermal vasculature	Microvascular leak	↓ Peri‐injection microvascular leakiness
4°	Disease Chronic in vivo	CCM2‐deficient knockout mice → Measure lesion burden with small‐animal MRI	MRI‐detectable cerebrovascular lesions	↓ Number and size of MRI‐detectable cerebrovascular lesions

Abbreviations: CCM, cerebral cavernous malformation; LoF, loss of function; MRI, magnetic resonance imaging; siRNA, small interfering RNA.

REC‐994 is a redox‐cycling nitroxide compound and free‐radical scavenger that reduces superoxide in an animal species–independent manner. Hypothesized to treat CCMs through a reduction in ROS, REC‐994 offers the potential to reduce debilitating CCM lesions through a safe molecular target. Herein, we present data on REC‐994 in a series of double‐blind, randomized, placebo‐controlled single‐ and multiple‐ascending dose studies to investigate the safety, tolerability, and pharmacokinetic (PK) profile of REC‐994 in healthy volunteers. Based on the results, REC‐994 demonstrates low potential for adverse side effects mediated through off‐target effects or negative pharmacodynamic effects in cardiovascular, respiratory, and central nervous systems.

2. METHODS

2.1. Study design

Single‐ascending dose (SAD) and multiple‐ascending dose (MAD) studies were conducted by the study physician at a single center in the United States (PRA Health Sciences, Lenexa, KS). Because the FDA did not require registration of phase 1 studies in healthy volunteers, these studies were not registered. Except MAD study Cohort 1, these were double‐blind, randomized studies, with randomization schedule produced by the PRA Health Sciences statistician prior to study start using a computer‐generated randomization scheme.

The SAD study evaluated PK and safety of REC‐994 and consisted of 4 cohorts of 8 participants, each receiving a single oral dose of 50, 100, 200, or 400 mg REC‐994 or matching placebo. Participants were screened between 28 and 3 days before REC‐994 administration (Days −28 to −3), admitted to the clinic on Day −2, and discharged on Day 4; a follow‐up visit was conducted between Days 18 and 25 (Figure 1a). Each cohort included 2 sentinel participants (1:1 placebo:REC‐994). The remaining 6 participants (1:5 placebo:REC‐994) in each cohort were dosed after at least 24 h had elapsed following dosing of the sentinel participants if safety and tolerability results were acceptable.

FIGURE 1.

Study designs for (A) SAD study and (B) MAD study Cohorts 2–7, and (C) formulation and food effect studies (MAD Cohort 1). *Cohort 4 of the MAD study was optional and not performed. MAD, multiple‐ascending dose; P, placebo; PK, pharmacokinetic; REC, REC‐994; SAD, single‐ascending dose.

In the MAD study, participants were screened between Days −28 and −2 and admitted to the clinic on Day −1. PK, safety, and tolerability of MADs of REC‐994 were investigated in Cohorts 2, 3, 5, and 7 (n = 8/cohort), in which participants were randomized 3:1 to multiple‐ascending oral doses of REC‐994 tablets at doses of 50, 200, 400, and 800 mg, respectively, or matching placebo, once daily (QD) for 10 days (Figure 1b). Cohorts 2–7 included sentinel participants (1 placebo, 1 active); remaining participants in each cohort were dosed after 24 h, given acceptable safety and tolerability results. Since the 800‐mg dose was not included in the SAD study, a single dose of 800‐mg REC‐944 tablets was evaluated in Cohort 6 prior to proceeding to Cohort 7 (800‐mg multiple‐dose cohort). Cohort 4 was an optional 800‐mg single oral dose cohort that was not conducted so that multiple doses of 400 mg (Cohort 5) could be evaluated first; Cohort 6 was conducted instead. Participants remained in the clinic until Day 14 (15 days) for Cohorts 2, 3, 5, and 7 and until Day 3 (4 days) for Cohort 6. Cohort 1 employed a single‐dose, randomized, open‐label, 3‐period crossover, formulation bridging, and food‐effect design (n = 12; Figure 1c). Participants were randomized to 1 of 3 treatment sequences (ABC, BCA, or CAB), during which they received a single oral dose of each treatment: 100‐mg REC‐994 solution in fasted state (treatment A), 100‐mg REC‐994 tablets in fasted state (treatment B), and 100‐mg REC‐994 tablets in fed state (treatment C). Treatments were separated by a minimum 72‐h washout period. Participants in Cohort 1 remained in the clinic until Day 9 (10 days). Participants in all cohorts returned 14–21 days after discharge for a follow‐up visit.

2.2. Participants

For both studies, healthy males and females ages 18–55 years with a body mass index (BMI) of 18–32 kg/m² and a minimum body weight of 50 kg were eligible to participate. All participants must have tested negative for hepatitis B and C and human immunodeficiency virus antibodies at screening and presented with laboratory test results that were within reference range and considered not clinically significant by the investigator. Negative screens for alcohol and drugs of abuse were also required, and participants must have refrained from taking all medications (including over‐the‐counter drugs) and herbal supplements for the duration of the study period. Exclusion criteria included a history of gastrointestinal, renal, hepatic, hematological, lymphatic, neurological, cardiovascular, psychiatric, musculoskeletal, genitourinary, immunological, dermatological, endocrine, or connective tissue disease or disorder. Participants with a medical history of chronic respiratory disorders were also excluded, as were those with hyperemesis, drug hypersensitivities, and a history of alcoholism or drug abuse.

2.3. Objectives

The SAD study primary objective was to assess safety and tolerability of REC‐994 after single oral doses. The secondary objective was to characterize the PK profile of REC‐994 after single oral doses.

The MAD study primary objectives were to assess safety and tolerability and characterize PK of multiple oral doses of REC‐994. Secondary objectives were to identify recommended multiple oral dose level(s) of REC‐994 using a tablet formulation and compare PK of a single oral dose of 100‐mg REC‐994 administered as a solution or tablets. The exploratory objective was to evaluate the effect of food on PK of a single dose of 100‐mg REC‐994 tablets.

2.4. PK and bioanalysis

Blood samples of 2 mL each were collected and plasma harvested for PK analysis of REC‐994 in plasma. Blood samples were taken via an indwelling intravenous catheter or by direct venipuncture into K₂‐ethylenediaminetetraacetic acid–containing tubes. REC‐994 in plasma was quantified by high‐performance liquid chromatography (HPLC) with tandem mass spectrometry. Sample processing was performed by protein precipitation using a sample volume of 50.0 μL. Separation between potential metabolites and interfering endogenous compounds was achieved by HPLC using an Acquity UPLC HSS T3 column (2.1 × 50 mm, 1.8 μm particles) at 35°C and isocratic elution with post‐elution wash‐off using 0.1 HCOOH in water as mobile phase A and 60:40:0.1 CAN:MeOH:HCOOH v/v/v as mobile phase B operating at a flow rate of 0.500 mL/min. A triple quadrupole mass spectrometer (API Triple Quad 5500) equipped with a Turbo‐ion spray source was used for detection in positive ion mode. The method was fully validated by assessment of precision, accuracy, sensitivity, and specificity of REC‐994 by the PRA North American Bioanalytical Laboratory. A total of 312 plasma samples from the SAD study and 1435 samples from the MAD study were successfully analyzed for REC‐994 by HPLC with tandem mass spectrometric detection and were deemed reliable.

Plasma PK parameters for the SAD study included maximum observed plasma concentration (C _max), time to attain C _max (T _max), area under the plasma concentration versus time curve (AUC) from time 0 to time of last quantifiable concentration (AUC_0–t and/or AUC_last), AUC from time 0 extrapolated to infinity (AUC_inf), terminal elimination half‐life (t _1/2), apparent oral clearance (CL/F), and apparent volume of distribution (V_z/F).

Plasma PK parameters estimated on Day 1 of the MAD study included C _max, T _max, AUC from time 0 to 24 h post‐dose (AUC_0–24), and trough plasma concentration observed at the end of dosing interval (C₂₄). Steady‐state plasma PK parameters estimated on Day 10 of the MAD study included C _max, T _max, AUC_0–24, C₂₄, t _1/2, CL/F, V_z/F, accumulation ratio (RA) for AUC_0–24, and RA for C _max.

Food effect was evaluated following a standardized breakfast consisting of a high‐fat (approximately 50% of total caloric content of meal) and high‐calorie (approximately 800–1000 calories) meal, which was to be finished within 20 min. Dosing occurred at 30 min after start of breakfast.

2.5. Data and statistical analyses

Sample size for the SAD study was based on empirical considerations. A total of 32 healthy participants enrolled into 4 cohorts were planned, and no formal statistical hypothesis testing or formal sample size calculation was performed. Approximately 52 healthy participants enrolled in up to 6 cohorts were planned for the MAD study. No prospective calculations of statistical power were made. Sample size was selected to provide information on safety, tolerability, and PK following single and multiple doses of REC‐994 and to evaluate the effect of food and formulations on REC‐994.

In both studies, the safety analysis set was defined as all participants who received at least 1 dose of REC‐994. The PK analysis set consisted of all participants in the safety set who provided sufficient PK data to calculate reliable estimates of parameters with no protocol deviation affecting PK evaluation. Descriptive statistics were used to summarize plasma concentrations by treatment at each scheduled timepoint. Plasma PK parameters for REC‐994 were estimated using noncompartmental methods with WinNonlin^® version 8.1. In the MAD study, parameters were estimated for each treatment period in the food‐effect cohort, for Day 1 and Day 10 in the MAD cohorts, and for Day 1 in the cohort that received a single dose of 800‐mg REC‐994. Plasma PK parameters were estimated from the concentration versus time profiles, and AUC values were calculated using a linear up/log down approach.

Effects of food and formulation on the PK of REC‐994 were assessed using the ratio and 90% confidence intervals (CIs) of the geometric least‐squares (LS) means of plasma PK parameters: C _max, AUC_0–t , and AUC_inf for REC‐994. A linear mixed‐effects model with fixed effects for treatment, period, and sequence and a random effect for participant nested in sequence was performed using natural log‐transformed parameters. Geometric LS means were determined for each treatment, and treatment B was used as the reference. For T _max, a Wilcoxon signed‐rank test was performed to compare treatments, with medians of the differences and p‐values provided for the comparisons (α = 0.1 for significance). Dose proportionality was evaluated using C _max and AUC with a power model as follows: ln(parameter) = α + β·ln(Dose) + ε, where parameter is the PK parameter value, α is the y‐intercept, β is the slope, and ε is the error term. A 90% CI that included 1 indicated dose proportionality. Evaluation of safety and tolerability occurred through a summary of adverse events (AEs), vital signs, clinical laboratory evaluations, electrocardiograms (ECGs), pulse oximetry, and physical and neurological examinations.

3. RESULTS

3.1. Participants

The SAD study, conducted between September 25, 2018, and February 4, 2019, enrolled 32 participants (16 male; 16 female) aged 18–54 years; 24 participants received REC‐994 and were included in the PK analysis set, and 8 received placebo. One participant withdrew from the study on Day 3 (family emergency) after receiving the 200‐mg dose of REC‐994; the remaining 31 (96.9%) completed the study. The MAD study, conducted between July 15, 2019, and January 14, 2020, enrolled 52 participants (27 female; 25 male) aged 18–53 years; 42 received REC‐994 and were included in the PK analysis set, and 8 received placebo. One participant in Cohort 2 discontinued REC‐994 after administration of 50‐mg REC‐994 on Day 8 and withdrew from the study on Day 16 (participant choice); the remaining 51 (98%) completed the study. Proportions of White and Black or African American participants were approximately equal across ascending‐dose cohorts in both studies, with 42% of Cohort 1 and 75% of Cohort 6 being White and the rest being Black or African American. The majority were not Hispanic or Latino. BMIs of all participants ranged from 19.4 to 32.0 kg/m² (Tables S1 and S2).

3.2. PK and bioanalysis

Mean REC‐994 (±standard deviation [SD]) plasma concentrations following single doses of REC‐994 at 50, 100, 200, and 400 mg were above the lower limit of quantitation (LLOQ; 5.0 ng/mL) in 23 of 24 participants through 24 h post‐dose (Figure 2A). Plasma concentrations increased with increasing doses, with similarly shaped profiles declining monoexponentially following C _max. REC‐994 exposures (C _max, AUC_last, and AUC_inf) generally increased in a dose‐proportional manner over the dose range of 50–400 mg (Table S3). Median T _max was short, ranging from 0.5 to 0.75 h, and geometric mean t _1/2 was approximately 6 h; both were independent of dose (Table 2).

FIGURE 2.

Mean plasma REC‐994 concentration versus time profiles by dose. (A) REC‐994 plasma concentrations at 50, 100, 200, and 400 mg in the SAD study. (B, C) Mean plasma concentration versus time profiles of REC‐994 on Day 1 and Day 10, respectively, of the MAD study. (D) Plasma concentration profiles following single‐dose administration of 100 mg REC‐994 in solution and tablet formulation and tablet in fed or fasted state. MAD, multiple‐ascending dose; SAD, single‐ascending dose.

TABLE 2.

SAD study PK parameters.

	SAD study
	50 mg (n = 6)	100 mg (n = 6)	200 mg (n = 6)	400 mg (n = 6)
C max (ng/mL)
Arithmetic mean (CV%)	114.3 (24.8)	290.8 (20.1)	659.7 (38.5)	918.7 (26.5)
Geometric mean	111.1	285.7	618.6	884.8
T max (h), median (range)	0.75 (0.33–1.10)	0.75 (0.75–1.00)	0.75 (0.25–2.00)	0.50 (0.25–6.00)
AUClast (h*ng/mL)
Arithmetic mean (CV%)	846.0 (29.8)	2212.2 (22.6)	4385.5 (19.5)	6139.0 (9.7)
Geometric mean	807.7	2163.3	4319.1	6116.0
AUCinf (h*ng/mL)
Arithmetic mean (CV%)	950.5 (29.6)	2400.2 (22.8)	4781.0 (18.9)	6751.5 (8.1)
Geometric mean	909.4	2348.1	4712.8	6733.5
t 1/2 (h)
Arithmetic mean (CV%)	6.337 (19.3)	6.163 (13.4)	6.428 (4.2)	6.592 (15.9)
Geometric mean	6.235	6.114	6.424	6.528
CL/F (L/h)
Arithmetic mean (CV%)	57.92 (37.8)	43.58 (23.9)	43.03 (17.7)	59.58 (7.8)
Geometric mean	54.98	42.61	42.45	59.43
Vz/F (L)
Arithmetic mean (CV%)	503.3 (20.2)	388.7 (30.9)	400.2 (19.9)	566.7 (17.4)
Geometric mean	494.8	375.7	393.3	559.5

Abbreviations: AUC, area under the curve; AUC_inf, AUC from time 0 extrapolated to infinity; AUC_last, AUC from time 0 until the time of last quantifiable concentration; CL/F, apparent oral clearance; C _max, maximum concentration; CV, coefficient of variation; h, hours; PK, pharmacokinetic; SAD, single‐ascending dose; t _1/2, half‐life; T _max, time to peak drug concentration; V_z/F, apparent volume of distribution.

As in the SAD study, plasma concentrations of REC‐994 increased with increasing dose following multiple doses of REC‐994 at 50, 200, 400, and 800 mg, with similarly shaped profiles (Figure 2B,C). Increases seen at 8 h post‐dose at 400 and 800 mg on Day 10 were due to anomalously high concentrations from 1 participant in each cohort. Steady state appeared to be reached by Day 3 of administration at all dose levels, based on the similarity in mean trough plasma concentrations on Days 2–9 (Figure S1). REC‐994 post‐dose plasma concentrations were above LLOQ (5.0 ng/mL) in all participants. Day 10 systemic exposures (C _max and AUC_0–24) increased in proportion to REC‐994 dose over the 50–800‐mg dose range (Table 3; Table S3); however, Day 1 exposure increases appeared slightly less than dose‐proportional. Median T _max appeared independent of dose and ranged from 0.88 to 1.53 h following administration of the first dose on Day 1 and 0.75–2 h following repeated administration on Day 10. Similarly, geometric mean t _1/2 was similar across doses following repeated administration (7.3–8.5 h). No accumulation of REC‐994 was observed after multiple doses, and both C _max and AUC_0–24 increased with increasing dose; however, exposures decreased by 26%–42% between Day 1 and Day 10 in the 800‐ and 50‐mg dose groups, respectively (Table 3).

TABLE 3.

MAD study PK parameters.

	MAD study
	Cohort 2: 50 mg (n = 6)	Cohort 3: 200 mg (n = 6)	Cohort 5: 400 mg (n = 6)	Cohort 7: 800 mg (n = 6)	Cohort 6 a : 800 mg (n = 6)
Day 1
C max (ng/mL)
Arithmetic mean (CV%)	222.7 (38.7)	626.7 (16.8)	1111.3 (43.7)	2686.7 (35.4)	1775 (32)
Geometric mean	207.4	618.2	1017.6	2527.8	1697.9
T max (h) b , median (range)	0.88 (0.50–2.00)	1.00 (1.00–2.00)	1.53 (0.75–4.00)	1.50 (0.75–4.00)	1.03 (0.75–4.00)
AUC0–24 (h*ng/mL)
Arithmetic mean (CV%)	1861.0 (34.2)	4939.0 (19.7)	8596.6 (37.3)	23 789.1 (32.3)	NA
Geometric mean	1763.1	4850.9	8143.4	22 565.6
C24 (ng/mL)
Arithmetic mean (CV%)	23.7 (44.9)	57.5 (39.1)	110.9 (69.6)	284.3 (39.8)	138 (36)
Geometric mean	21.63	53.35	95.93	265.0	130.0
Day 10
C max (ng/mL)
Arithmetic mean (CV%)	128.2 (16.3)	699.2 (24.9)	1138.0 (41.4)	1979.5 (55.2)	NA
Geometric mean	126.9	681.7	1062.5	1782.9
T max (h) b	0.75 (0.50–1.00)	1.50 (0.50–2.07)	2.00 (0.75–8.00)	1.50 (1.00–8.00)	NA
AUC0–24 (h*ng/mL)
Arithmetic mean (CV%)	1092.0 (15.2)	5038.4 (22.1)	9648.7 (26.4)	17 541.6 (47.7)	NA
Geometric mean	1081.0	4924.7	9375.3	16 200.4
C24 (ng/mL)
Arithmetic mean (CV%)	13.7 (26.3)	54.5 (38.2)	107.1 (33.0)	195.7 (62.1)	NA
Geometric mean	13.22	50.52	102.3	175.0
t 1/2 (h)
Arithmetic mean (CV%)	7.3 (17.6)	7.7 (15.4)	8.5 (10.7)	7.7 (13.1)	NA
Geometric mean	7.177	7.641	8.500	7.659
CL/F (L/h)
Arithmetic mean (CV%)	46.8 (17.3)	41.7 (26.3)	43.9 (26.7)	52.8 (37.0)	NA
Geometric mean	46.25	40.61	42.66	49.39
Vz/F (L)
Arithmetic mean (CV%)	483.3 (15.0)	450.6 (12.4)	538.2 (27.9)	583.6 (41.5)	NA
Geometric mean	479.0	447.9	523.2	545.8
Rac, C max
Arithmetic mean (CV%)	0.7 (41.9)	1.1 (21.6)	1.1 (37.0)	0.8 (46.8)	NA
Geometric mean	0.626	1.103	1.044	0.705
Rac, AUC0–24
Arithmetic mean (CV%)	0.7 (31.6)	1.0 (11.1)	1.2 (24.2)	0.7 (24.2)	NA
Geometric mean	0.630	1.015	1.152	0.718

Abbreviations: AUC, area under the curve; AUC_0–24, AUC from time 0 to 24 h post‐dose; C₂₄, trough plasma concentration observed at the end of the dosing interval; CL/F, apparent oral clearance; C _max, maximum concentration; CV, coefficient of variation; MAD, multiple‐ascending dose; PK, pharmacokinetic; Rac, drug accumulation ratio; t _1/2, half‐life; T _max, time to peak drug concentration; V_z/F, apparent volume of distribution.

^a Cohort 6 was a single‐dose cohort.

^b For T _max, the median (range) is presented instead of geometric mean (CV%).

3.3. Effects of formulation and food

Mean (±SD) plasma concentration versus time profiles of REC‐994 appeared similar following single‐dose, fasted state administration of 100‐mg solution (treatment A) and 100‐mg tablet (treatment B) (Figure 2D). T _max occurred 0.5 h earlier, C _max increased by 6.1%, and AUC_0–t and AUC_inf remained similar between solution and tablet formulations (Tables S4 and S5). No statistical differences for solution and tablet formulations were observed.

Following administration of tablet formulation with the standardized breakfast (fed state; treatment C), REC‐994 absorption appeared delayed, with a decrease in peak concentration compared to administration in fasted state. Median T _max was 1 h in the fasted state and 3 h in the fed state, also suggesting a delay in absorption with food (Table S4). Following a single dose of 100‐mg REC‐994 tablet under fed conditions, T _max with REC‐994 was delayed 1.25 h compared to fasted conditions, and C _max and AUC_0–t decreased by 28.1% and 10.1%, respectively; AUC_inf remained similar (Table S5).

3.4. Safety

A total of 4 treatment‐emergent AEs (TEAEs) were reported in 4 (12.5%) participants following administration of 50‐mg (n = 3; palpitations, headache, dizziness) and 200‐mg (n = 1; skin abrasion) REC‐994 in the SAD study (Table S6). One sentinel participant also reported a TEAE of headache following administration of 50‐mg REC‐994 that was considered potentially treatment‐related; all other TEAEs were considered unrelated to treatment. Following a single dose of 800‐mg REC‐994, 3 mild TEAEs were reported in 2 (25.0%) participants (Table S7). One participant reported TEAEs of contusion on the left leg and headache following a single administration of 800‐mg REC‐994 tablet, both of which were considered unrelated to REC‐994. The other participant reported a TEAE of headache following administration of placebo that was deemed possibly related to treatment. All TEAEs were mild, transient, and resolved by end of study.

In the MAD cohorts, 34 TEAEs were recorded in 14 (43.8%) participants (Table 4). Two participants reported moderate dizziness and headache, respectively, both of which were considered unrelated to REC‐994; all other reported TEAEs were mild. Three (9.4%) participants reported TEAEs considered likely related to REC‐994, 2 of which occurred following administration of multiple doses of 200‐mg REC‐994 (early satiety and headache), and the other (nausea) following administration of multiple doses of 800‐mg REC‐994. The most frequently reported (≥2) TEAEs were headache (15.6%), dizziness (12.5%), nausea (12.5%), abdominal pain (9.4%), back pain (9.4%), and hematochezia (6.3%). No AEs of special interest or TEAEs leading to discontinuation were reported during the study in any cohort, and no dose‐related trends in TEAEs, vital signs, ECGs, pulse oximetry, physical examination findings, or neurological examination findings were observed.

TABLE 4.

TEAEs in the MAD study.

Cohorts 2, 3, 5, and 7
Preferred term, n (%)	Placebo (n = 8)	REC‐994				Overall (n = 32)
		50 mg (n = 6)	200 mg (n = 6)	400 mg (n = 6)	800 mg (n = 6)
Headache	1 (12.5)	0	2 (33.3)	0	2 (33.3)	5 (15.6)
Dizziness	1 (12.5)	0	0	1 (16.7)	2 (33.3)	4 (12.5)
Nausea	0	0	1 (16.7)	0	3 (50.0)	4 (12.5)
Abdominal pain	1 (12.5)	0	1 (16.7)	0	1 (16.7)	3 (9.4)
Back pain	0	0	1 (16.7)	0	2 (33.3)	3 (9.4)
Hematochezia	1 (12.5)	0	1 (16.7)	0	0	2 (6.3)
Abnormal dreams	0	0	1 (16.7)	0	0	1 (3.1)
Anxiety	0	0	0	0	1 (16.7)	1 (3.1)
Constipation	0	0	0	1 (16.7)	0	1 (3.1)
Dermatitis atopic	0	0	0	1 (16.7)	0	1 (3.1)
Dry throat	0	0	0	0	1 (16.7)	1 (3.1)
Early satiety	0	0	1 (16.7)	0	0	1 (3.1)
Musculoskeletal chest pain	0	0	1 (16.7)	0	0	1 (3.1)
Orthostatic hypotension	0	0	1 (16.7)	0	0	1 (3.1)
Swelling of eyelid	0	0	0	0	1 (16.7)	1 (3.1)
Toothache	0	0	0	1 (16.7)	0	1 (3.1)
Vaginal infection	1 (12.5)	0	0	0	0	1 (3.1)
Vision blurred	0	0	0	0	1 (16.7)	1 (3.1)

Abbreviations: MAD, multiple‐ascending dose; TEAE, treatment‐emergent adverse event.

In the food‐effect portion of the study, 2 mild TEAEs of headache were reported in 2 (16.7%) participants (1 instance considered unlikely; 1 instance considered likely related to treatment), both following single administration of 100‐mg REC‐994 tablet under fed conditions.

4. DISCUSSION

4.1. Study overview

This study was conducted to evaluate the safety and PK of single‐ and multiple‐ascending oral doses of REC‐994 and the effect of food and formulation on REC‐994 PK. Overall, REC‐994 was well tolerated in healthy male and female participants. All TEAEs across both studies were transient and had resolved by end of study; most were mild, with only two that were considered moderate and none that were severe or led to treatment discontinuation. The most common TEAEs included headache and dizziness. There were no clinically meaningful findings or dose‐related trends in ECG, pulse oximetry, or neurological examinations and no meaningful hepatobiliary events or potential drug‐induced liver injuries.

Systemic exposure increased in proportion to REC‐994 dose after single doses of 50–800 mg (SAD and MAD Cohort 6) and after 10 days of dosing over the 16‐fold dose range of 50–800 mg QD (MAD). Median T _max and mean t _1/2 appeared to be independent of dose in both studies; solution formulation was more rapidly absorbed, while overall exposure was similar to that of tablet formulation. Taking REC‐994 with a high‐fat, high‐calorie meal delayed T _max and reduced C _max but had minimal effect on AUC compared to fasted conditions.

Increased levels of ROS are responsible for the pathogenesis of many human diseases, including cancer, neurodegenerative diseases, ²² and CCM. ⁷ , ⁸ REC‐994, a free‐radical scavenger, may treat CCM by reducing ROS. Capable of correcting a high‐dimensional morphological phenotype in CCM2‐deficient human dermal microvascular endothelial cells, REC‐994 ameliorates the endothelial hyperpermeability defect that is characteristic of CCM ¹³ , ¹⁸ , ²³ by decreasing ROS and oxidative stress, thereby improving endothelial function (Figure 3). Several drugs that are also reported to act by reducing ROS (some of which have chemical structures similar to that of REC‐994) have been tested and some approved for clinical use. Edaravone (Radicut^®), a free‐radical scavenger approved in Japan since 2001, has been used to treat acute ischemic stroke with an acceptable safety profile, ²⁴ demonstrating the utility, safety, and efficacy of reducing ROS therapeutically. More recently, edaravone has been approved in both Japan and the United States (Radicava^®, 2017) for the treatment of patients with amyotrophic lateral sclerosis, ²⁵ another disease with underlying pathophysiology mediated by ROS. NXY‐059, also a free‐radical scavenger, has been evaluated in two large phase 3 trials for the treatment of acute ischemic stroke. ²⁶ , ²⁷ REC‐994 is being studied in clinical trials for diseases with ROS‐related pathophysiology, including radiation‐induced mucositis, nephrotoxicity, and ototoxicity (NCT03480971); prostate cancer (NCT04337099, NCT04876755); and ataxia telangiectasia (NCT04887311).

FIGURE 3.

Cavernous angioma pathophysiology and mechanism of REC‐994 action in CCM. ⁷ , ⁹ , ¹³ , ²⁸ , ²⁹ , ³⁰ CCM, cerebral cavernous malformation; eNOS, endothelial nitric oxide synthase; FOX01, forkhead box O1; H₂O₂, hydrogen peroxide; KRIT1, Krev1 interaction trapped gene 1; O₂, oxygen; ONOO, peroxynitrite; ROS, reactive oxygen species; SOD, superoxide dismutase.

Reduction of ROS by REC‐994 is a direct chemical effect and therefore not anticipated to display a species dependence in reactivity. Although REC‐994 was not designed to bind to a specific molecular target, off‐target effects were ruled out using an in vitro pharmacological screen to identify potential secondary pharmacodynamic targets. In this SafetyScreen44™ panel, which includes 44 different enzymes, transporters, G‐protein‐coupled receptors, ion channels, and nuclear receptors, ³¹ REC‐994 did not inhibit any of these molecular targets by more than 40%. Notably, the clinical safety profile of REC‐994 in the current study and other compounds in clinical use suggests that reduction of ROS is not associated with target‐ or mechanism‐mediated toxicities.

4.2. REC‐994 and symptomatic CCM

The potential of REC‐994 as a treatment for symptomatic CCM has been assessed in vitro and in vivo. Increased vascular cell permeability in the brain or spinal cord is one hallmark of CCM disease, and an in vitro model of increased cell permeability performed in CCM2‐deficient cells (human dermal microvascular endothelial cells transfected with either a CCM2 small interfering ribonucleic acid [siRNA] or a scrambled control) demonstrated that REC‐994 reduced cellular permeability as measured by electrical resistance. ¹³ In an acute in vivo model evaluating dermal vascular permeability of a dye in CCM2‐siRNA knockout mice, REC‐994 reduced the permeability of vascular endothelial cells by 50%. Furthermore, a chronic animal model for CCM disease using CCM2 knockout mice demonstrated that REC‐994 in drinking water for 5 months led to a significant decrease in brain lesions as measured by magnetic resonance imaging (MRI), suggesting potential efficacy of REC‐994 in patients with symptomatic CCM. ¹³ However, the effects of REC‐994 on established, symptomatic CCM lesions have not yet been evaluated preclinically or clinically.

Preclinical efficacy data and safety data from SAD and MAD studies presented herein support conducting a phase 2 clinical trial in patients with symptomatic CCM. Therefore, the SYCAMORE trial, a multicenter, randomized, double‐blind, placebo‐controlled study, has initiated to investigate the safety, efficacy, and PK of REC‐994 compared to placebo in participants with symptomatic CCM (NCT05085561; recruitment is ongoing). Up to 60 enrolled participants will undergo screening and randomization, treatment during which participants will receive 200‐ or 400‐mg REC‐994 or placebo, and (potential) extension periods. Adults with anatomic CCM lesions demonstrated by brain MRI and symptomatic CCM are eligible. Expected outcomes include a better understanding of safety and tolerability of REC‐994 in patients with CCM and of interrelationships between treatment effect assessments and identification of assessments likely to predict treatment benefit in future studies.

AUTHOR CONTRIBUTIONS

Ron Alfa: Conceptualization, data curation, formal analysis, methodology, investigation, and supervision. Timothy Considine: Conceptualization, data curation, formal analysis, investigation, methodology, validation, writing—original draft, and writing—review & editing. Shafique Virani: Conceptualization, data curation, and methodology. Matt Pfeiffer: Conceptualization, methodology, writing—original draft, and writing—review & editing. Anthony Donato: Conceptualization, methodology, writing—original draft, and writing—review & editing. Daniel Dickerson: Conceptualization, data curation, formal analysis, investigation, methodology, supervision, validation, and resources. Diana Shuster: Formal analysis, validation, writing—original draft, and writing—review & editing. Joel Ellis: Conceptualization, methodology, writing—original draft, and writing—review & editing. Kristen Rushton: Data curation. Helen Wei: Formal analysis, validation, and writing—review & editing. Christopher Gibson: Identified REC‐994 molecule as potential treatment for CCM, conceptualization, data curation, methodology, visualization, and resources.

CONFLICT OF INTEREST STATEMENT

R. Alfa reports employment at Noetik Inc. and previous employment at Recursion Pharmaceuticals, Inc. T. Considine reports current employment at Model Medicines, previous employment at Recursion Pharmaceuticals, Inc., and consultancies for Forcyte Biotechnologies, Waypoint Bio, Ardigen, and Delcath Systems, Inc. S. Virani reports employment at Recursion Pharmaceuticals, Inc., and is a Non‐Executive Director at TauC3 Biologics. M. Pfeiffer reports employment at and holds stock in Recursion Pharmaceuticals, Inc. A. Donato reports serving as a Scientific Advisor for and stock options in Recursion Pharmaceuticals, Inc. and research funding from NIH grants for R01AG060395‐03, VA Merit CX002211, R21AG070740, R21/R33AG074498, and R01AG07775. D. Dickerson reports employment at ICON plc (Legacy PRA Health Sciences). J. Ellis and D. Shuster report former employment at Recursion Pharmaceuticals, Inc. K. Rushton and H. Wei report employment at Recursion Pharmaceuticals, Inc. C. Gibson is CEO and Cofounder of and holds stock in Recursion Pharmaceuticals, Inc.

Disclosure

Research and writing support were funded by Recursion Pharmaceuticals, Inc.

ETHICS STATEMENT

These studies were conducted in accordance with the Declaration of Helsinki and guidelines for Good Clinical Practice. Study protocols and amendments were approved by Midlands Independent Review Board (IORG0001486) prior to eligibility screening. All participants provided written informed consent.

Supporting information

Data S1.

Alfa R, Considine T, Virani S, et al. Clinical pharmacology and tolerability of REC‐994, a redox‐cycling nitroxide compound, in randomized phase 1 dose‐finding studies. Pharmacol Res Perspect. 2024;12:e1200. doi: 10.1002/prp2.1200

DATA AVAILABILITY STATEMENT

Data requests should be sent to the corresponding author. Recursion Pharmaceuticals, Inc., shares anonymized data upon request or as required by law or regulation. A qualified researcher must provide a curriculum vitae and demonstrated non‐conflict of interest. A statistician is also required for the approval. Approval of such requests is at Recursion's discretion and depends on the availability and intended use of the data, the nature of the request, and the merit of the research proposed.