Phase 1 Clinical Results for NP10679, a pH‐sensitive GluN2B‐selective N‐methyl‐d‐aspartate Receptor Inhibitor

NP10679 is a context‐dependent and subunit‐selective negative allosteric modulator of N‐methyl‐d‐aspartate (NMDA) receptors. It is a more potent inhibitor of GluN2B‐containing NMDA receptors at the acidic levels of extracellular pH (eg, 6.9) found in the penumbral regions associated with cerebral ischemia than at physiological pH. This property allows NP10679 to act selectively in ischemic tissue while minimizing the nonselective blockade of NMDA receptors in healthy brain, thereby reducing on‐target adverse effects. We report the results of a first‐in‐human pharmacokinetic and safety phase 1 clinical trial in healthy volunteers receiving single or multiple doses of NP10679 (NCT04007263). We found that NP10679 was well‐tolerated and with a half‐life of 20 hours, which is amenable to once per day dosing. The only notable side effect in this clinical trial was modest somnolence at higher doses, atypical in that the subject could easily be aroused. The overall results suggest that NP10679 is a candidate for further development for use in acute brain injury, such as ischemic stroke or aneurysmal subarachnoid hemorrhage, as well as for use in neuropsychiatric indications.

Cerebral ischemia is a central component of acute brain injuries such as stroke, traumatic brain injury, and subarachnoid hemorrhage (SAH) and may lead to significant subsequent neurological and cognitive deficits. Downstream mechanisms that are proposed to contribute to secondary tissue injury after ischemiarelated neuronal death include neuroinflammation, microglial activation, oxidative stress, 1 and excess activation of receptors recognizing glutamic acid due to the accumulation of this amino acid neurotransmitter in the extracellular space. 2,3 Previous neuroprotectant compounds have demonstrated promise but failed to achieve a tolerable safety profile or demonstrate significant efficacy in clinical investigations of stroke, spontaneous brain hemorrhage, or traumatic brain injury (TBI). 4,5 Proposed reasons for the clinical failures include delay in initiating treatment after insult, 6,7 patient heterogeneity, complex endpoints, dose-limiting on-target effects, and a failure to reach efficacious dose levels at the site of injury. 5,[8][9][10][11] Considering the potential reasons for past failures to identify an effective pharmacotherapy for ischemic brain injury, two elements appear necessary to provide the best chance to observe a meaningful positive signal in ameliorating the clinical manifestations of ischemic injury: a clinical plan that addresses pitfalls suggested from previous trials and a safe and effective neuroprotectant to implement that plan.
A major contributor to negative findings reported in stroke and TBI trials is the delay in time to treatment. 12 This latency to effective treatment would be reduced in clinical trials testing neuroprotectants in an indication such as aneurysmal SAH (aSAH), which is associated with the delayed onset of cerebral ischemia. Thus, this indication lends itself to early and prophylactic administration of a neuroprotectant. Referred to as delayed cerebral ischemia, the ischemia in aSAH appears to be driven by vascular events that occur 3-14 days after an aneurysmal hemorrhage. Dosing near the time of surgical repair of aneurysm would allow a neuroprotectant to be present in the brain prior to and essentially serve as prophylactic to a post-SAH ischemia. Furthermore, by dosing for 14 days after vessel rupture, the entire period of greatest ischemic risk would be covered with the neuroprotectant therapy. Additionally, by using imaging techniques such as magnetic resonance imaging, patient heterogeneity could be reduced.
While aSAH accounts for only 3%-5% of the overall stroke population, afflicted individuals tend to be younger and often have poorer long-term outcomes than those with ischemic stroke. 13,14 These factors lead to the loss of productive years in the general population similar to that of cerebral infarction, the most common form of stroke. 15 Despite surgical advances such as surgical clipping and endovascular coiling to address the initial and subsequent bleeds, aSAH is associated with a significant risk of delayed cerebral ischemia and related events (eg, vasospasm, infarct, and diffuse microvascular disease). 16 The underlying pathological mechanisms of delayed cerebral ischemia are still unclear and the calcium channel blocker nimodipine, introduced in 1988, remains the only approved pharmaceutical intervention to improve long-term functional outcomes after aSAH. The efficacy of nimodipine across a number of clinical studies has been inconsistent. While meta-analyses of nimodipine clinical studies indicate improved outcome, 17,18 the drug provides only 30%-40% rescue from neurological and cognitive deficits, rendering pharmacological strategies to reduce delayed cerebral ischemia following aSAH an important unmet clinical need. 19 A considerable amount of preclinical evidence demonstrates that prolonged exposure of neurons to extracellular glutamate released during cerebral ischemia leads to cell death that is largely mediated by calcium entry through N-methyl-D-aspartate receptors (NMDARs). 2,3,[20][21][22] NMDARs are composed of 2 GluN1 subunits and 2 GluN2 subunits, of which there are four subtypes (GluN2A-D). Blockade of NMDA receptors has been demonstrated to be neuroprotective in hundreds of preclinical studies in vivo, 23 leading to clinical trials of this strategy in a number of disorders. However, the dose-limiting side effects associated with nonselective NMDAR blockade have prevented successful translation into clinical trials. For the first two generations of NMDA antagonists (ie, nonselective channel blockers and competitive glutamate site antagonists), the level of occupancy needed for neuroprotection altered cardiovascular function and significantly disrupted cognition, leading to dose-limiting side effects. 9,24,25 To reduce the side effects associated with early NMDA inhibitors, modes of inhibition that provide selectivity for NMDAR subtypes were explored as early as the late 1980s. Noncompetitive inhibitors, such as ifenprodil, were identified that do not bind at either agonist binding sites or to the channel pore, [26][27][28] and are highly selective for GluN2B-containing NMDARs. 5,29 Ifenprodil and related GluN2B-selective inhibitors are better tolerated in animals and humans than previous NMDAR inhibitors, including high-affinity channel blockers and competitive antagonists. 30 However, limited efficacy and/or unfavorable safety margins due to untoward cardiovascular or central nervous system (CNS) effects, particularly at high doses, prevented latestage clinical development of GluN2B-NMDAR inhibitors such as traxoprodil (CP-101,606), 31 MK-0657 (CERC-301), 32 and EVT-101. 33 Although promising in a study of TBI patients, 34 ketamine-like dissociative effects were noted with CP-101,606 in a study in depressed individuals. 31 Selective GluN2B-NMDAR inhibitors with properties that should further reduce on-target side effects have been discovered by Neu-rOp Inc. These compounds have increasing potency as extracellular pH is incrementally lowered, which may lead to increased selectivity in the ischemic brain areas associated with tissue acidosis, such as occur in aSAH. 23 This strategy has the potential for reducing NMDAR-related side effects in humans. In particular, NP10679 ( Figure 1A) 35 is a potent, context-dependent, GluN2B-selective inhibitor of the NMDAR that is being developed by NeurOp Inc. with the aim of identifying agents that ameliorate neurological deficits resulting from brain ischemic events such as those associated with stroke, head trauma, and SAH.
In vitro, NP10679 has >100,000-fold selectivity for GluN2B over the other GluN2 subtypes of the NMDAR and is approximately 6-fold more potent at blocking the receptor at pH 6.9 than at pH 7.6. The in vitro metabolic stability profile of NP10679 is similar in mice, rats, dogs, and humans, with similar hepatic microsomal half-lives greater than 2 hours in all species (unpublished). NP10679 is substantially bound to plasma proteins with 97.4%-99.0% bound in the four species (unpublished). The compound readily distributes into brain tissues in mice as suggested by the ratio of brain to plasma levels observed at steady state (1.3-2.6), and is orally available (%F = 76% in mouse).
NP10679 is a weak (concentration that produces half maximal inhibition [IC 50 ] > 10 μM), reversible inhibitor of multiple human liver CYP enzymes studied to date and does not lead to significant time-dependent inhibition of CYP3A (unpublished data and Myers et al 35 ). These data suggest that the potential for pharmacokinetic (PK)-based drug-drug interactions between NP10679 and co-administered drugs is low. The compound is effective in reducing infarct volume in the murine transient middle cerebral artery occlusion (MCAO) model of ischemic brain injury when administered 15 minutes prior to occlusion. 35 The minimum effective dose of NP10679 in the MCAO model is 2 mg/kg for intraperitoneal (IP) administration.
NP10679 lacks demonstrable activity at over 40 CNS targets but does act as a histamine H1 antagonist (IC 50 73 nM), and in pre-clinical models leads to somnolence at doses above those necessary to achieve efficacy. 35 NP10679 also inhibits the human ether-ago-go-related-gene (hERG) channel with an IC 50 value (620 nM) approximately 30 times higher than its activity to inhibit GluN2B-containing NMDARs. 35 This is likely reflected in the activity of the compound to only minimally prolong the QT interval in dogs at doses several fold higher than those we predict will be associated with efficacy in rodent models (unpublished). The offtarget activity, particularly in the context of the safety data collected, appears to pose no serious issues that would limit the clinical development of NP10679. In repeat-dose toxicity studies of up to 14 days duration in rats and dogs, the only adverse effect directly related to intravenous (IV) administration of NP10679 was sedation, which was the dose-limiting effect in both species. In total, the data from good laboratory practice rat and dog safety studies indicate a good safety margin when viewed in relation to plasma levels in the mouse when administered at doses leading to efficacy in the MCAO model. 35 This report describes observations made from the initial phase 1 human clinical trials of NP10679. Results from both single ascending dose (SAD) and multiple ascending dose (MAD) studies are presented. The compound was found to be safe and has a PK profile in humans that lends itself to once per day dosing. The studies presented support further development of NP10679 in brain ischemic disorders such as aSAH.

Methods
These protocols as well as subject-informed consent packages were reviewed and approved by the institutional review board (IRB) for the study (IntegReview IRB, Austin, Texas); the clinical research organization (CRO) for both studies was Pharmaron CPC, Baltimore, Maryland. All subjects were informed of the na-ture and purpose of the study, and their written informed consent was obtained before any study-related procedures were performed. Studies were conducted in accordance with the principles set forth in the Declaration of Helsinki and the International Conference on Harmonization Tripartite Guidance on Good Clinical Practice. Protocols for both the SAD and MAD studies were reviewed and approved by the US Food and Drug Administration under an investigational new drug application.

Drug Substance and Product
The good manufacturing practice quality active pharmaceutical ingredient (API) of NP10679 for the manufacture of drug product to support phase I clinical development was synthesized through a contract with DavosPharma (Saddle River, New Jersey). The Figure 1A) is C 23 H 26 F 3 N 3 O (molecular weight = 449.5 g/mol).
Because NP10679 was to be administered by IV infusion in the SAD and MAD studies, the API was formulated into a clinical drug product for constitution into the IV solution at the clinical site. The manufacturing of drug product was performed by University of Iowa, Pharmaceuticals (UI-P) according to procedures established for the generation of lyophilized product. To formulate the drug product, the API was solubilized in a vehicle consisting of 25% hydroxypropyl-betacyclodextrin (HPBCD) solution in 50 mM potassium phosphate monobasic buffer (pH 6.0) to a concentration of 5 mg/mL. This solution was then filter sterilized and lyophilized into sterile vials each containing 50 mg of API. Subsequent dilutions of the appropriate quantity of the formulated material for each dose were made with 2.5% HPBCD in 0.9% saline up to 75 mL for IV infusion. The sterility of the vial contents was confirmed.

Inclusion and Exclusion Criteria
Healthy male and female subjects aged 18-55 years who were capable of providing consent and able to adhere to the visit schedule and other protocol requirements were eligible for the studies. If sexually active and of child-bearing potential (both men and women), volunteers were required to agree to use two forms of contraceptive methods (including one barrier) for the duration of the study.
Exclusion criteria were inadequate peripheral forearm vein access, pregnancy or lactation, use of nicotine-containing products during the study, current or recent (within 12 months) history of alcohol or drug abuse, recent blood donation within 90 days, and previous participation in a clinical trial withing 90 days. Subjects with excessive somnolence and those who had used medications or agents that might cause drowsiness within 7 days were also excluded. Volunteers with significant medical or psychiatric illness by history, examination, or clinical laboratory testing that would influence study results or preclude informed consent and study compliance were also excluded.

Clinical Study Designs
The SAD study (NP10679-101) was a single-center, randomized, double-blind, placebo-controlled, singledose, dose-escalation trial to investigate the safety, tolerability, and pharmacokinetics of NP10679 in healthy adult volunteers in six escalating dosing cohorts. The primary objective of the study was to assess the safety, tolerability, and pharmacokinetics of a single dose of NP10679 when delivered by IV infusion in comparison to placebo. Secondary objectives were to obtain a maximum tolerated dose of NP10679 in healthy adult volunteers and to establish a safe starting dose for the MAD study (NP10679-102).
The study consisted of a 30-day screening period, day 1 (single IV infusion of NP10679 or placebo, as randomized), day 2 in clinic/overnight assessments, and day 3 assessments. Subjects checked into the clinic on day 1 and remained in the clinic through the 48 hours post-dose blood draw on day 3, after which time they were discharged. Subjects returned to the clinic for a follow-up visit at day 8 after discharge.
There were six dosing cohorts studied in NP10679-101. Each cohort consisted of eight subjects. Six subjects of each cohort were administered NP10679 and two subjects received placebo. Doses were evaluated sequentially before escalating to the next dose level. Doses included in the study were 5, 15, 50, 100, 150, and 200 mg. Drug and placebo were administered by IV infusion in 75 mL of dosing vehicle over 30 minutes. A sentinel dosing, adaptive design approach was used for all cohorts, in which the first two subjects (1 active, 1 placebo) were dosed on day 1 and observed for 48 hours or until sufficient time had elapsed to review safety. If the safety committee (at a minimum, the Principal Investigator [PI]) and Medical Monitor (a subject matter medical doctor independent from the conduct of the study) agreed that it was safe to proceed, the remaining six subjects (5 active, 1 placebo) were dosed in that cohort at the same dose level. Safety/tolerability data as well as available PK data were reviewed prior to dosing in the next cohort of subjects. Acceptable results of the interim safety/tolerability review triggered enrollment into the next dosing cohort.
The purpose of the MAD study was to evaluate the safety and PKs of NP10679 on repeated dosing until steady state was reached. Based on results from the SAD study, it was determined that 5 days of once-daily dosing would lead to steady state. Subjects in the MAD study (NP10679-102) were treated in the same way as those in NP10679-101. The study consisted of a 30-day screening period, dosing days 1 through 5 (single 75-mL IV infusions of NP10679 or placebo over 30 minutes, as randomized), day 6 in clinic/overnight assessments, and day 7 assessments prior to discharge. Subjects checked into the clinic on day 1 and remained in the clinic through the 48 hours post-dose blood draw on day 7, after which time they were discharged. Subjects returned to the clinic for a follow-up visit at day 9. Three dosing cohorts of 8 subjects each (6 drug and 2 placebo) were recruited and dosing decisions were made as in NP10679-101. Dose levels included 25, 50, and 100 mg per day.

Safety Evaluations
Safety and tolerability parameters were assessed according to the protocol schedule of assessments and included assessment of treatment-emergent adverse events (TEAEs) based on physical examinations, infusion site examinations, laboratory findings, neuropsychiatric assessments, vital signs, and subject reported tolerability. End points also include hematology, chemistry, urinalysis, and 12-lead electrocardiogram. Since NP10679 is an NMDAR modulator and such compounds have a history of leading to neurobehavioral effects including cognitive and dissociative effects, the Hamilton Depression Rating Scale, the Mini-Mental Status Examination, the Suicide Behaviors Questionnaire-Revised, the seven-item General Anxiety Disorders scale, and the Clinician-Administered Dissociative States Scale (CADSS) were included as standard assessments. There were no detectable changes in mean values for these assessments. NP10679 also is a histamine H1 inhibitor and as such might lead to drowsiness, therefore the Modified Observer's Assessment of Alertness/Sedation (MOAA/S) and the Bond-Lader Visual Analogue Scales (VAS) sleepiness scale were also added.
All subjects who had at least one dose of the trial medication and a safety follow-up, whether withdrawn prematurely or not, were included in the safety analysis. Data were summarized by reporting the number and percentage of subjects in each category for categorical and ordinal measures, and mean, standard deviation (SD), median, and range for continuous measures. The safety endpoint included a summary of treatmentemergent clinical and laboratory-based adverse events and their severity. All adverse events were coded by System Organ Class and Preferred Term according to the Medical Dictionary for Regulatory Activities (Med-DRA). The TEAEs were tabulated by dose level, system organ class, and preferred term.

PK Measurements
For the SAD study, blood was drawn via a vein opposite the infusion arm (if possible) for determination of systemic NP10679 levels at predose and at the end of infusion (30 ± 5 minutes), and 0.5, 1, 2, 4, 6, 8, 10, 12, 18, 24, 36, 48, and 96 hours postdose. Collection tubes containing K 2 EDTA were used to collect 5 mL of whole-blood sample at each time point. Immediately after collection, tubes were inverted to mix the anticoagulant with the blood sample. Tubes were then centrifuged at a speed of approximately 3000 × g for 10 minutes at 4°C. Within 5 minutes of centrifugation the plasma fraction was transferred into two equal aliquots (1.25 mL each) into 2-mL cryovials and then frozen and stored at −70°C (±10°C) until shipment. For the MAD study, blood was also drawn via a vein opposite the infusion arm (if possible) for determination of systemic NP10679 levels at predose and at the end of infusion (30 ± 5 minutes), and at 0.5, 1, 2, 4, 6, 8, 10, 12, and 18 hours on days 1-5 and at 24, 36, 48, and 96 hours following the final dose on day 5.
Bioanalytical methods were developed and validated at TDM Pharmaceutical Research, Inc. to quantitate NP10679 for the SAD study. Standards, controls, and test plasma samples containing NP10679 were quantified by validated liquid chromatography tandem mass spectrometry (LC-MS/MS) assay(s) subsequent to protein precipitation with a detection range of 1-1000 ng/mL in a 50-μL volume. A close structural analog of NP10679 (NP10767) was used as the internal standard (IS). Each 50-μL aliquot of quality control sample, standard, or plasma sample was mixed with 250 μL of acetonitrile and 50 μL of working IS solution (50 ng/mL in methanol). The sample was vortexed and centrifuged at 1000 × g, and 150 μL of the resulting supernatant was transferred to autosampler vials for injection into an LC-MS/MS system for analysis. Chromatographic retention of NP10679 and the IS was obtained on an Agilent Poroshell 120 EC-C18, 2.1 × 30 mm, 2.7 μm column (Santa Clara, California) under gradient conditions with a flow rate of 0.3 mL/min (mobile phase A was 0.1% formic acid in water, mobile phase B was 0.1% formic acid in acetonitrile). Analytes were detected by multiple reaction monitoring using an MDS Sciex API 4000 mass spectrometer (AB Sciex, Framingham, Massachusetts) in positive multiple reaction monitoring mode (dwell time 100 ms, declustering potential 11 V, entrance potential 10 V, collision energy 43 V, collision cell exit potential 20 V). The analyte NP10679 had an m/z of 450.2 with transition to 243.2. The IS had an m/z of 464.2 and transitioned to 257.2. Plasma concentrations from the resulting LC-MS/MS data were calculated using a six-point calibration curve constructed from known concentrations of NP10679 in human plasma prepared fresh daily. The lower limit of quantitation (LLOQ) for NP10679 was 2 ng/mL in diluted plasma. The intra-run precision (coefficient of variation) was 1.3%-9.7% for NP10679 and the inter-run precision was 1.5%-10.0%. Mean absolute recovery was independent of concentration (94.2%-106.4%). The validation consisted of at least three consecutive assay runs for precision and accuracy, each including blanks, duplicate standard curves, and six replicates of quality controls at each concentration level.
The bioanalytical methods described above were adapted by Pharmaron ABS for the MAD study, with the following modifications. Each 50-μL aliquot of quality control sample, standard, or plasma sample was mixed with 25 μL of working IS solution (500 ng/mL in 5% acetonitrile in water) followed by addition of 500 μL of 50:50 acetonitrile:methanol. The sample was vortexed and centrifuged at 4800 × g, and 50 μL of the resulting supernatant mixed with 100 μL of mobile phase A was transferred to autosampler vials for injection into an LC-MS/MS system for analysis (Sciex 5500) under electrospray positive polarity MRM mode. The IS was deuterated NP10679-d8 with m/z of 458.2 and transitioned to 251.2. The intra-run precision (coefficient of variation) was 1.1%-4.9% for NP10679 and the inter-run precision was 2.7%-5.3%. Recovery was between 73% and 97% for NP10679 and 105% for the IS.
Descriptive PK parameters were calculated based on the plasma concentrations of NP10679. The PK analysis was performed based on a noncompartmental analysis 36 using MS Excel (Seattle, Washington).

Results
Forty-eight subjects were enrolled (Table S1a) into the seven cohorts of the NP10679-101 study. All subjects completed the study protocol, with the exception of one subject who left the study voluntarily due to personal reasons not related to the study. The median age for this study was 33.5 years (range 22-52 years). There were 30 males and 18 women enrolled into the study. Most subjects were Black (35) followed by White (13: 10 non-Hispanic and 3 Hispanic), and one patient was Asian. The NP10679-102 MAD study enrolled 24 subjects into its four cohorts (Table S1b). The median age for this study was 44.5 years (range 20-54 years). There were 15 males and nine women enrolled into the study. As in the SAD study, most subjects were Black (15) followed by White (8) and Asian (1). Table S2 summarizes the TEAEs by organ class and dose for the SAD study NP10679-101. The most common TEAE was somnolence, which appeared to be dose dependent (Table S3). The MOAA/S scale is scored from 0 to 5 with level 5 representing the lowest level of sedation. At level 5, a subject readily responds to normal spoken tones, level 4 indicates a lethargic response to voice, and level 3 requires a louder voice to elicit a response. Scores below 3 require increasing levels of physical stimuli to arouse subjects. NP10679 elicited moderate effects on the MOAA/S at higher dose levels. One of 6 subjects at doses of 5 and 15 mg, and 5 of 6 subjects at doses of 50-200 mg presented with somnolence. Two subjects in each of the 100 and 150 mg cohorts reached a transient level 3 score on the MOAA/S. However, no subject in the highest dose group (200 mg) achieved this score. There were also moderate increases in the Bond-Lader VAS scale starting at the 50 mg dose and continuing through to the 200 mg group. The somnolence observed in the study was viewed by the attending physician to be phenomenologically different from that observed with classic sedative hypnotics, which reduced confidence in the tools used to score it. Even at the highest dose tested, when subjects were stimulated, they quickly oriented to their environment within seconds and were able to perform relatively complex cognitive tasks such as the Digit Symbol Substitution Test (DSST). NP10679 does not have activity at gammaaminobutyric acid (GABA) receptors, 35 but does have some action at H1 histamine receptors. However, the somnolence observed with NP10679 was not accompanied by the impaired cognitive performance observed with some H1 blockers. 37 TEAEs of dizziness, headache, and tremor were noted in one or two subjects, and were not deemed to impact subject safety. Conjunctival and scleral hyperemia occurred in 3 of 6 subjects at the 200 mg dose, and may be related to a nonclinically significant lower blood pressure (both systolic and diastolic) observed at the highest two doses. However, there were no clinically significant changes in vital signs or ECGs in the study. The increases in QTc intervals or hypertension observed with previous GluN2B inhibitors 32,38 were not observed in the NP10679-101 SAD study.
No serious adverse events (SAEs) were observed in either the SAD or MAD studies. There was no evidence of dissociative symptoms or cognitive impairment related to NP10679 as assessed by the CADSS or the DSST in either the SAD or MAD studies, although one subject receiving a single 150 mg dose described intrusive thoughts. Consistent with the SAD study, the most common adverse effect in the MAD study was  Subjects with multiple occurrences of adverse events in the same preferred term are counted only once within that preferred term. Subjects with multiple occurrences of adverse events in the same system organ class are counted only once within that system organ class. System organ class as well as preferred terms under system organ class are sorted in descending order of frequency in combined NP10679 group first and then placebo.
somnolence, which was observed in 3 subjects in both the 50-and 100-mg groups and in three subjects in the placebo group. There were no signs of increased somnolence on repeat dosing. While there may have been some accommodation to the somnolence effect, since observations of this effect occurred for the most part only on the first and second day of dosing, there was not enough of a pattern to support a firm conclusion. Table 1 summarizes the TEAEs by organ class and dose for the MAD study NP10679-102.

Pharmacokinetics
In the NP10679-101 study, NP10679 plasma concentrations (see Figure 1A and Table 2) increased linearly with dose. The C max was approximately dose-linear, especially at doses of 100 mg or less ( Figure 1B). Like-wise the AUC inf was linear with dose ( Figure 1C). After the infusion period, plasma concentrations declined multi-exponentially with a terminal half-life of approximately 20 hours. The total clearance ranged from 9.8 to 12 L/h over the doses studied and, as expected, was approximated by dose/AUC (0-inf) .When compared to the hepatic blood flow in human of 87 L/h, NP10679 cleared the body slowly at less than 12% of the hepatic blood flow. NP10679 appears to distribute extensively throughout the body with a volume of distribution of more than 221 L, equating to 4.5 times the total body water space.
In the NP10679-102 MAD study, all subjects had quantifiable concentrations of NP10679 in plasma out to 24 hours (predose time point of following day) after the first four doses and out to 96 hours (last PK time  point) following the fifth dose (day 5), except for one subject in the 50 mg cohort. Mean C max increased with increasing dose (Figure 2A). Over the 25-100 mg dose range, there was a 5.9-and 2.7-fold increase in C max on days 1 and 5, respectively (for PK parameters see Table 3). Mean AUC 0-96 h determined on day 5 also increased linearly with increasing dose ( Figure 2B). Thus, both C max and AUC were linear with increases in dose. As expected from the residual drug remaining 24 hours after each dose, there was some accumulation of drug in the plasma from days 1 through 5, which was most pronounced at the highest dose (100 mg/kg/day; Figure 2C). Terminal half-life was similar across all doses and days studied with a mean range of 15-19, 16-21, and 13-2 hours for 25, 50, and 100 mg cohorts, respectively, excluding one value with the highest SD in each group, which was identified as an outlier by the Grubb's test. Clearance at steady state was similar across the doses studied with means of 12, 12, and 11 L/h for 25, 50, and 100 mg cohorts, respectively. As an internal check on the PK parameters derived from the MAD experiment, we verified that the calculated values of clearance were approximated by the ratio of dose to the day 5 AUC (0-24 h) , and that the 24 hours accumulation ratio at each dose (AUC (0-24 h, day 5) /AUC (0-24 h, day 1) ) was approximated by the predicted accumulation ratio calculated as 1/(1 − exp(−24 × β)), where β is the elimination rate constant determined from the terminal half-life. In this calculation we used the mean half-life at each dose, avoiding one outlier at each dose level with the largest standard deviation identified by a Grubb's test.

Discussion
This report describes observations from the first-inhuman administration of single and multiple doses of the novel pH context-dependent, GluN2B subtypeselective NMDAR inhibitor NP10679. Adverse events seen in the study were modest and primarily limited to an atypical somnolence characterized by ready arousal. PK data from the SAD and MAD studies indicate exposure was linear with dose and a half-life suitable for once-daily dosing.
These data support progression of NP10679 into phase 2 studies of the compound in an SAH patient population, given the potential for NP10679 to ameliorate neurological deficits arising from ischemia-induced neuronal damage during aSAH. This indication avoids a serious confounding variable of time to treatment that may have prevented positive outcomes in studies of neuroprotective agents in disorders like ischemic stroke.
The strategy to be used with NP10679 is to administer the compound immediately following presentation with aSAH to ensure that NP10679 treatment is on board throughout the time of highest risk (3-14 days post initial bleed) for delayed cerebral ischemia.
Along with issues related to timing of therapy, SAEs have prevented translation of previous NMDAR inhibitors for brain ischemia. These adverse events included psychotomimetic and dissociative effects, as well as off-target cardiovascular effects, such as increased blood pressure for MK-0657 (CERC-301) 32 and potential hERG channel-related QT interval prolongation for CP-101,606. 39 In the current study, no dissociative symptoms or reduction in cognitive performance was observed in either the SAD or MAD studies nor was there clinically significant cardiovascular events at the doses tested. Somnolence was the only prevalent TEAE noted in the studies. This appeared to be dosedependent, starting from the mid dose of 50 mg in the SAD study. The observation of somnolence did not appear to worsen over the course of 5 days of dosing in the MAD study. While there were signs of reduced somnolence on repeat dosing, this was inconclusive based on the number of subjects and relatively short duration of dosing. It is important to note that the somnolence observed was not similar to that observed with classical sedatives. Even at the highest dose in the SAD study (200 mg), subjects were readily aroused and were able to complete complex tasks such as the DSST. This is relevant to the proposed use of NP10679 in the intended brain injury and SAH population, where longitudinal neurological assessments are important for care.
The focus of the current report is NP10679, a representative of a class of highly selective GluN2B NMDAR inhibitors that possess pH-context dependent increased potency. The increase in potency of this class of compound under lower pH conditions has been described in multiple publications. 23,35,40,41 This selectivity should provide superior safety margins in the setting of cerebral ischemia associated with tissue acidosis.
The PK data from the NP10679 SAD and MAD studies indicated dose-linearity and a half-life (∼20 hours) consistent with the compound being dosed once daily. The half-life of NP10679 compares favorably to that of nimodipine, the pharmaceutical standard of care for SAH, since nimodipine is traditionally dosed at 4-hour intervals. The C max and AUC 0-inf values for NP10679 are in line with those that led to efficacy in the mouse MCAO model of stroke. For instance, at the 100 mg dose of NP10679 in the SAD study, the C max was 746 ng/mL and the AUC 0-inf was 9860 h ng/mL; at the minimum effective dose of 2 mg/kg IP in the mouse model, the C max was 581 ng/mL and the AUC 0-inf was 5690 h ng/mL. 35 The observation of somnolence in the phase 1 clinical studies of NP10679 also suggests that NP10679 is brain penetrant and interacts with a target, possibly the H1 histamine receptor, in sufficient occupancy to elicit the somnolence effect. Recent data suggest that the neuroprotective effects of NP10679 occur at lower doses than those associated with sedation in the mouse. 35 These observations along with that of somnolence observed in the human studies described in this report suggest that brain concentrations reached in humans with safe doses of NP10679 may be sufficient for efficacy in ischemic conditions such as the delayed cerebral ischemia associated with SAH.

Conclusions
The initial human studies NP10679-101 and NP10679-102 demonstrate that NP10679 at doses that are projected to be effective to treat the delayed cerebral ischemia associated with aSAH is safe in healthy humans. NP10679 administration to humans leads to dose linear pharmacokinetics with a half-life (20 hours) that is compatible with once daily dosing. The studies are encouraging for further development of NP10679 in the setting of acute ischemic and traumatic brain injuries, including aSAH.