Abstract
Historically, nitrogen oxides (NOx) in food, drinking water, as well as in the atmosphere have been believed to be associated with adverse health consequences. More recently, NOx have been implicated in normal homeostatic regulation, and exogenous administration has been associated with health benefits. One such potential health benefit is the prospect that inhaled nitrite will lower pulmonary blood pressure (BP) in patients with pulmonary arterial hypertension (PAH), a disease with poor prognosis due to the lack of effective treatment. To characterize potential chronic toxicity associated with inhaled AIR001 (sodium nitrite) for use in the treatment of PAH, 26-week exposures to AIR001 were carried out by inhalation administration in rats and by intravenous infusion in dogs. The studies revealed that methemoglobinemia was the primary adverse effect in both species. Methemoglobin levels less than 40% were well tolerated in both species, while levels greater than 50% methemoglobin caused death in some rats. Additionally, a decrease in systemic BP was also observed with inhaled AIR001 exposure in dogs. These acute secondary and exaggerated pharmacological effects occurred daily throughout the 26-week treatment period. Chronic exposure did not alter the magnitude of either methemoglobinemia or hypotension or result in additional toxicity or compensatory responses. Based on the exposure levels that produced these pharmacodynamic responses in animals, relative to those measured in early clinical studies, it appears that an adequate margin of safety exists to support the continued clinical development of inhaled AIR001.
Introduction
The potential toxic properties of nitrite (NO2) exposure continue to be discussed in the scientific literature almost 60 years after the discovery of hepatocarcinogenic effects of N-nitrosodimethylamine (NDMA). 1 N-Nitrosodimethylamine is one example of nitrosation of secondary aliphatic and aromatic amines leading to carcinogenic N-nitrosamines. Aside from concerns about potential mutagenesis and formation of carcinogens, breathing difficulties in infants with elevated methemoglobin levels have been associated with NO2 and nitrate (NO3) exposures. In humans, high levels of nitrogen oxides (NOx), including nitric oxide (NO), NO2 anion, and NO3 anion, have been associated with adverse inflammatory responses, especially during oxidative stress conditions or in low pH environments. 2 Furthermore, inhaled exposure to NOx atmospheric pollutants (primarily nitrogen dioxide) can cause airway irritation, lung inflammation, and exacerbate respiratory symptoms in asthmatics. 3
However, in the last 30 years, oxides of nitrogen including NO2 have been increasingly recognized for their essential role in biological processes. 2,4 Nitrites have long been known for their ability to cause vasodilation thereby reducing blood pressure (BP) 5,6 by a cyclic guanosine monophosphate-mediated mechanism. 5 This effect may be due to the conversion of NO2 to NO or possibly by NO2 acting as a reservoir for NO via nitrosation of thiols or nitrosylation of heme. 4 Studies have also demonstrated that NO2 may be therapeutically useful in ischemia–reperfusion injury of the heart, brain, liver, and peripheral limbs. 2 Although NO2 can come directly from dietary sources, most NO2 present in the body is believed to come from reduced NO3. 6 Dietary NO3 (from vegetables and cured meats) is concentrated in the salivary gland where it is converted into NO2 by commensal gram-negative bacteria on the tongue. Nitrate is also reduced by many mammalian enzymes and heme-containing proteins in the tissues to produce NO2. Nitrite can subsequently be reduced to NO and may serve as a storage reservoir for NO. 7 Nitric oxide can also be oxidized back to NO2. Recently, it has been shown that by enzymatic and non-enzymatic metabolism (dependent on local pH, oxygen tension, and redox status), NO2 may act directly in hypoxic signaling as well as in nitrosation of thiols and nitrosylation of hemes. 2,4 Thus, NO2 appears to have bioactivity on its own as well as being part of the mammalian nitrogen cycle, interconverting between the reduced (NO) and oxidized (NO3) forms, as well as being sequestered and stored by proteins.
Vasodilation, one of the primary pharmacological actions of NO2, has been demonstrated in multiple species, 4 including humans. 2 Vasodilation, leading to improvements in blood flow in animal models of vascular/ischemic injury, suggests the potential therapeutic value of NO2. In one such study, significant increases in pulmonary vascular resistance and decreases in cardiac index were markedly reversed by NO2 infusion in an acute pulmonary thromboembolism-induced pulmonary hypertension model in canines. 8 Nitrite may also decrease inflammation and reverse arterial wall remodeling observed in pulmonary arterial hypertension (PAH). For example, in a rat vascular injury model, sodium nitrite decreased intimal hyperplasia by 77% when administered orally, intraperitoneally, or by inhalation (IH). 9 In a dog coronary occlusion model, myocardial infarction size was reduced from 70% to 23%, and improved cardiac contractile function was reported. 10 In NO2-infused primates with experimentally induced subarachnoid hemorrhage, significant reduction in middle cerebral artery injury was observed. 11 In a hypoxemic mouse model of PAH, nebulized sodium nitrite administered either 1 or 3 times weekly resulted in the prevention of, and if administered after onset, the reversal of both pulmonary vasculature remodeling and the development of right ventricular hypertrophy. 12
The data from these animal models on the potential health benefits of NO2 have resulted in the development of AIR001 (sodium nitrite), to be delivered by IH, as a proposed treatment for PAH. Pulmonary arterial hypertension is a progressive disorder of increased pulmonary vascular resistance that results in right ventricular failure and ultimately death. 5 Currently available therapies for PAH, including prostanoids, endothelin receptor antagonists, guanylate cyclase stimulators and phosphodiesterase type 5 inhibitors, have led to significant improvements in the quality of life for many patients and improved survival; however, all these treatments possess limitations (ie, limited efficacy, safety concerns, and drug delivery challenges) supporting the need for alternative therapies to treat this life-threatening disease. 5
The enthusiasm for treatment with systemic sodium nitrite is dampened by the potential exaggerated pharmacological effects. Systemic exposure to high doses of NO2 can cause a significant decrease in systemic BP, a compensatory increase in heart rate (HR), and a decrease in the oxygen carrying capacity of blood due to oxidation of iron in deoxyhemoglobin to form NO and methemoglobin. Methemoglobinemia was the primary toxicity observed in 7- and 28-day repeat-dose IH toxicity studies of AIR001. These studies in rat (N = 5-15/sex/group) and dog (N = 3-5/sex/group) were conducted at doses up to maximum tolerated dose (MTD) using good laboratory practice (GLP) compliant protocols and study parameters consistent with International Conference on Harmonization (ICH) guidance. Additionally, decreased activity, modest decreases in body temperature, reduced body weight gain, and at high-dose levels associated with severe methemoglobinemia, compensatory responses commonly associated with chronic hypoxia (increased hematocrit and splenomegaly) were seen (data not presented). 13,14 Furthermore, when signs of cyanosis were observed and methemoglobin levels exceeded 50%, deaths occurred in some animals within 30 to 60 minutes post-IH exposure. At human therapeutic dose levels, IH administration of AIR001 would be expected to produce low systemic exposure while improving delivery of drug to the site of action in the pulmonary vasculature, therefore allowing for a favorable therapeutic index. To evaluate potential adverse effects beyond the acute findings described previously, a 26-week rat IH exposure study was performed to support the continued development of AIR001. Additionally, a 26-week dog intravenous (IV) infusion study was conducted to represent a worst case scenario in which inhaled AIR001 achieved 100% systemic bioavailability. Although previous data on chronic oral administrations of sodium nitrite exist, 13 no long-term exposure toxicology data are published for pulmonary or IV administration.
Materials and Methods
Chemical
AIR001 (sodium nitrite) was prepared as a solution in a phosphate buffer at the appropriate concentration for nebulization (rat) or IV infusion (dog). Phosphate buffers were used as the vehicle control in both species. The formulations used for the rat IH studies were not isotonic. The formulations used in the dog IV studies were adjusted to be isotonic. Formulated solutions were adjusted to pH 7.2 ± 0.3 and were filtered prior to nebulization or IV infusion. Samples of the dosing solutions were obtained at various intervals for analysis.
Validated methods were used for the determination of NO2 in the dose formulations for both the rat and the dog studies at several time points throughout each study. In addition, the aerosol concentration and particle size of AIR001 were frequently measured in the rat IH study using the same methods. Samples were analyzed on a reversed-phase high-performance liquid chromatography (HPLC) system with ultraviolet (UV) detection using an Agilent Eclipse XDB-C18 (Agilent Technologies, Santa Clara, CA) column with an isocratic mobile phase of 35:65 acetonitrile:0.01 mol/L trimethylstearylammonium chloride. The method was validated for an NO2 concentration range of 1 to 100 µg/mL.
Compliance
The studies were performed in compliance with both Food and Drug Administration (FDA) 15 and Organization for Economic Cooperation and Development 16 GLPs and followed ICH guidelines. 17 Characterization and stability of AIR001, in addition to filtered dose formulation sample analysis, were assessed per current Good Manufacturing Practice regulations. 19 The reference standard characterization of sodium nitrite was performed according to ISO 9001. 20 The study protocols were reviewed and assessed by the Animal Care Committee of ITR Laboratories (Baie d’Urfé, Canada). All animals used in these studies were cared for in accordance with standard principles 21,22 in an Association for Assessment and Accreditation of Laboratory Animal Care International-accredited facility.
Six-Month Rat IH Study
Animals
A total of 266 (133 males and 133 females) Sprague Dawley Crl: CD (SD) rats (Rattus norvegicus) were obtained from Charles River, Canada Inc. On arrival, all animals were weighed and deemed suitable for use after a detailed physical examination. Rats were housed individually in stainless steel wire mesh-bottom rodent cages equipped with an automatic watering system. The animal room environment was controlled (temperature 21°C ± 3°C, relative humidity 50% ± 20%, 12-hour light, 12-hour dark, and 10-15 air changes/h) and monitored at all times. A standard certified commercial rodent chow (Teklad Certified Rodent Diet (W) #8728C [Harlan Laboratories, Indianapolis, IN], NO2 content estimated < 2 ppm) and purified water (NO2 and NO3 levels < 0.09 mg/L) were provided to the animals ad libitum except during designated procedures. Additionally, rats were given nondietary items (e.g., toys) as part of an environmental enrichment program. Upon the onset of treatment, the rats were 8 to 9 weeks of age, with male rats weighing 258 to 366 g and female rats weighing 188 to 261 g. All rats were acclimatized to the laboratory environment for approximately 2 weeks prior to the start of the study.
Treatment
The study was designed with 5 treatment groups: air control, vehicle control, and 10, 30, and 90 mg/kg/d treatment with inhaled AIR001. The dose range was selected to achieve sufficient spacing between a high clinical dose and an acute MTD as determined in previous dose range-finding studies (data not presented). Twenty rats/sex/dose group were evaluated. Fifteen of those animals were sacrificed approximately 24 hours after the last exposure. The remaining 5 rats/sex/group were retained without further AIR001 treatment for 28 days, at which time they were sacrificed. Additionally, 3 rats/sex/group for air and vehicle control groups or 9 rats/sex/group for AIR001-treated groups were used for the collection of toxicokinetic (TK) samples. Sequential blood samples were collected from these satellite TK animals pre-IH, immediately post-IH and at 0.5, 1, 2, and 4 hours postend of IH on study days 1, 28, 85, and 176. Inhalation exposures were performed in a flow-past, radial, nose-only exposure system (CRE, Switzerland) with each animal confined to an individual restraint tube. AIR001 or the vehicle formulation were metered into clinical nebulizers (Sidestream; Respironics, Philips, Inc. Kennesaw, GA) using a syringe pump.The aerosol was produced by applying 10 L/min of medical grade air supplied by an air compressor to each of the 3 nebulizers. Rats in the air-only control group received the same medical grade air, at the same flow rate, as used for the vehicle and AIR001-exposed animals. Dilution air was added to provide a minimum of 1.0 L/min airflow to each animal exposure port (64 ports total) and was balanced to ensure a slight positive pressure. Determinations of aerosol concentration, particle size distribution, oxygen concentration, relative humidity, and temperature were performed on test atmosphere samples collected from the breathing zone area at a representative port of the exposure system. Aerosol concentration samples were collected onto filters from all groups, and the filter samples from the low-, mid-, and high-dose groups were analyzed chemically using a GLP-validated assay. Based on the data from approximately 38 samples from each dose group taken periodically over the 6 months of exposure, the average mg/kg dose, using standard equations, was calculated. 23 Air and vehicle control aerosol samples were similarly analyzed to confirm the absence of NO2 anion in the exposure. Aerosol concentration homogeneity among exposure ports was demonstrated prior to initiation of animal exposure. Samples were also collected into a 7-Stage Mercer Cascade Impactor (InTox Products, Moriarty, NM) to determine the distribution of particle size in the generated aerosols. The sample substrates from the low-, mid-, and high-dose groups were analyzed chemically for NO2 anion, while sample substrates from the vehicle control were evaluated gravimetrically after overnight drying. The mass median aerodynamic diameter (MMAD) and the geometric standard deviation (GSD) were calculated for each group based on the results obtained from the impactor filter samples using a log-probit transformation.
Before the animals were exposed to treatment, all animals (including controls and spares) were gradually accustomed to the IH restraint tubes over a period of at least 5 days for at least 120 minutes. After the equipment acclimation period, rats were exposed daily by nose-only IH to nebulized aerosol for 120 minutes/d for 26 consecutive weeks. There were 3 exceptions to this general exposure directive. In a previous pilot study, exposure to the highest dose on day 1 was associated with mortality. In an attempt to diminish this outcome, all rats in the high-dose group (90 mg/kg/d) were exposed for only 100 minutes (75 mg/kg) on day 1 and thereafter received the full (90 mg/kg/d, 120 minutes) treatment. Therefore, the initial TK samples were obtained on day 2 for the high-dose group, which was the first day they received the full 90 mg/kg dose. The second exception occurred on days 8 to 12 (depending on replicate) during which high-dose females demonstrated excess morbidity and mortality. For these high-dose female rats, exposure was reduced back to 100 minutes/d (75 mg/kg) for the remainder of the study. The final exception occurred on a few occasions (N = 18) when, due to perceived poor health, selected animals received a single-day AIR001-free holiday.
Six-Month Dog IV Infusion Study
Animals
Purpose bred and experimentally naive male and female beagle dogs (Marshall BioResources, North Rose, New York) were acclimated to laboratory conditions for 3 to 12 weeks. Seven dogs/sex were used in 4 dose groups for a total of 56 dogs. At the onset of treatment, the dogs were approximately 15 months old. Male dogs weighed between 6.5 and 10.3 kg, while females weighed 5.6 to 8.2 kg. The dogs were housed individually in appropriately sized stainless-steel cages equipped with an automatic watering system and were healthy and parasite free prior to dosing. The animal room environment was controlled (temperature 21°C ± 3°C, relative humidity 50% ± 20%, 12-hour light, 12-hour dark, and 10-15 air changes/h) and monitored. A standard certified commercial dog chow (400 g of Teklad Certified 25% Lab Dog Diet #8727C [Harlan Laboratories, Indianapolis, IN], ∼4 ppm NO2 content) was given to each dog at the end of the day and left overnight. Municipal tap water (NO2 and NO3 levels < 0.09 mg/L) was provided ad libitum except during designated procedures. Additionally, dogs were included in an environmental enrichment program that included low NO2/NO3 treats, nondietary items (e.g., toys), and exercise.
Treatment
In a previous 28-day IH toxicity study using dogs restrained in slings for up to 4 hours, no pulmonary toxicity was observed (data not presented). However, the IH procedure was deemed too stressful to the dogs to be extended for 26 weeks. Thus, the IV route of administration was used as a means of obtaining a range of plasma exposures adequate to address the potential chronic systemic toxicity of AIR001. The advantage of IV infusion was that a short drug administration period reduced the stress of physical restraint and provided simulation of a worst case systemic exposure after IH administration (100% bioavailability). AIR001 was administered by IV infusion at a dose rate of 10 mL/kg/h for 60 minutes (10 mL/kg total volume/animal) daily for 26 weeks. The infusion rate was controlled by an infusion pump while the dog was restrained in a sling. The actual volume infused per hour was calculated and adjusted based on the most recent body weight. The infusion pump was connected to an Angiocath (Becton Dickinson, Sandy, Utah) IV catheter that was inserted daily into one of the cephalic or saphenous veins on a daily rotation basis. The study design included 4 exposure groups: vehicle control and 7, 14, and 28 mg/kg/d of AIR001. Intravenous dose levels were selected to achieve a range between a high clinical dose and a MTD in dogs as determined in prior pilot studies (data not presented). Blood samples were also collected at various times for clinical pathology and for the determination of plasma NO2 and NO3 concentrations. Samples for plasma TKs were collected preinfusion, immediately postinfusion, and at 0.25, 0.5, 1, 2, and 4 hours postend of infusion on study days 1, 28, 91, and 182 from all treated dogs. An additional 3 blood samples were collected on day 182 at 8, 12, and 23 hours postend of infusion. These additional samples were used for NO3 evaluation only. Four dogs/sex/group were sacrificed approximately 24 hours after the last exposure on day 182. The remaining 3 dogs/sex/group were retained without further AIR001 treatment for 28 days after which time they were sacrificed to address potential reversibility of any findings.
Measurements: Rat and Dog Studies
Study parameters used in both species included detailed clinical examinations, measurement of body weight and food consumption, as well as an ophthalmological examination. All standard clinical pathology parameters (hematology, coagulation, and clinical chemistries) were evaluated from venous blood samples and collected from fasting animals prior to start of treatment, twice during treatment, and at termination of the treatment and recovery phases. In addition, blood was collected for measurement of troponin I by a validated enzyme-linked immunosorbent assay, and methemoglobin, oxygen saturation, and hemoglobin concentrations were measured by an ABL80 FLEX CO-OX, co-oximeter (Radiometer, Copenhagen, Denmark) at 30 minutes postexposure. The time course of methemoglobin appearance was investigated in previous Aires studies in both species, revealing that peak levels were observed at approximately 30 minutes postexposure. By 4 hours, levels were markedly decreased and returned to baseline levels prior to the next exposure (data not presented). Urinary chemistry was analyzed from urine samples collected from overnight fasting animals. For the dog studies only, BP (using cuff oscillometry), electrocardiograms (limb leads I, II, and III, and leads augmented vector right, augmented vector left, and augmented vector foot), and respiratory function (using a face mask pneumotachometer) were collected from all dogs at prescribed occasions throughout the study at approximately 30 minutes postexposure. For both species at termination, a complete gross pathology evaluation was conducted. Fourteen selected organs were weighed and histopathology on 46 tissues (dogs) or 50 tissues (rats) were evaluated by a board-certified pathologist. All dose groups were evaluated in the dog study. For the rat study, all control and high-dose tissues were initially examined; subsequently, target tissues (respiratory tract) were investigated in all dose groups. Additionally, the histology slides and pathology reports were externally peer reviewed by a board-certified veterinary pathologist. 24 Both pathologists were in agreement with the stated findings for both species.
Bioanalytical Methods
Validated methods were used for the determination of NO2 and the principal metabolite NO3 in heparinized plasma of both rat and dog. In brief, NO2 was extracted from plasma by protein precipitation with ammonium sulfate, followed by derivatization with the fluorometric reagent 2,3-diaminonaphthalene. After incubation, sodium hydroxide was added to enhance the detection of the fluorescent product, 1(H) naphthotriazole. Samples were analyzed on a reversed-phase HPLC system with fluorescence detection using a Pursuit C18 column (Agilent Technologies, Santa Clara, CA) with an isocratic mobile phase of 50:50 10.6 mmol/L sodium phosphate buffer:methanol. The method was validated for a plasma NO2 concentration range of 200 to 8000 ng/mL.
The NO3 metabolite was extracted from heparinized rat and dog plasma using a protein precipitation extraction. The extract was analyzed on an ion-pairing, reversed-phase HPLC system (Zorbax Eclipse XDB C18, Agilent Technologies, Santa Clara, CA) with an isocratic mobile phase of 10 mmol/L trimethylstearylammonium chloride in water:acetonitrile (65:35, v/v) with UV absorbance detection at 210 nm. The method was validated for a plasma NO3 concentration range of 3.10 to 310 µg/mL.
Statistics
Statistical evaluation was performed after an automatic transformation to analyze the data for homogeneity of variance using Levene tests and for normality using Kolmogorov-Smirnov tests. Parametric and nonparametric trends were analyzed using the Williams and the Shirley-Williams tests, respectively. Homogeneous data were analyzed using analysis of covariance/variance, and the significance of intergroup differences between the air control and all other groups was analyzed using Dunnett test. Heterogeneous data were analyzed using Kruskal-Wallis test, and the significance of intergroup differences between the air control and treated groups was assessed using Dunn test. Statistical significance was accepted if the corrected probability levels indicated P ≤ 0.05.
Results
Formulation Analysis
Dose groups and analytical information are presented in Table 1. Results of the dose formulation analysis for both studies were within the predefined limits for concentration (±10%) and pH (7.2 ± 0.3). For rat IH administration, achieved aerosol concentrations and estimated achieved dose were stable and acceptably close (±4% concentration, 93%-105% dose) to the target levels for all 3 dose groups over the 6-month exposure period. Between 77% and 97% of the AIR001 aerosol particle sizes were less than 3 μm, indicating the aerosol was in the respirable range for the rat at all concentration levels. Particle size increased as the concentration of AIR001 in the vehicle increased, ranging between 0.9 and 1.6 μm (MMAD) with a GSD of ≤2.07.
Dose Groups, Exposure, and Formulation Characteristics for Rats.
Abbreviations: GSD, geometric standard deviation; N/A, not applicable.
a Estimated based on dose equation of Alexander et al. 24
Mortality
In the rat IH study, 4 female rats and 3 male rats in the high-dose group died shortly after AIR001 exposure ended (Table 2). These deaths were associated with clinical signs of hypoxia and thus were considered to be secondary to methemoglobinemia. Animals that died within the first 2 weeks of exposure were replaced. Three other animals died on study due to causes unrelated to AIR001 exposure. No mortality was observed in the dog 26-week IV infusion study.
Group, Sex, Day, and Apparent Cause of Death in Rats Exposed to AIR001.
Abbreviation: TK, toxicokinetic.
a Day found dead.
Clinical Evaluations
Physical examinations
Pale/discolored skin and mucosa (cyanosis) was observed, generally on a daily basis, in all high-dose rats and dogs. Cyanosis occurred sporadically in the mid-dose group of both rats and dogs, but particularly in female rats that had significant elevations in methemoglobin. Generally, cyanosis began during drug administration but was usually not observed by 3 to 5 hours postexposure and never observed during the next day’s preexposure observation period. Neither tolerance nor increased sensitivity to these cyanotic changes was observed over the course of the study in either species.
Body weight determination
Body weights in female rats and dogs were unaffected by treatment with AIR001. Male rats in the high-dose group were 12% lighter than male air control rats, an effect that only reached significance (P = 0.01) during the last week of the 26-week study (Figure 1). There were no significant effects on food consumption or ophthalmoscopy in either species.
Body weight in male and female rats exposed once daily for 26 weeks to AIR001 by inhalation. N = 20/sex/group. The standard deviation during the exposure period never exceeded 14%. The gray area indicates the treatment-free recovery period, which began after the last exposure on week 26. N = 5/sex/group from week 26 until recovery group sacrificed at week 30. Dose groups per Table 1. *P = 0.01 at the time indicated.
Clinical pathology
No consistent, dose-related effects were observed in clinical chemistry, coagulation parameters, troponin I, or urinalysis except for a higher incidence of NO2 present in urine samples of treated rats and dogs. This finding was expected and interpreted as a reflection of excess urinary NO2 clearance rather than an indication of urinary tract infection for which there was no other supportive evidence.
Statistically significant changes in hematology were observed only at week 8 in the rat. The changes, suggestive of a small compensatory response to hypoxia, were well within the normal historical control range for this strain of rat, however. In male rats, red blood cell (RBC) counts were significantly decreased (5.8% [mid] and 4.4% [high]) and reticulocyte counts were significantly elevated (36.8% [mid] and 50.9% [high]) in the mid- and high-dose groups (Figure 2). In addition, white blood cell counts were significantly decreased (25.4%) in male rats due to decreases in lymphocytes (32.5%), monocytes (42.4%), and basophils (30%). In female rats, RBCs were not decreased; however, a similar significant increase in reticulocyte count (57.9%) was observed but only in the high-dose group. Total white blood cell counts were also similarly decreased (25.2%) in female rats but the reduction was associated with a decrease in monocytes (42.8%) and eosinophils (41.7%). These effects on white blood cell differential counts were generally not dose related, occurred at a similar magnitude in all dosed groups and were within the range of historical controls. Furthermore, by week 26, these hematological changes had resolved, and all hematological parameters were indistinguishable from controls in both male and female rats.
Red blood cell (left axis) and reticulocyte (right axis) counts in male rats exposed daily for 8 weeks to AIR001 by inhalation. N = 20/group. No significant hematology findings were observed when reassessed at 26 weeks. Dose groups per Table 1. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 difference by dose level.
For dogs, there were no changes in hematology parameters that were clearly attributed to treatment with AIR001. Blood cell morphology was normocytic and normochromic for all dogs, at each of the 4 sample collection times.
Co-oximetry
Increases in blood methemoglobin levels were observed in a dose-dependent manner in both rats and dogs. Pretreatment levels and control groups consistently had methemoglobin levels of 1.1% to 2.2% in rats and 2.0% to 2.6% in dogs.
Baseline methemoglobin levels were significantly different (P < 0.018) between male (1.42%) and female (1.50%) rats in the nonexposed control animals. A clear sex difference in methemoglobin levels was also observed in rats in response to AIR001 treatment (Figure 3). In the high AIR001 dose group, males averaged 29.1% while females averaged 42.8% (P < 0.001) methemoglobin. There were no obvious trends in the methemoglobin response with number of exposure days in any group and thus no evidence that repeated dosing with AIR001 caused either sensitization or tolerance to the development of acute-transient methemoglobinemia. No effect on oxygen saturation levels or hemoglobin by co-oximetry was observed in rats.
Percentage of methemoglobin in male and female rats (A) averaged over 24 weeks (N = 62-68/sex/group) and (B) at various intervals over 26 weeks to AIR001 by inhalation (N = 3-19/sex/group/time point). Dose groups per Table 1. *P ≤ 0.05 difference between males and females.
In dogs, methemoglobin levels obtained at 30 minutes postexposure were very stable day-to-day with the average methemoglobin varying between 34.7% and 38.5% in the high-dose group (Figure 4). No sex difference in this parameter was observed either before or after IV infusion with AIR001. No changes in oxygen saturation or hemoglobin by co-oximetry were observed in dogs.
Percentage of methemoglobin in male and female dogs (A) averaged over 26 weeks and (B) at various intervals over 26 weeks to AIR001 by intravenous infusion. N = 7/sex/group.
Cardiovascular parameters
In dogs, mean arterial BP decreased in a dose-related manner with males showing more of an effect than that observed in female dogs (Figure 5). Significant effects on mean arterial BP were seen in male dogs in all treatment groups, while females were significantly affected only in the high-dose group. A differential effect on systolic versus diastolic BP was not detected.
Mean arterial blood pressure in male and female dogs exposed daily for 26 weeks to AIR001 by intravenous infusion. Data are the average of all collection periods as there were no differences observed with respect to day of data collection. N = 7/sex/group. *P ≤ 0.05 difference from vehicle control response.
Heart rate increased following infusion of AIR001, achieving significance in the high-dose group as PR and QT shortened (not significant). Respiration and morphology of the electrocardiogram as well as quantitative measurements of the corrected QT (QTc; note 1) intervals and QRS complex duration in dogs were unaffected. As with BP, the observed dose-related increase in HR also varied depending on the sex of the dog. However, females showed a greater increase in HR than did males (Figure 6). Tolerance or increased sensitivity to the effects on BP and HR was not observed for either sex when comparing measurements made at the beginning, middle, and end of the study.
Pharmacokinetics
The plasma pharmacokinetics of NO2, and its major metabolite, NO3, was evaluated. Generally, NO2 Cmax values were observed immediately after the end of IH (rats) or infusion (dogs), while Cmax of the NO3 metabolite was delayed until approximately an hour after the end of dosing. Predose plasma concentrations of NO2 were generally below the lower limit of quantitation of the assay in all animals on study day 1 and in vehicle-treated animals throughout the study. No significant differences were observed in plasma Cmax and area under the curve (AUC) values for NO2 spanning the multiple assessments in both rats and dogs, indicating the reported values were at steady state. For the NO3 metabolite, plasma concentrations were consistently much greater (10 to 30 fold) than those of NO2 in all treated groups in both species, indicative of rapid in vivo conversion of NO2 to NO3.
In rat plasma, concentrations of NO2 decreased rapidly with a terminal half-life between 0.3 and 1.6 hours at steady state (Table 3), while the half-life of the NO3 metabolite was consistently longer (between 1.4 and 11.8 hours). Generally, plasma NO2 Cmax and AUC values were lower in male than in female rats despite males receiving 90 mg/kg while females received 75 mg/kg inhaled AIR001. Similarly, Cmax and AUC values of the NO3 metabolite were also lower in male than in female rats. Increases in Cmax and AUC were generally proportional to dose for both NO2 and NO3 in rats following IH administration. No accumulation of NO2 was observed, but slight accumulation of the NO3 metabolite was observed in rats at the end of the 26-week study.
Mean Steady State Plasma Nitrite TK Parameter Values on Day 176 Following a 2-Hour Inhalation Administration of AIR001 to Rats for 26 Consecutive Weeks.
Abbreviations: t1/2, half-life; Tmax, time to Cmax from the start of exposure; Cmax, maximum concentration; AUC, area under the curve; F, female; M, male; MS, missing (insufficient data points for determination); TK, toxicokinetic.
a Three animals/sex/time point.
In dogs, steady state concentrations of NO2 in plasma decreased rapidly with a terminal half-life between 0.2 and 0.4 hours (Table 4), while the half-life of the NO3 metabolite was much longer (between 9 and 20 hours). Predose plasma concentrations of NO2 were generally below the lower limit of quantitation of the assay in all animals on study day 1 and in vehicle-treated animals throughout the study. Dose-normalized Cmax and AUC values were comparable across all dose groups for both NO2 and NO3, indicating dose proportional increases in exposure following IV administration. No accumulation of NO2 was observed. Accumulation of the NO3 metabolite could not be assessed because plasma concentrations of this metabolite were only assessed at the end of the study. A sex difference in TK parameters was not observed in dogs for either NO2 or NO3.
Mean Steady State Nitrite TK Parameter Values on Day 182 Following a Once-Daily 1-Hour Intravenous Infusion of AIR001 to Dogs for 26 Consecutive Weeks.
Abbreviations: t1/2, half-life; Tmax, time to Cmax from the start of exposure; Cmax, maximum concentration; AUC, area under the curve; F, female; M, male; TK, toxicokinetic.
a Seven animals/sex/time point.
Organ Weights and Pathology
For rats and dogs, organ weights were within normal limits and there were no abnormal macroscopic findings at necropsy (data not presented). In rats, animals dying during the study had no histological abnormalities, and the deaths were ascribed to methemoglobinemia-related hypoxia. In dogs, there were no microscopic lesions that appeared to be treatment related. Nasal cavity lesions were observed in rat, but histopathology of the carina, nasopharynx, bronchial lymph node, larynx, trachea, and lungs was unremarkable as were all other nonrespiratory tissues. In the rat nasal olfactory epithelium, eosinophilic inclusions were often observed in all groups. An increased severity of these eosinophilic inclusions was noted both in the mid- and high-dose-treated rats compared to the background findings in air, vehicle control, and low-dose groups; and hence, the changes were considered treatment related. Table 5 presents the number of animals at the termination of the main study with eosinophilic granules in nasal standard sections T3 or T4 25 compared to the number of rats evaluated. A representative histology slide is presented in Figure 7. No other NO2-related histological changes were observed in either rats or dogs.
Histological section of nasal cavity in a high-dose male rat receiving 90 mg/kg/d of AIR001. H&E stain, 20×. Eosinophilic inclusions were observed as pink round aggregates inside the epithelial cells of the mucosa on the left (see arrows). H&E indicates hematoxylin and eosin.
Incidence of Increased Eosinophilic Inclusions in Nasal Sections 3 and 4 after 26 Weeks of Inhaled AIR001.
a As described by Harkema et al.25
Discussion
The safety of exposure to various NOx in humans remains a controversial topic. In more recent years, the health benefits, particularly those associated with NO, have further stimulated discussion of the balance between benefits versus risks. The primary purpose of this study was to provide insight into the potential chronic toxicity of sodium nitrite that might be relevant to the assessment of clinical risk. Although the chronic effects of sodium nitrite in drinking water have been studied in mice and rats, 13 this is the first study of the chronic effects of sodium nitrite by IH administration or IV infusion in rats and dogs, respectively.
Overall, exposure to inhaled or infused AIR001, at all but the highest doses tested, was toxicologically benign and physiologically predictable during and after 26 weeks of exposure to rats and dogs. The most consistent toxicologically adverse sign was an acute, reversible, dose-related increase in methemoglobin; a monitorable parameter in humans. High methemoglobin is known to produce hypoxia and death by reducing the ability of hemoglobin to transport oxygen to tissues. Indeed in the rat IH study, 7 rats in the high-dose exposure group were found dead soon after exposure ended. Similarly, lethality at very high methemoglobin levels was also observed in a previous dog 28-day IH study without other evidence of adverse effects (data not presented). The conclusion that these deaths were due to an acute intoxication rather than cumulative toxicity is supported by the data and the timing of death in relation to dosing. First, blood samples taken prior to death in some of these animals indicated methemoglobin levels of >50%. The lack of other toxicologically significant findings in the animals that died or in the surviving animals from the high-dose group also supports an acute mechanism. Furthermore, the TK data indicated that there was no NO2 accumulation and only slight accumulation of the NO3 metabolite with 26 weeks of exposure to near-lethal doses. Additionally, cyanosis and elevated methemoglobin levels were generally absent by 5 hours postexposure and were never present by the next day’s exposure. In the 26-week IV infusion study, no dogs died or appeared moribund despite some animals approaching methemoglobin levels of 50%.
The one persistent acute clinical sign was the presence of cyanosis (pale discoloration of the skin) that occurred daily throughout the 26 weeks of exposure to AIR001 in both rats and dogs. Although primarily observed in the high-dose animals, some animals in the mid-dose group also presented with this sign. The cyanosis, observed both during and postexposures, was considered to be secondary to methemoglobinemia. Although the cyanosis was initially interpreted as a life-threatening sign, it was an unreliable predictor of mortality. Only methemoglobin levels greater than 50%, but not cyanosis in either species, was reliably associated with acute lethality in rats. To put the animal data into perspective, in humans, cyanosis is initially observed at methemoglobin levels of 8% to 12% (methemoglobin concentrations exceeding 1.5 g/dL). 26 Under these circumstances, cyanosis occurs without changes in oxygen saturation. In contrast, cyanosis due to decreased hemoglobin oxygen saturation (hypoxia) occurs when the absolute level of deoxygenated hemoglobin (reduced hemoglobin) is much greater (4-5 g/dL) and is a life-threatening sign. In fact, other than cyanosis of the skin and mucous membranes, patients with type I congenital methemoglobinemia, who have chronically elevated methemoglobin concentrations as high as 40% of total hemoglobin, are typically asymptomatic and have normal oxygen saturation. Furthermore, in humans with hereditary type I methemoglobinemia, life expectancy is not shortened. 26 Similar to the case in humans, no change in oxygen saturation was observed in either species during these toxicological studies and no alterations in minute ventilation were seen in dogs. Even when high levels of methemoglobin (29%-42% in rats and dogs) were observed in these repeat-dose toxicology studies, only transient changes in compensatory hematological responses commonly associated with chronic hypoxia resulted. It is also notable that no changes indicative of platelet dysfunction (altered coagulation parameters, increased petechial hemorrhages, or decreased platelet numbers) were observed in either species, as currently the literature has conflicting evidence for a role of NO and NO-generating species affecting platelet function. 27,28
Sex-related differences in response to AIR001 exposure were observed. Primarily, female rats had higher methemoglobin levels than did males. At the start of the rat IH toxicology study, female rats exposed to 90 mg/kg had an average of 48% methemoglobin and several died. Because of this morbidity and mortality, the inhaled dose for this group was reduced early in the study (days 8-12 depending on replicate). Despite reducing the inhaled dose in females from 90 to 75 mg/kg/d, exposure in female rats (as measured by plasma AUC) was approximately 10% greater than it was in male rats exposed to 90 mg/kg (Table 3). This greater exposure was also evident in the observed higher methemoglobin and NO3 levels in female rats compared to these parameters in male rats. However, compared to more robust response in male rats, female rats demonstrated no significant decrease in RBCs and only a slight increase in reticulocytes at 8 weeks. Effect of AIR001 to decrease body weight gain at 26 weeks was also less marked in females (8%, P > .05) than in males (12%, P = .01). A sex difference in methemoglobin levels and in NO2 AUC was not observed in dogs. However, greater effects on BP and a lesser effects on HR were seen in male dogs compared to females. In both species, there was no evidence that repeated dosing with AIR001 caused either sensitization or tolerance to the development of methemoglobinemia or cyanosis (rats and dogs) or cardiovascular changes (dogs), unlike the tolerance observed in humans treated with organic nitrates and nitrites. 2
Interestingly, sex differences in the metabolic handling of NO2 and NO3 have also been observed in a clinical study. 29 In that study, baseline plasma NO2 levels in females were significantly higher compared to males and those higher baseline levels were associated with a lower baseline BP. In patients ingesting a NO3 load, resulting in similar NO3 exposure between the 2 sexes, plasma NO2 levels were higher in women than men. Although lower NO2 exposure was observed in men with NO3 administration, both diastolic and systolic BPs were decreased more in men than in women. 29 In the current rat study, although NO2 levels were below the limit of quantitation at predose baseline, methemoglobin levels were higher in females than in males, suggesting a possible difference in NO2 levels. Lower doses of AIR001 in females resulted in higher NO2 and NO3 exposures in females than in male rats. A similar enhanced lowering of BP was also seen in male dogs in the current study compared to female dogs receiving an equivalent exposure to infused AIR001. The mechanistic basis for these possible sex differences is currently unknown.
High-dose IH exposure to rats for 26 weeks revealed that respiratory tissues (other than nasal cavity) were unaffected by treatment. However, a finding of naturally occurring eosinophilic inclusions in the olfactory epithelium of the nasal cavity was observed with increased severity in the mid- and high-dose exposure groups. This finding persisted in the recovery animals as a greater incidence of eosinophil inclusions was observed in high-dose rats compared to the air controls. Importantly, there was no progression of this finding into more severe forms of epithelial damage (eg, necrosis, inflammation, or proliferation of the nasal mucosa) in the recovery rats. The finding of eosinophilic inclusions is consistent with what is considered to be a species-specific adaptive response of the rat nasal epithelium to a high concentration of mildly irritating particulate aerosol. As stated in the Society of Toxicologic Pathology nomenclature document, 30 these eosinophilic inclusions are a common spontaneous finding as rats and mice age and are considered a response to repeated environmental irritation of the nasal passages. 31 Additionally, it is a common response to irritant xenobiotics and at times coexists with more severe effects. 30,31 This finding is not considered relevant to human risk because such lesions in rodents are not commonly observed in similarly exposed dogs or primates due to differences in anatomy and because rats inhale exclusively via the nose while clinical administration of AIR001 will be by oropharyngeal IH, which mostly bypasses the nose. Other than the increased incidence of eosinophilic inclusions in the rat, no other adverse histological findings of any kind, including the rest of the respiratory system, were observed in the main study or recovery rats. There was no evidence of any treatment-related macroscopic or microscopic lesions in any dose group in the dog IV infusion study.
These studies were not designed to assess the risk of carcinogenicity. However, because the studies were for 26 weeks, a special effort was made to examine the tissue for preneoplastic changes and none were observed. Results are consistent with a recent review of studies examining the carcinogenic potential of NO2 in animals and humans. 6 This evaluation, by a diverse committee of scientists, rejected the conclusion that NO2 possessed “sufficient evidence of carcinogenicity,” as previously concluded by the International Agency for Research on Cancer. 32 The committee arrived at this divergent conclusion citing the inappropriate use of older animal studies with significant methodological deficiencies. 6 Indeed, because of these numerous methodological weaknesses in older studies, the US FDA requested that the National Toxicology Program (NTP) conduct a 2-year cancer bioassay of orally ingested sodium nitrite in both rats and mice using 50 animals/sex/group. 13 The results of this definitive study concluded that there “was no evidence of carcinogenic activity of sodium nitrite administered via drinking water in male and female rats and in male mice but there was equivocal evidence in female mice with regard to tumors in the forestomach.” This conclusion was based on the combined findings of both benign and malignant forestomach tumors, which showed a weak relationship with dose. The appropriateness of using the forestomach to extrapolate cancer risk to humans is suspect since humans don’t have an anatomically equivalent organ. Since the issuance of that NTP report, 10 recent animal toxicology studies using sound methodology have looked at the generation of specific tumors using a combination of NO2 and other compounds. In those studies, there was no evidence of tumor formation in the NO2-only control groups. 6
Given the significant acute toxicity and exaggerated pharmacological effects on the cardiovascular system, AIR001 must be administered with appropriate oversight, despite the lack of chronic toxicity findings in these studies. However, the risk of adverse consequences associated with exposure appears modest. In a recent phase 1 study in healthy volunteers, 33 an inhaled dose of 90 mg loaded into the nebulizer was determined to be the MTD. With that dose, Cmax and AUC in humans were 8- and 24-fold less, 34 respectively, than those associated with the no observed adverse effect level (NOAEL) in rats inhaling AIR001 (mid dose, 30 mg/kg/d). In dogs with IV dosing, simulating a worst case scenario of 100% bioavailability of an inhaled dose, there was still a 5-fold safety margin between exposure at the animal NOAEL and that in humans at MTD. 34 In the human study, methemoglobin levels reached only 2.5% in 1 individual at 1 time point when administered the MTD. This compares to average daily levels of 29% to 43% methemoglobin in high-dose-exposed rats and dogs with few, if any, adverse consequences.
In conclusion, the toxicologic findings in these chronic studies of inhaled or IV AIR001 were consistent with acute secondary or exaggerated pharmacologic effects, while effects suggestive of chronic toxicity were not observed. Cyanosis and methemoglobinemia were observed in both rats and dogs, primarily at the high dose. Lethality occurred in several of the high-dose rats. The lethality was considered subsequent to acute hypoxia secondary to methemoglobin levels greater than 50%. Other than acute methemoglobinemia, few significant findings were observed after 26 weeks of either IH (rat) or IV infusion (dog) exposure to AIR001. Given these data, the toxicity of AIR001 appears primarily due to methemoglobinemia; and therefore, the IH NOAEL, after 26 weeks of dosing, is considered to be 30 mg/kg/d in rats, while the IV NOAEL is considered to be 14 mg/kg/d in dogs. In comparison to early pharmacokinetic and pharmacodynamic data in humans, 34 it appears that a sufficient margin of safety exists to allow continued development of inhaled AIR001 in humans.
Footnotes
Acknowledgments
The authors would like to acknowledge the assistance of Adrian Bott and colleagues at BASi who helped develop the nitrite bioanalytical method and Lian Wen and colleagues at ABC Labs who helped develop the nitrate bioanalytical method.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors declared that none of the work was funded by government assistance. All work was solely supported by Aires Pharmaceuticals, Inc.