Preclinical Safety and Biodistribution Assessment of Ad-KCNH2-G628S Administered via Atrial Painting in New Zealand White Rabbits

Abstract

Post-operative atrial fibrillation (POAF) is the most common complication after cardiac surgery. Despite implementation of several pharmacological strategies, incidence of POAF remains at approximately 30%. An adenovirus vector encoding KCNH2-G628S has proven efficacious in a porcine model of AF. In this preclinical study 1.5x10¹⁰ or 1.5x10¹² Ad-KCNH2-G628S vector particles (vp) were applied to the atrial epicardium or 1.5x10¹² vp were applied to the whole epicardial surface of New Zealand White rabbits. Saline and vector vehicle served as procedure controls. Animals were followed for up to 42 days. Vector genomes persisted in the atria up to 42 days, with no distribution to extra-thoracic organs. There were no adverse effects attributable to test article on standard toxicological endpoints or on blood pressure, left atrial or ventricular ejection fractions, electrocardiographic parameters, or serum IL-6 or troponin concentrations. Mononuclear infiltration of the myocardium of the atrial free walls of low dose, but not high dose animals was observed at 7 and 21 days, but these changes did not persist or affect cardiac function. After scaling for heart size, results indicate the test article is safe at doses up to 25 times the maximum proposed for the human clinical trial.

INTRODUCTION

Approximately 300,000 to 600,000 cardiac surgeries are performed each year in the United States.^1,2 Post-operative atrial fibrillation (POAF) is the most common complication after cardiac surgery, with a widely varying incidence (10 – 60%), depending on the type of cardiac surgery (coronary artery bypass graft, valve replacement or repair, or a combination of the two procedures).^1,3,4,5 Incidence is also impacted by advanced age and pre-existing conditions including coronary artery disease, heart failure, and history of atrial fibrillation (AF).⁶ Timing to first incidence and window of occurrence may extend to two weeks post cardiac surgery.⁷ In some patients POAF may initially occur within the first 18 hours post-surgery, but peak time of initial occurrence is approximately 2-3 days after surgery.^4,5 Data indicate the incidence then steadily declines through 10 days post-surgery.⁵

Studies have consistently shown an association between POAF and subsequent increased risks of stroke, myocardial infarction, and other adverse health outcomes including death.^4,8,9,10 In one large international study, a single episode of POAF predisposed individuals to occurrences of myocardial infarction, congestive heart failure, pulmonary edema, and surgical site infections. Patients with more than one POAF occurrence had significant increases in incidences of myocardial infarction, encephalopathy, decline in mental state, increased stroke score, renal failure, adult respiratory distress syndrome, and infections.⁵

The underlying mechanisms of POAF are multifactorial. Several pharmacological strategies have been employed to prevent POAF, including beta blockers to counter post-operative catecholamine response, amiodarone to stabilize atrial electrical function, or various agents to reduce inflammation.⁶ These pharmacologic strategies have associated toxicities, requiring that they be implemented on a patient by patient basis. These strategies may counter the acute responses to cardiac procedures, but they do not address the pre-existing status of the atrial substrate per se. Many of these prophylactic measures have been used for several decades, but the incidence of POAF remains high.

We have developed an adenovirus vector encoding KCNH2-G628S, a dominant negative mutant of the I_Kr potassium channel α-subunit. When painted on the atrial epicardium in a porcine pacing model of AF, Ad-KCNH2-G628S decreased AF burden after surgery. Mechanistically, therapeutic efficacy comes through prolonging the atrial monophasic action potential duration and preventing re-entry.^7,11 Major advantages of the gene therapy approach in contrast to the pharmacological approaches include: 1) the ability to deliver the therapy specifically to the target tissue and thus reduce risk of off-target toxicity (e.g. ventricular pro-arrhythmia from antiarrhythmic drugs); 2) the possibility to attack the disease from inside target cells eliminating pharmacokinetic and pharmacodynamic factors that can reduce drug efficacy, and 3) the ability to directly target disease mechanisms in ways that are not possible with small molecules.

We performed an Investigational New Drug-enabling preclinical safety and biodistribution study in healthy New Zealand White rabbits to: 1) determine the safety of Ad-KCNH2-G628S administered by atrial painting at two dose levels; 2) assess the potential toxicity of Ad-KCNH2-G628S administered by painting the entire heart epicardium, at the high dose, representing the “worst case scenario” with respect to heart tissue exposure; and 3) evaluate vector retention in the atria and systemic distribution. Rabbits were used because the rabbit heart is large enough to reliably undergo the painting procedure using the same methods that would ultimately be used in the clinic. Further, we have demonstrated that rabbit ventricular myocytes are more easily infected by adenovirus than other large mammalian (e.g., pig and dog) ventricular myocytes, so the rabbit is an appropriately aggressive test subject for potential ventricular toxicity in the preclinical setting. In addition, rabbits have electrophysiological properties that are very similar to humans. Of particular relevance to this study, the role of KCNH2 in repolarization is the same. ^12–19

The rationale for the administered low and high doses in this study is as follows. A dose of 5 x 10¹¹ Ad-KCNH2-G628S vp/heart directly applied to the atria in the porcine model of atrial fibrillation was effective in reducing the incidence of AF – significantly increasing the time in sinus rhythm compared to controls.⁷ Since swine hearts are considered for human cardiac xenotransplation we implemented a 1:1 scaling for human to swine heart size.²⁰ Therefore, the human effective dose is expected to be 5 x 10¹¹ vp/heart. Based on data on human and rabbit body weight to heart weight data, the scaling factor of human to rabbit is 33.^21,22 Therefore, the effective dose in rabbits is 1.5 x 10¹⁰ vp/heart. This is the low dose in preclinical safety study described herein. The high dose in this preclinical study in rabbits is 100-fold higher than the expected effective dose and provides a large safety margin for dosing in the human clinical trial.

MATERIALS AND METHODS

Animals

New Zealand White rabbits (Strain Code:571 (OAKWOOD) were purchase from Charles River Laboratories and Covance Research Products (42 day euthanasia cohort males). All animal study procedures were approved by the Lovelace Biomedical Institutional Animal Care and Use Committee. The animals were quarantined and housed according to Standard Operating Procedures consistent with the Office of Laboratory Animal Welfare Guide for the Care and Use of Laboratory Animals.²³ The study was also conducted in accordance with the Basic&Clinical Pharmacology and Toxicology policy for experimental and clinical studies.²⁴ However, blinding was not incorporated into this pre-clinical safety study as the laboratories performing the endpoint analyses were Contract Research facilities with no vested interest in the therapy.

The animals were approximately 17 – 19 weeks old and weighed 2.34 – 4.26 kg on the day before vector application. Animals were received in 6 cohorts, by sex and euthanasia time point. Animals in each cohort were randomized by weight into dose groups. Animals were uniquely identified by alphanumeric numbers encoded in a transponding microchip (Trovan, Santa Barbara, CA) placed subcutaneously in the upper hindlimb.

Vector and Vehicle Controls

AdKCNH2-G628S is a first- generation adenovirus vector containing the cytomegalovirus immediate/early promoter/enhancer (CMV prom), the potassium channel mutant KCNH2-G628S, and the simian virus 40 (SV-40) poly-A signal. The vector backbone is a serotype 5 adenovirus, with the two inverted terminal repeats and the full wild type sequence deleted of E1 and E3 genes and with insertion into the vacated E1 site of the CMVprom-KCNH2-G628S-SV40 poly-A construct. The vector was produced at the Belfer Gene Therapy Core Facility, Weill Cornell Medical College, New York, NY. Initial and stability characterization endpoints included evaluation of particle concentration, sterility, endotoxin, pH, and mycoplasma. Vector was vialed at two concentrations needed to provide the desired applied dose to the rabbits.

The vector vehicle was 1x sucrose (88 mM sucrose, 150 mM NaCl, 21mM MgCl₂, 10 mM Tris HCl, pH 8.0) produced at the Belfer Gene Therapy Core. Pre-study and stability characterization endpoints included sterility, endotoxin, appearance, and pH.

Poloxamer Vehicle [Kolliphor^® P 407 28.75% (w/v) in saline] was formulated by SRI International, Menlo Park, CA. Sterile saline (0.9% Sodium Chloride, USP; Cat # 7908303; NDC No. 0409-7983) was purchased from Hospira, Inc., Lake Forest IL.

Test Article Formulation

The poloxamer vehicle and the required dose of AdKCNH2-G628S were prepared to a total volume of 1.0 ml for each rabbit just prior to application by the surgical team. The diluted poloxamer vehicle and the prepared vector solutions were warmed at 37±2 °C to become a firm gel for painting. The saline vehicle (1 mL) was applied directly without formulation.

Test Article Dosing

The rabbits were sedated, intubated and prepared for the sterile open-chest procedure. Median sternotomy and pericardiotomy were performed using standard techniques. During vehicle or vector application, the heart was manipulated to expose all epicardial surfaces of the atria. The vehicle and test articles were “painted” onto the atria using a sterile rectangular-tip foam applicator (Puritan Medical, product 25-1607 1PF SC). Each atrium (left and right) or each side of the whole heart was painted twice for 60 ± 5 seconds each with 80 ±10 seconds between each application. The 1.0 ml total volume of vector/poloxamer/saline was used, so each atrium (or each side of the whole heart) received approximately 0.50 ml of dose formulation. The vector was applied to the atria, the ventricles, and the great vessels of the whole heart dose group.

Pain management was as follows: The animals received a fentanyl patch (12.5 μg/hr - 25 μg/hr) that was applied one day prior to the cardiac surgery and was removed two days following the surgery. Prior to the animals recovering from surgical anesthesia, each animal received 0.2 mg/kg Meloxicam, subcutaneously. Meloxicam (0.2 mg/kg PO) was then administered once daily for 4 days. Buprenorphine (0.01 – 0.05 mg/kg, SC) was administered twice daily beginning on the day the fentanyl patch was removed and continuing for two additional days. After the epicardial surgery, animals were monitored for clinical signs at least twice daily for at least 6 days.

Study Design

Animals within each cohort were randomized by weight into dose groups as shown in Table 1. The day of vehicle or vector dosing was defined as Study Day 0 (SD0).

Table 1.

Experimental Design.

Group	Group Designation	Treatment (Dose)a	M/F per Euthanasia Cohort
			7 ± 1	21 ± 2	42 ± 2b
1	Saline Control	Saline Alone	3/3	3/3	3/3
2	Vehicle Control	20% Paloxamer	5/5	5/5	5/5
3	Low Dose	ADKCNH2-G628S in 20% poloxamer (1.5 x 1010 vp/animal)	5/5	5/5	5/5
4	High Dose	ADKCNH2-G628S in 20% poloxamer (1.5 x 1012 vp/animal)	5/5	5/5	5/5
5	Whole Heartc	ADKCNH2-G628S in 20% poloxamer (1.5 x 1012 vp/animal)			5/5

^aDosing day was designated as Study Day 0 (SD0).

^bIn 42 day male cohorts, 4 M in the vehicle control group, 5 M in the high dose group, and 4 M in the whole heart group received telemeters. For the 42 day female cohort, 4 F in vehicle control group, 3 F in low dose group, 4F in high dose group and 4 F in whole heart group received telemeters.

^cThis group of animals were dosed by painting vector onto the epicardial surface of the whole heart as a worst case scenario” for the vector dosing by atrial painting

For Groups 2-5 animals designated for telemetry, telemeters were surgically implanted subcutaneously in the abdominal region at least 10 days prior to cardiac surgery. After telemeter placement, animals were allowed to recover for at least 2 days and then pre-study baseline electrocardiograms were recorded continuously before animals were dosed. After the telemetry surgery, each rabbit was monitored twice daily for a minimum of 4 days by the veterinary staff.

Following the dosing, endpoint measurements included: echocardiography, electrocardiography, blood pressure, clinical signs, body weight, serum chemistry (at scheduled euthanasia time point, hematology (at scheduled euthanasia), the cardiac marker troponin (cTnI), interleukin -6 (IL-6).

On the day of scheduled euthanasia, terminal body weights were recorded and the animals were prepared for thoracotomy as described for the cardiac surgical procedure and open chest apical view echocardiograms were recorded. The animals were euthanized by intravenous injection of an overdose of Euthasol^®. Each animal received a complete necropsy with macroscopic lesions and organ weight recorded. Selected tissues were harvested for evaluation of histopathology and vector biodistribution. Animals dying before their scheduled euthanasia time point also received a complete necropsy.

Clinical Signs and Body Weight

Detailed clinical observations were performed before randomization, SD-1, twice daily on SD 1-6, every other day during the second and third week following cardiac surgery for the SD21 and SD 42 euthanasia cohorts, and then once per week until the SD42 cohort was euthanized. Observations included, but were not limited to, activity level, respiratory pattern, indicators of pain such as guarding and response to handling, licking or biting at the incision, abnormal posture, output and quality of urine and feces, and appetite (qualitative).

Body weight was collected at randomization, on SD-1, SD 2, SD 4, SD 6 and weekly thereafter until scheduled euthanasia.

Detailed observations and body weight data were entered into Provantis^™ preclinical software (Instem LSS ltd., Staffordshire, UK). This system is not set up to record data in a blinded fashion.

Blood Pressure

Blood pressure was measured using a DRE Waveline EZ Patient Monitor (DRE, Inc., Louisville, KY) equipped with Suntech #2 neonatal disposable blood pressure cuffs (DRE, Inc). Pressure was measured on animals in the 21-day (female) and 42-day cohorts (males and female). Baseline blood pressure was measured for a consecutive three days before epicardial painting and then once daily on SD 1 and SD3-SD6, and weekly thereafter. The blood pressure measurement was performed on awake animals manually restrained for the procedure.

Echocardiography

Echocardiography was performed using a GE Vividi electrocardiograph equipped with a 10S-RS-4.5MHz-11.5MHz probe. Open chest echocardiograms were obtained on the day of scheduled euthanasia. Multiple four chamber apical and two chamber apical views were obtained for each animal for analysis.

Left ventricular and left atrial end diastolic and end systolic volumes were measured. The values were used to left ventricular and atrial ejection fractions. In some cases, due to adhesions of the heart to the sternum or chest walls or anatomical distortion of the chamber walls during echo acquisition, readable (measurable) views of left ventricle, left atrium, or both could not be obtained. In those instances, the views (4 chamber or 2 chamber) were documented as “not interpretable”.

Telemeter Placement, Data Acquisition, and Analysis

Major adverse endpoints of concern for this therapy are the potential for QTc interval prolongation and the development of potentially lethal arrhythmias (i.e., ventricular fibrillation). Electrocardiographic waveforms were collected on the 42-day cohort rabbits via telemetry. A subset of the 42-day male and female cohorts were surgically implanted with easy TEL D L-HBTA telemetry devices (emka Technologies, Falls Church, VA). Data acquisition was achieved using emka IOX2 software.

Recordings were collected continuously prior to epicardial vector application through 21 days post-surgery (SD 21) and then for half hour intervals to the SD42 euthanasia time point.

The data for each animal were placed in individual electronic files by date. For one baseline day and SDs1, 2, 3, 5, 7, 10, 14, 17, 21, 24, 28, 32, 35, 39 and 42, RR interval (heart rate), PR interval, QRS interval and QT measurements were manually made on a target of 10 consecutive wave forms identified during a quiet period where no disturbances due to study activities were likely to occur. QT correction was performed using the formula (QTc = QT/RR^0.72) derived by Kijtawornrat et al. in rabbits.²⁵

For one pre-cardiac surgery day (target day −2 or day −1), and on SD1 through SD10 and SD14, 24-hours of waveforms were manually scanned for arrythmias. For SDs 21, 24, 28, 32, 35, 39 and 42, all waveforms in the 30 min acquisition window were scanned for arrythmias.

Clinical Pathology

For hematology, whole blood was collected and transferred into a K3-EDTA tube. Samples were analyzed by ADVIA^™ 120 Hematology System (Siemens Medical Solutions Diagnostics, Tarrytown, NY).

Serum chemistry parameters were measured using a Hitachi Modular Analytics Clinical Chemistry System (Roche Diagnostics, Indianapolis, IN).

Selected hematology and serum chemistry parameter endpoints and baseline values obtained from the saline and vehicle control animals are provided in Tables S4 and S5, respectively.

Interleukin 6

Blood was collected on SD1 and harvested serum was saved at ≤ −60 °C. The IL-6 was assayed by Charles Rivers Laboratories (Shrewsbury, MA) using a commercially available ELISA kit (cat# NBP2-62858, R&D system). Assay samples were run in duplicate. Data obtained from each euthanasia cohort were pooled by sex and dose group for statistical evaluation.

Troponin

Blood was collected for evaluation of cTnI concentration at baseline (SD-1), and all surviving animals at SD 1, SD7, SD 21 and SD42. Harvested serum was stored at ≤−60°C until submitted on dry ice to Charles River Laboratories (Shrewsbury, MA)

The samples were analyzed using a Rabbit cTnI Elisa kit (Life Diagnostics, Inc, West Chester PA; cat # CTN1-10-HS). The assay was qualified before use on study samples. The lower limit of quantitation (LLOQ) was 0.625 ng/mL. Data obtained from each cohort with quantifiable troponin concentrations were pooled by sex and dose group for statistical evaluation.

Euthanasia, Necropsy, Macroscopic and Microscopic Examination

At each scheduled euthanasia time point a terminal body weight was collected. The animals were anesthetized by IM injection of ketamine/xylazine and blood was collected from a peripheral vein for quantitation of vector content and clinical pathology endpoints. The animals were then intubated, thorax opened, and echocardiograms were recorded. Animals were then euthanized by intravenous administration of an overdose of Euthasol^®.

Animals were examined for macroscopic pathology. Tissues listed in Table S6 were examined histologically and assayed for vector content. Adrenals, brain, heart, lung, liver, kidney, spleen, testes and ovary, thymus and uterus/cervix weights were recorded.

Tissue samples for histological evaluation were fixed in 10% neutral buffered formalin or Bouins’ solution (testes and eyes). Tissues were trimmed, embedded in paraffin, processed to slides and stained with Hematoxylin and Eosin. Tissues listed in Table S6 from the high and vehicle control groups euthanized at 7 and 42 days were examined for microscopic changes. Hearts from the low dose groups were analyzed to identify any vector associated changes independent of the surgical procedure. Based upon the possible presence of Ad-KCNH2-G628S vector-related microscopic findings in the left and right atrial free wall on SD7, the left and right atrial free wall and left and right auricles from SD21 rabbits were also examined.

Vector Genomes in Blood and Tissues

Blood and tissue samples were snap frozen in liquid nitrogen and then stored at < −60° C until DNA extraction. Genomic DNA was extracted using the DNeasy Blood and Tissue Kit (Qiagen, Germantown, MD) according to the manufacturer’s instructions. Sample DNA concentration was evaluated by measuring the absorbance at 260nM (A260) and the DNA quality was determined by the A260/A280 ratio using a Biotech μQuant plate reader.

Vector genome content of the DNA was determined using a Taqman based assay qualified at Lovelace before use on samples. Qualification endpoints included standard curve, precision, accuracy, specificity and LLOQ. The assays were run in a background matrix of rabbit liver genomic DNA. Vector plasmid (pAd Herg G628 S) was used to generate the standard curve and positive controls run with each assay plate. The primers and probes were obtained from Ambion Life Technologies/ThermoFisher. The forward primer sequence was: 5’-GTCAATGGGTGGAGTATTTACGG-3’. The reverse primer sequence was: 5’-AGGTCATGTACTGGGCATAATGC-3’, and the probe Sequence was: CAAGTGTATCATATGCCAAGTACGCCCCC-3’.

The samples were assayed in triplicate using a Stratagene Mx 300Sp System. The Dye was FAM and quencher TAMRA. The cycling sequence was as follows: UDG Incubation: 50°C for 2:00 min; Initial Denaturation: 95°C for 10:00 min; followed by 45 cycles of denaturation (95°C, 00:15 min) and Anneal/Extension (60°C for 1:00 min). Sample and standard curves were un in triplicate. The Lower Limit of Quantitation was 50 vp/μg genomic DNA.

Statistical Analysis

Statistical calculations were performed using the SAS^® software system (version 9.4, Cary, NC), GraphPad Prism^® (version 5.04, La Jolla, CA), or using statistics programs in the Provantis^™ software package. For quantitative data (e.g. body weight, organ weight, clinical pathology, etc.), group mean and standard deviations were calculated and presented as appropriate. In general, for comparison of more than two groups, an overall trend among test groups were tested by an analysis of variance (ANOVA), and if significant, followed by a post-hoc test (e.g. Tukey’s or Dunnett’s test); for comparison of two groups, a t-test was performed. Significance levels were set at p < 0.05.

For assessment of the blood pressure and ECG parameters, a two-way ANOVA with terms of group, day, and group/day interaction, was used to assess differential exposure group effects over time by evaluating the interaction between exposure groups and day of blood pressure or ECG parameters. Generalized estimating equations (GEE) with exchangeable correlation structure were used to specify the correlation of repeated measurements for the same animal. If the p-value of the group/day interaction for the ANOVA model was significant (p ≤ 0.05), statistically significant trends in the parameter over time between the Vehicle Control group and other groups were tested.

Means and 95% confidence interval were used to report the biodistribution data.

RESULTS

Morbidity and Mortality:

Among 121 rabbits dosed, all but six of the animals survived to their scheduled euthanasia time point. Three of animals dying (one Group 2 male, one Group 4 female, and one Group 5 male) were replaced to restore group size. The early deaths or moribund euthanasia occurred between SD0 (day of cardiac surgery) and SD6. Grossly observable lesions of these animals at necropsy indicated that the early deaths were attributable to the invasive surgery, or early complications following surgery, not to the vector. One low dose female was euthanized on SD6 due to persistent self-directed destructive behavior.

Clinical Signs of Toxicity

A summary of the animal clinical status scores in the first 6 days post surgery is provided in Figure S1. Following 7 days post surgery, there were no clinical signs attributable to test article. During this period alopecia, skin lesions, and ear bruising (presumably from blood draws) were noted in a few male and female animals across the dose groups.

Body Weight

There were no statistically significant differences in body weights among the control and vector dose groups in the 7-, 21-, or 42-day male and female cohorts when one way ANOVA was applied at each measurement timepoint. Animals in all groups lost weight following the cardiac surgery, but slowly regained weight after SD6. Body weights for male and female animals in the 42-day cohorts are shown in Figure S2.

Blood Pressure

Following cardiac surgery, blood pressure data were highly variable. When group means were compared using one-way ANOVA and Dunnett’s test for all three cohorts by study day, occasional group differences were found in either systolic or diastolic measurements between vector dose animals compared to the saline or vehicle controls. The extent of changes in blood pressure values was not associated with vector dose group, and when differences were noted, they were not consistent from time point to time point.

When the data were analyzed using two-way ANOVA, there were no significant difference in systolic or diastolic blood pressure change trend for any vector dose group compared to either the saline control group or vehicle control group. Systolic and diastolic blood pressure values for are provided in Figures S3a and S3b (42 day females) and Figures S4a and S4b (males).

Echocardiography

Left ventricular and left atrial ejection fractions are summarized in Tables 2 and 3, respectively. Only the group mean left atrial ejections for females showed a decreasing trend among dose groups by ANOVA, but no individual group mean values were statistically different from that of the vehicle control. When data for males and females were combined to increase statistical power, group mean left atrial ejection fractions again showed a trend, with no individual group mean values being different from the vehicle control. There appears to be no statistical or biological effect of test article administration or transgene expression on heart function as measured by left atrial or left ventricular ejection fractions.

Table 2.

Summary of Left Ventricular Ejection Fractions

	Left Ventricular Ejection Fractions mean ± SD (n)
Dose Group	7 Day Males	7 Day Females	7 Day M/F	21 Day Males	42 Day Males	42 Day Females	42 Day M/F
Saline Control	42.68 ± 12.63 (2)	28.45 ± 1.67 (3)	34.14 ± 10.10 (5)	40.97 ± 8.59 (2)	34.36 ± 2.62 (2)	No dataa	34.36 ± 2.62 (2)
Vehicle Control	32.70 ± 11.25 (4)	33.66 ± 4.52 (4)	33.18 ± 7.96 (8)	23.47 ± 5.90 (2)	34.73 ± 9.99 (3)	43.58 ± 10.95 (4)	39.79 ± 10.06 (7)
Low Dose Vector	41.09 ± 9.02 (5)	40.12 ± 12.52 (4)	40.66 ± 9.99 (9)	30.51 ± 3.47 (3)	37.25 ± 3.35 (3)	40.31 ± 7.92 (5)	39.17 ± 6.45 (8)
High Dose Vector	30.37 ± 6.66 (4)	35.04 ± 10.51 (3)	32.37 ± 8.08 (7)	30.68 ± 1.43 (3)	39.29 ± 6.76 (3)	42.62 ± 10.35 (3)	40.11 ± 8.42 (6)
Whole Heart			NA		33.57 ± 8.40 (4)	38.76 ± 3.66 (3)	35.88 ± 9.06 (8)

^aNo interpretable echocardiographic views were available.

Table 3.

Summary of Left Atrial Ejection Fractions

	Left Atrial Ejection Fractions mean ± SD (n)
Dose Group	7 Day Males	7 Day Femalesa	7 Day M/F	21 Day Males	42 Day Males	42 Day Females	42 Day M/F
Saline Control	67.05 ± 16.69 (2)	68.12 ± 8.93 (3)	67.23 ± 9.54 (5)	41.37 ± 25.53 (2)	64.64 ± 10.93 (2)	No Datab	64.65 ± 10.92 (2)
Vehicle Control	62.19 ± 5.84 (5)	57.66 ± 4.60 (4)	60.18 ± 5.54 (9)	58.83 ± 10.85 (2)	48.92 ± 9.22 (3)	42.19 ± 12.13 (3)	45.56 ± 11.82 (6)
Low Dose Vector	69.32 ± 18.54 (5)	70.69 ± 13.42 (4)	69.93 ± 15.49 (9)	44.03 ± 10.41 (2)	54.06 ± 15.57 (3)	42.59 ± 11.26 (5)	46.85 ± 13.35 (8)
High Dose Vector	60.84 ± 4.89 (4)	46.39a± 2.77 (3)	54.62 ± 8.58 (7)	46.88 ± 10.79 (3)	62.17 ± 9.23 (4)	57.66 ± 1.91 (3)	60.24 ± 7.04 (7)
Whole Heart					42.20 ± 12.82 (4)	57.80 ± 10.59 (3)	47.97 ± 14.82 (7)

^aANOVA detected in vales among all dose group, but there were no group differences between vector dosed and vehicle control animals identified in the Dnnetts test (p <0.05).

^bNo interpretable echocardiographic views were available.

Electrocardiography

Baseline heart rate, PR interval, QRS interval, and QTc interval values for males and females are provided in Table S1. Group mean values for male and female animals combined for the vehicle control high dose and whole heart for heart rate, PR, QRS and QTc over time are shown in Figures 1 – 4.

Figure 1.

Group mean (± SD) heart rate (beats per minute) for male and female rabbits combined for the vehicle control, high dose and whole heart animals. N = 7 vehicle controls and whole heart groups. N = 9 for the high dose group.

Figure 4.

Group mean (± SD) QTc interval (msec) for male and female rabbits combined for the vehicle control, high dose and whole heart animals. N = 7 vehicle controls and whole heart groups. N = 9 for the high dose group.

Heart rate and the PR, QRS, and QTc intervals were quite variable among the animals. Among females, there were no statistical differences among the vector treated dose groups (low dose, high dose and whole heart) compared to the vehicle controls at each individual time point measured. For males, the group mean pre-operative baseline QRS and QTc values were significantly decreased in Group 5 animals (where vector was applied to the whole heart) compared to the Group 2 vehicle controls. However, there were no significant differences in any parameter among vector dosed and vehicle control animals at each time point measured after the cardiac surgery. The two- way ANOVA assessment of the ECG parameters also showed no significant difference in any parameter between the vehicle control, high dose and whole heart groups over time. Since data for the low dose group were only available for females, this group was not included in the two- way ANOVA analysis.

Approximately 220 hours of telemetry data were obtained from each of the Groups 2 – 5 animals and manually reviewed for the presence of arrhythmias. Primarily, normal sinus rhythm was observed. In some cases, other rhythms or waveforms were encountered at baseline or in vehicle control as well as vector dosed animals post dosing. In particular, there were extended periods of bradycardia noted both before and after surgery. Almost every rabbit had periods of abrupt onset bradycardia, with varying extent and duration of heart rate reduction that were not consistent from one episode to the next for any particular rabbit. Interestingly, these episodes often occurred between midnight and 6 AM. Rapid heart rates were common in the initial days post-surgery, with the PR interval notably shortened. Subsequently, rapid increases in heart rate were also recorded, and sometimes related to activity but not necessarily external stimulation. For example, the rabbits would spontaneously run in their cages in the absence of obvious cause.

Continuous telemetry showed periods of increased heart rate variability and sinus bradycardia before and after surgery in several animals from several groups, with no between-group differences. This rhythm was accentuated in one high dose female who had transient episodes of sinus bradycardia with ventricular escape rhythms on days one and two after vector administration. Transgene expression was unlikely to have caused these effects because the arrhythmias occurred before significant transgene expression would be expected, based on efficacy results in swine administered this vector by epicardial painting. ⁷ The rhythm was transient and there were no associated adverse events. Otherwise, there were no arrhythmias affecting survival in any animals of any group, including the whole heart painted animals. Specifically, there were no arrhythmias during the time period of expected transgene expression.

Serum IL-6

Data within the male and three female cohorts were pooled for statistical analysis, as all sample were collected on SD1. Statistical analysis (one way ANOVA) indicated no effect of vector administration on IL-6 concentration in either males or females. Data are provided in Table S2.

Serum Troponin

Prior to dosing (SD-1), all but 5 serum samples were below the LLOQ (0.625 ng/mL). Troponin concentrations in serum at SD1 were variably increased. Data within the male and female cohorts at SD1 were pooled for statistical analysis. At SD1, there was no statistical difference between groups for increasing troponin concentration for either males or females (one way ANOVA). Data for Day 1 post dosing are provided in Table S3.

On SD7, troponin serum concentrations were below the limit of detection in all animals. On SD21, five females in the vehicle control and low dose groups had low, but quantifiable troponin concentrations (0.660 – 0.880 ng/ml). At SD42, one low dose and one high dose female had 0.844 – 2.030 ng/mL troponin in serum. Since no female animal had quantifiable troponin concentrations at both SD21 and SD42, and there were no male animals with detectable troponin concentrations after SD1, the positive data in females at SD21 and SD42 were considered sporadic and not related to vector administration.

There was no scheduled euthanasia on SD1. Therefore, it is not possible to correlate the positive troponin values with microscopic findings in the heart. But since there were no statistical differences in troponin serum concentration among the exposure groups, it is likely that the positive values resulted from the surgical procedure.

Clinical Pathology

Control animal data for select hematology and serum chemistry parameters are provided in Tables S4 and S5, respectively. There were no changes attributable to the vector administration in hematology or clinical chemistry parameters in any cohort or dose group at any euthanasia time point examined.

Organ weights, Organ to Body Weight Ratio and Organ to Brain Weight Ratio

On SD7, male rabbits in the high dose group had a slight, but statistically significant, decrease in mean absolute spleen weight, spleen-to-brain and spleen-to-body weight ratios (Table 4). There were no macroscopic or microscopic correlates and the difference may have been due to variations in degree of exsanguination at necropsy.

Table 4.

Organs with Statistically Significant Changes in Weight, Organ to Body and/or Organ to Brain Weight. Mean ± SD (n)

		Male Spleen Weight 7D			Female Liver Weight at 21 D
	Spleen (g)	Spleen/Body (%)	Spleen/Brain (%)	Liver (g)	Liver/Body (%)	Liver/Brain (%)
Saline Control	1.125 ± 0.041 (3)	0.040 ± 0.003 (3)	11.90 ± 0.72 (3)	80.272 ± 11.612 (3)	2.203 ± 0.347 (2)	823.97 ± 152.37 (3)
Vehicle Control	1.477 ± 0.491 (5)	0.049 ± 0.014 (5)	14.80 ± 4.32 (5)	82.129 ± 13.809 (5)	2.297 ± 0.265 (4)	810.01 ± 174.88 (4)
Low Dose Vector	1.347 ± 0.189 (5)	0.046 ± 0.011 (5)	14.60 ± 2.16 (5)	89.349 ± 21.258 (4)	2.529 ± 0.410 (3)	909.30 ± 195.38 (4)
High Dose Vector	0.889 ± 0.139 (5)a	0.030 ± 0.005 (5)a	9.22 ± 1.38 (5)a	105.665 ± 7.943 (5)	2.920 ± 0.177 (5)b	1006.68 ± 72.14 (5)

^aSignificantly less than vehicle control. ANOVA with Dunnett’s post test,, p <0.05.

^bSignificantly greater than vehicle control. ANOVA with Dunnett’s post test,. p <0.05.

On Study SD21, there was a slight increase in mean absolute liver weight, and liver-to-body, and liver-to-brain weight rations for high dose females (Table 4). The increase was statistically significant only for the liver-to-body weight ratio. There were no macroscopic correlates or correlating alterations in clinical chemistry parameters. Therefore, the increased liver weight may have been unrelated to vector administration. Per the study protocol, the liver was not examined microscopically on SD21.

There were no Ad-KCNH2-G628S vector-related effects on organ weight parameters in male or female rabbits on SD42.

Macroscopic Pathology

All macroscopic observations at the scheduled euthanasia time points were related to the surgical procedure or were consistent with spontaneous background findings in New Zealand white rabbits.

Microscopic Observations

Microscopic findings in the heart on SD7 that may have been related to administration of Ad-KCNH2-G628S consisted of an increase in the incidence and severity mononuclear infiltrates in the myocardium of the left and right atrial free wall of some rabbits in the low dose vector group. The incidence and severity were greater in males than females. The findings are summarized in Table 5.

Table 5.

Microscopic Findings Possibly Related to Ad-KCNH2-G628S Administration on SD7.

Sex:	Male			Female
Group:	2	3	4	2	3	4
Dose (virus particles/animal):	0 (vehicle)	1.5x1010	1.5x1012	0 (vehicle)	1.5x1010	1.5x1012
Number examined:	5	5	5	5	5	4
Heart, Atrium, Left, Free Wall
Infiltration, Mononuclear, Myocardium
Mild	0	1	0	0	1	0
Moderate	0	1	0	0	0	0
Heart, Atrium, Right, Free Wall
Infiltration, Mononuclear, Myocardium
Minimal	1	1	0	0	1	1
Mild	0	2	0	0	1	0
Moderate	0	1	0	0	0	0

Minimal myocardial mononuclear infiltrates consisted of one or two focal aggregates of approximately 10 or fewer mononuclear cells; these are occasionally observed as a background finding in rabbits and are often exacerbated by stress.²⁶ Mild and moderate mononuclear cell infiltrates consisted of progressively increasing numbers of mononuclear cells involving a greater area of the myocardial tissue examined, with mild infiltrates involving approximately 5-20 % of the tissue examined, and moderate infiltrates involving approximately 21-40 % of the tissue examined. The reason for the lack of an increase in mononuclear cell infiltrates in the Group 4 animals given 1.5x10¹² virus particles/animal (atrial painting only) on SD7 is uncertain. Increased mononuclear cell infiltrates in the myocardium on SD7 was not adverse, as the severity was not sufficient to disrupt cardiac function.

Based upon the possible presence of Ad-KCNH2-G628S vector-related microscopic findings in the left and right atrial free wall on SD7, the left and right atrial free wall and left and right auricle were examined microscopically on SD21. Among the 21 day animals, there was an increase in the incidence and severity of mononuclear cell infiltrates in the myocardium of the right atrial free wall and right auricle of some male rabbits in the low dose group. These are summarized in Table 5.

There were no Ad-KCNH2-G628S vector-related microscopic findings in male or female rabbits on SD42.

Biodistribution

On SD1, quantifiable amounts of vector were present in the blood of high dose, but not low dose male and female rabbits. Vector concentrations were 3.12 to 3.67 log10 vp/μg DNA and 3.17 to 3.68 log10 vp/μg DNA in blood of males and females, respectively. Except for one low dose male (2.15 log10 vp/μg DNA) vector content in blood in low dose males and females was below the assay limit of quantitation (50 vp/μg DNA [1.7 log10 vp/ μg DNA]; BLQ). No detectable vector was present in high dose males or females by SD7.

Vector content in left and right atria as a function of time post cardiac painting is shown in Table 7. At SD7, vector concentrations in left and right atria of high and low dose males were similar to each other, with overlapping 95% confidence intervals. As with males, the vector content of the right atria were similar for females in the low and high dose groups.

Table 7.

Vector Copy Number in Auricles as a Function of Time Post Dosing

vp/μg DNA log 10 mean; 95% Confidence Interval (n = 5 unless otherwise noted)
Dose Group	SD7		SD21		SD42
	Right Auricle	Left Auricle	Right Auricle	Left Auricle	Right Auricle	Left Auricle
Males
Low	4.44 3.89 – 5.00	3.51 3.01 – 4.01	3.29 2.43 – 4.16	2.34 1.77 – 2.92	2.94a 2.52 – 3.36	1.94a 1.68 – 2.20
High	3.82 2.91 – 4.72	3.06 2.12 – 4.00	2.86 2.01 – 3.72	1.98 1.56 – 2.40	2.46 1.92 – 2.99	1.75 1.65 – 1.86
Whole Heart	NA	NA	NA	NA	2.20 1.38 – 3.02	1.75 1.66 – 1.84
Females
Low	4.92 4.72 – 5.12	3.96 3.58 – 4.35	3.09a 2.60 – 3.58	2.80a 2.57 – 3.03	2.81 2.52 – 3.10	2.23 1.94 – 2.52
High	4.27b 3.55 – 4.99	2.22a 1.78 – 2.65	2.75 2.14 – 3.36	2.22 1.79 – 2.65	2.02 1.75 – 2.28	1.77 1.63 – 1.92
Whole Heart	NA	NA	NA	NA	1.74a 1.69 – 1.79	1.77 1.64 – 1.89

^an = 4

^bn = 2

Vector content of the atria decreased by approximately an order of magnitude between SD7 and SD21. At SD42, quantifiable vector was present in some, but not all atria of both the low and high dose vector groups. Vector content of the group 5 whole heart animals at SD42 was, with few exceptions at or close to the lower limit of quantitation.

Quantifiable but variable amounts of vector were also present near the heart base of the left and right ventricles, the intraventricular septum, and lateral walls of the left and right ventricles of some, but not all, low and high dose group rabbits at SD7. Quantifiable amounts of vector were also present in the left and right ventricular apices of some high dose animals at SD7, although vector content was variable among animals. Quantifiable vector was present in the heart base of the left and right ventricles at 21 and to a lesser extent, at 42 days post dosing. The vector in these tissues may have resulted from overlap of the original vector application beyond the atria.

At SD7 there was no systemic distribution of vector to liver, kidney, spleen, or gonads among high dose male and female rabbits, or in the corresponding tissues of the low dose animals (Data not shown).

DISCUSSION

The purpose of this IND-enabling preclinical study was to assess the safety and biodistribution of Ad-KCNH2-G628S administered by atrial epicardial gene painting in New Zealand White rabbits. There were no Ad-KCNH2-G628S or transgene related effects based on evaluation of animal survival, clinical observations, body weights, clinical pathology, organ weights, macroscopic pathology, serum markers IL-6 and troponin, blood pressure and cardiac function (atrial and ventricular ejection fractions). There were no vector related adverse effects on electrocardiographic parameters, including QTc. No pathologic arrhythmias were noted, especially within the time frame of anticipated KCNH2-G628S gene expression.

Our echocardiographic data was obtained in the rabbits while fully anesthetized with the chest open. The left ventricle ejection fractions were somewhat low compared to those reported for awake normal New Zealand White Rabbits rabbits,²⁷ but were comparable to those reported in anesthetized rabbits.^28,29

Microscopic examination identified a slight increase in the incidence and severity (mild to moderate) of mononuclear infiltrates in the myocardium of the left and right atrial free wall of some rabbits (males > females) in the low vector dose group (1.5x10¹⁰ vp/animal, atrial painting) on SD 7. The increased incidence and severity (mild) of mononuclear cell infiltrates were also observed in the right atrial free wall and right auricle of some male rabbits in the vector low dose group (1.5x10¹⁰ vp/animal, atrial painting) on SD21. However, the increased incidence and severity of mononuclear cell infiltrates were not observed in rabbits in the vector high dose group (1.5x10¹² vp/animal, atrial painting) on SD7 or SD 21, even though similar levels of vector were detected in these high dose animals compared to the low dose animals. On SD42 there wer no microscopic findings possibly attributable to test article. The absence of the inflammatory findings in the vector high dose group suggests that the increased incidence and severity of mononuclear cell infiltrates was not related to vector administration. Minimal mononuclear cell infiltrates have been reported as background findings in rabbits and may be exacerbated by stress.²⁶ However, they are more commonly reported in the ventricles and the interventricular septum than in the atria. Regardless of the pathogenesis, the finding of increased mononuclear cell infiltrates in the myocardium of the atria/auricles in the vector low dose rabbits on SD7 and SD21 was not considered adverse due to the absence of correlating changes in the cardiac rhythm and function. The mononuclear cell infiltrates were not present of SD42, suggesting reversibility.

Vector DNA from Ad-KCNH-2-G628S applied directly to the epicardial surface of rabbit atria persisted in the auricles in most animals through 42 days. Copy numbers in left auricle tended to be lower than in the right auricle, possibly due to differences in accessibility of the applicator within the narrow rabbit chest wall. There was no dose- response in amount of vector in high dose versus low dose auricles at any time point, suggesting a possible saturation of uptake of the high dose of particles into the atrial tissue. The presence of large copy numbers of vector in blood in the high dose group rabbits suggests distribution of unabsorbed vector following cardiac painting in the high dose group. There was no evidence of systemic uptake by extra thoracic tissues.

In conclusion, the study evaluated potential toxicity in male and female New Zealand White rabbits that were administered vehicle control or Ad-KCNH2-G628S at 1.5x10¹⁰ and 1.5 x10¹² vp by epicardial atrial painting, or Ad-KCNH2-G628S at 1.5x10¹² vp by epicardial whole heart painting. There was no Ad-KCNH2-G628S related adverse effects in general toxicity endpoints or in specific cardiac endpoints: cardiac troponin in serum, blood pressure, left atrial and ventricle ejection fractions, heart rate, or QTc intervals. A non-adverse, transient, slight increase in mononuclear infiltrates in the myocardium was observed in the rabbits given 1.5x10¹⁰ vp, which was of uncertain relationship to administration of the vector.

Figure 2.

Group mean (± SD) PR interval (msec) for male and female rabbits combined for the vehicle control, high dose and whole heart animals. N = 7 vehicle controls and whole heart groups. N = 9 for the high dose group.

Figure 3.

Group mean (± SD) QRS interval (msec) for male and female rabbits combined for the vehicle control, high dose and whole heart animals. N = 7 vehicle controls and whole heart groups. N = 9 for the high dose group.

Table 6.

Microscopic Findings Possibly Related to Ad-KCNH2-G628S Administration on SD21

Sex:	Male			Female
Group:	2	3	4	2	3	4
Dose (virus particles/animal):	0 (vehicle)	1.5x1010	1.5x1012	0 (vehicle)	1.5x1010	1.5x1012
Number examined:	5	5	5	5	4	5
Heart, Auricle, Right
Infiltration, Mononuclear, Myocardium
Mild	0	3	1	0	1	0
Moderate	0	1	0	0	0	0
Heart, Atrium, Right, Free Wall
Infiltration, Mononuclear, Myocardium
Minimal	0	2	0	1	1	0
Mild	0	1	0	0	0	0