We studied time course of pathological remodeling occurring in the cynomolgus monkey hearts against persistent atrioventricular block condition (n = 10). The atrioventricular block induced the ventricular and atrial dilation followed by the ventricular hypertrophy. Interstitial fibrosis in the ventricle was also observed along with gradual increases in the plasma angiotensin II and aldosterone concentrations. These adaptations were associated with the changes in gene expression profiling reflecting fibrosis and hypertrophy. Atrioventricular block reduced the ventricular rate and cardiac output, but the ejection fraction and stroke volume increased, whereas the cardiac output was gradually restored to its basal level. Systolic/diastolic blood pressure after the atrioventricular block was kept equal to or lower than that before the block, according with lack of increase in the plasma catecholamine levels. Chronic atrioventricular block gradually prolonged the QRS width and JT interval, leading to the QT interval prolongation in conscious state. 10 mg/kg of dl-sotalol hydrochloride induced torsade de pointes (TdP) in 6 out of 10 animals by 15 months. Animals showing longer QTcF under anesthesia after the atrioventricular block developed dl-sotalol-induced TdP earlier. No marked difference was observed in pharmacokinetics of dl-sotalol between 1 and 7 months after the atrioventricular block. Each TdP spontaneously terminated, reflecting a monkey's relatively small "effective size of the heart (=∛(left ventricular weight)/wavelength of reentry)". These fundamental knowledge will help better utilize the chronic atrioventricular block monkeys as an in vivo proarrhythmia model for detecting drug-induced TdP.

Validation of risk-stratification method for the chronic atrioventricular block cynomolgus monkey model and its mechanistic interpretation was performed using 6 pharmacologically distinct drugs. The following drugs were orally administered in conscious state, astemizole: 1, 5, and 10 mg/kg (n = 6); haloperidol: 1, 10, and 30 mg/kg (n = 5); amiodarone: 30 mg/kg (n = 4); famotidine: 10 mg/kg (n = 4); levofloxacin: 100 mg/kg (n = 4); and tolterodine: 0.2, 1, and 4.5 mg/kg (n = 4). Astemizole of 5 and 10 mg/kg significantly prolonged Delta Delta QTcF, whereas no significant change was observed by the others. Torsade de pointes (TdP) was induced by astemizole of 5 and 10 mg/kg in 3/6 and 6/6, and by haloperidol of 10 and 30 mg/kg in 1/5 and 1/5, respectively, which was not observed in the others. Torsadogenic risk of the drugs was quantified using the criteria for the monkey model specified in our previous study. Namely, high-risk drugs induced TdP at <= 3 times of their maximum clinical daily dose. Intermediate-risk drugs did not induce TdP at this dose range, but induced it at higher doses. Low/no-risk drugs never induced TdP at any dose tested. The magnitude of risk was intermediate for astemizole and haloperidol, and low/no risk for the others. The prespecified, risk-stratification method for the monkey model may solve the issue existing between nonclinical models and patients with labile repolarization, which can reinforce the regulatory decision-making and labeling at time of marketing application of nondouble-negative drug candidate (hERG assay positive and/or in vivo QT study positive).

In order to bridge the gap of information between the in silico model and human subjects, we evaluated torsadogenic risk of cisapride, dl-sotalol, bepridil and verapamil selected from 12 training compounds in the comprehensive in vitro proarrhythmia assay using the chronic atrioventricular block monkeys. Cisapride (0, 1, and 5 mg/kg, n = 5 for each dose), dl-sotalol (0, 1, 3, and 10 mg/kg, n = 5 for each dose), bepridil (0, 10, and 100 mg/kg, n = 4 for each dose), verapamil (0, 1.5, 15, and 75 mg/kg, n = 4 for each dose) were orally administered to the monkeys in conscious state. Five mg/kg of cisapride, 1, 3, and 10 mg/kg of dl-sotalol and 100 mg/kg of bepridil prolonged ΔΔQTcF, which was not observed by verapamil. Torsade de pointes was induced by 5 mg/kg of cisapride in 2 out of 5 animals, by 10 mg/kg of dl-sotalol in 5 out of 5 and by 100 mg/kg of bepridil in 2 out of 4, which was not induced by verapamil. These torsadogenic doses were normalized by their maximum clinical daily ones to estimate torsadogenic risk. The order of risk was dl-sotalol >bepridil ≥cisapride >verapamil in our study. Since the order was bepridil ≥dl-sotalol >cisapride >verapamil in comprehensive in vitro proarrhythmia assay (CiPA) in silico mechanistic model validation, sympathetic regulation on the heart may play a pivotal role in the onset of torsade de pointes in vivo.

Abstract Objectives Chimeric antigen receptor (CAR)‐T cell therapy possesses the potential to cause unexpected on‐target toxicities that may be life‐threatening. Non‐human primates (NHPs) share considerable structural homology and expression profiles of most proteins with humans and are therefore utilised as an animal model for non‐clinical safety studies. We have developed a lymphodepleted NHP model by conditioning the animals with immunosuppressive chemotherapy designed to simulate clinical practice conditions, to induce transient mixed chimerism before the administration of human CAR‐T cells redirected to target Ephrin type‐B receptor 4 (EPHB4‐CAR‐T cells) to evaluate the toxicity of these cells. Methods We administered 60 mg m−2 day−1 of fludarabine for 4 days and 30 mg kg−1 day−1 of cyclophosphamide for 2 days intravenously to cynomolgus macaques for lymphodepletion; then, 3.3 × 106 kg−1 of non‐transduced or EPHB4‐CAR‐T cells was infused into the macaques, respectively. All macaques were closely monitored and evaluated for potential toxicity for 7 days. Results Lymphodepletion was successfully achieved on day −1 before T‐cell infusion and persisted over 7 days without severe organ toxicities. A single administration of human EPHB4‐CAR‐T cells did not induce overt organ toxicities, although EPHB4‐CAR‐T cells were activated in vivo as evidenced by the elevation in copy numbers of the CAR transgene 24 h after infusion. Conclusion Although this NHP model is limited for the full evaluation of toxicity of human CAR‐T cells and the conditioning protocol should be further optimised, this lymphodepleted NHP model could be used to assess acute on‐target/off‐tumor toxicities of CAR‐T cells.

Objectives.Chimeric antigen receptor (CAR)-T cell therapy redirected to specific antigens on tumor cells is a promising immunotherapy strategy for various cancers. Most target antigens are also expressed on normal tissues at varying levels, and therefore, a considerable challenge in the field is determining safety profiles, including life-threatening off-tumor and off-target toxicities. The granulocyte-macrophage colony-stimulating factor receptor (hGMR) is a promising target for CAR T-cell therapy for a subset of acute myelocytic leukaemia, although it is also expressed on normal cells including monocytes, macrophages, CD34-positive haematopoietic cells and vascular endothelial cells. hGMR and other immune-related proteins are highly conserved between humans and cynomolgus macaques (Macaca fascicularis). Therefore, in this study, we engineered cynomolgus T cells to express CAR molecules redirected to hGMR by piggyBac (PB) transposon-based gene transfer and adoptively transferred autologous hGMR-CAR T cells into cynomolgus macaques. Methods.We established PB-mediated human GMR (hGMR)-specific CAR T cells using cynomolgus peripheral blood mononuclear cells and transferred them into autologous individuals, and evaluated the potential toxicity related to hGMR-CAR T cells.Results. hGMR-CAR T cells did not exert overt organ toxicities such as bone marrow suppression, monocytopenia and vasculitis, although they recognised and killed cynomolgus monocytes and macrophages in vitro.Conclusion.Although our model did not simulate a tumor-bearing model, it supports the safety of hGMR-CAR T cells and demonstrates the usefulness of a non-human primate model to evaluate the safety of T-cell products by assessing off-tumor/off-target toxicity before clinical trials.

It has been difficult to experimentally reproduce synergistic effects of ketoconazole on terfenadine-induced torsade de pointes. We assessed proarrhythmic effects of terfenadine (30 mg/kg, p.o.) with/without ketoconazole (100 mg/kg, p.o.) pretreatment using the chronic atrioventricular block cynomolgus monkeys with repeated-measured design (n = 4). Terfenadine with ketoconazole pretreatment repeatedly induced non-sustained torsade de pointes in each animal, although terfenadine alone did not induce it at all. Thus, the chronic atrioventricular block cynomolgus monkeys can be used for studying drug interaction-associated torsade de pointes, providing a non-clinical strategy to circumvent untoward drug interactions in patients specially under polypharmacy.

We analyzed the effects of intravenously as well as orally administered moxifloxacin on the pharmacokinetic and electrocardiographic variables along with its torsadogenic action using the chronic atrioventricular block cynomolgus monkeys with a cross-over design. Initially, moxifloxacin was intravenously administered in doses of 60 mg/kg/2 h, 60 mg/kg/1 h and 105 mg/kg/1.75 h with an interval of >1 week (n = 3), which provided Cmax of 19.7, 25.4 and 37.8 μg/mL, and induced torsade de pointes in 1, 0 and 3 out of 3 animals, respectively. Next, moxifloxacin was orally administered in doses of 10, 30 and 100 mg/kg with an interval of >1 week (n = 6), which provided Cmax of 1.8, 4.2 and 8.9 μg/mL, and induced torsade de pointes in 0, 0 and 2 out of 6 animals, respectively. A close analysis of pharmacokinetic and electrocardiographic variables indicates that torsade de pointes was induced in animals that had experienced larger systemic exposure of moxifloxacin and/or greater peak QTcF, although Cmax by itself did not necessarily reflect the incidence of torsade de pointes when its administration route was different. These findings may provide a basic guide how to use moxifloxacin in safe for patients with labile repolarization process.

llogenic induced pluripotent stem cell-derived cardiomyocytes transplanted directly into infarcted cynomolgus monkey hearts show electrical coupling with host cardiomyocytes improve cardiac contractile function after mild immunosuppression.

The cynomolgus macaque (Macaca fascicularis) has emerged as an important experimental animal model for biomedical research in various domains, necessitating the more extensive characterization of the genetic backgrounds influencing the macaque’s response to drugs and sensitivity to experimental disease. The diversity of the variable mitochondrial DNA (mtDNA) D-loop region has been analyzed phylogenetically among geographically isolated populations or within subdivisions of the same regional population. However, the genetic differences among several substructures originating from a common population have not yet been investigated. By sequencing fragments of the mtDNA D-loop region from two subpopulations from the Indochinese region (Cambodian-Chinese and Vietnamese) along with two native Indonesian and Filipino populations, we identified 87 mtDNA D-loop haplotypes, of which 67 are new. The phylogenetic relationship suggests that the Indochinese haplotypes are intermingled in comparison to the distinct divergence of the Indonesian and Filipino lineages. The subpopulations were shown by estimation of evolutionary divergence and Wright’s F-statistic (Fst) to have little genetic differentiation. Altogether, the subpopulations may be used in biomedical research, even though a slight difference is observed in haplotype frequencies among them. Therefore, genetic diversity analyses will be necessary for the elucidation of genetic differences among the populations, as well as to obtain a better understanding of genetic diversity for biomedical research. This will involve the selection of macaques and the monitoring of genetic heterogeneity among and within breeding facilities.

We recently developed a new monkey model complicating chronic atrioventricular conduction block. In this study, we compared anatomical, neurohumoral and proarrhythmic profiles of the atrioventricular block monkeys with those of normal animals. The cynomolgus monkeys of either sex (3–5 kg) were used (n=12). The atrioventricular node of the 6 animals was ablated under anesthesia, and their anatomical, neurohumoral and proarrhythmic profiles were assessed more than 2 months after the surgery. The cardiothoracic ratio of the chronic atrioventricular block monkeys was 59±1% (n=6), which was greater than that of the normal group (47±1%, n=6, p<0.001). Plasma concentrations of atrial natriuretic peptide, brain natriuretic peptide and noradrenaline in the chronic atrioventricular block monkeys were 130±25, 29±8 pg/ml and 4.1±0.3 ng/ml, respectively (n=6), each of which was significantly higher than that in normal animals (25±9, 5±1 pg/ml and 1.6±0.2 ng/ml, respectively, n=6). Although no arrhythmia was detected in the normal monkeys after the oral administration of a class III antiarrhythmic drug sotalol (10 mg/kg, p.o., n=4), torsades de pointes arrhythmias were induced in all chronic atrioventricular block animals by the same dose of sotalol (n=6). All torsades de pointes spontaneously terminated. In conclusion, the chronic atrioventricular block monkey may have multiple risk factors for torsades de pointes arrhythmias, which makes it possible for the model to qualitatively and quantitatively predict clinical risks of the QT-interval prolonging drugs.