Angiotensin II unmasks immune checkpoint inhibitor-induced arrhythmias and conduction defects in mice

Background Immune checkpoint inhibitors (ICIs) can cause rare but potentially fatal cardiac toxicities. Although myocarditis has received the most attention, arrhythmias and conduction disorders occur more frequently and with variable timing. The mechanisms underlying these electrical abnormalities remain poorly defined, in part due to the absence of preclinical models designed to study ICI-associated conduction disturbances. Considering roughly one-third of patients with ICI-related arrhythmias have prior hypertension, we hypothesized that hypertensive stress sensitizes the heart to ICI-induced conduction abnormalities. Purpose To determine the extent to which angiotensin II (Ang II) influences ICI-associated conduction abnormalities in mice. Methods Adult male C57BL/6J mice (12-14 weeks) received continuous infusion of Ang II (1.0 mg/kg/day) or saline via osmotic minipumps for 14 days. After 7 days of infusion, mice were treated with two doses of combined anti-PD-1 and anti-CTLA-4 antibodies (25 mg/kg each, intraperitoneally, every 3 days). Ang II pre-treatment preceded ICI administration to minimize confounding early inflammatory effects, and ICI exposure was restricted to one week (2 injections) to avoid overt myocarditis. At study end, echocardiogram and surface electrocardiogram (ECG) were acquired under isoflurane anaesthesia. Electrical vulnerability was further evaluated using a bolus of epinephrine (2 mg/kg) and caffeine (120 mg/kg) during ECG recording. Results Surface ECG revealed that combined ICI treatment induced QRS prolongation and increased P-wave duration exclusively in Ang II-infused mice. High-grade atrioventricular block (AVB) occurred in 3 of 8 Ang II-treated mice (37.5%) without provocation, and 2 of these 3 mice developed ventricular tachycardia (VT) following catecholamine challenge. In contrast, no AVB or VT were observed in saline-infused mice. Echocardiogram revealed preserved left ventricular ejection fraction and fractional shortening in both groups. Ang II-treated mice exhibited modest increases in left ventricular mass and heart weight, consistent with hypertensive stress-induced cardiac hypertrophy. Histological analysis showed mild infiltrative lesions (<20% of longitudinal section area) in 3 of 8 Ang II-treated hearts; however, only one of these animals exhibited AVB, and arrhythmias occurred in the absence of detectable myocardial infiltration in the remaining cases. Conclusion Ang II unmasked ICI-associated conduction abnormalities and ventricular arrhythmias in the absence of myocardial dysfunction or overt myocarditis. These data provide proof-of-principle that Ang II-induced remodelling lowers the threshold for ICI-induced electrical toxicities within the cardiac conduction system. Our model enables interrogation of early conduction system-specific mechanisms of ICI cardiotoxicity and provides a platform for defining causal and temporal relationships between ICIs and arrhythmias.