A retrospective analysis of the concordance between cardiovascular effects in telemetry studies, toxicity studies, and early clinical trials

Abstract

This work aimed to compare the ability of several ECG technologies used in toxicity studies to identify cardiovascular effects. A retrospective analysis of 41 drug development projects across the Roche portfolio from 2005 to 2023 was conducted to evaluate the concordance of snapshot ECG (SS-ECG) and jacket external telemetry (JET, with or without blood pressure assessment) to detect effects on heart rate, ECG, and blood pressure in toxicity studies, when compared to implanted telemetry as the gold standard. Overall, 132 non-clinical studies were investigated. For 24 projects, translation to phase 1 safety data under consideration of plasma exposures could be followed up. Company-internal strategic decisions on advancing certain molecules despite cardiovascular findings were subject to a risk-benefit assessment and not evaluated in this investigation.

SS-ECGs never detected changes in heart rate (HR) previously revealed in telemetry studies, whereas JET identified these in all telemetry-positive cases. JET (or M11 implants) detected 44 % and SS-ECGs identified 33 % of QTc increases in telemetry-positive molecules, indicating that stand-alone telemetry is still the most reliable in vivo study in identifying QTc-effects. No clear species differences between beagle dogs, Göttingen minipigs or cynomolgus monkeys were evident when comparing heart rate and QTc detection. HR increases were noted clinically in about 50 % of the non-clinical positive cases, whereas HR decreases were never detected in clinical trials. Non-clinical blood pressure decreases showed higher translatability to humans (at least 80 %) than blood pressure increases (29 %), with a high prevalence of dogs and minipigs showing a signal. Non-clinical QTc increases detected by JET and/or telemetry were confirmed in 3 clinical studies analyzed with concentration/QTc modeling, however in 7 clinical studies employing other methodologies to measure QT, the non-clinical QTc effects did not translate.