BS21 Development of methodology for analysis of mouse echocardiograms to distinguish between the anti-hypertrophic and potential cardiotoxic effects of anti-cancer therapies that target the ERK1/2 cascade in mice treated with angiotensin II

Introduction

Several anti-cancer therapies target the ERK1/2 pathway in which RAF kinases phosphorylate/activate MEK1/2 which phosphorylate/activate ERK1/2. Trametinib inhibits MEK1/2 but is usually used with dabrafenib that inhibits BRAF. Trametinib has cardiotoxic effects, but dabrafenib could protect against hypertensive heart disease. However, dabrafenib inhibits kinases in addition to BRAF and the effects of the combination of trametinib/dabrafenib are unknown. Our hypothesis is that the combination treatment has greater cardiotoxicity potential than trametinib alone.

Experimental Approach

Male C57Bl/6J mice (10 wks) were treated with 0.8 mg/kg/d angiotensin II (AngII) or vehicle, alone or with 1 mg/kg/d trametinib or trametinib with 3 mg/kg/d dabrafenib using osmotic minipumps for drug delivery over 7 d. The effects on the heart were assessed at 3 d and 7 d by echocardiography using a Vevo 2100 system. Echocardiograms were analysed using VevoStrain software and data for segmental strain were exported for downstream analysis.

Results and Method Development

AngII promoted cardiac hypertrophy with an increase in estimated left ventricle (LV) mass, as determined using VevoStrain. This was prevented by trametinib or dabrafenib/trametinib, but we could detect no differences between the treatments using established measures of cardiac function/dimensions. However, visually, LV wall movements appeared different. We therefore undertook downstream analysis using segmental strain measurements (corresponding to 6 segments around the LV) and identified novel variables that incorporated longitudinal and radial displacement of opposing LV walls for a more comprehensive assessment of cardiac function. A large number of measurements were extracted from the time series for each segment. Feature engineering was performed and variable selection analysis undertaken. This identified two characteristics that enabled classification of the two treatment groups (trametinib and trametinib/dabrafenib each with AngII): one represents the difference in magnitude of radial displacement of the pair of ventricular middle segments; the second was developed to quantify the flattening of the longitudinal displacement time series of the outer ventricular segments. Thus, by analysing the entire temporal displacement series, nuanced differences in cardiac dynamics were revealed between the two treatment groups that were previously undetected.

Conclusions

In hypertension, trametinib/dabrafenib combination therapy has a different effect on LV performance compared with trametinib. This could not be distinguished with the usual parameters for cardiac function available from proprietary software for echocardiograms. However, novel variables allowed for variation in displacement of opposing LV walls. This approach introduces an alternative and complementary approach to evaluate cardiac function with potential applicability for early assessment of cardiotoxicity of drug therapies.

Conflict of Interest

no