A pooled analysis of nine studies in one institution to assess effects of whole heart irradiation in rat models

Abstract

Purpose Over the years, animal models of local heart irradiation have provided insight into mechanisms of and treatments for radiation-induced heart disease in human populations. However, it is not completely clear which manifestations of radiation injury are most commonly seen after whole heart irradiation, and whether certain biological factors impact experimental results. Combining 9 homogeneous studies in rat models of whole heart irradiation from one laboratory, we sought to identify experimental and/or biological factors that impact heart outcomes. We evaluated the usefulness of including (1) heart rate and (2) bodyweight as covariates when analyzing biological parameters, and (3) we determined which echocardiography, histological, and immunohistochemistry parameters are most susceptible to radiation effects. Finally, (4) as an educational example, we illustrate a hypothetical sample size calculation for a study design commonly used in evaluating radiation modifiers, using the pooled estimates from the 9 rat studies only for context. The results may assist investigators in the design and analyses of pre-clinical studies of whole heart irradiation. Materials and Methods We made use of data from 9 rat studies from our labs, 8 published elsewhere in 2008–2017, and one unpublished study. Echocardiography, histological, and immunohistochemical parameters were collected from these studies. Using mixed effects analysis of covariance models, we estimated slopes for heart rate and bodyweight and estimated the radiation effect on each of the parameters. Results Bodyweight was related to most echocardiography parameters, and heart rate had an effect on echocardiography parameters related to the diameter of the left ventricle. For some parameters, there was evidence that heart rate and bodyweight relationships with the parameter depended on whether the rats were irradiated. Radiation effects were found in systolic measures of echocardiography parameters related to the diameter of the left ventricle, with ejection fraction and fractional shortening, with atrial wall thickness, and with histological measures of capillary density, collagen deposition, and mast cells infiltration in the heart. Conclusion Accounting for bodyweight, as well as heart rate, in analyses of echocardiography parameters should reduce variability in estimated radiation effects. Several echocardiography and histological parameters were particularly susceptible to whole heart irradiation, showing robust effects compared to sham-irradiation. Lastly, we provide an example approach for a sample size calculation that will contribute to a rigorous study design and reproducibility in experiments studying radiation modifiers.