Modelling Tumor Growth Under Angiogenesis Inhibition with Mixed-effects Models

Angiogenesis inhibitors offer a promising new treatment modality in oncology. However, the optimal administration regimen is often not well-established, despite the fact that it might have substantial impact on the outcome. The aim of the present study was to investigate this issue. Eight weeks old male C57Bl/6 mice were implanted with C38 colon adenocarcinoma, and were given either daily (n = 9) or single (n = 5) dose of bevacizumab. Outcome was measured by tracking tumor volume; both caliper and magnetic resonance imaging was employed. Longitudinal growth curves were modelled with mixed-effects models (with correction for autocorrelation and heteroscedasticity, where necessary) to infer on population-level. Several different growth models (exponential, logistic, Gompertz) were applied and compared. Results show that the estimation of the exponential model is very reliable, but it prevents extrapolation in time. Nevertheless, it clearly established the advantage of the continuous regime.