Contents

📒 Fernandezchas 2018

Mechanism of doxorubicin cardiotoxicity evaluated by integrating multiple molecular effects into a biophysical model1

Sciwheel

Introduction

  • Doxorubicin (DOX) is the predominant anthracycline, but its use is limited due to cardiotoxicity.
  • acute and chronic DOX cardiotoxicity re fl ect a combination of direct DOX effects on channel, pump and exchanger function and indirect effects on expression levels
  • The metabolite doxorubicinol (DOXL) accumulating in cardiac tissue => contributor to the chronic toxicity
  • focused on modelling DOX and DOXL effects on electrophysiology and calcium handling in this study. Quantifying on ‘end’ effects (both direct and indirect)
  • biophysical modelling can replicate DOX cardiotoxicity well in both human and rabbit models
  • acute DOX and DOXL and chronic DOX exposure will increase the sarcoplasmic reticulum (SR) Ca2+ leak (ILeak)
  • reported increased K+ permeability in cells chronically exposed to DOX is consistent with an increase in the rapid delayed rectifying K+ current (IKr).
  • acute DOXL exposure is unlikely to cause inhibition of the Na+/Ca2+ exchangers (INaCa).

Methods

Cell models

  • ventricular myocyte EP model: rabbit (Morotti), human (TP)

  • Code in CellML. Converted to C / MATLAB by OpenOCR

  • The independent variables

    • IKr
    • L-type Ca2+ current (ICaL),
    • ILeak
    • Na+/K+ pump (INaK)
    • SR Ca2+ ATPase (IUp)
    • INaCa
    • ryanodine receptor (IRel)

Drug effects

  • Îą factor (effect of DOX): $ I_{i}=\alpha_{i} G_{i}\left(V-V_{i}\right) $
  • % change of APD, % change of systolic Ca2+concentration and % change of Ca2+relaxation time were determined by searching PubMed. experimental methods and data reporting are very heterogeneous.

Method 1

  • an assumption that the maximum concentration used in a given study is the concentration causing the maximum effect on each given target. => accuracy but uncertain precision
  • this method disregards the concentration dependence of drug effects.

Method 2

  • drug inhibition or activation of each channel can be described by a Hill equation (n=1). have precision, but uncertain accuracy. https://user-images.githubusercontent.com/40054455/86617392-bac55080-bfe9-11ea-94d9-62b90a7a31bd.png
  • Chronic DOX exposure was not evaluated by this approach since changes in current activity develop over time

Provenance of input data and methods for best estimation

  • For data lacked precision (too few studies and/or semi-quantitative estimates ) or accuracy (several studies reporting contradictory estimates)

Simulations

https://user-images.githubusercontent.com/40054455/86617395-bb5de700-bfe9-11ea-917d-408297936d98.png https://user-images.githubusercontent.com/40054455/86617401-bdc04100-bfe9-11ea-8990-643be36be170.png

Factorial analysis

  • APD, systolic Ca2+ concentration and Ca2+ relaxation time were generated by simulation with Îą values de fi ned either as 1 (drug free simulation) or equal to the consensus values derived by method 1 and then by method 2 (consensus altered simulation values)

https://user-images.githubusercontent.com/40054455/86617399-bc8f1400-bfe9-11ea-9910-7639c03aff04.png https://user-images.githubusercontent.com/40054455/86617403-be58d780-bfe9-11ea-8ced-4288eb184b3e.png

  • simulation of drug effects in rabbit and human models were comparable https://user-images.githubusercontent.com/40054455/86617400-bdc04100-bfe9-11ea-8a7a-03fbc97cf601.png

Testing model assumptions

  • measurements are very heterogeneous spanning a broad range of species, preparations and temperatures
  • default model used 1 Hz stimulation, too slow for rabbits
  • The level of inhibition of INaCa in the simulations of acute exposure to DOX was derived from a combination of the available experimental measurements from the three relevant publications. Rabbit 10%. Human: 40%.
  • Estimations for DOX inhibition of IKrwere based in part from guinea pig, with different IKs and IKr distributions from humans / rabbits https://user-images.githubusercontent.com/40054455/86617404-bef16e00-bfe9-11ea-8efd-788b63f7f2a2.png https://user-images.githubusercontent.com/40054455/86617408-bef16e00-bfe9-11ea-9418-0e63b948055e.png https://user-images.githubusercontent.com/40054455/86617411-bf8a0480-bfe9-11ea-86a9-046e91aafd33.png

Discussion

  • mathematical models of rabbit and human ventricular electrophysiology, when provided with appropriate scaling factors, to reproduce the consensus effects of acute DOX, chronic DOX and acute DOXL exposure on APD, systolic Ca2+ concentration and Ca2+ relaxation time

Inhibition of INaCa

  • inhibited by DOXL (Olson et al., 1988) by up to 100% may not be correct
  • the increase in Na+ alone would inhibit the activity of the Na+/Ca2+ exchanger

SR leak

  • The estimated increase in SR leak was approximated from qualitative observations
  • potentially due to an increase in ROS production
  • The wide range of SR leak predicted by the model may refl ect true differences in leak for different drugs under different conditions.

Action potential duration

  • The factorial analysis revealed that actions on IUp, INaK,IKr and ILeak were responsible for the majority of acute DOXL effects on APD, whereas actions on IUp, INaKand ILeakresulted in the changes in systolic Ca2+ concentration
  • DOX- and DOXL-induced increases in Îą for ILeak were necessary for optimal predictiveness
    • Depleting SR Ca
    • Shortening APD
  • INaK was found to play a role in the APD effects of acute exposure to DOXL.
  • the effect of acute DOX and chronic DOX on channels is similar, yet they have different effects on APD and systolic Ca2+concentration (Figure 3)

Clinical relevance

  • DOX peak plasma concentrations in the clinical setting are reported as 1 – 10 ΞM. value for DOXL is 0.2 ΞM.
  • intracellular drug accumulation may be the primary determinant of the cardiotoxicity,

Reference


  1. Fernandez-Chas M, Curtis MJ, Niederer SA. Mechanism of doxorubicin cardiotoxicity evaluated by integrating multiple molecular effects into a biophysical model. Br. J. Pharmacol. 2018;175(5):763-781. doi:10.1111/bph.14104. PMC5811623 ↩︎