University of Washington
Monte Carlo Damage Simulation (MCDS) Software

Robert D. Stewart, Ph.D.; Associate Professor of Radiation Oncology
School of Medicine, Department of Radiation Oncology, 1959 NE Pacific Street, Seattle, WA 98195


Results from the MCDS have been cited in at least 41 peer-reviewed studies and 2 book chapters.  In a 2012 Red Journal paper by Frese et al.2012, the MCDS was used in combination with the repair-misrepair-fixation (RMF) model to predict spatial variations in the relative biological effectiveness (RBE) of protons and carbon ions in representative spreadout Bragg peaks (SOBP).  See also RD Stewart, Models and Mechanisms Connecting Physics and Biology at Multiple Scales in the Biological Hierarchy (presented at the 2014 AAPM Meeting in Austin, TX, TU-A-BRE-3, July 22, 2014; Abstract #23482) for interesting background information and some of the theory behind the MCDS + RMF approach to modeling RBE effects.

Motivation

Many decades of biophysical research provide evidence suggesting that the number and spatial arrangement of energy deposits within and near the DNA produce many types of clustered DNA lesion, including single strand breaks (SSB), double strand breaks (DSB) and clusters of damaged bases.  Through a breakage and reunion process, double strand breaks (DSB) are converted to small- and larger-scale chromosomal exchanges with the potential to cause phenotypic alterations, neoplasia and cell death (see an idealized schematic showing time scales and selected events in radiation damage to cells and tissues). Other types of non-DSB cluster may also have significant biological consequences.  As the only method currently available to determine the number and spatial configuration of lesions forming a cluster, Monte Carlo simulations are a potentially useful adjunct to experiments probing the underlying basis for the effects of oxygen and radiation quality on cell death.  Estimates of DSB yields from Monte Carlo simulations can also be used in combination with kinetic reaction-rate models, such as the repair-misrepair-fixation (RMF) model (Carlson et al. 2008), to determine the relative biological effectiveness (RBE) for cell killing of different types of radiation.  An improved understanding of RBE and oxygen effects are needed to more fully exploit the biological potential of protons and carbon ions in radiation therapy, especially as high levels of pre-treatment tumor hypoxia have been implicated as a significant factor contributing to treatment failure.

General Features of the MCDS Software

The original MCDS (Semenenko and Stewart 2004, 2006) simulates the induction and clustering of DNA lesions in normoxic cells (O2 concentrations greater than about 21%) uniformly irradiated by monoenergetic electrons, protons and a particles with energies as high as 1 GeV. In the latest version of MCDS (Stewart et al. 2011), the allowed particle types have been expanded to include ions up to and including 56Fe.  The allowed range of particle energies has been expanded (see Table), and the induction of damage for arbitrary mixtures of charged particles with the same or different kinetic energies can be directly simulated.  Cluster yields for photons and other neutral particles can be computed for the distribution of secondary charged particles produced in monolayer cell or other geometries (e.g., see Hsiao and Stewart 2008).  Although not required for the simulation of damage induction, the MCDS reports information such as the charged particle stopping power in water, CSDA range, absorbed dose per unit fluence, frequency-mean specific energy, energy imparted per radiation event, and the lineal energy.

Software Availability

An executable of MCDS software (Version 3.10A, compiled December 5, 2011) for MS Windows and linux is freely available for commercial, educational or research purposes (see below). An executable of MCDS 2.01 (March 2006) is also available (see also the MCDS 2.01 website). Inquiries about the MCDS program and related publications or software should be directed to Dr. Rob Stewart.

Sample Input & Output Files

The MCDS reads particle and other simulation information from an ASCII input file(s) and then writes the results of the simulation to another ASCII (output) file.  For neutral particles (e.g., g-rays from 60Co and 137Cs) and for mixtures of different types of charged particles, the main input file (e.g., 60Co.inp) must include a line that points to the name of a secondary data file (e.g., 60Co.dat).  The .dat file specifies the types, energy and relative weight (fluence) to use to simulate a mixed radiation field.

NOTE: Many additional input and output files are included in the software distribution. 

RBE x Dose Tallies in MCNP 6.1 (and MCNPX)

A useful, and computationally very efficient, method to compute RBE-weighted dose in MCNP (or MCNPX) is to modify a standard (so-called F6) dose tally in MCNP by a user-defined dose-response function (DE DF option).  Below are links to some examples of MCNP tallies to record RBE weighted dose for the endpoint of DSB induction in well oxygenated and anoxic cells (links are to an ASCII file).  The RBE dose-response function (DE DF card) is based on 60Co g-rays as the reference radiation, i.e., 8.3 DSB Gy-1 Gbp-1 for well oxygenated cells and 2.86 DSB Gy-1 Gbp-1 for anoxic cells (Stewart et al. 2011, Kirkby et al. 2013).  See also Stewart et al., Rapid MCNP Simulation of DNA Double Strand Break (DSB) Relative Biological Effectiveness (RBE) for Photons, Neutrons, and Light Ions, Phys. Med. Biol. (2015).

    electron RBE (10 eV to 1 GeV)

    1H+ RBE (1 keV to 1 GeV)

    2H+ RBE (1 keV to 1 GeV)

    3H+ RBE (1 keV to 1 GeV)

    3He2+ RBE (1 keV to 1 GeV)

    4He2+ RBE (1 keV to 1 GeV)

 

In addition to sample dose tallies modified by the RBE for DSB induction, the above examples also includes tallies to record LET x dose, the frequency-mean specific energy (5 um biological target) x dose, and the mean specific energy times the square of the number of DSB Gy-1 Gbp-1.  The latter quantity is useful for estimating the RBE for reproductive cell death using the Repair-Misrepair-Fixation (RMF) model (Carlson et al. 2008), i.e., the linear quadratic (LQ) model parameters are approximately equal to a = qS + kzFS2 and b = kS2/2 (Carlson et al. 2008, Frese et al. 2012).  Here, q and k are are cell- and tissue- (or tumor-) specific adjustable (fitted) parameters that are independent of LET (as a first approximation), zF is the frequency-mean specific energy (in Gy), and S is the average number of DSB Gy-1 Gbp-1 as computed with the MCDS version 3.10A.

NOTE: Similar RBE x dose tallies can (could be) easily implemented in other Monte Carlo code systems (TOPAS\Geant4, EGSnrc, PENELOPE, FLUKA, ...)

Last updated: October 17, 2016
Contact Webmaster