gms | German Medical Science

48th Meeting of the Particle Therapy Co-Operative Group

Particle Therapy Co-Operative Group (PTCOG)

28.09. - 03.10.2009, Heidelberg

Decreasing the PTV margins in IMPT using Monte Carlo with incorporated random setup errors

Meeting Abstract

Search Medline for

  • M. Soukup - Elekta CMS Software, Freiburg
  • M. Alber - Section for Biomedical Physics, Radiooncological Clinic, Tübingen

PTCOG 48. Meeting of the Particle Therapy Co-Operative Group. Heidelberg, 28.09.-03.10.2009. Düsseldorf: German Medical Science GMS Publishing House; 2009. Doc09ptcog190

DOI: 10.3205/09ptcog190, URN: urn:nbn:de:0183-09ptcog1906

Published: September 24, 2009

© 2009 Soukup et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.



Background: To handle calculation and geometric uncertainties like random setup error, planning margins are added to the Clinical Target Volume (CTV) in common treatment planning. However, the margin concept can lead to an over-irradiation of healthy tissue and unnecessary trade-offs in target coverage. Its suitability for IMPT is limited, especially for geometries with lateral heterogeneities. New concepts handling calculation and geometric uncertainties are thus of great interest.

Material and methods: Maximum calculation accuracy can be achieved with Monte Carlo dose calculation. Moreover, any random setup error probability distribution can be implemented efficiently in Monte Carlo by sampling the isocenter position for each primary proton. As a direct consequence of the incorporation of the random setup error in the optimization, the planning margins can be decreased.

To study the effect of random setup errors, a heterogeneous patient geometry with a tumor in the vicinity of brain stem, chiasm and right optic nerve with one field (gantry 90 degree) treatment plan was used.

Two sets of treatments plans were created – one with the margin concept (CTV extended laterally to the field direction) and one without margins to CTV and with random-setup error included in the dose calculation for the optimization. Both sets were optimized for the same iso-toxicity with the Equivalent Uniform Dose (EUD) prescription for the target of 54 Gy.

The dosimetric effect of the random setup error was simulated by recalculation of the whole treatment plan for different isocenter positions spread around the original isocenter (49 instances). The recalculated dose distributions were weighted according to the random setup error probability distribution and accumulated to simulate the treatment course. Different random setup error distributions were used for the accumulation to evaluate the robustness of the treatment plans to random setup error changes.

The treatment planning system Hyperion with the Monte Carlo code VMCpro was used both for planning and recalculation.

Results: The plan optimized for the random setup errors results in basically the same CTV coverage if compared to the margin-based plan (EUD of 53.6 Gy and 53.8 Gy, respectively).

After performing the random setup error simulation, the dose to CTV deteriorated for the margin-based plan to 51.7 Gy.

Evaluating the sensitivity to a random setup error distribution other than the distribution used for the optimization (3 +/- 1 mm), the EUD to CTV for the random setup error optimized plan and for the margin-based plan ranged between 52.9 – 54.8 Gy and 50.9 – 52.8 Gy, respectively.

Conclusion: Optimizing the random setup errors with Monte Carlo is feasible and allows for IMPT plans with more robust CTV coverage than margin-based treatment planning.