Purpose: To evaluate the perturbation to dose distribution caused by utilizing a tungsten MLC instead of a brass aperture in passive scattering proton therapy. Method and Materials: A four‐dimensional GEANT4 Monte Carlo simulation of a proton therapy nozzle is used to simulate the behavior of a beam delivery system. Several datasets have been generated accounting for different configurations of system components including binary pre‐scatterer foils, double contoured scatterer, modulator wheel, and pre‐collimator jaws. The resulting beam profiles are subject to one of two different final collimator options: a tungsten MLC or a brass aperture. The performance of the beam delivery system is examined for different beam energies, intensity modulation patterns, field size, air gap between the MLC and the patient, and the introduction of a step in the design of the MLC. Results: We quantify dosimetric perturbations when conforming to a simple target volume. Variations in several factors were studied, including lateral field flatness and symmetry, penumbra, and flatness of the spread‐out Bragg peak. When compared to results obtained using a brass aperture, an MLC made of tungsten seems to produce similar results. Additionally, the respective neutron energy distributions were used to calculate the neutron equivalent dose absorbed outside of the treatment field in each case. Conclusion: The ability of an MLC to conform to target volume is a function of the specific contour shape. This is not necessarily true for a custom made aperture. However, our analysis of dosimetric differences between the two indicates that there are no significant differences for simple target volume contours. The performance of an MLC as a final collimator in passive scattering proton therapy is predicted to be comparable to the performance of a brass aperture. Furthermore, secondary neutron production is reduced by using a tungsten MLC rather than a brass aperture. © 2007, American Association of Physicists in Medicine. All rights reserved.
