Multileaf collimators, air gap, lateral penumbra, and range compensation in proton radiotherapy

Multileaf collimators (MLCs) are now employed at some proton therapy centers and may be under consideration by others. Technical problems such as proton leakage and neutron production have been solved1,2 and proton MLCs have been supplied by industry. Substituting a single MLC snout for the traditional multiple snouts, with their individually sized custom brass apertures and plastic range compensators, has obvious operational advantages. Patient specific machining is halved, technologist workload is reduced and throughput is increased. Though well aware of those potential advantages, early proton radiotherapy centers did not take this path, because of the following problem. A general purpose “universal” proton MLC, designed to shape large or small fields and to stop high or low energy proton beams—that is, to treat all disease sites—is necessarily large. It tends to interfere mechanically with patient anatomy or immobilization devices, increasing the air gap between the (aperture/range compensator) and the patient. Taken with the large effective source size of a typical double scattered beam, that leads to suboptimal transverse penumbra and (less obviously) to suboptimal performance of the range compensator, with over- and under-penetration of the beam in some parts of the field. The effect of air gap on range compensation was studied by Sisterson et al.7 Though the Harvard geometry used in that study is no longer relevant, the calculation is governed by the angular confusion θC (rms angular spread about the mean angle) of protons impinging on the range compensator, and θC happens to be the same (≈13 mRad) for the 3.55 g/cm2 preabsorber in Ref. 7 as it is for the deepest modulator step of a medium-range IBA option. (θC comes mostly from the second scatterer and varies little from one option to the next.) A 2.45 cm high, 60° apex angle wedge, with a 12 cm air gap between it and the compensating valley, yielded ≈2 mm maximum over- and under-penetration of the 50% dose level.7 By themselves these numbers prove nothing. One needs to know the clinical requirements on conformality, as well as the actual air gap in practice. However, they are of a potentially troublesome order-of-magnitude. Current practice at the Burr Proton Therapy Center, Massachusetts General Hospital (MGH) is as follows. There are three standard snout sizes, with custom milled apertures for each. The goal is an air gap less than 3 cm. 5 cm is undesirable, and 10 cm is unacceptable; the field is replanned. The snout extension (from which the air gap may be computed) is recorded for each treatment. If the gap achieved exceeds the planned gap by more than 2 cm the field is replanned and the new result is archived “as treated.” Reducing the cost of proton radiotherapy makes it available to more patients, and we do not, by any means, wish to discourage the development of proton MLCs towards that end. The technology is not yet mature, and more compact universal designs may evolve. Present day compact designs1 may yield dose distributions perfectly adequate to their clinical goals, especially in single-scattered beams with their much smaller θC. MLCs may play a role in improving transverse penumbra for magnetically scanned beams8,9 and in other ways not yet imagined. All that said, we urge caution in the wholesale replacement, at our present state of knowledge, of the traditional method by single universal MLC snouts. Often, the dose conformality in the high dose (target) region will be degraded. We have focused on air gap, but the 3D collimator effect (because the leaves always have their maximum thickness) and the scalloping effect (because the leaves only approximate the ideal shape) will also play a role. Whether the decrease in conformality is balanced by the increase in efficiency is a decision which should be informed, for each disease site, by clinical evaluation of treatment plans—as treated—for the two methods. We thank Judy Adams and Al Hernandez (MGH) for explaining the air gap policy at the Burr Center.