Currently in radiation therapy the beam is shaped to fit the physical volume of the tumour. This is great for sparing surrounding healthy tissue, but it doesn't account for the fact that tumours are inhomogenous, with some areas more resistant to radiation than others. In the ideal scenario, the radiation-resistant regions would receive a higher dose without increasing the patient's overall normal tissue exposure, thus making the treatment more effective and less toxic.
This is called biological optimization, and it should be possible now that combined PET/CT scanners are available that show the biological processes going on inside tumours (as well as their physical shape and location). "With biological optimization, the best possible compromise is made so tumour cure is maximized at the same time, as normal tissue damage is as low as clinically possible," explained Brahme, a professor in medical radiation physics at Karolinska.
So what's the catch? Plans designed to tailor the beam intensity to structures within tumours are necessarily more detailed and complex than plans that treat the tumour as a homogenous lump. Even the latest intensity-modulated radiation therapy (IMRT) systems weren't designed with such intricacy in mind (see also Streamlined thinking on IMRT planning on medicalphysicsweb). Because they rely on dynamic multileaf collimation to achieve the intensity modulation, with conventional linacs the irradiation time increases dramatically with the complexity of the desired dose distribution.
Heavy metal
The problem is that standard multileaf collimators (MLCs) need leaves made of thick (usually 5–7 cm) tungsten in order to block the powerful X-rays used for radiation therapy. This means that the MLCs are large, heavy and slow moving; it also means that plans that require several MLC rearrangements take significantly longer to implement. In a busy hospital environment, switching to biologically optimized treatments with existing linacs would be impractical and, in the long run, prohibitively expensive.
The way forward, claims Brahme, is an entirely new radiation-delivery system that uses area modulators or thin pencil beams of X-rays (or electrons or light ions) to build up the desired dose distribution by scanning across the target area. And because the beams are so thin (about 10 mm), only a lightweight MLC is needed to "trim" the edges. "The advantages include fast set-up and motion because of the low weight and high-speed dose delivery," Brahme told medicalphysicsweb.
At the Karolinska Institute, his group is working on an integrated treatment unit that will combine PET, CT and radiation therapy. Eventually, the team hopes that the device will even be capable of delivering the dose in synchronization with moving tumours in real time, using a new type of fast 4D laser camera. "This system allows radiobiologically based adaptive therapy in real time by altering the scanning pattern and multileaf settings in synchrony with the movements of the target tissue," Brahme and co-workers explained in a recent paper in Medical Physics (34 877).
So far, Brahme and his team have developed the fast-scanning pencil-beam technology, which can be used with photons, electrons or light ions. To increase precision and minimize the dose received by the healthy tissue surrounding tumours, however, some sort of collimator is required to trim and sharpen the edge of the beam. In their latest work, the researchers describe the design of a fast, lightweight MLC that simulations suggest should be perfect for the job.
"The task of the collimator is to protect the patient laterally and to sharpen and cut the tails of the elementary scanned beam," the team explained in Medical Physics. "With this new role, the thickness of the collimator can be decreased … and there is no need for focusing edges."
The MLC for X-rays will be mostly made of steel, which is much lighter than tungsten, and it will have leaves that are just 15 mm thick, bringing the weight down to 10 kg. These slimmed-down leaves will be capable of much faster rearrangement, making possible the rapid delivery of biologically optimized treatments.
Next the researchers have to actually build one, and that is exactly what the group will be doing for the next few years. "It will be used in two new treatment systems, one for scanned electron and photon beams, and the other for light ions, which we are developing for the planned light-ion centre at the Karolinska University Hospital," Brahme told medicalphysicsweb.