Very high-energy electrons (VHEEs), typically defined as those above 40 MeV, provide a potential new radiotherapy modality with dosimetric advantages. Beams of such electrons penetrate deep into the patient, enabling treatment of deep-seated tumours that photon-based irradiation may not reach.
Speaking at the Medical Physics & Engineering Conference (MPEC), Louie Hancock from the University of Manchester described the recent resurgence in interest in VHEE radiotherapy. “Over the last couple of decades, new linac designs mean that it’s now possible to produce roughly 200 MeV electrons in about two to three metres, whereas before it might have taken 20 metres or so,” he explained. “This has spurred interest in using these VHEEs for treating deep-seated tumours.”
Hancock noted that a linac-based VHEE treatment system should be compact enough to fit into a hospital bunker. “It’s much cheaper to put a machine in an existing bunker than to build a new building,” he pointed out. “I expect that VHEEs will probably be more expensive than photons to produce, but cheaper than protons.”
While there are currently no clinical systems available, there are electron accelerators for research use, such as the high-gradient X-band linac at CERN’s CLEAR facility, for example, and the CLARA electron accelerator at Daresbury Laboratory. In the meantime, Monte Carlo simulations can provide insight into VHEE treatments without having to actually build a machine.
Depth–dose curves show that, as well as delivering dose deep inside a patient, VHEEs should also be extremely resilient to changes in patient geometry. To confirm this, Hancock performed a back-of-the-envelope calculation for a simple block of water containing a cylindrical 5 cm tumour in its centre.
He simulated treatment of the tumour by rotating a beam around the target, and then introduced a cavity above the tumour (a 5 cm sphere of bone or air) to calculate its impact on the tumour dose.
For treatments simulated using 1.3 MeV X-rays, the introduction of an air bubble resulted in hot spots of about 15%, while an unexpected bony region led to cold spots of about 10%. For 250 MeV VHEEs, while hot and cold spots were seen, they were only of about 2%. “In this simple situation, VHEE appears to be nearly an order of magnitude more resilient to unexpected homogeneities when compared with X-rays,” said Hancock.
But will this advantage translate to patients? To find out, Hancock and colleagues examined a clinical case, comparing treatment plans for VHEE and volumetric-modulated arc therapy (VMAT).
VHEE treatment planning is a highly complex process, with millions of variables creating a huge optimization problem. As such, the team at Manchester has developed an open-source treatment planning system for VHEEs that incorporates various tools to create a full planning workflow.
This includes organ identification using Slicer 3D software, Monte Carlo dose calculations using Geant4, and then optimization of the generated dose profiles with Python software written by Hancock.
Using their VHEE code, the researchers created a treatment plan for a patient with cervical cancer, which included two large target sites to irradiate and many nearby organs to avoid. They compared this with a VMAT plan creating using Monaco.
The two plans delivered identical dose to the target, while nearby organs (the sigmoid colon, bowel and bladder) received slightly less dose from VHEE than VMAT. For organs further from the tumour, VHEE conferred significant dosimetric advantages, particularly to the femoral heads where it reduced the delivered dose from 35 to 15 Gy.
“We achieved coverage of all of the tumour with VHEEs, so can say for sure that these electrons are capable of treating a large tumour deep inside the patient,” said Hancock. “We could also see that the low-dose background was reduced compared with the VMAT plan.”
To examine the impact of geometric changes, the team re-simulated the treatment plans with the rectal cavity filled with air instead of water. The error between the two VHEE plans was roughly 0.15 Gy, while for the X-ray plans it was about 0.7 Gy. “In the simple water phantom, we saw a dose error nearly an order of magnitude lower with VHEEs, now we’ve seen this effect in silico in an actual patient case,” explained Hancock.
The next step, he said, will be to repeat the analysis with lung and brain cases, and “move towards putting real things in real beams”. The real thing is the MARVIN human head-and-neck phantom, while the real beams will be delivered by CLARA at Daresbury and CLEAR at CERN. These two facilities provide a wide span of electron energies and will enable VHEE measurements in realistic environments.
Initial experiments using a 45 MeV electron beam at CLARA to irradiate MARVIN demonstrated a near-uniform dose over a large area inside the phantom. Hancock noted that the study is currently on hold due to the pandemic.
Hancock concluded that the clinical test case, using the team’s VHEE treatment planning software, demonstrates that VHEE radiotherapy has the ability to treat tumours that are both deep and large. “VHEE might well also be more insensitive to inhomogeneities and changes in patient geometry than photons, which I think is likely to be clinically beneficial,” he added.