Reporter: J. Taylor Whaley, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 25, 2012
Moderator: Jay Flanz, Massachusetts General Hospital
Pedroni Eros, Paul Scherrer Institute
Clasie Ben, Massachusetts General Hospital
Zhu Ron, M.D. Anderson Cancer Center
Stephanie E Combs, University Hospital of Heidelberg
Kamada Tadashi, NIRS
This particular panel discussion focused on particle beam scanning, which is an innovative application of particle beam therapy. Particle beam therapy in today's realm generally refers to proton and carbon ion therapy. While particle beam scanning was initially realized to be a possibility approximately 30 years at the NIRS, it has taken several decades to translate this from the world of research and theory to clinical applications. From 2008-2012, interest in particle beam scanning has risen exponentially.
To begin, particle beam therapy utilizes the acceleration of heavy particles to a particular depth within a patient that is determined by the energy of the beam. When particle therapy is utilized, the vast majority of energy is deposited at the end of the beam; this phenomenon is known as the Bragg peak. By utilizing the advantages of particle therapy, including the lack of exit dose secondary to the Bragg peak, radiation treatments can spare more normal tissue than photon-based radiation.
Because the Bragg peak is a sharp narrow peak, this must be spread out to cover the entire tumor volumetrically. One mechanism of creating this Spread Out Bragg Peak is known as passive scattering. Passive scattering uses a range modulation wheel, compensators, and collimators to modify the beam to match that of the tumor volume.
Although nearly all particle therapy treatments over the past few decades have utilized passive scattering to shape the beam to match an individual's tumor, there are several disadvantages of passive scattering. The first is the creation of the individual compensators. These are brass structures that must be created for each patient and for each field that is treated. This is costly to produce and increases treatment time while the patient is on the table as the compensators must be changed for each field that is treated. The second disadvantage of passive scattering is the creation of secondary neutrons. As protons interact with the components in the beam line, nuclear reactions can occur and neutrons can be created. While the amount of neutrons is quite low, this remains a concern as neutrons are thought to potentially lead to secondary cancers.
The second mechanism for creating a SOBP to cover a tumor is known as active scanning. This involves changing beam energies electronically. It entails a very high level of technology that is only currently in development. While the beam remains on, the beam is delivered one "spot" at a time and the energy is changed dynamically as needed to cover the depth of a target. With active scanning, while the beam energy is adjusted to compensate for depth changes in the target, magnets are used to aim the narrow beam and "paint" the target in layers. Active scanning has the potential to create a much more conformal radiation plan than passive scattering. One of the disadvantages of active scanning is the vulnerability of accurately targeting a moving tumor, which can be problematic with the scanning beam. As proton therapy continues to evolve, most future treatments will be delivered with the active scanning method.
Before the panelists began discussing how their particular centers were incorporating scanning techniques, the moderator wanted to define several confusing terms. Spot scanning refers to the act of irradiation delivery one spot at a time. The beam is then temporarily stopped (for milliseconds) while the beam is moved to the next location. The beam is then turned back on. In contrast, raster scanning refers to irradiation of one spot; however, the beam remains on while moving to the next spot. Raster beam is obviously much faster but requires improved technology. Time driven scanning refers to a combination of time and intensity determining the dose at a given location. SFUD, or single field uniform dose, delivers a uniform dose to the target from each field. In contrast, IMPT, or intensity modulated particle beam therapy, involves modulation of the dose across each field- this could be called single field non-uniform dose.
So why has so much attention been placed on the scanning method to deliver particle therapy? Particle beam scanning is more conformal than passively scattered particle therapy. The treatment plans can decrease the proximal dose to normal structures and the conformity of the dose distribution can be tighter. A second reason scanning is preferred is the lack of patient specific hardware that must be created (because compensators and modulators are not needed) and fewer beam angles that may be needed in some tumor locations. While these advantages exist, for scanning to become practical, several new innovations from photon-based radiation will need to be adapted to work in particle therapy. These include adaptive therapy planning, organ motion management, more precise end of range accuracy, and increased real-time image guidance.
The panelists each discussed current innovations in particle beam scanning at their institution. These are briefly summarized below:
Dr. Eros from PSI in Switzerland reported on the initial experience at his institution. At PSI, they have now been treating humans with passively scattered proton therapy since the mid-90s and have accumulated nearly 20 years of experience on their Gantry 1. Today, one-third of the PSI patients are pediatric patients. PSI has now developed discrete pencil beam scanning (spot scanning). Again, this is a scanning technique in which the beam stops and then moves after delivering dose to each spot. Due to uncertainties with the continuous beam, they have yet to incorporate this. With the arrival of their second gantry which delivers a variable energy beam, they have attempted to tackle the problem of using scanning with organ motion. The largest issue they are currently addressing is the speed with which the beam energy can change. They have now reached a remarkable speed of only 80 milliseconds needed to change between various beam energies.
Dr. Ron from MDACC reported on their experience with proton therapy as well. Beginning in May 2008, they began using the scanning technique in treatments. Initially, a majority of patients were prostate (79%), with only 12 % pediatrics and 7% head and neck tumors. However, since 4/2012, they have limited prostate to 50% of patients. MDACC delivered its first IMPT plan in November 2011. For treatment planning, they typically use a single field optimization technique in which each field is optimized to deliver prescribed dose. From a research standpoint, they have focused on plan robustness which is critical for scanning beam. Currently, they have several tumor sites for which they utilize scanning therapy. These include high risk prostate cancers when the seminal vesicles will be included in the target volume. They are also treating base of skull tumors with the scanning beam. For both these tumor locations, they routinely using single field integrated boost. MDACC have just recently begun utilizing multi-field optimization for some H&N cancers. From a clinical standpoint, using IMPT has resulted in improved clinical outcomes with decrease acute toxicity. To compensate for uncertainties with plan robustness with respect to range and setup errors, they have assumed range uncertainty of 3.5%. In the near future, they hope to improve robustness optimization for SFO and MFO. At this point, they have now treated 700 patients with the scanning technique.
Dr. Combs from HIT in Heidelberg described their unique situation with both protons and carbon ions. She believes that in order to achieve our goal of improving local control with particle therapy, we must have proper technology. While most patients treated with particle therapy have been treated with passive scattering, it clearly has its limitations. With active beam scanning delivery, the dose distributions and tumor conformity can be optimized by superimposing thousands of Bragg peaks. At HIT, they utilize a pencil beam library with 253 energies and 7 spots and 15 intensity steps. They have developed the raster scanning method. Again, the raster scanning technique allows the beam to remain on while change energy and location. This requires faster scanning magnets to move the beam throughout the target volume. Since 2009, they have treated 900 patients with scanning beams. In order to compensate for moving targets, they use larger spot with their scanning technique. Also, because HIT only uses horizontal beams because they have no gantries, their beam angles for planning are limited. They are working to commission their first particle gantry. Finally, as they expand to tumors in various parts of the body, moving targets offers a new set of challenges with a particle scanning beam. They are currently studying gating and have now treated the first patient with gating on carbon ion therapy. They are currently accruing for a current dose escalation trial with hepatocellular carcinoma.
Dr. Clasie from MGH commented on why pencil beam scanning was developed at his institution. As covered above, the dose distributions are improved proximally, more uniform match lines (and therefore less cold spots), and deeper tumors can be treated compared to double scattering with less modulating material in the treatment nozzle.
Although double scattering proton therapy can provide excellent dose distributions, to reach dose distributions similar to active scanning requires more skill, effort, and time. This is because of the field size limits in double scattering required matched fields which must be patched together. In order to develop the scanning technique, they use an in-house Harvard specific treatment planning system, known as ASTROID. In their particle therapy clinic, they use a fraction group for daily field combinations. This is a technique where particular fields are only treated on certain days of the week. This allows improved throughput on treatment machines. Additionally, they can now change beam energies in 0.2 seconds. During treatment planning, a multi-criteria optimization is performed. With proton beam scanning, they are currently treating sarcomas, mesotheliomas, and chordomas. They have found that proton beam scanning is also an economically more efficient way of treating patients as planning takes only 20% of the time that double scattering plans require, no time is needed make individualized compensators, and finally, treatment time is shorter with no compensators needed to be inserted in the proton gantry. Future directions for Harvard's proton therapy team include organ motion managements, robust optimization, improved efficiency and simplified quality assurance procedures. They also plan to extend the scanning technique to treating H&N, gynecologic tumors, and medulloblastomas.
Dr. Tadashi from the NIRS in Japan remarked on his department's success in treating patients with carbon ions. To date, they have now treated more than 7000 patients with passively scattered carbon ions in the past few decades at National Institute of Radiologic Sciences in Chiba Japan. Their first patients were treated in 1979 with 2-dimensional therapy. They began carbon ion scanning treatments in 2011. Although they have been successful as many of these above programs, they too commented on the needed improvements in technology that are soon to be developed. Their future plans in scanning include the treatment moving targets as soon as 2012, and they hope to begin dose painting with carbon ion scanning in 2014. While one half of their treatments with passive carbon were delivered with gating in the past decade, they continue to search for best technique for hitting a moving target with the scanning technique. Clinically, they recently completed a hypofractionated trial with 46 patients using 51.6 in 12 fractions for prostate cancer. Their current limitations include the lack of a rotating gantry, similar to HIT. As all the speakers have stated, the future for particle beam scanning is very bright but much more research needs to be done for quality assurance. Additionally, more clinical trials need to occur for scanning beam as risks of missing the tumor certainly exist.