DOS 541 - Week 2 Discussion
Please summarize this information and contribute your thoughts in a short essay. Provide a reference or references to support your view of how Hall's information should be interpreted by the radiation oncology community.
Initial Post: Quantitative Risk Assessment in Radiation Therapy
Ionizing radiation is a double-edged sword. At the same time, it offers well-developed diagnostic and therapeutic utility, and it is a source of harm to the human body.1 As the usage of ionizing radiation in medicine increases, it is important to consider the harmful side effects, many of which take years or decades to manifest, and as such are difficult to model.
The end of World War II, brought about by the use of two atomic weapons on Japanese civilian populations, provided a large pool of information about the effects of radiation on the human body. Survivors at varying distances from the blasts received radiation doses in the range of a few Sieverts at the time of exposure. These survivors have been followed closely in the 70 years since the events. The conclusions about the relationship between dose and disease incidence rates have had to be revised multiple times over the decades because many effects do not manifest until old age. Survivors who were only children at the time of exposure may not have suffered the carcinogenic consequences of their exposure until they reached their senior years.
Current models suggest that the rate for radiation-induced cancer has a fairly linear relationship with dose in the range of 0.1 to 2.5 Gy, but data at lower doses and at higher doses is less well-modeled. As Hall points out, the ranges above and below are the most relevant to the practice of medicine, because diagnostic procedures tend to happen in the range of 10s of mGy, while radiation therapy tends to be in higher ranges of 20-80 Gy.
Hall points out that in the study of high dose effects on radiation-induced secondary malignancies, an interesting phenomenon appears relating to scatter dose. In a pool of 50,000 patients who received radiation therapy for prostate cancer, 34% of secondary malignancies occurred in the lungs, which are remote from the treatment site. The connection may be leakage and scatter dose from the treatment head of the linear accelerator (linac), which can be around 2 Gy for a full prostate treatment course. This surprising finding brings up important questions about the increased usage of intensity-modulated radiotherapy (IMRT) for prostate treatment, because IMRT uses complex arrangements of multileaf collimators (MLCs) to create tightly conforming dose profiles, at the expense of doubling or tripling scatter dose relative to standard 3D planning because of increased on-time while the MLC patterns are produced. Some options exist to reduce scatter, such as the use of flattening filter-free (FFF) treatment, which removes the high-density metallic flattening filter from the beam path, resulting in lower scatter at larger field sizes.2 Since MLCs are used to control beam fluence, a flattening filter is not necessary in IMRT treatment.1 FFF treatment can also help reduce on-time for IMRT treatments because the dose is not attenuated by the filter, and the desired dose can be delivered more quickly, especially at the center of the field. Even with these improvements, IMRT may still be a problem. Hall's paper states that the use of IMRT doubles the risk of secondary malignancies.
This doubling of risk must be taken in context. Most of the patients receiving IMRT prostate treatment are already elderly, and secondary malignancies may not have time to manifest before another cause of death claims the patient. Furthermore, doubling a low risk value is still a low risk value, although zero risk is always preferable. Risk must be evaluated against benefit, and curing primary disease weighs heavily on that equation.
For younger patients, the risk of secondary malignancies is much higher, both because of their longer predicted lifespans and because of a trend found in the atomic bomb data that shows higher sensitivity in younger people. For these patients, the tradeoffs considerations are more difficult.
One option may be to sidestep the problem altogether. Hall points out that many of the problems with radiation leakage and head scatter in X-Ray-based therapy can completely eliminated by a particle beam instead of an energy beam, such as in proton therapy. While the X-Ray scatter dose problems are eliminated, proton therapy has the potential to produce its own unique set of problems. The main problem that arises in proton therapy is neutron production. Neutron interaction can have extremely high radiobiological equivalent (RBE) depending on the energies involved. Hall suggests that scatter neutrons from proton therapy likely have an RBE in the neighborhood of 25, whereas X-Ray and electron contamination has an RBE of 1.0. This means neutron pack a heavier punch in the tissues where they interact.
In much the same way that the treatment head of a linear accelerator is the main source of scatter radiation, a proton beam nozzle can produce neutron contamination during the process of shaping the beam for treatment. Different designs have different levels of complexity and neutron production. The simplest design, used in older generation proton facilities, uses a passive scattering foil to expand a tiny proton beam into a clinically useful field size. A more complex design that uses steering magnets to scan the narrow beam uniformly across a designated field size is unsurprisingly called uniform scanning. This design dispenses with the scattering foil and its neutron contamination, but it shares another problem with passive scattering. Both passive scattering and uniform scanning use low-density compensators to control beam penetration depth and high-density apertures, usually made of brass, to control the shape of the field projection. Proton interactions with the brass aperture are the largest source of neutron contamination in both systems. A still more complex system called pencil beam scanning (PBS) uses steering magnets to shoot individual spots at designated locations and depths without the need for apertures and compensators. The dose from each spot adds to dose deposited by neighboring spots, allowing the creation of spread-out Bragg peaks and 3D dose sculpting (Kent McCune, Oral communication, March 2014). This system drastically reduces neutron contamination because the beam travels directly to the patient without having to first pass through any high-density material.
Since the time of Hall's publication in 2009, the number of operational proton centers in the United States has more than doubled, and more are still under construction.3 Hall expressed concern that proton therapy was trading one problem for another because passive scattering was the most common type of delivery used at the time. In the years since his publication, PBS has matured significantly and is becoming the standard operating mode for new centers because of its greater dose shaping capability, its significantly reduced neutron contamination, and the speed and cost advantages of not having to manufacture patient-specific apertures and compensators (Tony Wong, Oral communication, June 2014). At the Seattle Proton Therapy Center, more than two thirds of the patient load is planned and delivered with PBS, resulting in higher patient throughput and lower planning costs, plus the previously mentioned dose sculpting and low-dose scatter reduction benefits to the patient.
At the lower end of the dose scale, dose delivered through diagnostic imaging comes in many flavors. As Hall points out, the use of computed tomography (CT) and nuclear medicine (NM) is a significant source of low-dose exposure to the population, together accounting for 68% of diagnostic radiation dose. Barium enema studies of the GI tract and fluoroscopic interventional procedures account for another 22%, and the millions of other imaging studies only add up to another 10%, making it clear where the high-dose culprits are. While the risks incurred in the usage of these imaging modalities are small, they are not zero. Once again, the risk must be weighed against the benefit that the study will provide.
- Hall EJ. Is there a place for quantitative risk assessment?. J Radiol Prot. 2009;29(2A):A171-184. http://dx.doi.org/10.1088/0952-4746/29/2A/S12
- Richmond N, Allen V, Daniel J, Dacey R, Walker C. A comparison of phantom scatter from flattened and flattening filter free high-energy photon beams. Med Dosim. 2015;40(1):58-63. http://dx.doi.org/10.1016/j.meddos.2014.10.001
- Where to get PT? Proton Therapy Today Website. http://www.proton-therapy-today.com/where-to-get-pt/. Accessed Septemer 9, 2015.