Proton Radiotherapy

9.1

Introduction

First proposed by Robert Wilson in 1946, radiation therapy with proton beams quickly became a reality with the first patient treatments performed in 1954 at Lawrence Berkeley National Labs. Starting in 1961 the Harvard Cyclotron Lab began treating patients (and continued up to 2002), and in 1990 the first hospital-based system became operational at Loma Linda University Medical Center, California, which featured the first gantry unit. As of 2014 there are 42 facilities in operation worldwide (14 in the USA), 25 under construction (9 in the USA), and at least 11 in the planning stages, according to statistics from the Particle Therapy Cooperative Group (PTCOG). Further historical background can be found in International Commission on Radiation Units Report 78 (ICRU-78) and a personalized history in Smith.

Though numerous planning studies suggest dosimetric advantages to proton versus photon radiotherapy, support in the form of clinical trials has been lagging. 2007 saw the publication of three large systematic reviews of clinical experience. One of these was updated in 2012 and determined that the conclusions had not changed from the original report. Also published in 2012 was a report from the Emerging Technologies Committee of the American Society for Radiation Oncology (ASTRO).

One of the major issues identified in these reports is lack of prospective clinical trials. Randomized controlled trials (RCTs) are an important tool for evaluating health technology, and they address the inherent biases in retrospective studies. There are, however, few RCTs of proton radiotherapy. The review by Olsen et al. identified only four RCTs involving proton radiotherapy. RCTs in proton radiotherapy have included fewer than 700 patients or approximately 1% to 2% of all patients who have received proton radiotherapy. Further studies are under way, an example being Radiation Therapy Oncology Group (RTOG) 1308, which is a phase III trial of photon versus proton chemoradiotherapy in inoperable Stage II-IIIB non-small cell lung cancer with an endpoint of overall survival.

There is a lively debate as to the feasibility and even ethics of randomized trials with proton radiotherapy, given that a better dose distribution often can be achieved with particle beams compared to photons. The ethics of such trials are driven by the need for “clinical equipoise”—that is, a genuine uncertainty as to which method of treatment is superior. Although perhaps not everything need be tested in phase III trials, it has been suggested that more convincing data are needed for proton radiotherapy and that further clinical trials and/or registry data will be crucial.

Based on the balance of evidence, the review studies noted above conclude that proton radiotherapy has a rationale in some situations. The ASTRO emerging technologies report highlights the following disease sites where proton radiotherapy may provide an advantage:

• ▪

  Central nervous system (CNS): Potential sparing of critical structures and reduced overall integral dose to brain tissue are key advantages. Protons have been shown to be effective for meningiomas and skull base tumors such as chordomas and chondrosarcomas.

• ▪

  Ocular melanoma: Local control rates of approximately 95% have been shown and there is particular benefit for large and/or posterior lesions where brachytherapy is more challenging. Special low energy beam lines (≤70 MeV) have been developed for this application.

• ▪

  Lung cancer: Use of proton radiotherapy may reduce dose to normal lung (essentially no dose to the contralateral lung) and also spare the esophagus. It may offer some advantage for stereotactic body radiation therapy (SBRT). Respiratory motion and associated density changes are a particular challenge for proton radiotherapy of lung.

• ▪

  Gastrointestinal: Though largely untested outside of hepatocellular carcinoma, there is a rationale for proton radiotherapy in esophageal and pancreatic cancer for decreasing dose to duodenum, liver, stomach, and kidneys, since these are dose limiting structures.

• ▪

  Prostate cancer: Though this site has the most patients treated with proton radiotherapy, the possible advantages are still not proven in clinical trials. Two of the 4 phase III proton trials mentioned above involved prostate cancer, but they were not designed to directly compare proton and photon radiotherapy so no firm conclusions can be drawn. Proton radiotherapy represents an option for this site though it is not proven superior to photons.

• ▪

  Head and neck cancers: Proton radiotherapy offers potential sparing of critical structures, especially in targets near the base of the skull. It must be noted that air cavities/heterogeneities in the head and neck region make planning challenging.

• ▪

  Pediatric cancers: The reduction of integral dose makes proton therapy very attractive for pediatric patients because this may lower the rate of secondary malignancy and spare the developing tissues. Homogeneous doses can also prevent growth-related malformation (e.g., uneven vertebral growth). In the pediatric population over 50% of solid tumors arise within the CNS.

Given the lack of level I evidence for the use of proton therapy, much attention has been given to its cost-effectiveness. In a widely cited study in 2003, Goitein and Jermann estimated that proton therapy is 2.4 times as expensive per fraction as intensity-modulated radiation therapy (IMRT) using photons, with costs being dominated by business costs associated with facility construction (estimated to be 42% of operations costs versus 28% for a photon facility), though if the recovery of initial costs was not an issue the total cost ratio could drop to 1.6 to 1.3. In the USA, under the reimbursement rates currently dictated by insurance, a relatively limited number of proton facilities may be independently viable, and these may be restricted to major population centers. It must be noted, however, that the technology continues to evolve rapidly and compact single-room devices have now appeared. Costs may come down, which would change the equation.

Finally, in addition to proton beam radiotherapy, therapeutic beams with heavier ions are also in use (helium, carbon, and others). These may provide biological as well as dosimetric advantages due to the higher relative biological effectiveness (RBE) compared to protons and photons. As of this writing there are at least six facilities in operation, the newest being HIT in Heidelberg, Germany (2009), and CNAO in Pavia, Italy (2011). Compared to proton radiotherapy, however, there is less well-established clinical experience with heavy ion therapy.

9.2

Physics of Clinical Particle Beams

Unlike electrons, which are roughly 2000 times less massive, protons travel in relatively straight lines as they penetrate through tissue. As they travel they gradually slow down. As a proton slows down it loses energy, slowing down even further. (Recall that the energy loss per unit length, or stopping power, is proportional to 1/ v ² to first order.) Through this energy-loss process the proton slows down continuously as it penetrates through tissue. When the proton reaches the end of its range it becomes very slow, and the stopping power correspondingly becomes very large. This results in a sharp rise in deposited dose near the end of the proton range, called the “Bragg peak.” The energy of the proton determines where this Bragg peak is and how far the proton penetrates, or its range. Distal to this point essentially no dose is delivered. Data tables showing proton ranges and stopping powers for the continuous slowing down approximation can be found on the PSTAR section of the National Institute of Standards and Technology (NIST) website.

A typical “pristine” Bragg peak is shown in Figure 9.1 (colored curves). Because a single pristine Bragg peak is typically not large enough in extent to irradiate an entire target volume, clinical systems create a spread-out Bragg peak (SOBP) by essentially adding many pristine Bragg peaks together. This principle is illustrated in Figure 9.1 . The highest energy proton beam used determines the distal-most extent of the SOBP (red) while lower energy beams penetrate less deeply (e.g., violet). The end result of adding these beams together, each with the appropriate weight, is the SOBP (black). The SOBP irradiates a large “plateau” region to a uniform dose. From this figure the main advantage of proton beams is clear: that is, essentially no dose is deposited distal to the end of the SOBP. This results in greater normal tissue sparing in the intermediate to low dose range.

In actual practice the creation of an SOBP is achieved with a rotating stepped wheel or a filter with ridges (see the following description). The wheel acts to produce many pristine Bragg peaks with varying ranges. By selecting the appropriate maximum beam energy and modulation wheels one can control the shape of the SOBP as illustrated in Figure 9.2 . One disadvantage to note, however, is that skin sparing may be almost completely absent in a proton beam. The magnitude of skin dose depends on the modulation and other factors but the effect can be appreciated in Figure 9.2 . In some shallow target sites such as sacral chordoma, the skin dose is so high (close to 100%) that a combined proton plus photon plan is sometimes used.

The nomenclature of proton beam parameters is an important consideration, and a valuable reference is ICRU-78 Section 3.4. One key parameter is the range of the beam. The d90% depth is commonly used, though ICRU-78 advocates the use of d100%. Another parameter is the amount of ripple in the flat plateau portion of the SOBP and also potential bumps or dips on the leading or distal edge of the SOBP.

Beams of heavier ions such as carbon display many of the same characteristics as proton beams outlined above. They exhibit a Bragg peak that is even more enhanced than a proton beam, but they also deliver a small dose beyond the Bragg peak due to nuclear spallation products along the beam path. One important difference is the higher linear energy transfer (LET) throughout the entire depth that results in potentially more biological damage. Heavier ion beams also display a sharper penumbra due to less scatter (for a study of this see Kempe et al. ).