Carbon vs. Proton for Innovative Applications of Particle Beam Therapy
Reporter: Abigail Berman Milby, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 21, 2012
William Chu, Berkeley National Laboratory (Moderator)
Stephanie Combs, University Hospital of Heidelberg
Thomas DeLaney, Massachusetts General Hospital
Daniel Gomez, MD Anderson Cancer Center
Eugen Hug, ProCure Proton Therapy Center
Roberto Orecchia, ULICE
Koji Tsuboi, PMRC
Hirohiko Tsujii, NIRS
Kwan Ho Cho, NCC, Korea
A panel discussion ensued regarding the advantages and disadvantages of proton and carbon ion therapy. The most salient points are summarized below, followed by a description of specific questions that were raised and the discussion that followed each.
Proton therapy and carbon ion therapy are both forms of particle therapy that can be used to effectively treat tumors. Studies performed thus far have shown approximate equivalence of both modalities, although the studies are limited by comparison across multiple institutions and different patient populations. The following table summarizes the relative pros and cons of each modality.
Proton Beam Therapy
Carbon Ion Therapy
Proton Advantages over Carbon
Carbon Advantages over Proton
Similarities of Proton and Carbon
- There are several areas in need of future investigation:
- Continued optimization of both proton therapy and carbon ion therapy.
- For proton therapy, this includes implementation of pencil beam scanning and intensity modulated proton therapy.
- For carbon ion therapy, further investigations regarding the relative biologic effect (RBE) and the isoeffective dose should be performed.
- Further research should be performed to see if it is possible to select proton vs. carbon ion therapy based on the alpha/beta values of the target and surrounding normal tissue structures.
- Ongoing randomized trials of proton vs carbon ion therapy include:
- Two randomized trials of chordomas and chondrosarcomas
- Randomized CLEOPATRA trial for glioblastoma multiforme
- Randomized trial for recurrent GBM
- Randomized PINOCCHIO trial for skull base meningiomas
Questions and Answers Addressed during Panel of Proton vs. Carbon Ion
Question 1: Do physical and biological differences between proton and carbon ions influence clinical results?
Answer 1 (Tsuboi): Biology does not sufficiently explain the differences between protons and carbon ions. Biologically, carbon ions have greater potential to cause greater DNA damage, i.e. a greater isoeffective dose. Isoeffective dose can be calculated by absorbed dose multiplied by the weighting factor, which includes radiation quality (linear energy transfer, LET, and relative biologic effect, RBE). Carbon ions are thought to cause less indirect effect (base damage) than protons; however, the double strand breaks may be increased. Therefore, the irreparable cluster damage may be greater with carbon, as well as apoptosis induction, and loss of clonogenicity. Overall, this may lead to greater tumor cell kill.
For skull base chordosomas and chondrosarcomas, results suggest similarity between treatment modalities. Overall survival is similar between fractionated RT, gamma knife, protons, and carbon, with survival approximately 74 to 88% at 5 years. There is, however, a suggestion that local control is higher with carbon at 73% at 5 years (DiMaio J Neurosurg 2011). Given that the patients were treated at different institutions and were different patient populations. Of note, this study excluded data from Massachusetts General Hospital and Paul Scherer Institute.
Continued questions about the biology of carbon ion therapy include optimizing the isoeffective dose and recalculating the quality of radiation based on what is known as microdosimetry. Future research directions include possibly deciding on the optimal ion based on alpha/beta values of the target and normal tissue.
Question 2: What is the radiobiologic effect (RBE) of proton vs. carbon ion therapy?
Answer 2 (Cho): Protons have a better dose distribution but a lower RBE (1.0-1.1) than carbon ions (RBE 1.5-3.4). However, RBE depends on many aspects of a delivery including radiation quality, LET, fraction size, and the biological aspects of the target. Uncertainty in the RBE can cause large variations in the actual delivered dose. While higher RBE is good for tumor control, it is bad for normal tissue toxicity. We learned this in the setting of neutron therapy (J Oral Maxillofacial Surg 2005) where the effects of normal tissues and variation of RBE were underestimated and there were many grade 3 and 4 toxicities.
The theoretical advantages of the carbon ion therapy have not been proven in the clinical setting. The uncertainties of RBE of the carbon beam can be a burden for physicians to make a clinical decision. Proton beam therapy has practical advantages including lower cost and precise geometric delivery of beams by using a gantry.
Counterpoint: The increased RBE with carbon ion therapy is seen in the Bragg peak itself, where the tumor is, and usually not in the lateral penumbra or areas of normal tissue.
Question 3: What is the role of carbon vs. proton for innovative clinical applications?
Answer 3 (Tsujii): Dose distribution may be better with carbon ion therapy, particularly if tumor sites that can be treated with fixed beam lines are compared. In addition, radiobiologic properties of carbon may improve local control of photon-resistant tumors. NIRS in Japan has extensive experience using carbon ion therapy for a wide range of malignancies. Examining carbon ion beam for hypofractionation in prostate cancer, the rates of GI and GU toxicity were very low. Over 1400 patients were treated to 66 GyE in 20 fractions with hormone therapy. Overall survival was found to be comparable to standard RTOG studies, although these are not identical patient populations. In addition, NIRS has treated patients with sarcoma with excellent local control and survival. The speaker states given the excellent local control, carbon ion therapy could theoretically replace surgical resection. NIRS has also started using carbon ion therapy for single fraction SBRT for stage I (T1-2) lung cancer with good early results. The discussants recommended a randomized clinical trial of carbon ion vs. proton therapy within Japan.
Question 4: What are the practical aspects of deciding whether to treat with proton or carbon?
Answer 4 (Hug): Many more patients have been treated with proton therapy (~95,000 patients) versus carbon ion therapy (~13,000 patients). Nonetheless, a greater percent of patients treated with carbon ion therapy have been treated on studies and reported in published manuscripts. All studies on both proton and carbon ion therapy are small and therefore difficult to compare. There are several sites that are suitable for randomized trials between proton and carbon. Sites should be chosen where the optimum treatment schema for both modalities has been established, not sites where the optimal treatment is debated.
Question 5: What is the role of randomized trials in proton and carbon ion therapy?
Answer 5 (DeLaney): Protons have a physical advantage over photons with a reduced integral dose by ~60%. Because the RBE is similar to photons, protons can be easily combined with photons in a course of treatment. While protons are accepted for pediatric patients, the role in the adult setting is more controversial due to the high cost of proton beam therapy. There are two ongoing randomized clinical trials of photon vs. proton therapy in the United Sates, one randomizing patients with locally advanced non-small cell lung cancer (MGH, MDACC); in this study, pneumonitis and local control are the primary endpoints. There is also a randomized trial of protons vs. IMRT for prostate cancer with the primary endpoint as EPIC bowel scores at 6 months (MGH, University of Pennsylvania).
One argument for both proton and carbon ion therapy is a reduction in secondary tumors over photon therapy. However, it may not be possible to perform a randomized trial to determine this answer as very large numbers of patients would be needed. Assuming a 60% decrease in 0.5% incidence of secondary malignancies at 15 years, approximately 6000 patients in each arm would be needed. In Ewing's sarcoma, where the absolute rate of secondary malignancies is more common, approximately 600 patients would be needed in each arm to demonstrate a decreased risk in proton vs. photon. This likely is not a practical study to perform prospectively, although a retrospective, multi-institutional study could be performed to compare the rates of secondary malignancies in Ewing's sarcoma with proton vs. photon therapy.
Randomized trials comparing proton and carbon ion therapy are possible, and, as further discussed, ongoing in Europe. In addition, it may be feasible to design a trial randomizing patients to a carbon ion boost.
Question 6: What is the state of randomized trials in particle therapy at the University of Heidelberg?
Answer 5 (Combs): There are currently two phase III randomized trials of proton vs. photon at HIT, one for chondrosarcoma and one for chordomas. Both of these trials are designed as non-inferiority trials. There is also the CLEOPATRA trial which includes GBM patients randomized to a standard arm of proton beam therapy of 60 Gy in 5 fractions versus 6 fractions of carbon ion therapy to the macroscopic tumor of T1-contrast enhancement. There is also a randomized trial of carbon ion radiotherapy versus stereotactic radiotherapy in recurrent GBMs (over 400 patients). The first part of the study will be a phase I dose escalation carbon ion therapy study with 10 Gy (RBE) x 3 fractions or 16 Gy (RBE) x 3 fractions. The winner of that arm will then be randomized against fractionated stereotactic RT to 36 Gy in 2 Gy fractions. Primary outcome will be overall survival. The IPI trial will look at low risk prostate cancer and randomize to protons (20 fractions x 3.3 Gy (RBE)) versus carbon ion (20 fractions x 3.3 Gy (RBE)). Primary endpoint will be toxicity. The PINOCCHIO trial will randomize proton and carbon ion therapy in skull base meningiomas with a primary endpoint of toxicity. This trial will also include a standard fractionation photon arm and a hypofractionated photon arm.