Clinical Trials in Particle Therapy

Reviewer: Eric Shinohara MD, MSCI
Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 26, 2008

Presenter: Bhadrasen Vikram, MD
Presenter's Affiliation: National Cancer Institute
Type of Session: Reporting


Dr. Vikram started his presentation with several pictures which demonstrated the precision with which radiation can be delivered to single cells. He showed a picture of several carbon ion beams hitting a single cell within a 10 micron area. He also showed pictures of an argon beam which was able to hit a single cell multiple times. He then stated that perhaps in the future we will be able to target and destroy individual tumor stem cells in vivo without damaging individual normal cells.

Dr. Vikram then went on to discuss the role of clinical trials in proton therapy. He states that the debate regarding the need for clinical trials should not focus on cost but rather on whether patients live longer or better with such treatment. He questioned whether there is any evidence that protons are any better than conventional radiation at present. He went on to discuss results from a panel of experts who evaluated available data on proton therapy in 2006. They found that there was no level I evidence suggesting a survival benefit or a decrease in toxicity in prostate, breast, head and neck, CNS, lung, cervix, and colorectal cancers treated with protons. He does point out that the lack of evidence does not mean that there is no benefit. He notes that the strongest argument for the use of protons can be made in pediatric patients but the majority of patients that are treated with protons are prostate patients.

Dr Vikram then goes on to discuss results from three major prospective randomized prostate trials. He states that the major conclusions of these trials are that protons and dose escalation have worse toxicity. He notes that neither protons nor dose escalation improved survival. He presented toxicity data from Massachusetts General Hospital:

Uretheral Stricture

Rectal Toxicity

3D Conformal






He argues that theses toxicities are likely to continue to increase as the time from treatment increases.

He discussed data from MD Anderson that demonstrated less GI toxicity in patients treated with 70 Gy compared with 78 Gy and noted that there was better survival in the 70 Gy arm. He then presented data from Memorial Sloan Kettering as follows:

70 Gy

79 Gy

Overall Survival



GI Toxicity



GU Toxicity



He again notes that the toxicity is likely to increase further out from treatment. He also noted that survival was similar despite the higher toxicity with higher dose.

He mentions several implications based on his interpretation of the data presented above. First, he believes that the use of PSA as an endpoint in prostate cancer is dangerous. He believes that using PSA as a surrogate marker for survival has provides a false positive signal for these studies. He believes that the Phenoix and ASTRO definitions of PSA failure are invalid and should not be used. Secondly, he believes that the use of dose escalation in the treatment of prostate cancer is likely hurting patients. He then mentions that in 71% of randomized controlled trials run by the RTOG results favored the standard arm rather than the experimental one. He notes the same case in 53% of studies run by the COG. He therefore concluded that the value of a new treatment can not be predicted accurately in advance. He then mentions that a trial that he believes is worth running is to have patients randomized to high dose proton therapy without androgen ablation compared with patients treated with androgen ablation and 3D conformal radiation.

He discusses the process by which a new technology should be evaluated with prior to its adoption. First, data showing that the new technology enables patients to live longer or better should be provided. In the absence of this, if the technology is already FDA licensed, immediate trials should be performed. He concludes that manufacturers can not be held accountable for this and ultimately physicians must demand this data from the manufacturers or demand funding to conduct such trials. He believes that new technology in radiation oncology should first be shown to have a benefit “in silica” (dosimetrical). The next step is to ensure that there is consistent planning and execution of the plan (he believes that IMRT is at this point). Lastly, the new technology must demonstrate the ability to improve survival or decrease toxicities. Problems Dr. Vikram described regarding the application of new technologies included: First, advanced techniques are less tolerant of poor administration. Second, misadministration is harder to detect. Third, in vivo dosimetry is not possible at present. He notes that any new technique with increased conformality (i.e. IMRT/stereotactic radiosurgery/proton therapy) increases your risk of missing the tumor.

Dr. Vikram then describes papers which have studies consistency in radiation administration. He describes a recent paper by Das et al. (JNCI 2008) which collected data from 803 patients treated at five institutions where the IMRT treatment plans were done by experienced physicists. They found that in 46% of patients the maximum dose the CTV received was at least 10% higher than that prescribed by the physician and the largest deviation seen was 40% greater than prescribed. 63% of patients received at least 10% less than the prescribed dose to the CTV with the greatest deviation being 0% of the prescribed dose being delivered to the CTV. He then describes a paper that studied phantoms. He described phantoms as the “ideal patient” as there is no respiratory motion or changes in tumor volume. 128 institutions who were members of the ROTG were studied. They were assigned the task of designing a treatment plan for the phantom and delivering dose to the CTV to within 7% of the prescribed dose. 1/3 of institution failed to do this in an “ideal patient.”

Given the difficulty in achieving proper coverage with IMRT it is critical, if margins are to be tightened with proton therapy, to know the position, size and depth of the target. Dr. Vikram feels that current imaging modalities are inadequate and the highest priority in radiation oncology should be improved imaging. He points out that target delineation can often feel arbitrary based on data from prior studies. He mentions that there have been head and neck studies where eight experts in the field drew volumes for the same patient and the overlap in volumes between the experts was only 50%. A similar study of axillary fields for breast treatment demonstrated a 45% overlap between eight experts. Prostate iliac node volumes for a patient ranged in volume from 82 to 877 cc with only 30 cc’s overlapping between physicians. He points out that with this disparity in tumor volumes is it safe to tighten margins? Do our current wide margins save us? He wonders if atlases or a credentialing mechanism would help decrease variability. He points out that the NCI developed a protocol outlining a minimal standard for IMRT in 2004. Any institution who wishes to enroll patients on an NCI protocol much be certified. However, he points out, at present, the FDA requires no such credentialing.

Author's Conclusions

In summary Dr. Vikram feels that particle therapy has the potential to help patients but clinical trials with a good QA (Quality and Assurance) mechanism are needed. Currently, the NCI has a credentialing mechanism but centers which are not part of the NCI system are not held to a standard QA system. Without a good QA system, the development of particle therapy could be delayed, or even worse, given a bad name based on faulty data from a poor QA system. He ends by stating that a good clinical trial should have a clear hypothesis and sample size calculation. He does note that there is a role for retrospective analysis in generating hypotheses.

Clinical/Scientific Implications

IMRT has developed in large part without rigorous clinical trials. This may be due to the general feeling that it is the same radiation that is used in 3D conformal radiation, just with different delivery. However, there has been a large shift in how tumor volumes are defined in 3D conformal radiation compared with IMRT. Dr. Vikram does cite studies which have shown that the definition of targets in IMRT can be inconsistent and that the delivery of the IMRT can also be inadequate even at top institutions. Given that volumes are likely to shrink even further with proton therapy it is important to set up some basic standards for proton based clinical trials. Due to the cost of protons their use will likely need to be justified by either an improvement in survival or a decrease in side effects. Without a strict QA program it may be impossible to detect such benefits. Designing such standards will require collaboration across numerous countries and organizations but may be critical in the proper development of particle therapy.


Cancer on Your Own Terms
by Bob Riter
October 29, 2014

Cancer Center Advertising: Doing More Harm Than Good?
by Rodney Warner, JD
July 21, 2016