Randomized trials comparing proton beams with standard radiation for the treatment of prostate cancer and other common tumor types are years from completion, but healthcare providers around the nation are betting billions of dollars that the greater accuracy of proton beam therapy will justify the greater costs.
Anthony L. Zietman, MD
Randomized trials comparing proton beams with standard radiation for the treatment of prostate cancer and other common tumor types are years from completion, but healthcare providers around the nation are betting billions of dollars that the greater accuracy of proton beam therapy will justify the greater costs. The nation’s first proton beam center opened at Loma Linda Medical Center in 1990, but the building boom that has brought the total number of American facilities to 20—most of which can treat more than 100 patients a day—began less than a decade ago, just as some private insurers began citing new research to stop covering proton therapy for prostate cancer treatment.
Despite the financial risk of losing most patients with prostate cancer to conventional radiation therapy, the building boom continues. Another 16 proton beam facilities are either under actual construction or in the advanced stages of planning, according to the National Association for Proton Therapy1—a fact that observers attribute to everything from the pressure that major cancer centers feel to offer all treatment options to the belief that research will ultimately vindicate a treatment option that irradiates significantly less healthy tissue than any other approach.
Anthony L. Zietman, MD, a noted prostate cancer researcher who is the Shipley Professor of Radiation Oncology at Massachusetts General Hospital, offered a thumbnail perspective on the challenges facing the implementation of the technology.
“Virtually no one was building proton beam facilities when they were seen as primarily offering value for rare pediatric cancers. These had little profit potential because of the slow and complex nature of the treatment,” said Zietman in an interview with OncologyLive. “Some facilities therefore began recruiting prostate cancer patients, who are substantially more numerous, and quickly realized that prostate cancer was a gold mine. Prostate treatment requires no anesthesia, so they could treat six prostate patients in the time it took to treat one child, but they could still bill the same amount for each treatment.
“But the prostate-cancer model is falling apart. New guidelines that call for less aggressive screening and, in many cases, no immediate treatment when cancer is detected have reduced the number of prostate cancer patients getting any sort of radiation,” Zietman said. “What’s more, Medicare could also follow private insurers and either reduce or eliminate the premium it’s willing to pay for proton beam therapy in prostate cancer payments."
“If we are to support even the current number of facilities going forward, some combination of two things will need to happen. Proton beam therapy will need to demonstrate itself superior to the best photon beam treatment in some common forms of cancer—and there are trials underway that could do that—or centers will have to find a way to survive on standard radiation reimbursement rates,” Zietman noted.The idea of using protons to irradiate tumors dates back to 1946, when the physicist Robert R. Wilson, PhD, suggested it to the research community, which began performing some human experiments with the approach by the 1950s. It took nearly 40 years, however, for the technology to mature enough for that first clinical treatment facility at Loma Linda.
That facility, and those that followed, were both enormous and enormously expensive. Each of those first-generation centers covers nearly as much land as a football field and houses a cyclotron that speeds protons up to more than half the speed of light before using enormous magnets that direct them to gantries, through nozzles that weigh more than 10 tons each and into patients. Construction costs generally ranged from $100 million to $200 million for those 4-gantry facilities, and operating costs added millions more each year.
The advantage to all this expenditure was accuracy. Proton beam therapy allowed for much more granular control over the placement of radiation. It greatly reduced radiation exposure not only for healthy organs surrounding the target in a 2-dimensional plane but also for healthy tissue below the target. Unlike radiation from x-rays, which steadily attenuate after hitting the tumor at full strength, proton beams could be set to penetrate no farther than a target depth.
“When I started in radiation oncology 30 years ago, the things that proton beam therapy can do would be considered absolutely mind-blowing. Standard photon radiation then was effective in treating many tumor types, but the radiation hit so much healthy tissue that its benefits often came at the price of crippling toxicity and a severe risk of secondary tumors,” said Zietman.
“Had photon radiation stagnated at that point, there’d be no question that proton beam would be the way to go for most patients, and we would be building even more facilities,” he added. “As things turned out, regular radiation improved far beyond what nearly anyone thought possible. The development of first SBRT [stereotactic body radiation therapy] and then IMRT [intensity- modulated radiation therapy] have narrowed the accuracy gap, and proton beam must prove itself against them.”It was widely assumed, even when proton beam therapy was still under development, that its accuracy would provide the biggest benefits to children because they have so much more time to develop secondary tumors from stray radiation. Research has vindicated that theory in the treatment of pediatric brain and head and neck tumors, mostly by comparing how much radiation hits healthy tissue with IBRT versus proton beam therapy and using models to estimate the risk of everything from developmental disabilities to secondary cancers.2
A 2011 study published in Radiotherapy and Oncology developed excess relative and absolute risk models for radiotherapy treatments given either to the cranium or spine in 6 different treatment volumes to patients of 6 different ages. The investigators found that risks of secondary cancers from proton treatment were generally a full order of magnitude smaller than those from IMRT.3
In January, the technology received a boost from a study into proton radiotherapy among approximately 60 patients with medulloblastoma with a median age of 6.6 years who were treated at Massaschusetts General Hospital’s center.4 After a median follow-up of 7 years, researchers found that survival outcomes were similar to what would be expected with conventional radiotherapy but that toxicities were lower.
Differences in intensity-modulated radiation therapy (IMRT) and proton beam therapy are illustrated in these images from the Northwestern Medicine Chicago Proton Center. The top image shows radiation exposure anticipated in treatment plans for IMRT, left panel, and proton therapy. In the bottom image, the radiation exposure to healthy tissue around the tumor and critical organs with proton therapy, left, and conventional radiotherapy is compared.
Specifically, the incidence of grade 3-4 hearing loss at 5 years was 16% with proton therapy versus the 24%-25% incidence reported in studies that utilized standard radiotherapy. “Other late effects common in photon-treated patients, such as cardiac, pulmonary, and gastrointestinal toxic effects, were absent,” Yock et al reported.4 The study, reported in The Lancet Oncology, resonated in the United Kingdom, where journal editors noted in an accompanying press release the controversy that ensued 2 years ago when the parents of a 5-year-old boy removed their son from a British hospital without his doctor’s permission so that he could be treated for a brain tumor with proton therapy in Prague.5 Health officials are now planning England’s first two proton therapy centers, which are scheduled to open in 2018.Such benefits, in and of themselves, would only justify the construction of a small number of proton beam facilities, for the pediatric cancers in question are rare. US doctors are expected to diagnose approximately 10,400 cases of cancer among children (ages 0-14 years) this year, with 26% of those cancers involving brain and other central nervous system tumors, according to the American Cancer Society.6
Proton beam advocates have long argued, however, that the technology’s accuracy, which remains better than even the most accurate form of photon radiation, would translate into better outcomes for patients with a wide variety of more common tumor types. Randomized trials have yet to test this hypothesis very thoroughly, but the growing number of proton beam facilities is finally encouraging the sort of research that could answer these questions.
Randomized trials and other types of studies are currently investigating whether proton beam therapy offers any advantage to patients with breast cancer7,8, adult brain cancer9,10, sarcomas11,12, lung cancer,13 and other tumor types.
Most of these tumors were treated with conventional radiation before proton beam therapy became an option, but the properties of proton beams expand the use of radiation into areas such as liver cancer,14,15 where radiotherapy has traditionally been rare.
Much of the research completed to date has focused on comparing proton to photon radiation in patients with prostate cancer. None of those studies have found reason to believe that proton therapy treats primary tumors any more effectively than photon therapy. The only real question is whether proton beam therapy reduces complications and, if so, how much those reductions are worth.
Some studies have concluded that there’s no significant evidence that proton beam therapy produces less radiation-related toxicity in the healthy organs that surround the pancreas. For example, a retrospective analysis of 12,976 men that appeared in JAMA found that IMRT was associated with less gastrointestinal toxicity than proton beam therapy (absolute risk, 12.2 vs 17.8 per 100 person-years; relative risk, 0.66; 95% CI, 0.55-0.79). There were no significant differences in rates of other morbidities or additional treatments between IMRT and proton therapy.16
Other studies have concluded that proton beam therapy does produce significantly less toxicity, but many such studies have also found that the differences last only a few months and may, therefore, be too transitory to justify the extra cost.
For example, a retrospective analysis of 27,647 Medicare beneficiaries that appeared in the Journal of the National Cancer Institute found that patients who received proton beam therapy patients were less prone than those who received IMRT to suffer genitourinary toxicity at 6 months (5.9% vs 9.5%; odds ratio, 0.60, 95% CI, 0.38 to 0.96; P = .03). The difference was gone by 12 months, however, and there was no statistically significant difference in gastrointestinal toxicity or any other toxicity at either 6 months or 12 months.17None of the research to date on this very important question, which concerns the best way to treat a large fraction of the nearly 181,000 US men who will be diagnosed with prostate cancer this year,6 has come from randomized controlled trials.
That should change in the next few years, however. Massachusetts General and the University of Pennsylvania, which both have proton beam facilities, hope to recruit 400 men for a prostate cancer trial that will compare outcomes of IMRT and proton beam therapy over 2 years of follow- up. Investigators currently hope that the last data will be collected by 2018 and the results will follow shortly thereafter.18
“If we were dealing with medications, the randomized trial would have to come first. No expensive new prostate cancer treatment would become widely available to patients, covered by Medicare and largely covered by private insurance, without first demonstrating some degree of superiority over existing treatments,” said James B. Yu, MD, associate professor of Therapeutic Radiology at Yale University, in an interview. “The regulation for medical technology is totally different. Usually, all it needs for regulatory approval is to demonstrate its safety. And whether Medicare or private insurance pays for it is sometimes hard to predict.
“As a result, technology manufacturers have very little incentive to undertake comparative trials, and the money to do these trials isn’t coming in very large amounts from anyone else either,” Yu said. “About half of all cancer patients undergo radiotherapy as part of their treatment but less than 2% of all cancer research dollars from the National Institutes of Health fund radiation studies.”
Insurance coverage for proton beam therapy varies. The Centers for Medicare & Medicaid Services does not have a national policy for covering proton therapy treatment, according to a search of the CMS.gov database. There are local determinations for Alabama, Wisconsin, Florida, and Illinois indicating that coverage is most likely to be authorized for a short list of cancer types including brain tumors, head and neck cancers, skull-based tumors, and intraocular melanomas. Additional tumor types, including lung and prostate cancers, may be covered depending upon the circumstances and a demonstration of medical necessity. Private insurers have been more restrictive.
Yet financial hurdles, including the difficulty of obtaining insurance authorizations, are not the only obstacles to randomized controlled trials that compare medical devices.
“Enrolling patients on radiation trials may be more challenging than pharmaceutical trials because of the incentive. The only opportunity a patient may have to receive a new drug may be to participate on a trial. Comparing a newer technology can be challenging since participants may feel that they are not receiving the newest and best therapy.” said Andrew K. Lee, MD, MPH, the medical director of the recently opened Texas Center for Proton Therapy. “Physicians may also be reluctant to enroll patients on such trials especially if they are concerned about a lack of equipoise between the two arms. Would I really want to randomize half my patients to photon radiation knowing that those patients will receive more radiation to normal tissues?” he said. “Ultimately, the trial may or may not show a statistical difference for specific endpoints with short follow-up, but a physician may still be concerned about exposing patients to unnecessary radiation due to possible long-term side effects,” Lee said. “This may be particularly the case when the ‘new’ technology has already successfully treated thousands of patients for decades.”
Lee thinks the ongoing prostate cancer trial will provide better evidence than any other research to date on the comparative merits of proton therapy and IMRT in that tumor type, but he doubts that the results will provide anything like the definitive answers that people want. “The trial compares IMRT, which is the current state-of-the-art for photon radiation, to all forms of proton radiation including conventional passive-scattered proton therapy rather than just pencil-beam, which is the current state-ofthe- art in proton therapy,” Lee said. “Even if the comparison were perfectly apt, the experience and skill of the practitioners and treatment centers needs to be considered, especially for a newly acquired technology like proton therapy. “Furthermore, the continuing progress in both types of treatments may make the results of a multiyear study largely irrelevant 10 years from now.” said Lee. “This trial was powered to detect a difference in a single, very specific gastrointestinal endpoint at only 2 years after treatment. However, it’s entirely possible that other significant differences could go undetected or simply fail to reach the level of significance with the relatively limited number of patients expected to be enrolled.
“The short follow-up is also a concern, especially in a disease that requires many years to determine the ultimate outcomes and side effects of treatment. I am not aware of any prostate cancer randomized trials that use only a 2-year primary endpoint,” Lee said.Lee’s experience with prostate cancer treatment has shown him that proton beam therapy damages surrounding organs less than even the best photon therapy, but the facility he runs isn’t counting on huge numbers of patients with prostate cancer to keep its doors open.
Unlike many of the older proton beam facilities, which were built around a business model that relied on patients with prostate cancer filling a significant majority of their daily openings, the Texas Center for Proton Therapy is intended to treat a wide variety of tumor types.
Lee estimates that prostate cancer cases have accounted for approximately 10% of all cases since the facility’s first gantry opened in November. Other patients have come with a wide variety of cancers, everything from pediatric cancers to head and neck cancers to cancers of the upper gastrointestinal tract, and lung and breast cancers.
The Texas Center for Proton Therapy has one big advantage over many of the other proton beam centers that are opening around the country: market size. The Dallas-Ft. Worth metro area is the fourth largest in the nation and was the largest market in the nation without any proton beam facility nearby.
Other new proton beam facilities, like the one at Ackerman Cancer Center in Jacksonville, Florida, are bringing the technology to smaller markets by keeping costs down. They employ single-gantry designs alongside newer equipment that is somewhat more compact to keep construction costs down around $30 million. These designs also have lower operating costs. Mevion Medical Systems, which built the Jacksonville facility, says its newest facilities are 75% smaller than traditional proton therapy centers and uses as much as 90% pless energy, significantly lowering capital and operating costs. The other driver of growth is major cancer centers with large endowments that allow them to offer all types of treatment, even if some of them lose money.
In order to track outcomes, the University of Pennsylvania has charged the same for proton and photon radiation treatments from two insurers despite the extra costs associated with building and operating its proton beam facility. 19 The Mayo Clinic is defraying the cost of its proton beam facilities—one that has opened in Rochester, Minnesota, and one that’s under construction in Phoenix—with financial gifts.20 The hospital received a $100 million donation earmarked to help with their construction.
Still, the addition of new facilities may come along with the loss of old facilities, even when operated by organizations with relatively deep pockets. Indiana University has already closed its money- losing facility near Bloomington, and some other big facilities, particularly those in smaller markets, are thought to be operating well below capacity.