This news release summarizes findings from an observational study that looks at the long-term side effects of treating childhood brain tumors with proton radiotherapy compared with the more standard photon therapy. The premise of the study and its follow-up release is that proton beam therapy, by damaging less normal brain tissue, leaves a young patent with fewer long-term future complications from side effects.
The study, published in Lancet Oncology, reports on the follow-up of 59 children treated at one institution with proton beam therapy between 2003 and 2009. The release imparts a lot of information about the complicated study in a brief, readable way, but we have several concerns with missing information and implied results. All of the patients in the study had additional treatments including surgery and chemotherapy, so the benefits and harms cannot be fully attributed to the proton beam therapy. The release also does not address costs (which are significantly higher than standard therapy) or a very important limitation explained in the study which is that, for ethical reasons, there is no randomized control group treated in a similar way except for receiving conventional radiation therapy.
Treating cancer at any age is a challenging task for doctors, but when it involves children the stakes can’t be higher. This is particularly the case when the cancer is in a sensitive location, in this case the child’s brain, and when it can be life threatening, such as medulloblastoma. The treatments need to be strong enough to provide long-term management and at the same time spare normal surrounding tissue. When that surrounding tissue is the brain of a young child, damage can lead to life-long, serious complications. Studies that purport to offer improved outcomes need attention and scrutiny.
The release does not address the costs of the therapy. This is especially important to include in any discussion of proton therapy because its costs are astronomically high — often twice as much as standard therapy.
The news release quantifies some outcomes, as in this statement: “Of 59 patients who enrolled, 12 died from their tumor during the study period and one from a traumatic brain injury. Significant hearing loss was seen in 12 percent of patients three years after treatment and in 16 percent at five years, which compares with around 25 percent reported in studies using photon radiotherapy.” The release later indicates that key outcomes are comparable to photon therapy: “Survival rates and the incidence and type of tumor recurrence all were similar to what has been reported for photon radiotherapy.” We’ll call this good enough for a Satisfactory rating, although the release certainly could have provided more detail on what previous studies have found with respect to photon therapy.
The release addresses some of the harms associated with the therapy and provides specific numbers that are easy for readers to understand. It says, for example, “Of 59 patients who enrolled, 12 died from their tumor during the study period and one from a traumatic brain injury. Significant hearing loss was seen in 12 percent of patients three years after treatment and in 16 percent at five years, which compares with around 25 percent reported in studies using photon radiotherapy.”
The release didn’t point out that all of the patients had additional treatments — including surgery and chemotherapy — so it’s not possible to pinpoint years later where to place the blame for any hearing loss or cognitive deficits.
We think some of the harms from the therapy were given short shrift, but we address this in the “Quality of Evidence” measure.
We have two main complaints about this release.
1. It does not make clear that the study did not directly compare therapies side-by-side in a randomized, controlled trial. It studied one therapy and then made comparisons to other published studies. There’s one allusion to that in the release, but the nuance would be missed by most readers. And the release should have explicitly acknowledged that this was a limitation of the study design.
2. It does not delve deeply enough into the evidence actually presented in the study to make the case that risks associated with the therapy are low. The actual study clearly states, “Full Scale Intelligence Quotient decreased by 1·5 points (95% CI 0·9–2·1) per year after median follow-up up of 5·2 years (IQR 2·6–6·4), driven by decrements in processing speed and verbal comprehension index.” Losing intelligence over the course of five years, especially in young children, is a critical side effect to be measured and managed. We don’t think it’s adequate to say, “The cognitive effects of proton therapy – primarily affecting verbal comprehension and processing speed – were also less serious than what has been reported with photon radiotherapy, and those effects were primarily seen in children who were under the age of 8 when treated.”
There is no disease mongering in the release.
The funding sources and conflicts are clearly explained.
The release does make multiple comparisons to studies of other therapies. We do not feel that this was framed properly in the release, but we address this in the quality of evidence section.
The release makes it clear that the therapy is not widely available and not often used. The first quote says, “Proton radiotherapy is still not widely available in the U.S. or around the world, but it is increasingly recognized for its potential to reduce the side effects of treatment, particularly in the pediatric population.”
Both the release and the study make a claim of novelty, meaning that this type of analysis had not been done before. The study establishes that this type of research is new based on its literature review and extensive citations of previous studies. It says, “To the best of our knowledge, no other data for the late effects or disease control of proton therapy for patients with medulloblastoma are as complete and mature as ours.” The release adequately summarizes the state of current evidence by saying, “While proton therapy may appear to be ideal for reducing these adverse effects, no previous study had prospectively examined the long-term results of children treated for medulloblastoma with proton therapy.”
This is a tough call, but we believe the release uses unjustifiable language with this quote from one of the researchers: “While we are currently investigating quality of life differences between proton and photon treatment, I truly believe that – particularly for the youngest children – the ability to offer them proton therapy can make a big difference in their lives.”
This statement plays to emotions, and the researcher essentially acknowledges that the statement isn’t justified since the appropriate comparative research hasn’t been done. The families of pediatric cancer patients want to believe, but scientists need to base their statements, not on beliefs, but on data from the highest quality comparative studies we can ethically perform.