Browse through our gallery of featured videos and stories. Protons for Kids Protons for Kids Stories Proton for Kids Videos Coloring Book
ASCO: Proton RT Less Harmful to Kids' Brains

By Charles Bankhead, Staff Writer, MedPage Today
Published: June 06, 2013
Reviewed by F. Perry Wilson, MD, MSCE; Instructor of Medicine, Perelman School of Medicine at the University of Pennsylvania


CHICAGO -- Children with brain tumors had stable cognitive function with proton irradiation, whereas photon radiotherapy was associated with significant decline in IQ, a retrospective study showed.

IQ of children treated with proton irradiation averaged 99.8 before treatment and 98.0 afterward. By comparison, children began photon irradiation with a mean IQ of 90.6, which declined to 86.6 by the end of treatment, according to Lisa Kahalley, PhD, of Baylor College of Medicine in Houston, and colleagues.

After controlling for differences in baseline characteristics and craniospinal irradiation (CSI), investigators found significantly greater decline in IQ with photon irradiation (P<0.001), they reported here at the American Society of Clinical Oncology.

"IQ remained generally stable for the proton group in the first 3 years' post-radiation therapy," Kahalley said. "In contrast, photon is associated with significant cognitive risk, with IQ scores declining by more than half a standard deviation with each additional year post photon.

"Our preliminary findings suggest proton may spare cognitive functioning."

An assessment of long-term outcomes is ongoing, since the first children treated with proton radiation therapy have just reached 5-year follow-up, she added.

Although an essential treatment modality for many pediatric brain tumors, radiation therapy carries a risk of neurocognitive decline as a late adverse effect. A child's IQ declines by 2 to 4 points per year after radiation therapy, and the risk increases with younger age at treatment, higher radiation therapy doses, and larger irradiated brain volumes, said Kahalley.

In contrast to conventional photon radiation therapy, proton-beam radiotherapy deposits its maximum dose at the greatest penetration depth, resulting in a reduced entrance dose and no exit dose. Several studies have described potential neuroprotective benefits of proton-beam radiotherapy, but no previous studies have compared actual neurocognitive outcomes in patients treated with proton versus photon irradiation, Kahalley continued.

Since 2006, patients at Texas Children's Hospital in Houston have undergone proton beam radiation therapy at MD Anderson Cancer Center, which has one of 11 proton beam radiation therapy centers that offer treatment for pediatric brain tumors. The collaborative arrangement provided investigators an opportunity to examine and compare the relative neurocognitive effects of brain irradiation by the type of energy source.

The study included 22 children treated by photon radiation therapy from 2000 to 2007 and 31 children treated by proton-beam irradiation from 2007 to 2011. Patients with poor-prognosis tumors were excluded. Kahalley and colleagues compared the patients' baseline IQ scores with scores on follow-up tests conducted within 3 years of radiation therapy.

Children treated by proton-beam radiation therapy were significantly older at the time of treatment (10.3 versus 7.5 years, P<0.01), and they received a lower total dose of radiation (50.4 versus 55.8 Gy, P<0.01).

In the proton group, tumor histology was germ cell in 10 cases, followed by medulloblastoma/PNET (nine), glioma (seven), and "other" (five). In the photon group, medulloblastoma/PNET accounted for half of the cases, followed by glioma and ependymoma (four each), "other" (two), and germ cell (one).

Tumor location was supratentorial in 22 patients in the proton group and 12 in the photon group, and infratentorial in nine and 10 cases, respectively.

Comparison of IQ averages in the two groups showed a significantly higher mean on follow-up in the proton group compared with the photon group (P<0.05). The difference in IQ change from baseline to follow-up also differed significantly between the groups (P<0.05).

"Multiple linear regression showed that the proton group exhibited essentially stable IQ over time (0.1 IQ point per year post-RT), whereas the photon group exhibited a significant IQ decline over time" of 10.3 IQ points per year (P<0.01), said Kahalley. "The 10.3 decline per year in the photon group is way higher than would be expected, and we intend to drill down to see whether we can identify factors to explain this.

"But the point is that IQ decline was significantly greater in the photon group."

The fully adjusted model accounted for 78% of the variance in follow-up IQ scores. In contrast to previous studies, age at the time of radiation therapy was not a predictor of neurocognitive outcomes. Consistent with the literature, craniospinal irradiation did predict lower follow-up IQ scores (P<0.05).

Invited discussant Gregory Armstrong, MD, of St. Jude Children's Research Hospital in Memphis, Tenn., noted that the primary advantage of proton-beam radiotherapy is a reduction in radiation dose to normal tissues. The newer radiotherapy modality also has several potential disadvantages: uncertain radiobiological effect, unknown toxicity profile, lack of comparison populations for key outcomes, and low enrollment in front-line clinical trials.

Armstrong also cautioned against assuming that the two study groups were similar. Although the 9-point difference in baseline IQ did not reach statistical significance, such a difference is "a clinically significant, measurable difference."

"I must emphasize that there is a gap in knowledge regarding the benefit and risk of proton-beam therapy," said Armstrong. "This study is the first to step into that gap and compares IQ against a control group. Taken alone, a single retrospective study cannot file the gap, nor would we expect it to do so."

Future studies should be prospective, longitudinal, and focus on systematic assessment, he added. The studies should be limited to a single tumor type to reduce heterogeneity, and the comparison population should be a well-established control group or a concurrent, contemporaneous group treated by photon therapy at a different institution. Finally, future studies should include a comprehensive evaluation of the late effects of brain irradiation in children.

Kahalley and coinvestigator had no relevant disclosures.

Armstrong had no relevant disclosures.

Primary Source: American Society of Clinical Oncology
Source reference:
Kahalley LS, et al "IQ change within three years of radiation therapy in pediatric brain tumor patients treated with proton beam versus photon radiation therapy." ASCO 2013; Abstract 10009.




> Send mail to with questions or comments about this web site.