RT Journal
A1 Mendenhall NP, Schild S, Slater J
T1 RAdiation therapy modalities for prostate cancer
JF JAMA
JO JAMA
YR 2012
FD August 1
VO 308
IS 5
SP 450
OP 450
DO 10.1001/jama.2012.8112
UL http://dx.doi.org/10.1001/jama.2012.8112
AB 
The authors characterized the SEER-Medicare data as “ . . . a large population-based sample and are therefore more likely to reflect results from treatments widely available in the community . . . .” More than 12 000 geographically distributed patients receiving IMRT and conformal radiation therapy were included, and that statement might be correct for these therapies. However, there are significant limitations of the SEER-Medicare database in the evaluation of proton therapy. Only 684 patients receiving proton therapy were identified, and the authors failed to disclose that all were treated at a single institution, Loma Linda University Medical Center—the only proton facility treating patients with prostate cancer and submitting data to the SEER registry during the timeframe of the study. Although the authors performed a propensity score adjustment, that is insufficient to eliminate single institution bias—particularly with respect to radiation dose, a key predictor for both disease control and toxicity, and surrogate outcomes (eg, proactive surveillance with assessments such as colonoscopy).