Eclipse fail.

JCO missed the once in a generation opportunity to publish an ECLYPS study during an actual total eclipse by just over a week. SMH. This report prospectively evaluates the ability of FDG-PET scans to detect recurrence of locally advanced head and neck cancer when performed 12 weeks after definitive chemoradiation. Last year's PET-NECK trial showed that this approach is safe and that doing a planned neck dissection on PET-negative patients does not improve outcomes. ECLYPS went a step further by using a more standardized PET evaluation criteria and describing the clinical accuracy of the approach: Overall accuracy was 86% with a negative predictive value of 92%. However, sensitivity was only 65%. Why? As you might guess, 12-week PET is good at detecting disease that will clinically re-emerge within the first few months (hence the 83% sensitivity at 3 months). However, sensitivity fell below 60% among recurrences at 12 months. Maybe we should just re-image at 12 months, you say? We agree. This study is informative, but it could work on its timing.