First in-room PET studies for range verification of proton therapy

Objectives PET imaging of proton induced activity (11C, 15O, 13N) is the only practical approach for in vivo range verification in proton therapy. The use of an in-room rather than an off-line scanner allows the collection of short half-life 15O, in addition to 11C, signal by greatly reducing the delay between irradiation and imaging. We report on the first ever in-room proton therapy patient studies with a mobile NeuroPET scanner.

Methods Phantom studies: Elementally tissue-equivalent gel phantom was irradiated with a circular spread-out Bragg-peak (SOBP) field, 14.1 cm water equivalent (WE) range, 10 cm WE modulation, and 2 Gy dose. After 1.5 min the phantom was imaged for 45 min in listmode with NeuroPET in-room and reconstructed for the 45 min, the first 5 min, the first 30 min, and the last 30 min. Range/lateral field width uncertainties were determined as the standard deviations of 50% distal fall-off depth/lateral field width. Patient studies: Two patients undergoing fractionated proton beam treatment were scanned twice 1-week apart (total 4 scans) for 30 min each. Reproducibility of activity distal 50% fall-off depths was determined for 20 profiles.

Results In the phantom study, the total coincidence counts in a 5-min in-room scan were 2.3 times that of a 30-min offline equivalent scan, and the range uncertainty was 40% lower. The lateral field width uncertainties were comparable at smaller beam penetration depths, but 8.2 times larger near the end of SOBP track for offline equivalent. The reproducibility in repeated patient studies was 5.4±7.5 mm for 5-min and 3.8 ± 15.2 mm for 30-min reconstructions.

Conclusions In-room PET studies were performed ~2.5 minutes after proton irradiation resulting in significantly improved counting statistics. Both range and lateral field width uncertainties were better in a 5-min in-room scan compared to a 30-min offline PET, especially for the lateral field width near the end of proton track.

Research Support NIH R21-CA134812, T32-EB00210