Purpose: Temporal lobe (TL) parenchyma toxicity constitutes one of the most frequent late adverse event in high-dose proton therapy (PT) for tumors of the skull base. We analyzed clinical events with dosimetric parameters in our patients treated for skull base tumors with spot-scanning PT.
Methods and materials: Between 1998 and 2005, a total of 62 patients received PT to a median dose of 71.7 Gy (relative biologic effectiveness [RBE]) (range, 63-74 Gy). The dose-volume histogram of each TL and the entire brain parenchyma (BP) were analyzed according to maximum, mean, and minimum dose as well as doses to 0.5, 1, 2, and 3 cc of brain volume (D(0.5), D(1), D(2), D(3)) and correlated with clinical events. Generalized equivalent uniform dose (gEUD) values were calculated.
Results: At a mean follow-up of 38 months (range, 14-92 months), 2 patients had developed symptomatic Grade 3 and 5 patients asymptomatic Grade 1 TL toxicity. Mean doses to a 2-cc volume of BP increased from 71 ± 5 Gy (RBE) for no toxicity to 74 ± 5 Gy (RBE) for Grade 1 and to 76 ± 2 Gy (RBE) for Grade 3 toxicity. TL events occurred in 6 of 7 patients (86%) at or above dose levels of ≥ 64 Gy (RBE) D(3), ≥ 68 Gy (RBE) D(2), ≥ 72 Gy (RBE) D(1), and ≥ 73 Gy (RBE) D(0.5), respectively (p = NS). No statistically significant dose/volume threshold was detected between patients experiencing no toxicity vs. Grade 1 or Grade 3. A strong trend for Grade 1 and 3 events was observed, when the gEUD was 60 Gy.
Conclusions: A statistically significant normal tissue threshold dose for BP has not been successfully defined. However, our data suggest that tolerance of TL and BP to fractionated radiotherapy appears to be correlated with tissue volume included in high-dose regions. Additional follow-up time and patient accrual is likely needed to achieve clinical significance for these dose-volume parameters investigated. Our findings support the importance of establishing an organ-at-risk maximally permissible dose for BP.
Copyright © 2012 Elsevier Inc. All rights reserved.