Treatment Interruptions in Stereotactic Body Radiotherapy (SBRT) and Stereotactic Radiosurgery (SRS): Influence and Compensation
F.Q. Guo1*, V Chiang1, 2, J Yu1, B Chang1, J Bond1, Z Chen1, R Nath1, J Deng1, (1) Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, CT, (2) Department of Neurosurgery, Yale University School of Medicine, New Haven, CTSU-E-T-389 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: Unscheduled treatment interruptions in SBRT/SRS can occur for a variety of reasons. As the fractional dose is much larger and the total treatment time much shorter, treatment prolongation may have larger influence in SBRT/SRS than in conventional radiotherapy.
Methods: A biologically effective dose (BED) model was used to quantify the impact of interruptions with a daily equivalent repopulation constant K. The BEDs were calculated for regular daily treatment and interrupted treatments with various interruption schemes and K values. The dose required to compensate for the BED reduction due to interruption was then calculated.
Results: For a typical five-fraction SBRT assuming (α/β) 10Gy for tumor and 2Gy for normal tissue, K 0.5Gy/day, and a seven-day gap, the dose per fraction required for the remainder of the treatments to compensate for treatment breaks (dnew) would be 103%, 104%, 106%, or 111% of the originally prescribed dose per fraction (d) with the gap after 1, 2, 3, or 4 fraction(s). For a four-fraction SBRT, the dnew would be 104%, 106%, or 111% with the gap after 1, 2, or 3 fraction(s). For a three-fraction SBRT, the dnew would be 106% or 111% with the gap after 1 or 2 fraction(s). With a fourteen-day gap and same K, the increase Δd=[(dnew-d)/d]x100% would double for the various scenarios respectively. If K increases from 0.5Gy/day to 1.0Gy/day, the Δd would roughly double for the various scenarios accordingly.
Conclusion: According to our BED model, significantly higher dnew would be needed to compensate for the reduction in local control with interruptions, particularly for the large gaps and the late occurrings. Therefore, for anticipated interruptions, it is better to interrupt the treatment sooner than later. As our model depends on K, more clinical data would be needed to substantiate its validity for various lesions treated by SBRT/SRS.