Spatiotemporal Optimization of Pelvic Radiation Therapy

Abstract

Conformal and adaptive Radiation Therapy (RT) remains critical in the treatment of locally advanced pelvic cancers despite acute and late bowel complications. This work introduces spatiotemporal optimization of pelvic RT including mathematical modelling of intestinal radiobiology to evaluate a novel treatment planning approach for fractionated pelvic RT. A single-plan fractionated RT delivery has n-fractions at dose/fraction = d and has biologically effective dose BED = nd ( 1 + d / [α/β] ) - ln(2) ([T - Tk] / [αTp]), where Tk and Tp are the kickoff time and doubling time of repopulation, and T is the time to deliver all fractions (including off-days). We present a parameterization of the repopulation BED-term to estimate a range of daily recovered BED including uncertainties in α and Tp. Recovery of BED as currently formulated is independent of the delivered dose; however, the assumption of spatiotemporal optimization is that a regional reduction of dose during RT by mixing treatment plans is critical to allow for normal tissue healing. We evaluate an achievable spatiotemporal optimization treatment planning strategy for an advanced stage prostate and a cervical cancer patient using 25-fraction delivery over 5-weeks and 4-weekends (T = 33 days) and propose 3 treatment plans: (1) a conformal "clinical standard" plan, (2) a right-bowel sparing plan, and (3) a left-bowel sparing plan. The plans are optimized to ensure no increased dose in opposing bowel or normal organs at risk (OARs). For radiobiological parameters ranging from α = 0.2-0.35/Gy, α/β = 2-3 Gy, Tk = 5-10 days, and Tp = 3-7 days, parameterization shows daily BED recovery of 0.44±0.08 Gy (range = 0.28-0.69 Gy), or an average of 4Gy BED recovery for 5-days treatment surrounded by 2 weekends. Conservatively evaluating the 3.6Gy BED distribution in treatment planning may identify regions for complete recovery (after Tk days). In a prostate 3-plan set, the right- and left- bowel sparing plans deliver equivalent mean dose to targets and OARs to within 5cGy/fx. The clinical plan includes 298cc of bowel under 14Gy, but this same bowel will be spared throughout treatment. Delivery of the clinical plan in week 1 to achieve Tk days, followed by alternating the left and right plan each week has the potential to reduce BED to zero for an additional 116cc of bowel using the left-sparing plan in weeks 2 and 4, and 91cc using the right-bowel sparing plan in weeks 3 and 5. The cervical cancer plan-set was more challenging due to right-sided nodal volumes and higher clinical bowel dose; OAR mean dose variations were >8cGy/fx in a few organs, but the left- and right- sided sparing plans results in 394 cc and 139cc, respectively of unique spared bowel at the 3.6Gy/week threshold to potentially reach zero BED during treatment. A novel spatiotemporal optimization of pelvic RT has the potential to preserve bowel, allow for increased intra-treatment intestinal tissue regeneration, and reduce radiation-induced complications.