Session Item

Clinical track: Lower GI (colon, rectum, anus)
9306
Poster
Clinical
10:25 - 10:30
Online-adaptive versus robust IMPT planning for prostate: how much do we gain?
PD-0312

Abstract

Online-adaptive versus robust IMPT planning for prostate: how much do we gain?
Authors: Jagt|, Thyrza(1)*[t.jagt@erasmusmc.nl];Breedveld|, Sebastiaan(1);van Haveren|, Rens(1);Heijmen|, Ben(1);Hoogeman|, Mischa(1,2);
(1)Erasmus MC Cancer Institute, Department of Radiation Oncology, Rotterdam, The Netherlands;(2)HollandPTC, Department of Medical Physics & Informatics, Delft, The Netherlands;
Show Affiliations
Purpose or Objective

Intensity-modulated proton therapy (IMPT) is highly sensitive to anatomical variations, which can cause inadequate target coverage during treatment. Robust treatment planning preemptively includes error scenarios to account for these variations, but may also increase OAR doses. Online-adaptive planning aims at avoiding target underdose caused by daily variations and may simultaneously reduce OAR doses. This study compares a non-adaptive robust planning strategy to two online-adaptive IMPT strategies to determine the benefit of online adaptation.

Material and Methods

The study cohort consisted of CT data of 11 prostate cancer patients, including 8-10 weekly repeat CTs (rCTs) for each patient. Taking the first rCT as planning CT (pCT), 88 rCTs remained for evaluation. Dose was prescribed according to a simultaneously-integrated boost scheme, with 74 Gy to the high-dose target (prostate) and 55 Gy to the low-dose target (lymph nodes and seminal vesicles).

We compared a robust planning strategy against  two adaptive strategies:

  1. Robust: First, we determined the minimum degree of robustness and internal margin around the CTV, needed for the pCT to ensure adequate target coverage (V95%≥98% for both targets) in the rCTs for >90% of the patients. To this end, the pCT based plans were recalculated and evaluated on the rCTs.

  2. Restoration: For each rCT, the dose distribution of a prior plan optimized on the corresponding pCT was restored. This was done structure-wise, using the pCT contours projected onto the rCT. The restoration method uses an energy-adaptation followed by a fast spot-intensity re-optimization with focus on the targets.

  3. Full adaptation: For each rCT, the prior plan optimized on the corresponding pCT was used as a warm-start for adaptation. The method uses an energy-adaptation followed by the addition of 2500 new spots and a spot-intensity optimization based on the rCT contours, using the Reference Point Method.

All evaluations were done on the rCT scans and contours.

Results

Margins of 4 and 8 mm around the high- and low-dose target regions, a setup error of 6 mm and a range error of 3% were found to obtain adequate target coverage for all rCTs in 10/11 patients. Plan restoration and full adaptation both yielded V95%≥98% for all rCTs in all patients.

For all patients, the online-adaptive planning strategies could better spare the OARs while also better conforming to the target planning criteria. Largest improvements were seen for the V45Gy of rectum and bladder (Figure 1) with differences up to 11%-point between robust planning and restoration and up to 18%-point between robust planning and full adaptation. Figure 2 shows an example of the compared dose distributions.

Restoration and full adaptation took on average 1.7 and 4.7 minutes per rCT, excluding the final dose calculation.


Conclusion

Full adaptive planning is a time-feasible approach resulting in superior plan quality compared to non-adaptive robust planning and simple dose restoration, both in target coverage and OAR sparing.