Coverage‐based treatment planning to accommodate deformable organ variations in prostate cancer treatment

Abstract

Purpose: To compare two coverage-based planning (CP) techniques with standard fixed marginbased planning (FM), considering the dosimetric impact of interfraction deformable organ motion exclusively for high-risk prostate treatments. Methods: Nineteen prostate cancer patients with 8-13 prostate CT images of each patient were used to model patient-specific interfraction deformable organ changes. The model was based on the principal component analysis (PCA) method and was used to predict the patient geometries for virtual treatment course simulation. For each patient, an IMRT plan using zero margin on target structures, prostate (CTV prostate ) and seminal vesicles (CTV SV ), were created, then evaluated by simulating 1000 30-fraction virtual treatment courses. Each fraction was prostate centroid aligned. Patients whose D 98 failed to achieve 95% coverage probability objective D 98,95 ≥ 78 Gy (CTV prostate ) or D 98,95 ≥ 66 Gy (CTV SV ) were replanned using planning techniques: (1) FM (PTV prostate = CTV prostate + 5 mm, PTV SV = CTV SV + 8 mm), (2) CP OM which optimized uniform PTV margins for CTV prostate and CTV SV to meet the coverage probability objective, and (3) CP COP which directly optimized coverage probability objectives for all structures of interest. These plans were intercompared by computing probabilistic metrics, including 5% and 95% percentile DVHs (pDVH) and TCP/NTCP distributions. Results: All patients were replanned using FM and two CP techniques. The selected margins used in FM failed to ensure target coverage for 8/19 patients. Twelve CP OM plans and seven CP COP plans were favored over the other plans by achieving desirable D 98,95 while sparing more normal tissues. Conclusions: Coverage-based treatment planning techniques can produce better plans than FM, while relative advantages of CP OM and CP COP are patient-specific. C 2014 American Association of Physicists in Medicine. [http://dx

Coverage‐based treatment planning to accommodate delineation uncertainties in prostate cancer treatment

“…More details of this patient cohort have been described in earlier published work. 20,27,28 A single physician at VCU contoured the structures of prostate (CTV prostate ), seminal vesicles (CTV SV ), rectum, bladder, left femur, and right femur on each image. Contours of CTV SV , rectum, and bladder were modified slightly to eliminate overlapping region with CTV prostate or with other structures.…”

“…The in-house IMRT criteria used for CP OM , CP COP , and FM were the same as a recent published work. 20 For the CP techniques, the coverage probability objectives 34 were CTV prostate D 98,95 ≥ 78 Gy, D 2,5 ≤ 81 Gy, CTV SV D 98,95 ≥ 66 Gy, bladder D 70,5 ≤ 18 Gy, D 50,5 ≤ 36 Gy, D 30,5 ≤ 57 Gy, D 20,5 ≤ 66 Gy, D 14,5 ≤ 69 Gy, D 9,5 ≤ 75 Gy, D 2,5 ≤ 81 Gy, and rectum D 50,5 ≤ 36 Gy, D 30,5 ≤ 51 Gy, D 20,5 ≤ 66 Gy, D 5,5 ≤ 69 Gy, D 2,5 ≤ 75 Gy. To achieve the static dose-volume criteria (used for FM or CP OM ) for 95% of the alternative structure sets simulated, q = 95% and q = 5% coverage values were selected for minimum (target) or maximum (OAR) percentile DVH (pDVH) criteria used for CP COP .…”

“…From the DVH distribution, a percentile DVH at a specific coverage probability q was constructed by finding the line with equal coverage probability as a function of dose. Readers are referred to the previous 19,20 work for details. The pDVHs were used in not only plan evaluation to estimate the chance that dose volume criteria may be met, but also in the CP COP dose optimization to drive the pDVHs toward the desired coverage metrics.…”

“…CP COP included all the voxels that covered the dynamic ROI as sampled from the DU model for the different alternative structures, and weighted these voxels according to their potential contribution to the ROI dose. Then, the ROI dose was optimized based on Newton's method and minimizing the sum of a series of pDVH objectives [e.g., like Xu et al 20 but with the DU model (see Sec. 2.C) used in the pDVH computation as opposed to the PCA organ deformation T I. SD values in mm of in RL(x), PA(y), and SI(z) directions used in ASSD model for ROIs of prostate cancer patients.…”

“…CP OM iteratively increases PTV margins until the predefined coverage criteria are satisfied, and CP COP directly modifies the fluence to achieve the specified coverage criteria. Our previous studies [18][19][20] showed that both CP OM and CP COP can spare more normal tissue dose relative to conventional PTV-based planning when (rigid-body) setup errors or organ deformable motions were accounted for. However, the application of CP OM or CP COP to accommodate the delineation uncertainties has yet to be studied.…”