Purpose: This study evaluated the dosimetry of prostate VMAT plans optimized by the photon optimization (PO) algorithm, which is introduced to replace the progressive resolution optimization (PRO) algorithm in the Varian Eclipse treatment planning system.
Methods: The dosimetric evaluation was carried out by studying ten prostate patients with VMAT plans created by the 6 MV photon beams and dual-arc technique using the Eclipse treatment planning system. The prescription of the dose was equal to 78 Gy/39 fractions. For each patient, two plans optimized by the PO and PRO algorithm, as per the same dose-volume criteria for the target and organs-at-risk, were created. Dose-volume histograms (DVHs) for the planning target volume and critical organs, namely, rectum, bladder, left and right femur, were determined. From the DVHs, dose-volume index such as conformity (CI) and homogeneity index (HI) were calculated, together with radiobiological variables such as tumour control probability (TCP) and normal tissue complication probability (NTCP).
Results: Comparing the dosimetry of VMAT plans optimized by the PO and PRO algorithm, it is found that the average CI values of plans optimized by the PO algorithm (0.912) was slightly higher than that by the PRO (0.908). On the other hand, the average rectal NTCP of plans using the PO algorithm was higher than that using the PRO (0.054 vs. 0.052). For the average HI and TCP of the prostate, both algorithms produced the same values of 0.054 and 0.955 from their VMAT plans, respectively.
Conclusion: It is concluded that prostate VMAT plan optimized by the PO algorithm had a slightly higher average CI and rectal NTCP value compared to the PRO, with other plan dosimetry remaining almost the same. Considering other advantages of the PO algorithm, namely, less multileaf collimator aperture complexity and faster optimization speed, the PO algorithm is a good successor for the PRO.
Prostate Therapy, Optimization, Treatment Planning
TH- External Beam- Photons: IMRT/VMAT dose optimization algorithms