Plan quality and robustness in field junction region for craniospinal irradiation with VMAT

PurposeTo propose a “staggered overlap” technique in volumetric modulated arc therapy (VMAT) for craniospinal irradiation (CSI) and compare the dose distribution and plan robustness with “overlap” technique and “gradient optimization” approach.Methods and Materials6 patients previously treated in our clinic were retrospectively selected. 9 VMAT plans of each patient were optimized with “staggered overlap”, “overlap” and “gradient optimization” in overlapping region of 3 cm, 6 cm, and 9 cm separately. For the “staggered overlap” plan, adjacent field sets were intentionally overlapped by staggering field edges in an appropriate step size to avoid sharp dose gradient. Evaluation metrics including V_95%, D_2%, D_98%, conformity number (CN) and homogeneity index (HI) were employed to evaluate the dose distribution. Moreover, shifts of the upper spinal field isocenter in each direction were performed to simulate junction errors for robustness analysis.ResultsThe CN and HI of VMAT plans with “staggered overlap” were 0.82 (0.811–0.822) and 0.113 (0.112–0.114), while they were 0.778 (0.776–0.782) and 0.131 (0.130–0.131) for plans with “gradient optimization”. In the robustness study, <3% dose deviations were found for 5 mm shifts in lateral and vertical directions with all techniques. In cranial-caudal direction, “overlap” technique created hot spots (D_2% > 170%) and cold spots (D_98% < 44%) in the junction region with 10 mm shifts. The dose deviations were decreased to 22% for plans with “staggered overlap” and 9 cm overlapping region.Conclusion“Staggered overlap” technique provides better plan quality as compared to “gradient optimization” approach and makes the plan more robust against junction errors as compared to “overlap” technique.     

Dosimetric comparison of analytic anisotropic algorithm and Acuros XB algorithm in VMAT plans for high-grade glioma

PurposeTo investigate the dosimetric impact between the anisotropic analytical algorithm (AAA) and the Acuros XB (AXB) algorithm in volumetric-modulated arc therapy (VMAT) plans for high-grade glioma (HGG).MethodsWe used a heterogeneous phantom to quantify the agreement between the measured and calculated doses from the AAA and from the AXB. We then analyzed 14 patients with HGG treated by VMAT, using the AAA. We newly created AXB plans for each corresponding AAA plan under the following conditions: (1) re-calculation for the same number of monitor units with an identical beam and leaf setup, and (2) re-optimization under the same conditions of dose constraints. The dose coverage for the planning target volume (PTV) was evaluated by dividing the coverage into the skull, air, and soft-tissue regions.ResultsCompared to the results obtained with the AAA, the AXB results were in good agreement with the measured profiles. The dose differences in the PTV between the AAA and re-calculated AXB plans were large in the skull region contained in the target. The dose difference in the PTV in both types of plan was significantly correlated with the volume of the skull contained in the target (r = 0.71, p = 0.0042). A re-optimized AXB plan's dose difference was lower vs. the re-calculated AXB plan's.ConclusionsWe observed dose differences between the AAA and AXB plans, in particular in the cases in which the skull region of the target was large. Considering the phantom measurement results, the AXB algorithm should be used in VMAT plans for HGG.     

Optimization of treatment planning parameters used in tomotherapy for prostate cancer patients

Background and purposeTomotherapy treatment planning depends on parameters that are not used conventionally such as: field width (FW), pitch factor (PF) and modulation factor (MF). The aim of this study is to analyze the relationship between these parameters and their influence on the quality of treatment plans and beam-on time.Material and methodsTen prostate cancer patients were included in the study. For each patient, two cases of irradiation were considered depending on the target volume: PTV1 included the prostate gland, seminal vesicles, pelvic lymph nodes and a 1 cm margin, whereas PTV2 included only the prostate gland with a 1 cm margin. For each patient and each case of irradiation (PTV1 and PTV2) 8 treatment plans were created – all consisted of a different combination of planning parameters (FW = 1.05, 2.5, 5 cm; PF = 0.107, 0.215, 0.43; MF = 1.5, 2.5, 3.5). Default values used in this study were FW = 2.5 cm, PF = 0.215 and MF = 2.5. Hence, for plans with different FWs, parameters of PF and MF were 0.215 and 2.5, respectively; for different PFs, FW and MF were 2.5 and 2.5, respectively; finally for different MFs, FW and PF were 2.5 and 0.215, respectively. The reference plan was optimized for FW = 1.05 cm, PF = 0.107 and MF = 3.5, which was assumed to result in the best dose distribution and the longest treatment time. As a result, 160 plans were created. Each plan was analyzed for dose distribution and execution time.Results and conclusion: Treatment plans with FW of 5 cm resulted in the shortest execution time compromising the dose distribution. Moreover, the dose fall off in the longitudinal direction was not sharp. FW of 1.05 cm and PF of 0.107 were not recommended for routine prostate plans due to long execution time, which was 3 times longer than for plans with FW = 5 cm. There was no substantial decrease of irradiation time when PF was increased from 0.215 to 0.43 for both cases (PTV1 and PTV2); however, the dose distribution was slightly compromised. Finally, decreasing MF from 2.5 to 1.5 was useless because it did not change the beam-on time; however, it did remarkably decrease the dose distribution. Nevertheless, increasing MF up to 3.5 could be considered. The lowest EUD for the rectum and intestines, could be observed for PF = 0.107. For the other plans the differences were rather small (the EUD was almost the same). By reducing PF from 0.43 to 0.107 or FW from 5 to 1.05 the EUD for bladder (in PTV1 case) decreased by 3.13% and 2.60%. When PTV2 was a target volume, the EUD for bladder decreased by 4.54% and 3.43% when FW was changed from 5 to 1.05 and MF from 1.5 to 3.5, respectively. For optimal balance between beam-on time and dose distribution in OARs for routine patients, the authors would suggest to use: FW = 2.5, PF = 0.215 and MF = 2.5.     

Dose-escalated volumetric modulated arc therapy for total marrow irradiation: A feasibility dosimetric study with 4DCT planning and simultaneous integrated boost

PurposeTo evaluate the planning feasibility of dose-escalated total marrow irradiation (TMI) with simultaneous integrated boost (SIB) to the active bone marrow (ABM) using volumetric modulated arc therapy (VMAT), and to assess the impact of using planning organs at risk (OAR) volumes (PRV) accounting for breathing motion in the optimization.MethodsFive patients underwent whole-body CT and thoraco-abdominal 4DCT. A planning target volume (PTV) including all bones and ABM was contoured on each whole-body CT. PRV of selected OAR (liver, heart, kidneys, lungs, spleen, stomach) were determined with 4DCT. Planning consisted of 9–10 full 6 MV photon VMAT arcs. Four plans were created for each patient with 12 Gy prescribed to the PTV, with or without an additional 4 Gy SIB to the ABM. Planning dose constraints were set on the OAR or on the PRV. Planning objective was a PTV D_mean < 110% of the prescribed dose, a PTV V_110% < 50%, and OAR D_mean ≤ 50–60%.ResultsPTV D_mean < 110% was accomplished for most plans (n = 18/20), while all achieved V_110%<50%. SIB plans succeeded to optimally cover the boost volume (median ABM D_mean = 16.3 Gy) and resulted in similar OAR sparing compared to plans without SIB (median OAR D_mean = 40–54% of the ABM prescribed dose). No statistically significant differences between plans optimized with constraints on OAR or PRV were found.ConclusionsAdding a 4 Gy SIB to the ABM for TMI is feasible with VMAT technique, and results in OAR sparing similar to plans without SIB. Setting dose constraints on PRV does not impair PTV dosimetric parameters.     

VMAT-like plans for magnetic resonance guided radiotherapy: Addressing unmet needs

PurposeVMAT delivery technique is currently not applicable to Magnetic Resonance-guided radiotherapy (MRgRT) hybrid systems. Aim of this study is to evaluate an innovative VMAT-like (VML) delivery technique.Material and methodsFirst, planning and dosimetric evaluation of the MRgRT VML treatment have been performed on 10 different disease sites and the results have been compared with the corresponding IMRT plans. Then, in the second phase, 10 of the most dosimetrically challenging locally advanced pancreas treatment plans have been retrospectively re-planned using the VML approach to explore the potentiality of this new delivery technique. Finally, VML robustness was evaluated and compared with the IMRT plans, considering a lateral positioning error of ± 5 mm.ResultsIn phase one, all VML plans were within constraint for all OARs. When PTV coverage is considered, in the 50% of the cases VML PTV coverage is equal or higher than in IMRT plan. In the remaining 50%, the highest target under coverage difference in comparison with IMRT plan is −1.71%. The mean and maximum treatment time differences (VML-IMRT) is 0.2 min and 3.1 min respectively. In phase two, the treatment time variation (VML-IMRT), shows a mean, maximum and minimum variations of 1.3, 4.6 and −0.6 min respectively. All VML plans have a better target coverage if compared with IMRT plans, keeping in any case the OARs constraints within tolerance. VML doesn’t increase plan robustness.ConclusionVMAT-like treatment approach appeared to be an efficient planning solution and it was decided to clinically implement it in daily practice, especially in the frame of hypo fractionated treatments.     

Adaptive volumetric modulated arc treatment planning for esophageal cancers using cone beam computed tomography

PurposeTo assess the potential of cone beam CT (CBCT) derived adaptive RapidArc treatment for esophageal cancers in reducing the dose to organs at risk (OAR).Methods and materialsTen patients with esophageal cancer were CT scanned in free breathing pattern. The PTV is generated by adding a 3D margin of 1 cm to the CTV as per ICRU 62 recommendations. The double arc RapidArc plan (Clin_RA) was generated for the PTV. Patients were setup using kV orthogonal images and kV-CBCT scan was acquired daily during first week of therapy, then weekly. These images were exported to the Eclipse TPS. The adaptive CTV which includes tumor and involved nodes was delineated in each CBCT image set for the length of the PTV. The composite CTV from first week CBCT was generated using Boolean union operator and 5 mm margin was added circumferentially to generate adaptive PTV (PTV1). Adaptive RapidArc plan (Adap_RA) was generated. NTCP and DVH of the OARs of the two plans were compared. Similarly, PTV2 was generated from weekly CBCT. PTV2 was evaluated for the coverage of 95% isodose of Adap_RA plan.ResultsThe PTV1 and PTV2 volumes covered by 95% isodose in adaptive plans were 93.51 ± 1.17% and 94.59 ± 1.43% respectively. The lung V_10Gy, V_20Gy and mean dose in Adap_RA plan was reduced by 17.43% (p = 0.0012), 34.64% (p = 0.0019) and 16.50% (p = 0.0002) respectively compared to Clin_RA. The Adap_RA plan reduces the heart D_35% and mean dose by 17.35% (p = 0.0011) and 17.16% (p = 0.0012). No significant reduction in spinal cord and liver doses were observed. NTCP for the lung (0.42% vs. 0.08%) and heart (1.39% vs. 0.090%) was reduced significantly in adaptive plans.ConclusionThe adaptive re-planning strategy based on the first week CBCT dataset significantly reduces the doses and NTCP to OARs.     

Efficacy of tangential irradiation with volumetric modulated arc therapy on scalp angiosarcoma using medical linac

PurposeTo increase the superficial dose and reduce the brain dose for radiotherapy of scalp angiosarcoma, we propose a novel irradiation technique of tangential irradiation volumetric modulated arc therapy (TI-VMAT).MethodsTI-VMAT and the conventional VMAT treatment plans for thirteen scalp angiosarcoma patients were created with a prescribed dose of 70 Gy. Each treatment was normalized to cover 95% of the planning target volume (PTV) with its prescribed dose. To realize TI-VMAT, an avoidance structure (AS) function was applied. AS was defined as a contour subtracted PTV by a certain space from the brain contour. TI-VMAT treatment plans for six different spaces between PTV and AS were developed and compared with the conventional VMAT treatment plan with respect to the following dosimetric parameters: homogeneity index (HI) and conformity index (CI) of the PTV, mean brain dose, and brain volume irradiated with 20% (V_20% [cc]), 40% (V_40% [cc]), 60% (V_60% [cc]), 80% (V_80% [cc]), and 100% (V_100% [cc]) of the prescribed dose.ResultsHI and CI were comparable between TI-VMAT and the conventional VMAT, the mean brain dose for TI-VMAT with AS defined by a space of 2.0 cm and jaw tracking was 14.27 Gy, which was significantly lower than that for the conventional VMAT (21.20 Gy). In addition, dosimetric parameters such as V_20% [cc] were significantly suppressed compared to those for high doses.ConclusionOur proposed irradiation technique TI-VMAT shows the potential to reduce radiation doses in the brain with maintaining higher dose coverage on the PTV.     

Patient-specific quality assurance and plan dose errors on breast intensity-modulated proton therapy

PurposeTo investigate, in proton therapy, whether the Gamma passing rate (GPR) is related to the patient dose error and whether MU scaling can improve dose accuracy.MethodsAmong 20 consecutively treated breast patients selected for analysis, two IMPT plans were retrospectively generated: (1) the pencil-beam (PB) plan and (2) the Monte Carlo (MC) plan. Patient-specific QA was performed. A 3%/3-mm Gamma analysis was conducted to compare the TPS-calculated PB algorithm dose distribution with the measured 2D dose. Dose errors were compared between the plans that passed the Gamma testing and those that failed. The MU was then scaled to obtain a better GPR. MU-scaled PB plan dose errors were compared to the original PB plan.ResultsOf the 20 PB plans, 8 were passed Gamma testing (G_pass_group) and 12 failed (G_fail_group). Surprisingly, the G_pass_group had a greater dose error than the G_fail_group. The median (range) of the PTV DVH RMSE and PTV ΔDmean were 1.36 (1.00–1.91) Gy vs 1.18 (1.02–1.80) Gy and 1.23 (0.92–1.71) Gy vs 1.10 (0.87–1.49) Gy for the G_pass_group and the G_fail_group, respectively. MU scaling reduced overall dose error. However, for PTV D99 and D95, MU scaling worsened some cases.ConclusionFor breast IMPT, the PB plans that passed the Gamma testing did not show smaller dose errors compared to the plans that failed. For individual plans, the MU scaling technique leads to overall smaller dose errors. However, we do not suggest use of the MU scaling technique to replace the MC plans when the MC algorithm is available.     

Characterization of robust optimization for VMAT plan for liver cancer

PurposeWe investigated the feasibility of robust optimization for volumetric modulated arc therapy (VMAT) stereotactic body radiation therapy (SBRT) for liver cancer in comparison with planning target volume (PTV)-based optimized plans. Treatment plan quality, robustness, complexity, and accuracy of dose delivery were assessed.MethodsTen liver cancer patients were selected for this study. PTV-based optimized plans with an 8-mm PTV margin and robust optimized plans with an 8-mm setup uncertainty were generated. Plan perturbed doses were evaluated using a setup error of 8 mm in all directions from the isocenter. The dosimetric comparison parameters were clinical target volume (CTV) doses (D_98%, D_50%, and D_2%), liver doses, and monitor unit (MU). Plan complexity was evaluated using the modulation complexity score for VMAT (MCSv).ResultsThere was no significant difference between the two optimizations with respect to CTV doses and MUs. Robust optimized plans had a higher liver dose than did PTV-based optimized plans. Plan perturbed dose evaluations showed that doses to the CTV for the robust optimized plans had small variations. Robust optimized plans were less complex than PTV-based optimized plans. Robust optimized plans had statistically significant fewer leaf position errors than did PTV-based optimized plans.ConclusionsComparison of treatment plan quality, robustness, and plan complexity of both optimizations showed that robust optimization could be feasibile for VMAT of liver cancer.     

Potential interest of developing an integrated boost dose escalation for stereotactic irradiation of primary prostate cancer

IntroductionThe stereotactic irradiation is a new approach for low-risk prostate cancer. The aim of the present study was to evaluate a schema of stereotactic irradiation of the prostate with an integrated-boost into the tumor.Material and methodsThe prostate and the tumor were delineated by a radiologist on CT/MRI fusion. A 9-coplanar fields IMRT plan was optimized with three different dose levels: 1) 5 × 6.5 Gy to the PTV1 (plan 1), 2) 5 × 8 Gy to the PTV1 (plan 2) and 3) 5 × 6.5 Gy on the PTV1 with 5 × 8 Gy on the PTV2 (plan 3). The maximum dose (MaxD), mean dose (MD) and doses received by 2% (D2), 5% (D5), 10% (D10) and 25% (D25) of the rectum and bladder walls were used to compare the 3 IMRT plans.ResultsA dose escalation to entire prostate from 6.5 Gy to 8 Gy increased the rectum MD, MaxD, D2, D5, D10 and D25 by 3.75 Gy, 8.42 Gy, 7.88 Gy, 7.36 Gy, 6.67 Gy and 5.54 Gy. Similar results were observed for the bladder with 1.72 Gy, 8.28 Gy, 7.01 Gy, 5.69 Gy, 4.36 Gy and 2.42 Gy for the same dosimetric parameters. An integrated SBRT boost only to PTV2 reduced by about 50% the dose difference for rectum and bladder compared to a homogenous prostate dose escalation. Thereby, the MD, D2, D5, D10 and D25 for rectum were increased by 1.51 Gy, 4.24 Gy, 3.08 Gy, 2.84 Gy and 2.37 Gy in plan 3 compared to plan 1.ConclusionsThe present planning study of an integrated SBRT boost limits the doses received by the rectum and bladder if compared to a whole prostate dose escalation for SBRT approach.     

A study on conventional IMRT and RapidArc treatment planning techniques for head and neck cancers

AimTo evaluate the performance of volumetric arc modulation with RapidArc against conventional IMRT for head and neck cancers.BackgroundRapidArc is a novel technique that has recently been made available for clinical use. Planning study was done for volumetric arc modulation with RapidArc against conventional IMRT for head and neck cancers.Materials and methodsTen patients with advanced tumors of the nasopharynx, oropharynx, and hypopharynx were selected for the planning comparison study. PTV was delineated for two different dose levels and planning was done by means of simultaneously integrated boost technique. A total dose of 70 Gy was delivered to the boost volume (PTV boost) and 57.7 Gy to the elective PTV (PTV elective) in 35 equal treatment fractions. PTV boost consisted of the gross tumor volume and lymph nodes containing visible macroscopic tumor or biopsy-proven positive lymph nodes, whereas the PTV elective consisted of elective nodal regions. Planning was done for IMRT using 9 fields and RapidArc with single arc, double arc. Beam was equally placed for IMRT plans. Single arc RapidArc plan utilizes full 360° gantry rotation and double arc consists of 2 co-planar arcs of 360° in clockwise and counter clockwise direction. Collimator was rotated from 35 to 45° to cover the entire tumor, which reduced the tongue and groove effect during gantry rotation. All plans were generated with 6 MV X-rays for CLINAC 2100 Linear Accelerator. Calculations were done in the Eclipse treatment planning system (version 8.6) using the AAA algorithm.ResultsDouble arc plans show superior dose homogeneity in PTV compared to a single arc and IMRT 9 field technique. Target coverage was almost similar in all the techniques. The sparing of spinal cord in terms of the maximum dose was better in the double arc technique by 4.5% when compared to the IMRT 9 field and single arc techniques. For healthy tissue, no significant changes were observed between the plans in terms of the mean dose and integral dose. But RapidArc plans showed a reduction in the volume of the healthy tissue irradiated at V_15 Gy (5.81% for single arc and 4.69% for double arc) and V_20 Gy (7.55% for single arc and 5.89% for double arc) dose levels when compared to the 9-Field IMRT technique. For brain stem, maximum dose was similar in all the techniques. The average MU (±SD) needed to deliver the dose of 200 cGy per fraction was 474 ± 80 MU and 447 ± 45 MU for double arc and single arc as against 948 ± 162 MU for the 9-Field IMRT plan. A considerable reduction in maximum dose to the mandible by 6.05% was observed with double arc plan. Double arc shows a reduction in the parotid mean dose when compared with single arc and IMRT plans.ConclusionRapidArc using double arc provided a significant sparing of OARs and healthy tissue without compromising target coverage compared to IMRT. The main disadvantage with IMRT observed was higher monitor units and longer treatment time.     

Impact of patient setup error in the treatment of head and neck cancer with intensity modulated radiation therapy

PurposeTo study the impact of setup errors on the dose to the target volume and critical structures in the treatment of cancer of nasopharynx with intensity modulated radiation therapy (IMRT).Methods and materialsTwelve patients of carcinoma of nasopharynx treated by IMRT with simultaneous integrated boost technique were enrolled. The gross tumor volume, clinical target volume and low-risk nodal region were planned for 70, 59.4 and 54 Gy, respectively, in 33 fractions. Based on the constraints, treatment plans were generated. Keeping it as the base plan, the patient setup error was simulated for 3, 5 and 10 mm by shifting the isocenter in all three directions viz. anterior, posterior, superior, inferior, right and left lateral. The plans were evaluated for mean dose, maximum dose, volume of PTV receiving >110% and <93% of the prescribed dose. For both the parotids, the mean dose and the dose received by >50% of the parotid were evaluated. The maximum dose and dose received by 2 cc of spinal cord were also analyzed.ResultsThe dose to the target volume decreases gradually with increase in setup error. The superior and inferior shifts play major role in tumor under-dosage. A setup error of 3 mm along the posterior and lateral directions significantly affects the dose to the spinal cord. Similarly, setup error along lateral and anterior directions affects the dose to both parotids.ConclusionsThe isocenter position should be verified regularly to ensure that the goal of IMRT is achieved.     

Multi-institutional comparison of simulated treatment delivery errors in ssIMRT,manually planned VMAT and autoplan-VMAT plans for nasopharyngeal radiotherapy

PurposeTo quantify the impact of simulated errors for nasopharynx radiotherapy across multiple institutions and planning techniques (auto-plan generated Volumetric Modulated Arc Therapy (ap-VMAT), manually planned VMAT (mp-VMAT) and manually planned step and shoot Intensity Modulated Radiation Therapy (mp-ssIMRT)).MethodsTen patients were retrospectively planned with VMAT according to three institution’s protocols. Within one institution two further treatment plans were generated using differing treatment planning techniques. This resulted in mp-ssIMRT, mp-VMAT, and ap-VMAT plans. Introduced treatment errors included Multi Leaf Collimator (MLC) shifts, MLC field size (MLCfs), gantry and collimator errors. A change of more than 5% in most selected dose metrics was considered to have potential clinical impact. The original patient plan total Monitor Units (MUs) were correlated to the total number of dose metrics exceeded.ResultsThe impact of different errors was consistent, with ap-VMAT plans (two institutions) showing larger dose deviations than mp-VMAT created plans (one institution). Across all institutions’ VMAT plans the significant errors included; ±5° for the collimator angle, ±5 mm for the MLC shift and +1, ±2 and ±5 mm for the MLC field size. The total number of dose metrics exceeding tolerance was positively correlated to the VMAT total plan MUs (r = 0.51, p < 0.001), across all institutions and techniques.ConclusionsDifferences in VMAT robustness to simulated errors across institutions occurred due to planning method differences. Whilst ap-VMAT was most sensitive to MLC errors, it also produced the best quality treatment plans. Mp-ssIMRT was most robust to errors. Higher VMAT treatment plan complexity led to less robust plans.     

Modulated volumetric arc therapy for total marrow irradiation: A feasibility study in the oncology hospital of CMN SXXI from a medical physics approach

AimThe objective of this study is to explore the use of volumetric arc therapy (VMAT) to perform total marrow irradiation (TMI) and compare its results to the standard TBI technique in the Mexican public health system.BackgroundThe standard total body irradiation (TBI) technique is used with chemotherapy as a method of a pre-transplant conditioning of the bone marrow. In this technique, the whole body of the patient is considered to be PTV and irradiated generating toxicities and raising concerns about possible development of radio-induced tumors.Materials and methodsThrough the use of simulation tomography of 12 patients previously treated with TBI, twelve different treatment plans were created with the proposed TMI technique and compared with the conventional protocol, the treatment plans were evaluated with a dose volume histogram analysis and quality assurance was evaluated with a portal dosimetry system using the gamma index criteria 3%/3 mm.ResultsExperimental results show an increasing dose to 99% of PTV of up to 41.1% by using TMI with the VMAT technique. The mean average dose to PTV was increased up to 19.3%. The use of the new TMI technique caused an improvement in the mean average dose to 99% of the PTV as well the homogeneity of the dose distribution prescribed at the PTV while leading to a better reproducibility of the treatment. The Qa of all the plans met the criterion of gamma index 3 mm-3%.ConclusionThe results analysis shows that the proposed TMI technique is feasible and applicable in the Mexican public health system.     

Quality assurance-based optimization (QAO): Towards improving patient-specific quality assurance in volumetric modulated arc therapy plans using machine learning

IntroductionPrevious literature has shown general trade-offs between plan complexity and resulting quality assurance (QA) outcomes. However, existing solutions for controlling this trade-off do not guarantee corresponding improvements in deliverability. Therefore, this work explored the feasibility of an optimization framework for directly maximizing predicted QA outcomes of plans without compromising the dosimetric quality of plans designed with an established knowledge-based planning (KBP) technique.Materials and MethodsA support vector machine (SVM) was developed – using a database of 500 previous VMAT plans – to predict gamma passing rates (GPRs; 3%/3mm percent dose-difference/distance-to-agreement with local normalization) based on selected complexity features. A heuristic, QA-based optimization (QAO) framework was devised by utilizing the SVM model to iteratively modify mechanical treatment features most commonly associated with suboptimal GPRs. Specifically, leaf gaps (LGs) <50 mm were widened by random amounts, which impacts all aperture-based complexity features. 13 prostate KBP-guided VMAT plans were optimized via QAO using user-specified maximum LG displacements before corresponding changes in predicted GPRs and dose were assessed.ResultsPredicted GPRs increased by an average of 1.14 ± 1.25% (p = 0.006) with QAO using a 3 mm maximum random LG displacement. There were small differences in dose, resulting in similarly small changes in tumor control probability (maximum increase = 0.05%) and normal tissue complication probabilities in the bladder, rectum, and femoral heads (maximum decrease = 0.2% in the rectum).ConclusionThis study explored the feasibility of QAO and warrants future investigations of further incorporating QA endpoints into plan optimization.     

A convolution neural network for higher resolution dose prediction in prostate volumetric modulated arc therapy

PurposeThis study aims to investigate the feasibility of using convolutional neural networks to predict an accurate and high resolution dose distribution from an approximated and low resolution input dose.MethodsSixty-six patients were treated for prostate cancer with VMAT. We created the treatment plans using the Acuros XB algorithm with 2 mm grid size, followed by the dose calculated using the anisotropic analytical algorithm with 5 mm grid with the same plan parameters. U-net model was used to predict 2 mm grid dose from 5 mm grid dose. We investigated the two models differing for the training data used as input, one used just the low resolution dose (D model) and the other combined the low resolution dose with CT data (DC model). Dice similarity coefficient (DSC) was calculated to ascertain how well the shape of the dose-volume is matched. We conducted gamma analysis for the following: DVH from the two models and the reference DVH for all prostate structures.ResultsThe DSC values in the DC model were significantly higher than those in the D model (p < 0.01). For the CTV, PTV, and bladder, the gamma passing rates in the DC model were significantly higher than those in the D model (p < 0.002–0.02). The mean doses in the CTV and PTV for the DC model were significantly better matched to those in the reference dose (p < 0.0001).ConclusionsThe proposed U-net model with dose and CT image used as input predicted more accurate dose.     

Impact of interfractional changes in head and neck cancer patients on the delivered dose in intensity modulated radiotherapy with protons and photons

PurposeTo investigate the influence of interfractional changes on the delivered dose of intensity modulated proton (IMPT) and photon plans (IMXT).Methods and materialsFive postoperative head and neck cancer patients, previously treated with tomotherapy at our institute, were analyzed. The planning study is based on megavoltage (MV) control images. For each patient one IMPT plan and one IMXT plan were generated on the first MV-CT and recalculated on weekly control MV-CTs in the actual treatment position. Dose criteria for evaluation were coverage and conformity of the planning target volume (PTV), as well as mean dose to parotids and maximum dose to spinal cord.ResultsConsiderable dosimetric changes were observed for IMPT and IMXT plans. Proton plans showed a more pronounced increase of maximum dose and decrease of minimum dose with local underdosage occurring even in the center of the PTV (worst IMPT vs. IMXT coverage: 66.7% vs. 85.0%). The doses to organs at risk (OARs) increased during the treatment period. However, the OAR doses of IMPT stayed below corresponding IMXT values at any time. For both modalities treatment plans did not necessarily worsen monotonically throughout the treatment.ConclusionsAlthough absolute differences between planned and reconstructed doses were larger in IMPT plans, doses to OARs were higher in IMXT plans. Tumor coverage was more stable in IMXT plans; IMPT dose distributions indicated a high risk for local underdosage during the treatment course.