The Development and External Validation of Artificial Intelligence-Driven MRI-Based Models to Improve Prediction of Lesion-Specific Extraprostatic Extension in Patients with Prostate Cancer.

Adequate detection of the histopathological extraprostatic extension (EPE) of prostate cancer (PCa) remains a challenge using conventional radiomics on 3 Tesla multiparametric magnetic resonance imaging (3T mpMRI). This study focuses on the assessment of artificial intelligence (AI)-driven models with innovative MRI radiomics in predicting EPE of prostate cancer (PCa) at a lesion-specific level. With a dataset encompassing 994 lesions from 794 PCa patients who underwent robot-assisted radical prostatectomy (RARP) at two Dutch hospitals, the study establishes and validates three classification models. The models were validated on an internal validation cohort of 162 lesions and an external validation cohort of 189 lesions in terms of discrimination, calibration, net benefit, and comparison to radiology reporting. Notably, the achieved AUCs ranged from 0.86 to 0.91 at the lesion-specific level, demonstrating the superior accuracy of the random forest model over conventional radiological reporting. At the external test cohort, the random forest model was the best-calibrated model and demonstrated a significantly higher accuracy compared to radiological reporting (83% vs. 67%, p = 0.02). In conclusion, an AI-powered model that includes both existing and novel MRI radiomics improves the detection of lesion-specific EPE in prostate cancer.

Is the introduction of more advanced radiotherapy techniques for locally-advanced head and neck cancer associated with improved quality of life and reduced symptom burden?

Technical improvements in head and neck cancer radiotherapy over the last decade have resulted in substantial reductions in dose to organs-at-risk. For a mix of tumors, we saw less xerostomia moving from 3D-conformal to more advanced techniques. For oropharynx-only there were additional improvements, including in global quality-of-life and sticky saliva.

Analysis of EORTC-1219-DAHANCA-29 trial plans demonstrates the potential of knowledge-based planning to provide patient-specific treatment plan quality assurance.

INTRODUCTION: Radiotherapy treatment plan quality can influence clinical trial outcomes and general QA may not identify suboptimal organ-at-risk (OAR) sparing. We retrospectively performed patient-specific quality assurance (QA) of 100 head-and-neck cancer (HNC) plans from the EORTC-1219-DAHANCA-29 study. MATERIALS AND METHODS: A 177-patient RapidPlan (Varian Medical Systems) model comprising institutional HNC plans was used to QA trial plans (Ptrial). RapidPlan plans (Prapidplan) were created using RapidPlan and Eclipse scripting to achieve a high degree of automation. Comparison between Prapidplan mean predicted/achieved OAR doses, and Ptrial mean OAR doses was made for parotid/submandibular glands (PGs/SMGs) and swallowing muscles (SM). RESULTS: OAR predictions were made within 2 min per patient. Averaged PG/SMG/SM mean doses were 2.0/9.0/3.8 Gy lower in Prapidplan. Using predicted Prapidplan combined mean OAR dose as the benchmark, a total of 60/27/4 trial plans could be improved by 3/6/9 Gy respectively. DISCUSSION: Individualized QA indicated that OAR sparing could frequently be improved in EORTC-1219 study plans, even though they met the trial's generic plan criteria. Automated, patient-specific QA can be performed within a few minutes and should be considered to reduce the influence of planning variation on trial outcomes.

Personalized automated treatment planning for breast plus locoregional lymph nodes using Hybrid RapidArc.

PURPOSE: Breast cancer patients who require locoregional lymph node (LLN) irradiation can be treated using a hybrid RapidArc technique combining 2 tangential and 3 RapidArc fields. Because the creation of hybrid RapidArc plans is complex and labor-intensive, we developed an automated treatment planning workflow using the scripting application programming interface of the Eclipse treatment planning system. METHODS AND MATERIALS: Fifteen patients (5 right- and 10 left-sided) previously treated with breast + LLN radiation therapy were replanned using the script. The automated workflow included1 optimal placement of the tangential fields based on the planning target volume and organ-at-risk contours, followed by optimization of field weights and beam energy2; positioning of the RapidArc fields; and3 subsequent RapidArc optimization using the RapidPlan knowledge-based planning solution. RESULTS: Average total planning times were 163 ± 97 and 33 ± 5 minutes for the manual and automated plans, respectively, with approximately 130 and 5 minutes of user interaction. Dosimetrically, both sets of plans were very similar, with comparable planning target volume dose homogeneity values and organ-at-risk mean dose differences of ≤1.9 Gy. In 14/15 patients, the physician judged that the automated plan was either preferred (n = 4) or equal (n = 10) to the manual plan. CONCLUSIONS: The complex hybrid RapidArc planning process for patients requiring breast + LLN irradiation was automated by optimizing the tangential field setup and integrating RapidPlan. The quality of the automated and manual plans was comparable, whereas automated planning times were substantially shorter. The principles described here could be used to automate other planning workflows.

Using a knowledge-based planning solution to select patients for proton therapy.

BACKGROUND AND PURPOSE: Patient selection for proton therapy by comparing proton/photon treatment plans is time-consuming and prone to bias. RapidPlan™, a knowledge-based-planning solution, uses plan-libraries to model and predict organ-at-risk (OAR) dose-volume-histograms (DVHs). We investigated whether RapidPlan, utilizing an algorithm based only on photon beam characteristics, could generate proton DVH-predictions and whether these could correctly identify patients for proton therapy. MATERIAL AND METHODS: ModelPROT and ModelPHOT comprised 30 head-and-neck cancer proton and photon plans, respectively. Proton and photon knowledge-based-plans (KBPs) were made for ten evaluation-patients. DVH-prediction accuracy was analyzed by comparing predicted-vs-achieved mean OAR doses. KBPs and manual plans were compared using salivary gland and swallowing muscle mean doses. For illustration, patients were selected for protons if predicted ModelPHOT mean dose minus predicted ModelPROT mean dose (ΔPrediction) for combined OARs was ≥6Gy, and benchmarked using achieved KBP doses. RESULTS: Achieved and predicted ModelPROT/ModelPHOT mean dose R2 was 0.95/0.98. Generally, achieved mean dose for ModelPHOT/ModelPROT KBPs was respectively lower/higher than predicted. Comparing ModelPROT/ModelPHOT KBPs with manual plans, salivary and swallowing mean doses increased/decreased by <2Gy, on average. ΔPrediction≥6Gy correctly selected 4 of 5 patients for protons. CONCLUSIONS: Knowledge-based DVH-predictions can provide efficient, patient-specific selection for protons. A proton-specific RapidPlan-solution could improve results.

Effect of Dosimetric Outliers on the Performance of a Commercial Knowledge-Based Planning Solution.

PURPOSE: RapidPlan, a commercial knowledge-based planning solution, uses a model library containing the geometry and associated dosimetry of existing plans. This model predicts achievable dosimetry for prospective patients that can be used to guide plan optimization. However, it is unknown how suboptimal model plans (outliers) influence the predictions or resulting plans. We investigated the effect of, first, removing outliers from the model (cleaning it) and subsequently adding deliberate dosimetric outliers. METHODS AND MATERIALS: Clinical plans from 70 head and neck cancer patients comprised the uncleaned (UC) ModelUC, from which outliers were cleaned (C) to create ModelC. The last 5 to 40 patients of ModelC were replanned with no attempt to spare the salivary glands. These substantial dosimetric outliers were reintroduced to the model in increments of 5, creating Model5 to Model40 (Model5-40). These models were used to create plans for a 10-patient evaluation group. Plans from ModelUC and ModelC, and ModelC and Model5-40 were compared on the basis of boost (B) and elective (E) target volume homogeneity indexes (HIB/HIE) and mean doses to oral cavity, composite salivary glands (compsal) and swallowing (compswal) structures. RESULTS: On average, outlier removal (ModelC vs ModelUC) had minimal effects on HIB/HIE (0%-0.4%) and sparing of organs at risk (mean dose difference to oral cavity and compsal/compswal were ≤0.4 Gy). Model5-10 marginally improved compsal sparing, whereas adding a larger number of outliers (Model20-40) led to deteriorations in compsal up to 3.9 Gy, on average. These increases are modest compared to the 14.9 Gy dose increases in the added outlier plans, due to the placement of optimization objectives below the inferior boundary of the dose-volume histogram-predicted range. CONCLUSIONS: Overall, dosimetric outlier removal from or addition of 5 to 10 outliers to a 70-patient model had marginal effects on resulting plan quality. Although the addition of >20 outliers deteriorated plan quality, the effect was modest. In this study, RapidPlan demonstrated robustness for moderate proportions of salivary gland dosimetric outliers.

A longitudinal evaluation of improvements in radiotherapy treatment plan quality for head and neck cancer patients.

PURPOSE: To investigate changes in head-and-neck cancer (HNC) plan quality following the introduction of new technologies and planning techniques in the last decade. METHODS AND MATERIALS: Thirty plans were selected from each of four successive periods (P). P1: 7-field static intensity-modulated radiotherapy (IMRT) with parotid gland sparing; P2: dual-arc volumetric-modulated arc therapy (VMAT, similar to P3-P4), including submandibular gland sparing; P3: inclusion of individual swallowing muscles and attempts to further reduce parotid and oral cavity doses through manual interactive optimization; P4: containing the same organs-at-risk (OARs) as P3, but automatically interactively optimized. Plan benchmarking included mean salivary gland/swallowing muscle/oral cavity (Dsal/Dswal/Doc) doses. Differences in mean doses between the periods were analyzed by an ANCOVA, taking geometric differences across periods into account. RESULTS: Compared to P1, P2 plans improved Dsal by 3.4Gy on average. P3 improved Dsal/Dswal/Doc by 6.9/11.5/7.2Gy over P2, showing that Dswal and Dsal could be improved simultaneously. In P4, Doc/Dswal slightly improved over P3 by 1.7/3.8Gy. Improved OAR sparing in P3/P4 did not come at the cost of increased dose deposition elsewhere and planning target volume (PTV) dose homogeneity was similar. CONCLUSIONS: New technologies and planning techniques were successfully implemented into routine clinical care and resulting in improved HNC plan quality.

Detailed evaluation of an automated approach to interactive optimization for volumetric modulated arc therapy plans.

PURPOSE: Interactive optimization during treatment planning requires intermittent adjustment of organ-at-risk (OAR) objectives relative to the dose-volume histogram line. This is a labor-intensive process and the resulting plans are prone to variations in quality. The authors' in-house developed approach to automated interactive optimization (AIO) automatically moves the mouse cursor to adjust the position of on-screen optimization objectives. This allows for the use of more objectives per OAR and results in a more frequent and consistent adjustment of these objectives during optimization. The authors report a detailed evaluation of AIO performance in support of its implementation for routine head and neck cancer (HNC) planning and an evaluation for locally advanced lung cancer (LC) planning which requires a different optimization strategy. METHODS: Volumetric modulated arc therapy AIO plans (APs) were created for 70 HNC patients with a simultaneously integrated boost and 20 LC patients and benchmarked against their respective manually interactively optimized plans (MPs). The same set of optimization objectives and priorities was used for all APs, although planning target volume (PTV) optimization priorities could be increased manually in a subsequent "continue previous optimization" calculation. HNC plans were benchmarked using mean dose to individual and composite OARs and elective/boost PTV (PTVE/PTVB) volumes receiving 95% and 107% of the prescription dose (V95% and V107%, respectively). A clinician performed blinded comparison of 20 APs and respective MPs. LC plans were compared using PTV V95%/V107%, contralateral lung (CL) volume receiving 5 Gy (V5Gy), total lung (TL)-PTV V5Gy/V20Gy, and esophagus and heart V40Gy/V60Gy/mean doses. RESULTS: For HNC, statistically significant improvements in sparing of all OARs, except for the ipsilateral submandibular gland and trachea, were obtained in the APs compared to MPs. Average mean dose to oral cavity, composite salivary, and swallowing structures were 25.4/23.8, 24.2/23.2, and 29.5/25.5 Gy, respectively, for the MPs/APs. PTV heterogeneity was similar: in the APs, PTVB V95% was 0.2% higher while PTV B/PTV E V107% was 0.4%/1.0% lower. In 19 out of 20 HNC patients, the clinician preferred the AP, mainly because of better OAR sparing and PTV dose homogeneity. For LC, APs had a significantly lower CL V5Gy (6.1%), heart mean dose/V60Gy (0.9 Gy/1.2%) and esophagus mean dose/V60Gy (0.9 Gy/2.8%), a nonsignificantly higher TL V20Gy (1.4%), and a slight, but significantly higher dose deposition to the body. PTV dose coverage and homogeneity were similar in the APs and MPs. AIO was considered sufficiently robust for clinical use in LC. CONCLUSIONS: HNC and LC APs were at least as good as, and often of improved quality over MPs. To date, AIO has been clinically implemented for HNC planning.

Can knowledge-based DVH predictions be used for automated, individualized quality assurance of radiotherapy treatment plans?

BACKGROUND: Treatment plan quality assurance (QA) is important for clinical studies and for institutions aiming to generate near-optimal individualized treatment plans. However, determining how good a given plan is for that particular patient (individualized patient/plan QA, in contrast to running through a checklist of generic QA parameters applied to all patients) is difficult, time consuming and operator-dependent. We therefore evaluated the potential of RapidPlan, a commercial knowledge-based planning solution, to automate this process, by predicting achievable OAR doses for individual patients based on a model library consisting of historical plans with a range of organ-at-risk (OAR) to planning target volume (PTV) geometries and dosimetries. METHODS: A 90-plan RapidPlan model, generated using previously created automatic interactively optimized (AIO) plans, was used to predict achievable OAR dose-volume histograms (DVHs) for the parotid glands, submandibular glands, individual swallowing muscles and oral cavities of 20 head and neck cancer (HNC) patients using a volumetric modulated (RapidArc) simultaneous integrated boost technique. Predicted mean OAR doses were compared with mean doses achieved when RapidPlan was used to make a new plan. Differences between the achieved and predicted DVH-lines were analyzed. Finally, RapidPlan predictions were used to evaluate achieved OAR sparing of AIO and manual interactively optimized plans. RESULTS: For all OARs, strong linear correlations (R(2) = 0.94-0.99) were found between predicted and achieved mean doses. RapidPlan generally overestimated the amount of achievable sparing for OARs with a large degree of OAR-PTV overlap. RapidPlan QA using predicted doses alone identified that for 50 % (10/20) of the manually optimized plans, sparing of the composite salivary glands, oral cavity or composite swallowing muscles could be improved by at least 3 Gy, 5 Gy or 7 Gy, respectively, while this was the case for 20 % (4/20) AIO plans. These predicted gains were validated by replanning the identified patients using RapidPlan. CONCLUSIONS: Strong correlations between predicted and achieved mean doses indicate that RapidPlan could accurately predict achievable mean doses. This shows the feasibility of using RapidPlan DVH prediction alone for automated individualized head and neck plan QA. This has applications in individual centers and clinical trials.

Evaluation of a knowledge-based planning solution for head and neck cancer.

PURPOSE: Automated and knowledge-based planning techniques aim to reduce variations in plan quality. RapidPlan uses a library consisting of different patient plans to make a model that can predict achievable dose-volume histograms (DVHs) for new patients and uses those models for setting optimization objectives. We benchmarked RapidPlan versus clinical plans for 2 patient groups, using 3 different libraries. METHODS AND MATERIALS: Volumetric modulated arc therapy plans of 60 recent head and neck cancer patients that included sparing of the salivary glands, swallowing muscles, and oral cavity were evenly divided between 2 models, Model(30A) and Model(30B), and were combined in a third model, Model60. Knowledge-based plans were created for 2 evaluation groups: evaluation group 1 (EG1), consisting of 15 recent patients, and evaluation group 2 (EG2), consisting of 15 older patients in whom only the salivary glands were spared. RapidPlan results were compared with clinical plans (CP) for boost and/or elective planning target volume homogeneity index, using HI(B)/HI(E) = 100 × (D2% - D98%)/D50%, and mean dose to composite salivary glands, swallowing muscles, and oral cavity (D(sal), D(swal), and D(oc), respectively). RESULTS: For EG1, RapidPlan improved HI(B) and HI(E) values compared with CP by 1.0% to 1.3% and 1.0% to 0.6%, respectively. Comparable D(sal) and D(swal) values were seen in Model(30A), Model(30B), and Model60, decreasing by an average of 0.1, 1.0, and 0.8 Gy and 4.8, 3.7, and 4.4 Gy, respectively. However, differences were noted between individual organs at risk (OARs), with Model(30B) increasing D(oc) by 0.1, 3.2, and 2.8 Gy compared with CP, Model(30A), and Model60. Plan quality was less consistent when the patient was flagged as an outlier. For EG2, RapidPlan decreased D(sal) by 4.1 to 4.9 Gy on average, whereas HI(B) and HI(E) decreased by 1.1% to 1.5% and 2.3% to 1.9%, respectively. CONCLUSIONS: RapidPlan knowledge-based treatment plans were comparable to CP if the patient's OAR-planning target volume geometry was within the range of those included in the models. EG2 results showed that a model including swallowing-muscle and oral-cavity sparing can be applied to patients with only salivary gland sparing. This may allow model library sharing between institutes. Optimal detection of inadequate plans and population of model libraries requires further investigation.

Comparison of organ-at-risk sparing and plan robustness for spot-scanning proton therapy and volumetric modulated arc photon therapy in head-and-neck cancer.

PURPOSE: Proton radiotherapy for head-and-neck cancer (HNC) aims to improve organ-at-risk (OAR) sparing over photon radiotherapy. However, it may be less robust for setup and range uncertainties. The authors investigated OAR sparing and plan robustness for spot-scanning proton planning techniques and compared these with volumetric modulated arc therapy (VMAT) photon plans. METHODS: Ten HNC patients were replanned using two arc VMAT (RapidArc) and spot-scanning proton techniques. OARs to be spared included the contra- and ipsilateral parotid and submandibular glands and individual swallowing muscles. Proton plans were made using Multifield Optimization (MFO, using three, five, and seven fields) and Single-field Optimization (SFO, using three fields). OAR sparing was evaluated using mean dose to composite salivary glands (CompSal) and composite swallowing muscles (CompSwal). Plan robustness was determined for setup and range uncertainties (±3 mm for setup, ±3% HU) evaluating V95% and V107% for clinical target volumes. RESULTS: Averaged over all patients CompSal/CompSwal mean doses were lower for the three-field MFO plans (14.6/16.4 Gy) compared to the three-field SFO plans (20.0/23.7 Gy) and VMAT plans (23.0/25.3 Gy). Using more than three fields resulted in differences in OAR sparing of less than 1.5 Gy between plans. SFO plans were significantly more robust than MFO plans. VMAT plans were the most robust. CONCLUSIONS: MFO plans had improved OAR sparing but were less robust than SFO and VMAT plans, while SFO plans were more robust than MFO plans but resulted in less OAR sparing. Robustness of the MFO plans did not increase with more fields.

Automatic interactive optimization for volumetric modulated arc therapy planning.

BACKGROUND: Intensity modulated radiotherapy treatment planning for sites with many different organs-at-risk (OAR) is complex and labor-intensive, making it hard to obtain consistent plan quality. With the aim of addressing this, we developed a program (automatic interactive optimizer, AIO) designed to automate the manual interactive process for the Eclipse treatment planning system. We describe AIO and present initial evaluation data. METHODS: Our current institutional volumetric modulated arc therapy (RapidArc) planning approach for head and neck tumors places 3-4 adjustable OAR optimization objectives along the dose-volume histogram (DVH) curve that is displayed in the optimization window. AIO scans this window and uses color-coding to differentiate between the DVH-lines, allowing it to automatically adjust the location of the optimization objectives frequently and in a more consistent fashion. We compared RapidArc AIO plans (using 9 optimization objectives per OAR) with the clinical plans of 10 patients, and evaluated optimal AIO settings. AIO consistency was tested by replanning a single patient 5 times. RESULTS: Average V95&V107 of the boost planning target volume (PTV) and V95 of the elective PTV differed by ≤0.5%, while average elective PTV V107 improved by 1.5%. Averaged over all patients, AIO reduced mean doses to individual salivary structures by 0.9-1.6Gy and provided mean dose reductions of 5.6Gy and 3.9Gy to the composite swallowing structures and oral cavity, respectively. Re-running AIO five times, resulted in the aforementioned parameters differing by less than 3%. CONCLUSIONS: Using the same planning strategy as manually optimized head and neck plans, AIO can automate the interactive Eclipse treatment planning process and deliver dosimetric improvements over existing clinical plans.

Toward optimal organ at risk sparing in complex volumetric modulated arc therapy: an exponential trade-off with target volume dose homogeneity.

PURPOSE: Conventional radiotherapy typically aims for homogenous dose in the planning target volume (PTV) while sparing organs at risk (OAR). The authors quantified and characterized the trade-off between PTV dose inhomogeneity (IH) and OAR sparing in complex head and neck volumetric modulated arc therapy plans. METHODS: Thirteen simultaneous integrated boost plans were created per patient, for ten patients. PTV boost(B)/elective(E) optimization priorities were systematically increased. IHB and IHE, defined as (100% - V95%) + V107%, were evaluated against the average of the mean dose to the combined composite swallowing and combined salivary organs (D-OAR(comp)). To investigate the influence of OAR size and position with respect to PTVB/E, OAR dose was evaluated against a modified Euclidean distance (DMB/DME) between OAR and PTV. RESULTS: Although the achievable D-OAR(comp) for a given level of PTV IH differed between patients, excellent logarithmic fits described the D-OAR(comp)/IHB and IHE relationship in all patients (mean R(2) of 0.98 and 0.97, respectively). Allowing an increase in average IHB and IHE over a clinically acceptable range, e.g., from 0.4% ± 0.5% to 2.0% ± 2.0% and 6.9% ± 2.8% to 14.8% ± 2.7%, respectively, corresponded to a decrease in average dose to the composite salivary and swallowing structures from 30.3 ± 6.5 to 23.6 ± 4.7 Gy and 32.5 ± 8.3 to 26.8 ± 9.3 Gy. The increase in PTVE IH was mainly accounted for by an increase in V107, by on average 5.9%, rather than a reduction in V95, which was on average only 2%. A linear correlation was found between the OAR dose to composite swallowing structures and contralateral parotid and submandibular gland, with DME (R(2) = 0.83, 0.88, 0.95). Only mean ipsilateral parotid dose correlated with DMB (R(2) = 0.87). CONCLUSIONS: OAR sparing is highly dependent on the permitted PTVB/E IH. PTVE IH substantially influences OAR doses. These results are relevant for clinical practice and for future automated treatment-planning strategies.

Different treatment planning protocols can lead to large differences in organ at risk sparing.

BACKGROUND AND PURPOSE: Different planning protocols may define varying planning target volume (PTV) dose criteria. We investigated the hypothesis that this could result in differences in organ-at-risk (OAR) sparing. MATERIAL AND METHODS: Volumetric modulated arc therapy plans were created for ten locally advanced head and neck cancer patients following PTV criteria specified by the RTOG, EORTC and institutional (VUmc) protocols. Resulting plans were evaluated on the basis of the homogeneity index, calculated for the boost/elective PTVs as HIB/HIE=100%*(D2%-D98%)/D50% and mean dose to individual and composite salivary (compsal) and swallowing (compswal) OARs. RESULTS: RTOG plans were the most homogeneous, with mean HIB of 8.2±0.9%, compared to 9.5±1.0%/11.6±1.5% for the VUmc/EORTC plans. EORTC plans provided most OAR sparing, with compsal/compswal doses of 24.6±7.7/22.9±4.2Gy, compared to 32.2±9.7/29.9±4.2Gy and 28.4±8.1/24.7±5.3Gy for RTOG and VUmc, respectively. EORTC provided 7.2/7.7Gy mean dose reductions to the contra/ipsilateral parotid glands compared to RTOG. CONCLUSIONS: Different planning protocols resulted in different levels of PTV dose homogeneity. We observed differences of up to ⩾7Gy in composite and individual mean OAR doses. This could influence rates of toxicity and should be taken into account when comparing clinical studies. A consensus should be reached between major trial groups on appropriate PTV parameters.