Consistency in contouring of organs at risk by artificial intelligence vs oncologists in head and neck cancer patients.

BACKGROUND: In the Danish Head and Neck Cancer Group (DAHANCA) 35 trial, patients are selected for proton treatment based on simulated reductions of Normal Tissue Complication Probability (NTCP) for proton compared to photon treatment at the referring departments. After inclusion in the trial, immobilization, scanning, contouring and planning are repeated at the national proton centre. The new contours could result in reduced expected NTCP gain of the proton plan, resulting in a loss of validity in the selection process. The present study evaluates if contour consistency can be improved by having access to AI (Artificial Intelligence) based contours. MATERIALS AND METHODS: The 63 patients in the DAHANCA 35 pilot trial had a CT from the local DAHANCA centre and one from the proton centre. A nationally validated convolutional neural network, based on nnU-Net, was used to contour OARs on both scans for each patient. Using deformable image registration, local AI and oncologist contours were transferred to the proton centre scans for comparison. Consistency was calculated with the Dice Similarity Coefficient (DSC) and Mean Surface Distance (MSD), comparing contours from AI to AI and oncologist to oncologist, respectively. Two NTCP models were applied to calculate NTCP for xerostomia and dysphagia. RESULTS: The AI contours showed significantly better consistency than the contours by oncologists. The median and interquartile range of DSC was 0.85 [0.78 - 0.90] and 0.68 [0.51 - 0.80] for AI and oncologist contours, respectively. The median and interquartile range of MSD was 0.9 mm [0.7 - 1.1] mm and 1.9 mm [1.5 - 2.6] mm for AI and oncologist contours, respectively. There was no significant difference in ΔNTCP. CONCLUSIONS: The study showed that OAR contours made by the AI algorithm were more consistent than those made by oncologists. No significant impact on the ΔNTCP calculations could be discerned.

Individual lymph node position variation for rectal cancer patients treated with long course chemoradiotherapy.

Background and purpose: Delivery of high precision radiotherapy lymph node boosts requires detailed information on the interfraction positional variation of individual lymph nodes. In this study we characterized interfraction positional shifts of suspected malignant lymph nodes for rectal cancer patients receiving long course radiotherapy. Furthermore, we investigated parameters which could affect the magnitude of the position variation. Materials and Methods: Fourteen patients from a prospective clinical imaging study with a total of 61 suspected malignant lymph nodes in the mesorectum, presacral, and lateral regions, were included. The primary gross tumor volume (GTVp) and all suspected malignant lymph nodes were delineated on six magnetic resonance imaging scans per patient. Positional variation was calculated as systematic and random errors, based on shifts of center-of-mass, and estimated relative to either bony structures or the GTVp using a hierarchical linear mixed model. Results: Depending on location and direction, systematic and random variations (relative to bony structures) were within 0.6-2.8 mm and 0.6-2.9 mm, respectively. Systematic and random variations increased when evaluating position relative to GTVp (median increase of 0.6 mm and 0.5 mm, respectively). Correlations with scan time-point and relative bladder volume were found in some directions. Conclusions: Using linear mixed modeling, we estimated systematic and random positional variation for suspected malignant lymph nodes in rectal cancer patients treated with long course radiotherapy. Statistically significant correlations of the magnitude of the lymph node shifts were found related to scan time-point and relative bladder volume.

Treatment planning for patients with low rectal cancer in a multicenter prospective organ preservation study.

BACKGROUND: Non-surgical management of rectal cancer relies on (chemo)radiotherapy as the definitive treatment modality. This study reports and evaluates the clinical high dose radiotherapy treatment plans delivered to patients with low resectable rectal cancer in a Danish multicenter trial. METHODS: The Danish prospective multicenter phase II Watchful Waiting 2 trial (NCT02438839) investigated definitive chemoradiation for non-surgical management of low rectal cancer. Three Danish centers participated in the trial and committed to protocol-specified treatment planning and delivery requirements. The protocol specified a dose of 50.4 Gy in 28 fractions to the elective volume (CTV-/PTV-E) and a concomitant boost of 62 Gy in 28 fractions to the primary target volume (CTV-/PTV-T). RESULTS: The trial included 108 patients, of which 106 treatment plans were available for retrospective analysis. Dose coverage planning goals for the main target structures were fulfilled for 94% of the treatment plans. However, large intercenter differences in doses to organs-at-risk (OARs) were seen, especially for the intestines. Five patients had a V60Gy>10 cm3 for the intestines and two patients for the bladder. CONCLUSION: Prescribed planning goals for target coverage were fulfilled for 94% of the treatment plans, however analysis of OAR doses and volumes indicated intercenter variations. Dose escalation to 62 Gy (as a concomitant boost to the primary tumor) introduced no substantial high dose volumes (>60 Gy) to the bladder and intestines. The treatment planning goals may be used for future prospective evaluation of highdose radiotherapy for organ preservation for low rectal cancer.

Evaluation of decentralised model-based selection of head and neck cancer patients for a proton treatment study. DAHANCA 35.

INTRODUCTION: Proton treatment can potentially spare patients with H&N cancer for substantial treatment-related toxicities. The current study investigated the reproducibility of a decentralised model-based selection of patients for a proton treatment study when the selection plans were compared to the clinical treatment plans performed at the proton centre. METHODS: Sixty-three patients were selected for proton treatment in the six Danish Head and Neck Cancer (DAHANCA) centres. The patients were selected based on normal tissue complication probability (NTCP) estimated from local photon and proton treatment plans, which showed a ΔNTCP greater than 5%-point for either grade 2 + dysphagia or grade 2 + xerostomia at six months. The selection plans were compared to the clinical treatment plans performed at the proton centre. RESULTS: Of the 63 patients, 49 and 25 were selected based on an estimated benefit in risk of dysphagia and xerostomia, respectively. Eleven patients had a potential gain in both toxicities. The mean ΔNTCP changed from the local selection plan comparison to the clinical comparison from 6.9 to 5.3 %-points (p = 0.01) and 7.3 to 4.9 %-points (p = 0.03) for dysphagia and xerostomia, respectively. Volume differences in both CTV and OAR could add to the loss in ΔNTCP. 61 of the 63 clinical plans had a positive ΔNTCP, and 38 had a ΔNTCP of 5%-points for at least one of the two endpoints. CONCLUSION: A local treatment plan comparison can be used to select candidates for proton treatment. The local comparative proton plan overestimates the potential benefit of the clinical proton plan. Continuous quality assurance of the delineation procedures and planning is crucial in the subsequent randomised clinical trial setting.

High-dose thoracic radiation therapy for non-small cell lung cancer: a novel grading scale of radiation-induced lung injury for symptomatic radiation pneumonitis.

BACKGROUND: Symptomatic radiation pneumonitis (RP) may be a serious complication after thoracic radiation therapy (RT) for non-small cell lung cancer (NSCLC). This prospective observational study sought to evaluate the utility of a novel radiation-induced lung injury (RILI) grading scale (RGS) for the prediction of RP. MATERIALS AND METHODS: Data of 41 patients with NSCLC treated with thoracic RT of 60-66 Gy were analysed. CT scans were scheduled before RT, one month post-RT, and every three months thereafter for one year. Symptomatic RP was defined as Common Terminology Criteria for Adverse Events grade ≥ 2. RGS grading ranged from 0 to 3. The inter-observer variability of the RGS was assessed by four senior radiologists. CT scans performed 28 ± 10 days after RT were used to analyse the predictive value of the RGS. The change in the RGS severity was correlated to dosimetric parameters. RESULTS: The CT obtained one month post-RT showed RILI in 36 (88%) of patients (RGS grade 0 [5 patients], 1 [25 patients], 2 [6 patients], and 3 [5 patients]). The inter-observer agreement of the RGS grading was high (Kendall's W coefficient of concordance = 0.80, p < 0.01). Patients with RGS grades 2-3 had a significantly higher risk for development of RP (relative risk (RR): 2.4, 95% CI 1.6-3.7, p < 0.01) and RP symptoms within 8 weeks after RT (RR: 4.8, 95% CI 1.3-17.6, p < 0.01) compared to RGS grades 0-1. The specificity and sensitivity of the RGS grades 2-3 in predicting symptomatic RP was 100% (95% CI 80.5-100%) and 45.4% (95% CI 24.4-67.8%), respectively. Increase in RGS severity correlated to mean lung dose and the percentage of the total lung volume receiving 5 Gy. CONCLUSIONS: The RGS is a simple radiologic tool associated with symptomatic RP. A validation study is warranted.

A phantom study of dose compensation behind hip prosthesis using portal dosimetry and dynamic MLC.

BACKGROUND AND PURPOSE: A dose compensation method is presented for patients with hip prosthesis based on Dynamic Multi Leaves Collimator (DMLC) planning. Calculations are done from an exit Portal Dose Image (PDI) from 6 MV Photon beam using an Electronic Portal Imaging Device (EPID) from Varian. Four different hip prostheses are used for this work. METHODS: From an exit PDI the fluence needed to yield a uniform dose distribution behind the prosthesis is calculated. To back-project the dose distribution through the phantom, the lateral scatter is removed by deconvolution with a point spread function (PSF) determined for depths from 10 to 40 cm. The dose maximum, D(max), is determined from the primary plan which delivers the PDI. A further deconvolution to remove the dose glare effect in the EPID is performed as well. Additionally, this calculated fluence distribution is imported into the Treatment Planning System (TPS) for the final calculation of a DMLC plan. The fluence file contains information such as the relative central axis (CAX) position, grid size and fluence size needed for correct delivery of the DMLC plan. GafChromic EBT films positioned at 10 cm depth are used as verification of uniform dose distributions behind the prostheses. As the prosthesis is positioned at the phantom surface the dose verifications are done 10 cm from the prosthesis. CONCLUSION: The film measurement with 6 MV photon beam shows uniform doses within 5% for most points, but with hot/cold spots of 10% near the femoral head prostheses.

A new lung stent tested as fiducial marker in a porcine model.

INTRODUCTION: The aim of the present study was to test the feasibility of a Nickel-Titanium (Ni-Ti) stent technique (Memocore™) in a porcine model. The stent is intended as a new fiducial for gated image guided radiotherapy in the lung. The study included test of an improved insertion system and respiratory gated treatments with this new technique. METHODS AND MATERIALS: Tests were carried out in a porcine model using Göttingen mini-pigs. The study included 10 animals. Planning CT was performed as 4 dimensional CT (4DCT) using the Varian RPM system. Respiratory gated radiotherapy treatments were simulated using the Brainlab ExacTrac system. Reproducibility of stent position during treatment was analyzed off-line using an experimental version of the ExacTrac software. The experimental version has a dedicated algorithm for segmentation of the stent in the planning CT and subsequent registration to X-ray position images. RESULTS: A total of 23 stents were inserted in the 10 animals. Stents could be placed in all parts of the lungs. No stent migrated within the four weeks the experiment lasted. Stent trajectories in the lung were not reproducible, even though respiration was highly standardized using a respirator. The best accuracy of stent position in the gating window was obtained using gating at the half_max amplitude as reference level. The smallest stent movement within the gating window was observed in the exhale phase. Further success of human application will depend on the possibility to insert the stent within or close to lung tumors. CONCLUSIONS: This new technique based on the Memocore™ lung stent used in connection with respiratory gated radiotherapy was demonstrated to be feasible in a porcine model. The study demonstrated lack of reproducibility in lung trajectories of inserted stents. The technique gave the best accuracy when applied to the exhale phase of respiration.