Agreement among Radiation Oncologists and Radiologist in Contouring GTV on Planning and Mid-Treatment MRI for Locally Advanced Oropharynx Cancer

PURPOSE/OBJECTIVE(S)

Over the 7-week radiotherapy course, patients with head and neck cancers undergo significant anatomical changes, including weight loss and tumor shrinkage (with complete response at mid-treatment in as high as 50% of patients). The current standard of care is to maintain the same plan for the entire treatment duration, unless major dosimetric deviations are suspected. The use of MRI for treatment adaptation has the advantage of increased soft tissue contrast and is being integrated into several clinical practices with the recent implementation of the MR-Linac technology. In this interim secondary analysis of a phase 2, randomized trial, we assessed inter-observer variability (IOV) in the delineation of GTV at planning and mid-treatment in locally advanced oropharynx cancer (OPC).

MATERIALS/METHODS

Three expert radiation oncologists (RO) and 1 expert radiologist, with a minimum of 5 years' experience in MRI contouring, blinded to each other's volumes, were provided with planning and mid-treatment MRI (T1 gadolinium, T2-weighted and DWI) of 9 patients with locally advanced OPC. Primary gross tumor volume (GTV) was independently contoured by all 3 ROs and the radiologist. Contouring agreement was analyzed using the dice similarity coefficient (DSC) index and the average Hausdorff distance.

RESULTS

Nine cases of locally advanced OPC were included. Fifty percent of the patients were stage 3 (AJCC 8th edition) and all were p16+ OPC. There was moderate IOV in the delineation of GTV on planning MRI as evidenced by mean DSC index and Hausdorff distance of 0.75 (range 0.53 - 0.87) and 0.30 cm (max 2.14 cm). At mid-treatment, the GTV IOV had a mean DSC index and Hausdorff distance of 0.30 (range 0.19 - 0.76) and 0.55 cm (max 2.83), respectively. The contours of the radiologist compared to RO shown a mean DSC index and Hausdorff distance on planning MRI 0.63 (range 0.53 - 0.69) and 0.41 cm (max 2.14 cm), and on mid-treatment MRI of 0.36 (range 0.19 - 0.59) and 0.68 cm (max 1.08 cm).

CONCLUSION

Although this study demonstrates the presence of moderate IOV between 3 experienced head and neck ROs and one radiologist in an academic institution for the delineation of GTV on planning MRI, we observed a higher IOV between the 4 experts on mid-treatment MRI. We also experienced a higher IOV between the radiologist and RO on planning and mid-treatment images. These results show high degree of GTV volume variability on mid-treatment MRI. Therefore, peer review of contours is important and guidelines for MR adaptive tumor delineation are needed.

Stereotactic Boost and Short-Course Radiotherapy for p16-Associated Oropharynx Cancer (SHORT-OPC): First Planned Interim Safety Analysis from a Randomized Phase II Trial

PURPOSE/OBJECTIVE(S)

There is a need for safe treatment de-intensification in p16+ oropharynx cancer (OPC). The standard of care (SOC) radiotherapy (RT) regimen is cumbersome and associated with high toxicity. Stereotactic radiotherapy (SBRT) and multimodality image guidance is an opportunity to precisely target the gross tumor while safely reducing elective irradiation dose. We aim to assess the safety and efficacy of a short course RT for p16+ OPC, consisting of an SBRT boost to the gross tumor volume (GTV) followed by de-escalated elective irradiation.

MATERIALS/METHODS

In this randomized phase II trial, patients with p16-positive, stage I-II OPSCC with primary tumor <30 cc (8th Ed AJCC) are planned with combined CT, MRI and FDG-PET, and randomized to 1) SBRT boost (14 Gy in 2 fractions) to the GTV followed with de-escalated RT (+/- Cisplatin) to a dose of 40 Gy in 20 fractions, or 2) SOC RT (+/- Cisplatin) to a dose of 70 Gy in 33 fractions to the GTV and 59.4-54Gy (or equivalent) to the intermediate-to-low dose elective region. Patients are stratified by stage (I vs. II) and use of chemotherapy. The primary endpoint of the trial is locoregional control at 2 years, powered for a sample size of 100 patients. A Bayesian adaptive design includes 2 planned safety interim analysis using grade ≥ 3 subacute toxicities >40% as a stopping criterion, and 1 planned futility analysis. Acute adverse events (AE) are defined as those occurring ≤ 60 days from RT, subacute AE between 60-180 days after RT, and late AE >180 days from RT. This is the first planned toxicity analysis.

RESULTS

Twenty-one patients were randomly assigned and eligible (11 in SOC and 10 in experimental arm). Median age was 69 years (range 49-84); 29% and 71% had stage T1 and T2, while 10%, 85% and 1 patient had N0, N1 and N2 disease, respectively. RT alone and chemoradiation was administered in 67% and 33% of patients, respectively. At a median follow-up of 11 months (range 1.7-17.6), there was 1 local recurrence at the primary tumor site in the SOC arm (at 10 month) and no recurrence in the experimental arm. All enrolled patients remain alive at the time of analysis. There was a 54.5% rate of grade 3 acute AE in the SOC arm and 30.0% rate of grade 3 acute AE in the experimental arm. More specifically, 1, 5 (45%), 2 (18%), and 2 (18%) versus 0, 1, 1 and 1 patient developed acute grade 3 dysphagia, mucositis, pain and dermatitis in the SOC and experimental arm, respectively. There was no acute grade 4 or 5 toxicity. There was no grade ≥ 3 subacute toxicity or late toxicity in both arms.

CONCLUSION

This primary safety analysis showed that SBRT boost followed by a short course of de-escalated elective irradiation in p16+ OPC has limited early toxicity and meets criteria for study continuation.

Dynamic Prediction of Toxicities in Head and Neck Cancer Radiotherapy by 3D Convolutional Neural Network Using Daily Cone-Beam CTs

PURPOSE/OBJECTIVE(S)

Radiotherapy (RT) is essential in head and neck cancer (HNC) treatments, but often causes significant toxicity. Different machine learning models have shown promise in predicting RT-induced toxicity, but none have yet integrated the fluctuating anatomical changes. By integrating daily cone-beam CTs (CBCT) allowing sequential anatomical views, our aim is to build a dynamic predictive model for three major HNC RT toxicities: reactive feeding tube placement, hospitalization and radionecrosis (RN).

MATERIALS/METHODS

292 HNC cases treated with curative RT between 2017 and 2019 at our institution were retrospectively analyzed for clinical and radiological data. VoxelMorph, a deep deformable registration model, integrated the daily anatomical deformations between each CBCT and the planning CT, then converted them to Jacobian determinant matrix (Jf). Resnet, a convolutional neural network with multiple layers was trained using a 5-fold cross validation to integrate both radiological and clinical data. Each toxicity was classified as a binary decision using the cross-entropy loss to account for a class imbalance. Its predictive performance was compared to the baseline model using only clinical data.

RESULTS

The cohort included 78% men and 22% women, with a median age of 63 years (range 35-84). Primary cancer sites were 46% oropharynx, 19% larynx, 14% oral cavity, 7.5% nasopharynx, 5% hypopharynx, 4% unknown primary and 5% others; and stage ranged between Tx-4b N0 and 3b M0 (AJCC 8th Ed). Induction chemotherapy, concurrent chemotherapy, and adjuvant RT was used in 9%, 57% and 20% of patients, respectively. The incidence of feeding tube, hospitalization and RN was 19.9%, 7.2%, and 3.8%, respectively. Integrating Jf from the 10th RT CBCT showed better accuracy for each toxicity prediction: feeding tube (69.1% > 57.2%), hospitalization (75.3% > 63.1%) and RN (85.8% > 75.7%). Integrating both the raw CBCT and Jf improved hospitalization prediction (79.0% > 73.6%). Substituting Jf for the raw CBCT improved the prediction for RN (79.7% > 74.7%) and hospitalization (73.6% > 64.4%). For feeding tube, predictive performance of the Jf model trained against deformations showed a positive correlation between its performance and the RT received (r2 > 0.9) with increasing RT fractions, with a maximum accuracy of 83.1% at the 25th fraction. No such correlation was found for RN or hospitalization prediction.

CONCLUSION

To our knowledge, this is the first study showing promising results to predict HNC RT toxicities using daily per-treatment CBCT. Next steps involve integrating both the radiomic and the dosimetric inputs to build a more powerful model. This could expand to predict therapeutic outcomes and, ultimately, could guide decisions in individualized RT.

The Canadian Association of Radiation Oncology scope of practice guidelines for lung, liver and spine stereotactic body radiotherapy

AIMS

The Canadian Association of Radiation Oncology-Stereotactic Body Radiotherapy (CARO-SBRT) Task Force was established in 2010. The aim was to define the scope of practice guidelines for the profession to ensure safe practice specific for the most common sites of lung, liver and spine SBRT.

MATERIALS AND METHODS

A group of Canadian SBRT experts were charged by our national radiation oncology organisation (CARO) to define the basic principles and technologies for SBRT practice, to propose the minimum technological requirements for safe practice with a focus on simulation and image guidance and to outline procedural considerations for radiation oncology departments to consider when establishing an SBRT programme.

RESULTS

We recognised that SBRT should be considered as a specific programme within a radiation department, and we provide a definition of SBRT according to a Canadian consensus. We outlined the basic requirements for safe simulation as they pertain to spine, lung and liver tumours, and the fundamentals of image guidance. The roles of the radiation oncologist, medical physicist and dosimetrist have been detailed such that we strongly recommend the development of SBRT-specific teams. Quality assurance is a key programmatic aspect for safe SBRT practice, and we outline the basic principles of appropriate quality assurance specific to SBRT.

CONCLUSION

This CARO scope of practice guideline for SBRT is specific to liver, lung and spine tumours. The task force recommendations are designed to assist departments in establishing safe and robust SBRT programmes.