Three-year outcomes of uveal melanoma treated with intra-operative ultrasound-guided iodine-125 brachytherapy using custom-built eye plaques.

Purpose: The aim of this study was to demonstrate that uveal melanoma (UM) treated with eye plaque brachytherapy (EPB) with intra-operative ultrasound (IOUS) guidance results in increased local control. Material and methods: A retrospective study was conducted among 212 patients with 214 UM tumors treated by iodine-125 EPB with IOUS guidance from 2013 to 2019. 85 Gy was prescribed to tumor apical height or 5 mm from inner sclera, whichever was greater. Lesions were treated to 95% of 85 Gy at 2 mm margin from tumor edge. Local failure (LF), distant metastasis (DM), and radiation-related toxicity were recorded. Results: Median tumor apical height was 3.3 mm. COMS stage was 90 small (42.1%), 81 medium (37.9%), and 43 large (20.1%). Most patients had gene expression profile (GEP) class available, with 119 (55.6%), 30 (14.0%), 55 (25.7%) cases classified as 1A, 1B, and 2, respectively. Median dose at apex for tumor height > 5 mm and ≤ 5 mm was 85.0 Gy and 120.6 Gy, respectively. Outcomes data for 180 patients with over 12 months follow-up were reported. Mean follow-up was 37.3 months. Rates of LF and DM were 0.0% and 12.2%, respectively. Actuarial estimates of 5-year DM for class 1A, 1B, and 2 tumors were 2.5%, 0.0%, and 57.8%, respectively. 87 patients (48.3%) developed radiation-related toxicities. Conclusions: The excellent local control rate amongst lesions ranging across all sizes and GEP classes emphasizes the importance of image-guided brachytherapy with IOUS. We report favorable 5-year DM rates compared to established rates. Acceptable rate and severity of radiation-related toxicities were observed.

Machine Learning Applications in Head and Neck Radiation Oncology: Lessons From Open-Source Radiomics Challenges.

Radiomics leverages existing image datasets to provide non-visible data extraction via image post-processing, with the aim of identifying prognostic, and predictive imaging features at a sub-region of interest level. However, the application of radiomics is hampered by several challenges such as lack of image acquisition/analysis method standardization, impeding generalizability. As of yet, radiomics remains intriguing, but not clinically validated. We aimed to test the feasibility of a non-custom-constructed platform for disseminating existing large, standardized databases across institutions for promoting radiomics studies. Hence, University of Texas MD Anderson Cancer Center organized two public radiomics challenges in head and neck radiation oncology domain. This was done in conjunction with MICCAI 2016 satellite symposium using Kaggle-in-Class, a machine-learning and predictive analytics platform. We drew on clinical data matched to radiomics data derived from diagnostic contrast-enhanced computed tomography (CECT) images in a dataset of 315 patients with oropharyngeal cancer. Contestants were tasked to develop models for (i) classifying patients according to their human papillomavirus status, or (ii) predicting local tumor recurrence, following radiotherapy. Data were split into training, and test sets. Seventeen teams from various professional domains participated in one or both of the challenges. This review paper was based on the contestants' feedback; provided by 8 contestants only (47%). Six contestants (75%) incorporated extracted radiomics features into their predictive model building, either alone (n = 5; 62.5%), as was the case with the winner of the "HPV" challenge, or in conjunction with matched clinical attributes (n = 2; 25%). Only 23% of contestants, notably, including the winner of the "local recurrence" challenge, built their model relying solely on clinical data. In addition to the value of the integration of machine learning into clinical decision-making, our experience sheds light on challenges in sharing and directing existing datasets toward clinical applications of radiomics, including hyper-dimensionality of the clinical/imaging data attributes. Our experience may help guide researchers to create a framework for sharing and reuse of already published data that we believe will ultimately accelerate the pace of clinical applications of radiomics; both in challenge or clinical settings.

Patterns of locoregional failure following post-operative intensity-modulated radiotherapy to oral cavity cancer: quantitative spatial and dosimetric analysis using a deformable image registration workflow.

BACKGROUND: We sought to identify spatial/dosimetric patterns of failure for oral cavity cancer patients receiving post-operative IMRT (PO-IMRT). METHODS: Two hundred eighty-nine OCC patients receiving PO-IMRT were retrospectively reviewed from 2000 to 2012. Diagnostic CT documenting recurrence (rCT) was co-registered with planning CT (pCT) using a validated deformable image registration software. Manually segmented recurrent gross disease (rGTV) was deformed to co-registered pCTs. Mapped rGTVs were compared dosimetrically to planned dose and spatially to planning target volumes using centroid-based approaches. Failures types were classified using combined spatial/dosimetric criteria: A (central high-dose), B (peripheral high-dose), C (central intermediate/low-dose), D (peripheral intermediate/low-dose), and E (extraneous-dose). RESULTS: Fifty-four patients with recurrence were analyzed; 26 local recurrence, 19 regional recurrence, and 9 both local and regional recurrence. Median time to recurrence was 4 months (range 0-71). Median rGTVs volume was 3.7 cm3 (IQR 1.4-10.6). For spatial and dosimetric analysis of the patterns of failure, 30 patients (55.5%) were classified as type A (central high-dose). Non-central high dose failures were distributed as follows: 2 (3.7%) type B, 10 (18.5%) type C, 1 (1.8%) type D, and 9 (16.7%) type E. Non-IMRT failure in the matching low-neck field was seen in two patients. No failures were noted at the IMRT-supraclavicular field match-line. CONCLUSIONS: Approximately half of patients with local/regional failure had non-central high dose recurrence. Peripheral high dose misses were uncommon reflecting adequate delineation and dose delivery. Future strategies are needed to reduce types C and E failures.

Outcomes of oral cavity cancer patients treated with surgery followed by postoperative intensity modulated radiation therapy.

OBJECTIVES: Although treatment paradigms have not changed significantly, radiotherapy, surgery, and imaging techniques have improved, leading us to investigate oncologic and survival outcomes for oral cavity squamous cell cancer (OCSCC) patients treated with surgery followed by postoperative IMRT. MATERIAL AND METHODS: Records of patients with pathological diagnosis of OCSCC treated between 2000 and 2012 were retrospectively reviewed. Patients' demographic, disease, and treatment criteria were extracted. Kaplan-Meier method was used to calculate survival curves. RESULTS: Two hundred eighty-nine patients were analyzed. Median follow-up was 35months. Two hundred sixty-eight had neck dissections (93%), of which 66% had nodal involvement, and 51% of those positive dissections had extracapsular extension. Forty patients received induction chemotherapy and 107 received concurrent chemotherapy. Median dose to high risk clinical target volume was 60Gy/30 fractions. The 5-year locoregional control and overall survival rates were 76% and 57%, respectively. Tumors with >1.5cm depth of invasion had significantly higher risk of local failure compared with ≤1.5cm (p<0.001). In multivariate analysis, positive and no neck dissection (p=0.01), positive lymphovascular invasion (p=0.006) and >1.5cm depth of invasion (p=0.003) were independent predictors of poorer survival. CONCLUSIONS: Disease outcomes were consistent with historical data and did not appear compromised by the use of IMRT.

Prospective Qualitative and Quantitative Analysis of Real-Time Peer Review Quality Assurance Rounds Incorporating Direct Physical Examination for Head and Neck Cancer Radiation Therapy.

PURPOSE: Our department has a long-established comprehensive quality assurance (QA) planning clinic for patients undergoing radiation therapy (RT) for head and neck cancer. Our aim is to assess the impact of a real-time peer review QA process on the quantitative and qualitative radiation therapy plan changes in the era of intensity modulated RT (IMRT). METHODS AND MATERIALS: Prospective data for 85 patients undergoing head and neck IMRT who presented at a biweekly QA clinic after simulation and contouring were collected. A standard data collection form was used to document alterations made during this process. The original pre-QA clinical target volumes (CTVs) approved by the treating-attending physicians were saved before QA and compared with post-QA consensus CTVs. Qualitative assessment was done according to predefined criteria. Dice similarity coefficients (DSC) and other volume overlap metrics were calculated for each CTV level and were used for quantitative comparison. Changes are categorized as major, minor, and trivial according to the degree of overlap. Patterns of failure were analyzed and correlated to plan changes. RESULTS: All 85 patients were examined by at least 1 head and neck subspecialist radiation oncologist who was not the treating-attending physician; 80 (94%) were examined by ≥3 faculty members. New clinical findings on physical examination were found in 12 patients (14%) leading to major plan changes. Quantitative DSC analysis revealed significantly better agreement in CTV1 (0.94 ± 0.10) contours than in CTV2 (0.82 ± 0.25) and CTV3 (0.86 ± 0.2) contours (P=.0002 and P=.03, respectively; matched-pair Wilcoxon test). The experience of the treating-attending radiation oncologist significantly affected DSC values when all CTV levels were considered (P=.012; matched-pair Wilcoxon text). After a median follow-up time of 38 months, only 10 patients (12%) had local recurrence, regional recurrence, or both, mostly in central high-dose areas. CONCLUSIONS: Comprehensive peer review planning clinic is an essential component of IMRT QA that led to major changes in one-third of the study population. This process ensured safety related to target definition and led to favorable disease control profiles, with no identifiable recurrences attributable to geometric misses or delineation errors.

Toward a model-based patient selection strategy for proton therapy: External validation of photon-derived normal tissue complication probability models in a head and neck proton therapy cohort.

OBJECTIVE: To externally validate head and neck cancer (HNC) photon-derived normal tissue complication probability (NTCP) models in patients treated with proton beam therapy (PBT). METHODS: This prospective cohort consisted of HNC patients treated with PBT at a single institution. NTCP models were selected based on the availability of data for validation and evaluated by using the leave-one-out cross-validated area under the curve (AUC) for the receiver operating characteristics curve. RESULTS: 192 patients were included. The most prevalent tumor site was oropharynx (n=86, 45%), followed by sinonasal (n=28), nasopharyngeal (n=27) or parotid (n=27) tumors. Apart from the prediction of acute mucositis (reduction of AUC of 0.17), the models overall performed well. The validation (PBT) AUC and the published AUC were respectively 0.90 versus 0.88 for feeding tube 6months PBT; 0.70 versus 0.80 for physician-rated dysphagia 6months after PBT; 0.70 versus 0.68 for dry mouth 6months after PBT; and 0.73 versus 0.85 for hypothyroidism 12months after PBT. CONCLUSION: Although a drop in NTCP model performance was expected for PBT patients, the models showed robustness and remained valid. Further work is warranted, but these results support the validity of the model-based approach for selecting treatment for patients with HNC.

Radiomics in head and neck cancer: from exploration to application.

In the context of clinical oncology, a fundamental goal of radiomics is the extraction of large amounts of quantitative features whose subsequent analysis can be used for decision support towards personalized and actionable cancer care. Head and neck cancers present a unique set of diagnostic and therapeutic challenges by nature of its complex anatomy and heterogeneity. Radiomics holds the potential to address these barriers, but only if as a collective field we direct future effort towards investigating specific oncologic function and oncologic outcomes, with external validation and collaborative multi-institutional efforts to begin standardizing and refining radiomic signatures. Here we present an overview of radiomic texture analysis methods as well as the software infrastructure, review the developments of radiomics in head and neck cancer applications, discuss unmet challenges, and propose key recommendations for moving the field forward.