PurposeThe purpose of this study is to investigate whether machine learning with dosiomic, radiomic, and demographic features allows for xerostomia risk assessment more precise than normal tissue complication probability (NTCP) models based on the mean radiation dose to parotid glands.Material and methodsA cohort of 153 head-and-neck cancer patients was used to model xerostomia at 0–6 months (early), 6–15 months (late), 15–24 months (long-term), and at any time (a longitudinal model) after radiotherapy. Predictive power of the features was evaluated by the area under the receiver operating characteristic curve (AUC) of univariate logistic regression models. The multivariate NTCP models were tuned and tested with single and nested cross-validation, respectively. We compared predictive performance of seven classification algorithms, six feature selection methods, and ten data cleaning/class balancing techniques using the Friedman test and the Nemenyi post hoc analysis.ResultsNTCP models based on the parotid mean dose failed to predict xerostomia (AUCs < 0.60). The most informative predictors were found for late and long-term xerostomia. Late xerostomia correlated with the contralateral dose gradient in the anterior–posterior (AUC = 0.72) and the right–left (AUC = 0.68) direction, whereas long-term xerostomia was associated with parotid volumes (AUCs > 0.85), dose gradients in the right–left (AUCs > 0.78), and the anterior–posterior (AUCs > 0.72) direction. Multivariate models of long-term xerostomia were typically based on the parotid volume, the parotid eccentricity, and the dose–volume histogram (DVH) spread with the generalization AUCs ranging from 0.74 to 0.88. On average, support vector machines and extra-trees were the top performing classifiers, whereas the algorithms based on logistic regression were the best choice for feature selection. We found no advantage in using data cleaning or class balancing methods.ConclusionWe demonstrated that incorporation of organ- and dose-shape descriptors is beneficial for xerostomia prediction in highly conformal radiotherapy treatments. Due to strong reliance on patient-specific, dose-independent factors, our results underscore the need for development of personalized data-driven risk profiles for NTCP models of xerostomia. The facilitated machine learning pipeline is described in detail and can serve as a valuable reference for future work in radiomic and dosiomic NTCP modeling.

No abstract

No abstract

1 J. Magn. Reson. Imaging 2017;45:889-898.

PurposeTo access the long-time outcome and patterns of failure in patients with advanced head and neck squamous cell carcinoma (HNSCC).Methods and materialsBetween 1992 and 2005 127 patients (median age 55 years, UICC stage III n = 6, stage IV n = 121) with primarily inoperable, advanced HNSCC were treated with definite platinum-based radiochemotherapy (median dose 66.4 Gy). Analysed end-points were overall survival (OS), disease-free survival (DFS), loco-regional progression-free survival (LPFS), development of distant metastases (DM), prognostic factors and causes of death.ResultsThe mean follow-up time was 34 months (range, 3-156 months), the 3-, 5- and 10-year OS rates were 39%, 28% and 14%, respectively. The median OS was 23 months. Forty-seven patients achieved a complete remission and 78 patients a partial remission. The median LPFS was 17 months, the 3-, 5- and 10-year LPFS rates were 41%, 33% and 30%, respectively. The LPFS was dependent on the nodal stage (p = 0.029). The median DFS was 11 months (range, 2-156 months), the 3-, 5- and 10-year DFS rates were 30%, 24% and 22%, respectively. Prognostic factors in univariate analyses were alcohol abuse (n = 102, p = 0.015), complete remission (n = 47, p < 0.001), local recurrence (n = 71, p < 0.001), development of DM (n = 45, p < 0.001; median OS 16 months) and borderline significance in nodal stage N2 versus N3 (p = 0.06). Median OS was 26 months with lung metastases (n = 17). Nodal stage was a predictive factor for the development of DM (p = 0.025). Cause of death was most commonly tumor progression.ConclusionsIn stage IV HNSCC long-term survival is rare and DM is a significant predictor for mortality. If patients developed DM, lung metastases had the most favourable prognosis, so intensified palliative treatment might be justified in DM limited to the lungs.

BackgroundGoal of this retrospective analysis was to evaluate the role of repeat whole brain radiotherapy in the palliative care of patients with brain metastases due to solid tumors.MethodsData regarding demographic data, primary tumor, metastasis, radiotherapy and symptoms were compiled on 134 patients with cerebral metastases that received repeat whole brain radiotherapy (WBRT) in our clinic between 2002 and 2011.ResultsThe analyzed group consisted of 63 (47%) women and 71 (53%) men with a median age of 57 at the start of re-irradiation. Most frequent primary site was the lung (87%).Sixty patients with lung cancer received the first WBRT prophylactically. At the time of re-WBRT 81% of all patients suffered from additional extracerebral metastases.Time between first and second WBRT was a median of 13.4 months. Full dose for the first WBRT was 30 Gy in 2.0 Gy single dose, for the second 20 Gy in 2.0 Gy single dose.At the start of the Re-WBRT 81 patients (60.4%) had mild, 32 (23.9%) severe neurological symptoms, 21 patients (15.7%) were asymptomatic. The median Karnofsky performance status was 70%. Overall, re-WBRT was tolerated satisfactorily. Main side effects were fatigue, erythema and focal alopecia, 10% of patients discontinued treatment before reaching the planned dose. Median survival was 2.8 months since the end of the re-WBRT with good performance status at the start of the re-irradiation being a key indicator for longer survival.Fifty-two patients (39%) showed a clinical improvement of neurological symptoms after the therapy, 59 patients (44%) remained stable, 23 patients (17%) showed worse symptoms.ConclusionsFrom this large patient collective we were able to show that re-WBRT can be an important therapeutic option with low rate of acute side effects for patients in adequate general condition.

No significant difference between carbon ions and protons in the therapy of skull base chondrosarcoma could be detected in these initial retrospective results. Cancer 2018;124:2036-44. © 2018 American Cancer Society.

Reirradiation with intensity-modulated radiotherapy in recurrent head and neck cancer is feasible with acceptable toxicity and yields encouraging rates of local control and overall survival.