Dataset Information

Normal tissue complication probability (NTCP) modelling using spatial dose metrics and machine learning methods for severe acute oral mucositis resulting from head and neck radiotherapy.

ABSTRACT: Severe acute mucositis commonly results from head and neck (chemo)radiotherapy. A predictive model of mucositis could guide clinical decision-making and inform treatment planning. We aimed to generate such a model using spatial dose metrics and machine learning.Predictive models of severe acute mucositis were generated using radiotherapy dose (dose-volume and spatial dose metrics) and clinical data. Penalised logistic regression, support vector classification and random forest classification (RFC) models were generated and compared. Internal validation was performed (with 100-iteration cross-validation), using multiple metrics, including area under the receiver operating characteristic curve (AUC) and calibration slope, to assess performance. Associations between covariates and severe mucositis were explored using the models.The dose-volume-based models (standard) performed equally to those incorporating spatial information. Discrimination was similar between models, but the RFCstandard had the best calibration. The mean AUC and calibration slope for this model were 0.71 (s.d.=0.09) and 3.9 (s.d.=2.2), respectively. The volumes of oral cavity receiving intermediate and high doses were associated with severe mucositis.The RFCstandard model performance is modest-to-good, but should be improved, and requires external validation. Reducing the volumes of oral cavity receiving intermediate and high doses may reduce mucositis incidence.

SUBMITTER: Dean JA

PROVIDER: S-EPMC5021201 | biostudies-literature | 2016 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Normal tissue complication probability (NTCP) modelling using spatial dose metrics and machine learning methods for severe acute oral mucositis resulting from head and neck radiotherapy.

Dean Jamie A JA Wong Kee H KH Welsh Liam C LC Jones Ann-Britt AB Schick Ulrike U Newbold Kate L KL Bhide Shreerang A SA Harrington Kevin J KJ Nutting Christopher M CM Gulliford Sarah L SL

Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology 20160527 1

<h4>Background and purpose</h4>Severe acute mucositis commonly results from head and neck (chemo)radiotherapy. A predictive model of mucositis could guide clinical decision-making and inform treatment planning. We aimed to generate such a model using spatial dose metrics and machine learning.<h4>Materials and methods</h4>Predictive models of severe acute mucositis were generated using radiotherapy dose (dose-volume and spatial dose metrics) and clinical data. Penalised logistic regression, suppo ...[more]

PMID: 27240717

Similar Datasets

Project description:PURPOSE:Current normal tissue complication probability modeling using logistic regression suffers from bias and high uncertainty in the presence of highly correlated radiation therapy (RT) dose data. This hinders robust estimates of dose-response associations and, hence, optimal normal tissue-sparing strategies from being elucidated. Using functional data analysis (FDA) to reduce the dimensionality of the dose data could overcome this limitation. METHODS AND MATERIALS:FDA was applied to modeling of severe acute mucositis and dysphagia resulting from head and neck RT. Functional partial least squares regression (FPLS) and functional principal component analysis were used for dimensionality reduction of the dose-volume histogram data. The reduced dose data were input into functional logistic regression models (functional partial least squares-logistic regression [FPLS-LR] and functional principal component-logistic regression [FPC-LR]) along with clinical data. This approach was compared with penalized logistic regression (PLR) in terms of predictive performance and the significance of treatment covariate-response associations, assessed using bootstrapping. RESULTS:The area under the receiver operating characteristic curve for the PLR, FPC-LR, and FPLS-LR models was 0.65, 0.69, and 0.67, respectively, for mucositis (internal validation) and 0.81, 0.83, and 0.83, respectively, for dysphagia (external validation). The calibration slopes/intercepts for the PLR, FPC-LR, and FPLS-LR models were 1.6/-0.67, 0.45/0.47, and 0.40/0.49, respectively, for mucositis (internal validation) and 2.5/-0.96, 0.79/-0.04, and 0.79/0.00, respectively, for dysphagia (external validation). The bootstrapped odds ratios indicated significant associations between RT dose and severe toxicity in the mucositis and dysphagia FDA models. Cisplatin was significantly associated with severe dysphagia in the FDA models. None of the covariates was significantly associated with severe toxicity in the PLR models. Dose levels greater than approximately 1.0 Gy/fraction were most strongly associated with severe acute mucositis and dysphagia in the FDA models. CONCLUSIONS:FPLS and functional principal component analysis marginally improved predictive performance compared with PLR and provided robust dose-response associations. FDA is recommended for use in normal tissue complication probability modeling.

Project description:PurposeMultiparametric magnetic resonance tomography (mpMRI) and prostate specific membrane antigen positron emission tomography (PSMA-PET/CT) are used to guide focal radiotherapy (RT) dose escalation concepts. Besides improvements of treatment effectiveness, maintenance of a good quality of life is essential. Therefore, this planning study investigates whether urethral sparing in moderately hypofractionated RT with focal RT dose escalation influences tumour control probability (TCP) and normal tissue complication probability (NTCP).Patients and methods10 patients with primary prostate cancer (PCa), who underwent 68Ga PSMA-PET/CT and mpMRI followed by radical prostatectomy were enrolled. Intraprostatic tumour volumes (gross tumor volume, GTV) based on both imaging techniques (GTV-MRI and -PET) were contoured manually using validated contouring techniques and GTV-Union was created by summing both. For each patient three IMRT plans were generated with 60 Gy to the whole prostate and a simultaneous integrated boost up to 70 Gy to GTV-Union in 20 fractions by (Plan 1) not respecting and (Plan 2) respecting dose constraints for urethra as well as (Plan 3) respecting dose constraints for planning organ at risk volume for urethra (PRV = urethra + 2mm expansion). NTCP for urethra was calculated applying a Lyman-Kutcher-Burman model. TCP-Histo was calculated based on PCa distribution in co-registered histology (GTV-Histo). Complication free tumour control probability (P+) was calculated. Furthermore, the intrafractional movement was considered.ResultsMedian overlap of GTV-Union and PRV-Urethra was 1.6% (IQR 0-7%). Median minimum distance of GTV-Histo to urethra was 3.6 mm (IQR 2 - 7 mm) and of GTV-Union to urethra was 1.8 mm (IQR 0.0 - 5.0 mm). The respective prescription doses and dose constraints were reached in all plans. Urethra-sparing in Plans 2 and 3 reached significantly lower NTCP-Urethra (p = 0.002) without significantly affecting TCP-GTV-Histo (p = p > 0.28), NTCP-Bladder (p > 0.85) or NTCP-Rectum (p = 0.85), resulting in better P+ (p = 0.006). Simulation of intrafractional movement yielded even higher P+ values for Plans 2 and 3 compared to Plan 1.ConclusionUrethral sparing may increase the therapeutic ratio and should be implemented in focal RT dose escalation concepts.

Project description:Background The aim of the study was to evaluate the changes in ?-H2AX expression in peripheral blood lymphocytes (PBL) according to severity of radiation-induced mucositis. Patients and method Fifty patients with head and neck cancer treated with radiotherapy (RT) or chemoradiation were included in the study. Blood samples were collected before treatment to measure baseline ?-H2AX levels. Second sample was taken 45 minutes after the first RT fraction and then once a week, 45 min after irradiation. In patients treated with chemoradiation the blood sample was taken the day after chemotherapy. Mucositis was evaluated once a week and reported according to CTCAE v4 and RTOG/EORTC scales. PBL were analyzed with flow cytometry and level of H2AX phosphorylation at every time point was evaluated. Results In 35 patients mild to moderate (grade 1-2) mucositis was observed and 15 patients developed severe (grade 3) mucositis. No cases of grade 4 mucositis were observed. The difference in baseline levels of ?-H2AX between groups with mild and severe mucositis was statistically insignificant (p = 0.25). The statistically significant difference in ?-H2AX level was observed in week 7 of treatment (p = 0.01). No significant differences in ?-H2AX level were found neither between group treated with concomitant chemoradiation or RT alone neither between groups with and without common comorbidities. In the analysis of the kinetics of ?-H2AX during treatment, a statistically significant difference (p = 0.0088) between groups with mild and severe mucositis was observed. After fourth week of treatment levels of ?-H2AX decreased significantly in the group with severe mucositis and increased in patients with mild side effects. The observed difference was not caused by the decrease in peripheral lymphocyte count, which was similar in both groups. Conclusions Presented results indicate that severity of radiation-induced mucositis does not correlate directly with ?-H2AX levels measured in vivo in PBL. Prediction of mucositis grade based on ?-H2AX level is not yet possible, either before treatment or early during treatment, but preliminary results, indicating significant differences in ?-H2AX kinetics between groups, encourage further studies.

Dataset Information

Normal tissue complication probability (NTCP) modelling using spatial dose metrics and machine learning methods for severe acute oral mucositis resulting from head and neck radiotherapy.

Publications

Normal tissue complication probability (NTCP) modelling using spatial dose metrics and machine learning methods for severe acute oral mucositis resulting from head and neck radiotherapy.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure