Search

• The miscalibration of pulmonary and esophageal toxicities in patients with lung cancer treated by (chemo)-radiotherapy is frequent. • Usual dose-volume histograms do not account for dose spatial distribution. • Radiomics based models contribute to a significant improvement in acute and late pulmonary toxicities prediction. For acute esophageal toxicity, dosimetric and radiomics based models achieve similar results. (Chemo)–radiotherapy (RT) is the gold standard treatment for patients with locally advanced lung cancer non accessible for surgery. However, current toxicity prediction models rely on clinical and dose volume histograms (DVHs) and remain unsufficient. The goal of this work is to investigate the added predictive value of the radiomics approach applied to dose maps regarding acute and late toxicities in both the lungs and esophagus. Acute and late toxicities scored using the CTCAE v4.0 were retrospectively collected on patients treated with RT in our institution. Radiomic features were extracted from 3D dose maps considering Gy values as grey-levels in images. DVH and usual clinical factors were also considered. Three toxicity prediction models (clinical only, clinical + DVH and combined , i.e., including clinical + DVH + radiomics) were incrementally trained using a neural network on 70% of the patients for prediction of grade ≥2 acute and late pulmonary toxicities (APT/LPT) and grade ≥2 acute esophageal toxicities (AET). After bootstrapping (n = 1000), optimal cut-off values were determined based on the Youden Index. The trained models were then evaluated in the remaining 30% of patients using balanced accuracy (BAcc). 167 patients were treated from 2015 to 2018: 78% non small-cell lung cancers, 14% small-cell lung cancers and 8% other histology with a median age at treatment of 66 years. Respectively, 22.2%, 16.8% and 30.0% experienced APT, LPT and AET. In the training set (n = 117), the corresponding BAcc for clinical only/clinical + DVH/combined were 0.68/0.79/0.92, 0.66/0.77/0.87 and 0.68/0.73/0.84. In the testing evaluation (n = 50), these trained models obtained a corresponding BAcc of 0.69/0.69/0.92, 0.76/0.80/0.89 and 0.58/0.73/0.72. In patients with a lung cancer treated with RT, radiomic features extracted from 3D dose maps seem to surpass usual models based on clinical factors and DVHs for the prediction of APT and LPT. [ABSTRACT FROM AUTHOR]

Poster: Physics track: Radiobiological and predictive modelling, and radiomics PO-1530: Pulmonary toxicity in lung cancer treated by (chemo)-radiotherapy: a radiomics-based NTCP. V. Bourbonne, F. Lucia, G. Dissaux, B. Julien, D. Visvikis, O. Pradier, M. Hatt, U. Schick, PO-1530: Pulmonary toxicity in lung cancer treated by (chemo)-radiotherapy: a radiomics-based NTCP. [Extracted from the article]

Poster: Physics track: Radiobiological and predictive modelling, and radiomics PO-1562: Radiomics applied to dose distributions to predict toxicity after radiotherapy in cervical cancer F. Lucia, V. Bourbonne, G. Dissaux, O. Miranda, G. Dissaux, U. Schick, D. Visvikis, J. Bert, M. Hatt, R. Abgral. [Extracted from the article]

Poster discussion: CL: Gynaecology PD-0658: Suboptimal dosimetric coverage of PET/CT hotspots is associated with recurrence for cervical cancer F. Lucia, D. Visvikis, V. Bourbonne, J. Bert, R. Abgral, G. Dissaux, O. Miranda, O. Pradier, M. Hatt, U. Schick. [Extracted from the article]

Poster: Clinical track: Prostate PO-1199: Rectal toxicity prostate cancer treated with Brachytherapy: a radiomics-machine learning based NTCP G. Dissaux, M. Ibrahim, F. Lucia, V. Bourbonne, N. Boussion, O. Pradier, D. Visvikis, A. Valeri, J. Bert, M. Hatt, U. Schick. [Extracted from the article]

Poster: Clinical track: Prostate PO-1198: Comparison between built custom linked seeds and loose seeds in prostate brachytherapy L. Ollivier, F. Lucia, C. Lucas, V. Bourbonne, J. Marolleau, N. Boussion, G. Goasduff, G. Fournier, G. Dissaux, O. Pradier, A. Valeri, U. Schick. [Extracted from the article]