A joint physics and radiobiology DREAM team vision - Towards better response prediction models to advance radiotherapy

Radiotherapy developed empirically through experience balancing tumour control and normal tissue toxicities. Early simple mathematical models formalized this practical knowledge and enabled effective cancer treatment to date. Remarkable advances in technology, computing, and experimental biology now create opportunities to incorporate this knowledge into enhanced computational models. The ESTRO DREAM (Dose Response, Experiment, Analysis, Modelling) workshop brought together experts across disciplines to pursue the vision of personalized radiotherapy for optimal outcomes through advanced modelling. The ultimate vision is leveraging quantitative models dynamically during therapy to ultimately achieve truly adaptive and biologically guided radiotherapy at the population as well as individual patient-based levels. This requires the generation of models that inform response-based adaptations, individually optimized delivery and enable biological monitoring to provide decision support to clinicians. The goal is expanding to models that can drive the realization of personalized therapy for optimal outcomes. This position paper provides their propositions that describe how innovations in biology, physics, mathematics, and data science including AI could inform models and improve predictions. It consolidates the DREAM team's consensus on scientific priorities and organizational requirements. Scientifically, it stresses the need for rigorous, multifaceted model development, comprehensive validation and clinical applicability and significance. Organizationally, it reinforces the prerequisites of interdisciplinary research and collaboration between physicians, medical physicists, radiobiologists, and computational scientists throughout model development. Solely by a shared understanding of clinical needs, biological mechanisms, and computational methods, more informed models can be created. Future research environment and support must facilitate this integrative method of operation across multiple disciplines.

Interpretability of a Deep Learning-Based Prediction Model for Mandibular Osteoradionecrosis

PURPOSE/OBJECTIVE(S)

The development of radiation-induced toxicities is a multifactorial process. Existing DVH-based prediction models use traditional multivariate analysis to combine all the potential risk factors. Recently, deep learning (DL) has been proposed for predicting mandibular osteoradionecrosis (ORN) directly from 3D dose distribution maps [1]. However, with this approach, incorporating non-imaging data such as potential risk factors presents challenges. We investigate the use of DL-based multimodality fusion for the purpose of radiation-induced ORN toxicity prediction.

MATERIALS/METHODS

This study explores early and late fusion strategies for combining 3D radiation dose distribution maps and clinical and demographic variables in the prediction of mandibular ORN incidence in head and neck cancer patients treated with radiotherapy. The results are compared to single-modality predictions with a random forest (RF) trained only on clinical variables and a 3D DenseNet40 trained on dose maps alone. We investigate two different fusion approaches. In the first, the image features extracted from the radiation dose maps using a 3D DenseNet40 were concatenated with the clinical variables into one single vector using a type II early fusion strategy. The combined feature vector was input into a fully connected layer for classification of ORN vs. controls. A final softmax activation layer was added to obtain the class predicted probabilities. The second approach used a late fusion strategy, in which the outputs from the 3D DenseNet40 and the RF model were combined by averaging the predicted classification probabilities for each of the two classes (ORN and no ORN) to obtain the final class decision on a case-by-case basis.

RESULTS

The AUROC values obtained for the late and early fusion models and the single-modality 3D DenseNet40 and RF models were 0.70, 0.68, 0.69 and 0.60, respectively. The highest AUC ROC was observed with the late fusion approach, which was statistically significantly different to that of the RF single-modality model with a significance level of 0.05. However, after Bonferroni correction (Altman 1999) for multiple comparisons was applied, resulting in a corrected significance level of 0.05/6 = 0.008 for each comparison, no statistically significant difference was observed between any of the models' AUROC values. This is most likely due to the lack of discriminative contribution observed from clinical variables, which on their own resulted in a poorly predictive RF model.

CONCLUSION

To our knowledge, no previous work has been published on the use of multimodal fusion DL methods to combine dose distribution maps and clinical variables in the prediction of mandibular ORN. Although non-conclusive results were obtained, this study demonstrates the potential of DL in the prediction of the multifactorial side effects resulting from radiotherapy treatments. [1] Humbert-Vidan L et al. Prediction of Mandibular ORN Incidence from 3D Radiation Dose Distribution Maps Using DL (2022).

Dynamic nomogram for long-term survival in patients with locally advanced oropharyngeal cancer after (chemo)radiotherapy

PURPOSE

This study aimed to establish a nomogram for predicting overall survival (OS) in oropharyngeal cancer patients treated with curative (chemo)radiotherapy.

MATERIALS AND METHODS

The dynamic nomogram was constructed on 273 patients with oropharyngeal squamous cell carcinoma treated in a Tertiary Head and Neck Cancer Unit. The clinical features that were previously reported to be associated with OS were analyzed. The performance of the nomogram was assessed using concordance index (C-index) and calibration curves.

RESULTS

The nomogram incorporated three explanatory variables derived from a decision tree approach including HPV status, N classification according to 8th edition TNM and early response to (chemo)radiotherapy. The nomogram was capable to predict OS with a validation C-index of 0.768. The proposed stratification in risk groups allowed significant distinction between Kaplan-Meier curves for OS outcome (p < 0.0001).

CONCLUSIONS

The nomogram provided an accurate evaluation of OS for oropharyngeal cancer patients treated with curative (chemo)radiotherapy.