Multimodal Data Integration to Predict Severe Acute Oral Mucositis of Nasopharyngeal Carcinoma Patients Following Radiation Therapy

Abstract

(1) Background: Acute oral mucositis is the most common side effect for nasopharyngeal carcinoma patients receiving radiotherapy. Improper or delayed intervention to severe AOM could degrade the quality of life or survival for NPC patients. An effective prediction method for severe AOM is needed for the individualized management of NPC patients in the era of personalized medicine. (2) Methods: A total of 242 biopsy-proven NPC patients were retrospectively recruited in this study. Radiomics features were extracted from contrast-enhanced CT (CECT), contrast-enhanced T1-weighted (cT1WI), and T2-weighted (T2WI) images in the primary tumor and tumor-related area. Dosiomics features were extracted from 2D or 3D dose-volume histograms (DVH). Multiple models were established with single and integrated data. The dataset was randomized into training and test sets at a ratio of 7:3 with 10-fold cross-validation. (3) Results: The best-performing model using Gaussian Naive Bayes (GNB) (mean validation AUC = 0.81 ± 0.10) was established with integrated radiomics and dosiomics data. The GNB radiomics and dosiomics models yielded mean validation AUC of 0.6 ± 0.20 and 0.69 ± 0.14, respectively. (4) Conclusions: Integrating radiomics and dosiomics data from the primary tumor area could generate the best-performing model for severe AOM prediction.

Predictive Factors for Chemoradiation-Induced Oral Mucositis and Dysphagia in Head and Neck Cancer: A Scoping Review

“…The remaining ten studies that reported genetic factors each returned a different factor. Beyond the previously mentioned categories, one study crafted a prediction model employing radiomics and dosiomics features derived from the primary tumor volume [23]. Two studies highlighted a significant correction between bioelectrical impedance measurements and OM in univariate analyses [34,72], while another study identified perfusion parameters as a significant determinant in a univariate analysis [69].…”

“…All the validated predictive models for OM were specific to the acute period. Severeor-higher OM (≥grade 3), as scored by RTOG, CTCAE, or WHO, was used as an outcome by 6 out of 8 models [17][18][19][20][21]23]. Alternatively, an increase in RTOG grade from mild (grade 1) to moderate (grade 2) was used by Zhu et al [16] and DAHANCA grade 3+ was used by Hansen et al [22].…”

“…Zhu et al, reported a model using genetic information from oral bacteria [16]. Dong et al, reported a model using MR and CECT radiomic features extracted from the gross tumor volume [23]. The model size ranged between 2 and 19 features.…”

“…Predictive models mostly focused on severe toxicity, indicating their intended use for identifying high-risk patients who can be targeted for closer monitoring and more aggressive preventative measures. For acute OM, eight models were identified with AUCs ranging from 0.65 to 0.81 [16][17][18][19][20][21][22][23]. Five models emerged for acute dysphagia (AUC: 0.60-0.82) [93][94][95][96][97], and nine for late dysphagia (AUC: 0.70-0.85) [98][99][100][101][102][103][104][105][106].…”

“…The best performing models tended to incorporate multiple predictor types. For example, for OM, Hansen et al included treatment acceleration alongside DVH parameters of the extended oral cavity [22], and Dong et al, included texture features from multiple imaging modalities [23]. For dysphagia, Dean et al, included patient, tumor, treatment, and DVH parameters of the pharyngeal mucosa [97].…”