Explainable Boosting Machine Model with a Parallel Ensemble Design Predicts Local Failure for Head and Neck Cancer With Clinical, CT, and Delta CBCT Radiomic Features

Objectives To identify and validate baseline magnetic resonance imaging (b-MRI) radiomic features (RFs) as predictors of disease outcomes in effectively cured head and neck squamous cell carcinoma (HNSCC) patients. Materials and methods Training set (TS) and validation set (VS) were retrieved from preexisting datasets (HETeCo and BD2Decide trials, respectively). Only patients with both pre- and post-contrast enhancement T1 and T2-weighted b-MRI and at least 2 years of follow-up (FUP) were selected. The combination of the best extracted RFs was used to classify low risk (LR) vs. high risk (HR) of disease recurrence. Sensitivity, specificity, and area under the curve (AUC) of the radiomic model were computed on both TS and VS. Overall survival (OS) and 5-year disease-free survival (DFS) Kaplan–Meier (KM) curves were compared for LR vs. HR. The radiomic-based risk class was used in a multivariate Cox model, including well-established clinical prognostic factors (TNM, sub-site, human papillomavirus [HPV]). Results In total, 57 patients of TS and 137 of VS were included. Three RFs were selected for the signature. Sensitivity of recurrence risk classifier was 0.82 and 0.77, specificity 0.78 and 0.81, AUC 0.83 and 0.78 for TS and VS, respectively. VS KM curves for LR vs. HR groups significantly differed both for 5-year DFS (p<.0001) and OS (p=.0004). A combined model of RFs plus TNM improved prognostic performance as compared to TNM alone, both for VS 5-year DFS (C-index: 0.76 vs. 0.60) and OS (C-index: 0.74 vs. 0.64). Conclusions Radiomics of b-MRI can help to predict recurrence and survival outcomes in HNSCC.

To determine whether radiomics features derived from diffusion-weighted imaging (DWI) and arterial spin labeling (ASL) can improve the differentiation between radiation-induced brain injury (RIBI) and tumor recurrence (TR) in glioma patients. A total of 4199 radiomics features were extracted from conventional MRI, apparent diffusion coefficient (ADC), and cerebral blood flow (CBF) maps, obtained from 96 pathologically confirmed WHO grade 2~4 gliomas with enhancement after standard treatment. The intraclass correlation coefficient (ICC) was used to test segmentation stability between two doctors. Radiomics features were selected using the Mann-Whitney U test, LASSO regression, and RFE algorithms. Four machine learning classifiers were adopted to establish radiomics models. The diagnostic performance of multiparameter, conventional, and single-parameter MRI radiomics models was compared using the area under the curve (AUC). The models were evaluated in the subsequent independent validation set (n = 30). Eight important radiomics features (3 from conventional MRI, 1 from ADC, and 4 from CBF) were selected. Support vector machine (SVM) was chosen as the optimal classifier. The diagnostic performance of the multiparameter MRI radiomics model (AUC 0.96) was higher than that of the conventional MRI (AUC 0.88), ADC (AUC 0.91), and CBF (AUC 0.95) radiomics models. For subgroup analysis, the multiparameter MRI radiomics model showed similar performance, with AUCs of 0.98 in WHO grade 2~3 and 0.96 in WHO grade 4. The incorporation of noninvasive DWI and ASL into the MRI radiomics model improved the diagnostic performance in differentiating RIBI from TR; ASL, especially, played a significant role. • The multiparameter MRI radiomics model was superior to the conventional MRI radiomics model in differentiating glioma recurrence from radiation-induced brain injury. • Diffusion and perfusion MRI could improve the ability of the radiomics model in predicting the progression in patients with glioma. • Arterial spin labeling played an important role in predicting glioma progression using radiomics models.

6046 Background: SOC immunotherapy (IO) for head and neck squamous cell carcinoma (HNSCC) has limited efficacy with inadequate biomarkers (BMs), necessitating improved strategies. Routine radiology and pathology scans provide underutilized tumor data that can address this need. BMs built from these scans can be integrated - along with existing BMs and clinical data - into multimodal predictors for improved stratification by clinical benefit. We evaluated (1) separate BMs using radiomics, pathomics, or clinical data, (2) their cross-modality correlations, and (3) a fused multimodal predictor of overall survival (OS). Methods: 100 HNSCC patients (96% male; mean age: 55 yrs; mean BMI: 22.6) treated primarily with pembrolizumab or nivolumab ± chemotherapy in 1L setting were analyzed. Radiomic and pathomic features were extracted via the Picture Health Px Platform. Tumors and adjacent vessels were segmented on pre- and on-treatment CT. Radiomic features (shape, texture, quantitative vessel tortuosity) were extracted, and longitudinal changes were distilled into feature clusters. Pathomics features of tumor and immune cell nuclei morphology and spatial interactions were extracted from baseline H&amp;E whole-slide images. Clinical variables included: PD-L1 CPS, local/regional/distant recurrence, M stage, oral/non-oral cavity site, P16 status, and BMI. Uni- and multimodal models were trained and evaluated by cross-validation to predict OS. Results: Radiomics and pathomics models outperformed the clinical model, P16 (HR=0.9, p=0.82) and PD-L1 status (HR=0.7, p=0.08) alone. A fused multimodal model of 6 radiomic clusters, 10 pathomic features, and 6 clinical variables achieved strongest OS stratification (Table 1). High pathomic tumor-immune cell interaction - indicating immune activation - was associated with PD-L1 (p=0.020) and P16 status (p&lt;1e-5), and tied to on-treatment decreases in radiomic wavelet-entropy features of heterogeneity (r=-0.30, p=0.044) and increases in radiomic structural homogeneity (r=0.27, p=0.024). Despite these correlations, each modality contributed independently to the multimodal prediction (R2&lt;0.03), emphasizing complementarity. Conclusions: A multimodal BM integrating radiology and pathology to capture tumor properties (heterogeneity, angiogenesis, immune infiltration, nuclei morphology) refined understanding of tumor behavior and IO outcome. Findings suggest an interpretable multimodal BM may better identify high-risk patients who would benefit from alternative therapies, enabling more personalized and effective HNSCC management. Model summaries by modality. Modality N Low risk N High risk N High Risk HR (p-value) Median Shortened OS (yrs) Clinical 100 47 53 1.8 (0.049) 1.09 Pathology 85 62 23 2.4 (0.012) 3.04 Radiology 68 54 14 3.2 (0.0004) 3.06 Multimodal 57 33 24 6.3 (0.0001) 4.20

ObjectiveThis study aims to create a reliable framework for grading esophageal cancer. The framework combines feature extraction, deep learning with attention mechanisms, and radiomics to ensure accuracy, interpretability, and practical use in tumor analysis.Materials and methodsThis retrospective study used data from 2,560 esophageal cancer patients across multiple clinical centers, collected from 2018 to 2023. The dataset included CT scan images and clinical information, representing a variety of cancer grades and types. Standardized CT imaging protocols were followed, and experienced radiologists manually segmented the tumor regions. Only high-quality data were used in the study. A total of 215 radiomic features were extracted using the SERA platform. The study used two deep learning models—DenseNet121 and EfficientNet-B0—enhanced with attention mechanisms to improve accuracy. A combined classification approach used both radiomic and deep learning features, and machine learning models like Random Forest, XGBoost, and CatBoost were applied. These models were validated with strict training and testing procedures to ensure effective cancer grading.ResultsThis study analyzed the reliability and performance of radiomic and deep learning features for grading esophageal cancer. Radiomic features were classified into four reliability levels based on their ICC (Intraclass Correlation) values. Most of the features had excellent (ICC > 0.90) or good (0.75 < ICC ≤ 0.90) reliability. Deep learning features extracted from DenseNet121 and EfficientNet-B0 were also categorized, and some of them showed poor reliability. The machine learning models, including XGBoost and CatBoost, were tested for their ability to grade cancer. XGBoost with Recursive Feature Elimination (RFE) gave the best results for radiomic features, with an AUC (Area Under the Curve) of 91.36%. For deep learning features, XGBoost with Principal Component Analysis (PCA) gave the best results using DenseNet121, while CatBoost with RFE performed best with EfficientNet-B0, achieving an AUC of 94.20%. Combining radiomic and deep features led to significant improvements, with XGBoost achieving the highest AUC of 96.70%, accuracy of 96.71%, and sensitivity of 95.44%. The combination of both DenseNet121 and EfficientNet-B0 models in ensemble models achieved the best overall performance, with an AUC of 95.14% and accuracy of 94.88%.ConclusionsThis study improves esophageal cancer grading by combining radiomics and deep learning. It enhances diagnostic accuracy, reproducibility, and interpretability, while also helping in personalized treatment planning through better tumor characterization.Clinical trial numberNot applicable.

Background: Radiomics is a promising methodology for quantitative analysis and description of radiological images using advanced mathematics and statistics. Tumor delineation, which is still often done manually, is an essential step in radiomics, however, inter-observer variability is a well-known uncertainty in radiation oncology. This study investigated the impact of inter-observer variability (IOV) in manual tumor delineation on the reliability of radiomic features (RF).Methods: Three different tumor types (head and neck squamous cell carcinoma (HNSCC), malignant pleural mesothelioma (MPM) and non-small cell lung cancer (NSCLC)) were included. For each site, eleven individual tumors were contoured on CT scans by three experienced radiation oncologists. Dice coefficients (DC) were calculated for quantification of delineation variability. RF were calculated with an in-house developed software implementation, which comprises 1404 features: shape (n = 18), histogram (n = 17), texture (n = 137) and wavelet (n = 1232). The IOV of RF was studied using the intraclass correlation coefficient (ICC). An ICC >0.8 indicates a good reproducibility. For the stable RF, an average linkage hierarchical clustering was performed to identify classes of uncorrelated features.Results: Median DC was high for NSCLC (0.86, range 0.57–0.90) and HNSCC (0.72, 0.21–0.89), whereas it was low for MPM (0.26, 0–0.9) indicating substantial IOV. Stability rate of RF correlated with DC and depended on tumor site, showing a high stability in NSCLC (90% of total parameters), acceptable stability in HNSCC (59% of total parameters) and low stability in MPM (36% of total parameters). Shape features showed the weakest stability across all tumor types. Hierarchical clustering revealed 14 groups of correlated and stable features for NSCLC and 6 groups for both HNSCC and MPM.Conclusion: Inter-observer delineation variability has a relevant influence on radiomics analysis and is strongly influenced by tumor type. This leads to a reduced number of suitable imaging features.

To explore the clinical value of contrast-enhanced computed tomography (CECT) radiomics in predicting primary site response to neoadjuvant chemotherapy in high-risk neuroblastoma. Seventy patients were retrospectively included and separated into very good partial response (VGPR) group and non-VGPR group according to the changes in primary tumor volume. The clinical features with statistical difference between the two groups were used to construct the clinical models using a logistic regression (LR) algorithm. The radiomics models based on different radiomics features selected by Kruskal-Wallis (KW) test and recursive feature elimination (RFE) were established using support vector machine (SVM) and LR algorithms. The radiomics score (Radscore) and clinical features were integrated into the combined models. Leave-one-out cross-validation (LOOCV) was used to validate the predictive performance of models in the entire dataset. The optimal clinical model achieved an area under the curve (AUC) of 0.767 [95% confidence interval (CI): 0.638, 0.896] and an accuracy of 0.771 after LOOCV. The AUCs of the best KW + SVM, KW + LR, RFE + SVM, and RFE + LR radiomics models were 0.816, 0.826, 0.853, and 0.850, respectively, and the corresponding AUCs after LOOCV were 0.780, 0.785, 0.755, and 0.772, respectively. The AUC and accuracy after LOOCV of the optimal combined model was 0.804 (95% CI: 0.694, 0.915) and 0.814, respectively. The Delong test showed a statistical difference in predictive performance between the optimal clinical and combined models after LOOCV (Z = 2.003, P = 0.045). The decision curve analysis showed that the combined model performs better than the clinical model. The CECT radiomics models have a favorable predictive performance in predicting VGPR of high-risk neuroblastoma to neoadjuvant chemotherapy. When integrating radiomics features and clinical features, the predictive performance of the combined models can be further improved.

PurposeThis study aimed to construct a risk prediction model based on radiomics, circulating tumor cells (CTCs), and dual-center clinical data to predict the invasiveness of lung adenocarcinoma, specifically for discriminating between minimally invasive adenocarcinoma (MIA) and invasive adenocarcinoma (IAC). The clinical value of this model in the precise diagnosis of early-stage lung adenocarcinoma was investigated to provide a reference for formulating reasonable treatment plans.Patients and MethodsClinical data, imaging data, CTCs, and pathological information from 202 patients with lung adenocarcinoma were retrospectively collected and analyzed from two medical centers between May 2022 and July 2023. The 146 cases from medical center 1 were randomly divided into a development set and an internal test set at a 3:2 ratio. The 56 cases from medical center 2 served as an external validation set. Machine learning was employed to analyze preoperative CTC counts and CT radiomic features. A feature selection method based on LASSO regression (with λ determined by the minimum criterion) was used to screen out 12 radiomic features. These features were subsequently incorporated into logistic regression to construct three prediction models: (1) a radiomics model based on radiomic features; (2) a CTCs-clinical data model based on the total development set; and (3) a composite clinical data-radiomics-CTCs model integrating the former two. The optimal model was selected to construct a nomogram. Its goodness-of-fit was assessed using a calibration curve (Hosmer-Lemeshow goodness-of-fit test), and its predictive performance was validated in the external validation set.ResultsA total of 107 radiomic features were extracted and categorized into 7 groups: 18 (16.8%) first-order features, 24 (22.4%) gray-level co-occurrence matrix (GLCM) features, 14 (13.1%) gray-level dependence matrix (GLDM) features, 16 (15.0%) each for gray-level run length matrix (GLRLM) and gray-level size zone matrix (GLSZM) features, 5 (4.7%) neighboring gray-tone difference matrix (NGTDM) features, and the remaining (13.1%) were shape-based features. In the total development set, significant differences were observed in clinical-imaging semantic features including CEA, CK19, CTC count, and lesion diameter, which were used to construct the clinical model. The area under the curve (AUC) for the radiomics model was 0.896 95% CI:0.832–0.960. The CTCs-clinical model demonstrated superior performance AUC:0.960, 95% CI:0.926–0.994. The composite clinical-radiomics-CTCs model showed the highest predictive accuracy AUC:0.980, 95% CI:0.960–1.000. According to decision curve analysis and the Akaike information criterion, the composite clinical-radiomics-CTCs model outperformed any single clinical or radiomic feature in terms of clinical predictive capability.ConclusionFor assessing the invasiveness of early-stage lung adenocarcinoma, the radiomics approach can effectively discriminate between MIA and IAC. However, compared to single-modality methods, the composite clinical-radiomics-CTCs model offers a novel auxiliary diagnostic method for evaluating the risk of invasiveness in early-stage lung cancer.

To develop a model for differentiating the predominant subtype-based prognostic groups of lung adenocarcinoma using CT radiomic features, and to validate its performance in comparison with radiologists' assessments. A total of 993 patients presenting with invasive lung adenocarcinoma between March 2010 and June 2016 were identified. Predominant histologic subtypes were categorized into three groups according to their prognosis (group 0: lepidic; group 1: acinar/papillary; group 2: solid/micropapillary). Seven hundred eighteen radiomic features were extracted from segmented lung cancers on contrast-enhanced CT. A model-development set was formed from the images of 893 patients, while 100 image sets were reserved for testing. A least absolute shrinkage and selection operator method was used for feature selection. Performance of the radiomic model was evaluated using receiver operating characteristic curve analysis, and accuracy on the test set was compared with that of three radiologists with varying experiences (6, 7, and 19years in chest CT). Our model differentiated the three groups with areas under the curve (AUCs) of 0.892 and 0.895 on the development and test sets, respectively. In pairwise discrimination, the AUC was highest for group 0 vs. 2 (0.984). The accuracy of the model on the test set was higher than the averaged accuracy of the three radiologists without statistical significance (73.0% vs. 61.7%, p = 0.059). For group 2, the model achieved higher PPV than the observers (85.7% vs. 35.0-48.4%). Predominant subtype-based prognostic groups of lung adenocarcinoma were classified by a CT-based radiomic model with comparable performance to radiologists. • A CT-based radiomic model differentiated three prognosis-based subtype groups of lung adenocarcinoma with areas under the curve (AUCs) of 0.892 and 0.895 on development and test sets, respectively. • The CT-based radiomic model showed near perfect discrimination between group 0 and group 2 (AUCs, 0.984-1.000). • The accuracy of the CT-based radiomic model was comparable to the averaged accuracy of the three radiologists with 6, 7, and 19years of clinical experience in chest CT (73.0% vs. 61.7%, p = 0.059), achieving a higher positive predictive value for group 2 than the observers (85.7% vs. 35.0-48.4%).

PurposeThe aim of this study was to evaluate the value of different multiparametric MRI-based radiomics models in differentiating stage IA endometrial cancer (EC) from benign endometrial lesions.MethodsThe data of patients with endometrial lesions from two centers were collected. The radiomics features were extracted from T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC) map, and late contrast-enhanced T1-weighted imaging (LCE-T1WI). After data dimension reduction and feature selection, nine machine learning algorithms were conducted to determine which was the optimal radiomics model for differential diagnosis. The univariate analyses and logistic regression (LR) were performed to reduce valueless clinical parameters and to develop the clinical model. A nomogram using the radscores combined with clinical parameters was developed. Two integrated models were obtained respectively by the ensemble strategy and stacking algorithm based on the clinical model and optimal radiomics model. The area under the curve (AUC), clinical decisive curve (CDC), net reclassification index (NRI), and integrated discrimination index (IDI) were used to evaluate the performance and clinical benefits of the models.ResultsA total of 371 patients were incorporated. The LR model was the optimal radiomics model with the highest average AUC (0.854) and accuracy (0.802) in the internal and external validation groups (AUC = 0.910 and 0.798, respectively), and outperformed the clinical model (AUC = 0.739 and 0.592, respectively) or the radiologist (AUC = 0.768 and 0.628, respectively). The nomogram (AUC = 0.917 and 0.802, respectively) achieved better discrimination performance than the optimal radiomics model in two validation groups. The stacking model (AUC = 0.915) and ensemble model (AUC = 0.918) had a similar performance compared with the nomogram in the internal validation group, whereas the AUCs of the stacking model (AUC = 0.792) and ensemble model (AUC = 0.794) were lower than those of the nomogram and radiomics model in the external validation group. According to the CDC, NRI, and IDI, the optimal radiomics model, nomogram, stacking model, and ensemble model achieved good net benefits.ConclusionsMultiparametric MRI-based radiomics models can non-invasively differentiate stage IA EC from benign endometrial lesions, and LR is the best machine learning algorithm. The nomogram presents excellent and stable diagnostic efficiency.

Cyclin-dependent kinase inhibitor 2A/B (CDKN2A/B) homozygous deletion has been verified as an independent and critical biomarker of negative prognosis and short survival in isocitrate dehydrogenase (IDH)-mutant astrocytoma. Therefore, noninvasive and accurate discrimination of CDKN2A/B homozygous deletion status is essential for the clinical management of IDH-mutant astrocytoma patients. To develop a noninvasive, robust preoperative model based on MR imagefeatures for discriminating CDKN2A/B homozygous deletion status of IDH-mutant astrocytoma. Retrospective. Two hundred fifty-one patients: 107 patients with CDKN2A/B homozygous deletion and 144 patients without CDKN2A/B homozygous deletion. 3.0 T/1.5 T: Contrast-enhanced T1-weighted spin-echo inversion recovery sequence (CE-T1WI) and T2-weighted fluid-attenuation spin-echo inversion recovery sequence (T2FLAIR). A total of 1106 radiomics and 1000 deep learning features extracted from CE-T1WI and T2FLAIR were used to develop models to discriminate the CDKN2A/B homozygous deletion status. Radiomics models, deep learning-based radiomics (DLR) models and the final integrated model combining radiomics features with deep learning features were developed and compared their preoperative discrimination performance. Pearson chi-square test and Mann Whitney U test were used for assessing the statistical differences in patients' clinical characteristics. The Delong test compared the statistical differences of receiver operating characteristic (ROC) curves and area under the curve (AUC) of different models. The significance threshold is P < 0.05. The final combined model (training AUC = 0.966; validation AUC = 0.935; test group: AUC = 0.943) outperformed the optimal models based on only radiomics or DLR features (training: AUC = 0.916 and 0.952; validation: AUC = 0.886 and 0.912; test group: AUC = 0.862 and 0.902). Whether based on a single sequence or a combination of two sequences, radiomics and DLR models have achieved promising performance in assessing CDKN2A/B homozygous deletion status. However, the final model combining both deep learning and radiomics features from CE-T1WI and T2FLAIR outperformed the optimal radiomics or DLR model. 4 TECHNICAL EFFICACY: Stage 2.

To explore a novel approach for the early prediction of malignant cerebral edema (MCE) in stroke patients using radiomics features extracted from dual-energy computed tomography angiography (DE-CTA) reconstructed images of infarcted brain tissue. This retrospective study enrolled 398 stroke patients who underwent DE-CTA between April 2016 and November 2022 from three medical centers. Patients were allocated into a training cohort (n = 227) and a test cohort (n = 171) based on their source institution. Radiomics features were extracted from DE-CTA reconstructions of the infarct tissue. Radiomics models were built for each reconstruction scan images and for a combined model using features from all images. Additionally, a nomogram model integrating significant radiomics features and clinical characteristics was developed. The diagnostic performance of all models was evaluated using receiver operating characteristic (ROC) curve analysis, calculating the area under the curve (AUC). Radiomics models based on each individual DE-CTA reconstruction demonstrated robust predictive performance. In the test cohort, the combined radiomics model integrating features from all DE-CTA reconstructions significantly superior predictive performance compared to most single-image models. In the test cohort, both the combined radiomics model (AUC = 0.950) and the nomogram model (AUC = 0.935) significantly outperformed the clinical model (AUC = 0.574), both p < 0.001. No significant difference in performance was observed between the nomogram and the radiomics model (p = 0.30). Radiomics models derived from DE-CTA reconstructed images of infarcted tissue provide a reliable method for the early prediction of MCE development in stroke patients.

To evaluate the value of a machine learning model using enhanced CT radiomics features in the prediction of lymphovascular invasion (LVI) of esophageal squamous cell carcinoma (ESCC) before treatment. We reviewed and analyzed the enhanced CT images of 258 ESCC patients from June 2017 to December 2019. We randomly assigned the patients in a ratio of 7:3 to a training set (182 cases) and a validation (76 cases) set. Clinical risk factors and CT image characteristics were recorded, and multifactor logistic regression was used to screen independent risk factors of LVI of ESCC patients. We extracted the CT radiomics features using the FAE software and screened radiomics features using maximum relevance and minimum redundancy (MRMR) and least absolute shrinkage and selection operator (LASSO) algorithms, and finally, the radiomics labels of each patient were established. Five machine learning algorithms, namely, support vector machine (SVM), K-nearest neighbor (KNN), logistic regression (LR), Gauss naive Bayes (GNB), and multilayer perceptron (MLP), were used to construct the model of radiomics labels, and its clinical features were screened. The predictive efficacy of the machine learning model for LVI of ESCC was evaluated using the receiver operating characteristic (ROC) curve. Tumor thickness [OR = 1.189, 95% confidence interval (CI) 1.060-1.351, P = 0.005], tumor-to-normal wall enhancement ratio (TNR) (OR = 2.966, 95% CI 1.174-7.894, P = 0.024), and clinical N stage (OR = 5.828, 95% CI 1.752-20.811, P= 0.005) were determined as independent risk factors of LVI. We extracted 1,316 features from preoperative enhanced CT images and selected 14 radiomics features using MRMR and LASSO to construct the radiomics labels. In the test set, SVM, KNN, LR, and GNB showed high predictive performance, while the MLP model had poor performance. In the training set, the area under the curve (AUC) values were 0.945 and 0.905 in the KNN and SVM models, but these decreased to 0.866 and 0.867 in the validation set, indicating significant overfitting. The GNB and LR models had AUC values of 0.905 and 0.911 in the training set and 0.900 and 0.893 in the validation set, with stable performance and good fitting and predictive ability. The MLP model had AUC values of 0.658 and 0.674 in the training and validation sets, indicating poor performance. A multiscale combined model constructed using multivariate logistic regression has an AUC of 0.911 (0.870-0.951) and 0.893 (0.840-0.962), accuracy of 84.4% and 79.7%, sensitivity of 90.8% and 87.1%, and specificity of 80.5% and 79.0% in the training and validation sets, respectively. Machine learning models can preoperatively predict the condition of LVI effectively in patients with ESCC based on enhanced CT radiomics features. The GNB and LR models exhibit good stability and may bring a new way for the non-invasive prediction of LVI condition in ESCC patients before treatment.

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer, with approximately 650,000 new cases diagnosed annually and 400,000 HNSCC-related deaths worldwide each year. HNSCC is typically diagnosed when already at an advanced stage. Despite advancements in surgery, radiation therapy, and chemotherapy, patients with advanced HNSCC have a poor prognosis. The median overall survival time for patients with recurrence and metastasis is 10-13 months in the setting of first-line chemotherapy and 6 months in the second-line setting. Many studies have shown that noncoding RNAs encoded by the human genome are functional and play critical roles in various cellular processes, e.g., cell growth, migration, invasion, and apoptosis. Among noncoding RNAs, microRNAs (miRNAs) are endogenous single-stranded RNA molecules comprising 19-22 nucleotides that function as fine-tuners of RNA expression. A single miRNA regulates many RNA transcripts, and bioinformatics studies have shown that more than half of protein-coding genes are controlled by miRNAs. Aberrantly expressed miRNAs are closely associated with cancer pathogenesis via the disruption of RNA networks within cancer cells. To identify novel oncogenic targets in head and neck squamous cell carcinoma (HNSCC), we have analyzed antitumor microRNAs (miRNAs) and their controlled molecular networks in HNSCC cells. Based on our miRNA signature in HNSCC, both strands of the miR-99a-duplex (miR-99a-5p: the guide strand, and miR-99a-3p: the passenger strand) are downregulated in cancer tissues. Moreover, low expression of miR-99a-5p and miR-99a-3p significantly predicts poor prognosis in HNSCC, and these miRNAs regulate cancer cell migration and invasion. We previously showed that passenger strands of miRNAs have antitumor functions. Here, we screened miR-99a-3p-controlled oncogenes involved in HNSCC pathogenesis. Thirty-two genes were identified as miR-99a-3p-regulated genes, and 10 genes (STAMBP, TIMP4, TMEM14C, CANX, SUV420H1, HSP90B1, PDIA3, MTHFD2, BCAT1, and SLC22A15) significantly predicted 5-year overall survival. Notably, among these genes, STAMBP, TIMP4, TMEM14C, CANX, and SUV420H1 were independent prognostic markers of HNSCC by multivariate analyses. We further investigated the oncogenic function of STAMBP in HNSCC cells using knockdown assays. Our data demonstrated that the aggressiveness of phenotypes in HNSCC cells was attenuated by siSTAMBP transfection. Moreover, aberrant STAMBP expression was detected in HNSCC clinical specimens by immunochemistry. Our strategy, i.e., identification of antitumor miRNAs and their targets, maybe an attractive tool to reveal novel prognostic and therapeutic targets in HNSCC. Citation Format: Reona Okada, Keiichi Koshizuka, Yasutaka Yamada, Shogo Moriya, Naoko Kikkawa, Takashi Kinoshita, Toyoyuki Hanazawa, Naohiko Seki. Regulation of oncogenic targets by miR-99a-3p (the passenger strand of the miR-99a-duplex) in head and neck squamous cell carcinoma [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3711.

Background: Head and neck squamous cell carcinoma (HNSCC) is a disastrous disease with substantial morbidity and mortality. This study aims to explore the effective diagnostic and prognostic biomarker for HNSCC. Methods: MiRNA expression data and corresponding clinical information of HNSCC from The Cancer Genome Atlas (TCGA) database were analyzed comprehensively to identify the differentially expressed miRNAs. The predictive ability of different classification was analyzed of the three-miRNA combinations. Then, diagnostic and prognostic value were evaluated and verified in clinical patients. Results: A total of 486 HNSCC samples and 44 normal samples were obtained from the TCGA database. After analysis, 128 differentially expressed miRNAs in HNSCC tissues were identified, and 10 miRNAs were finally selected for further study. Classification analysis developed a threemiRNA signature of hsa-mir-383, hsa-mir-615 and hsa-mir-877 with an average AUC (Area Under the Curve) of 99.4% in differentiating tumor tissues from normal tissues. The combination of the three miRNAs in serum yielded an AUC value of 0.986 (95%CI, 0.974-0.998; P＜0.0001) with 97.5% sensitivity and 94.6% specificity, which was superior to that of a single or a two-miRNA signature. One-way ANOVA test results suggested the expression levels of hsa-mir-615 and hsa-mir-877 were significantly changed with different tumor grades, with the p-values of 0.001 and 0.056, respectively, while hsa-mir-383 showed no significant correlation with tumor grade (P=0.18). And no significant correlations were found between hsa-mir-383, hsa-mir-615 or hsa-mir-877 and tumor stage with the p-values of 0.14, 0.95 and 0.81, respectively. Survival analysis indicated different expression levels of hsa-mir-383 led to significantly different survival rates (P<0.022), while no significant survival differences were observed between the high and low expression set of hsamir-615 (P<0.401) or hsa-mir-877 (P<0.403). This result was verified in 81 validation patients. The low expression level of hsa-mir-383 was significantly correlated with poor prognosis (P=0.0404), while the expression level of hsa-mir-615 or hsa-mir-877 showed no significant relationship with the prognosis (hsa-mir-615, P=0.6247; hsa-mir-877, P=0.5199). Conclusion: A three-miRNA signature of hsa-mir-383, hsa-mir-615, and hsamir-877 may serve as an excellent diagnostic biomarker for HNSCC, and therein hsa-mir-383 shows potential prognostic significance for HNSCC patients. Funding Statement: Shandong Provincial Health Commission Foundation (No. 2016WS0410), China postdoctoral Scinence Foundation Grant (No. 2019M652380), Fundamental Research Funds of Shandong University (No.2018CJ047). This work is also supported by Qilu Young Scholar Program of Shandong University. Declaration of Interests: The authors declare that they do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted. Ethics Approval Statement: All patients and normal controls gave informed consent. This study was approved by the Medical Institutional Ethical Committee of the Provincial Hospital Affiliated to Shandong University.

ObjectiveTo evaluate the clinical utility of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI)-derived clinicoradiological characteristics and intratumoral/peritumoral radiomic features in predicting pathological upgrades (malignant transformation) in high-risk breast lesions.Materials and methodsRetrospectively collected the data of 174 patients with high-risk breast lesions who underwent preoperative breast MRI examinations and were confirmed by biopsy pathology in Shenzhen People’s Hospital between January 1, 2019 and January 1, 2024. The dataset was randomly divided into a training set (n = 121) and a test set (n = 53) at a ratio of 7:3. Initially, during the second stage of DCE-MRI, the region of interest (ROI) was delineated along the maximum cross-section of the lesion, and then automatically expanded outward by 3 mm, 5 mm, and 7 mm as the peritumoral ROIs. The intratumoral, each peritumoral, and the combined intratumoral and peritumoral radiomic models were established respectively. Independent risk factors predictive of malignant upgrades in high-risk lesions were identified through univariate and multivariable logistic regression analyses, which were subsequently incorporated as clinical and imaging characteristics. Finally, a combined model was established by integrating the intratumoral and peritumoral radiomic features with the clinical and imaging features. The performance of each model was analyzed using the receiver operating characteristic (ROC) curves, and the area under the curve (AUC) was calculated.ResultsThe peritumoral 3 mm radiomics model achieved the highest diagnostic performance among all the peritumoral models, with the AUC values of 0.704 and 0.654 for the training and test sets, respectively. In the training set, the combined model showed the highest diagnostic performance (AUC = 0.883), which was superior to that of the clinical and imaging features model (AUC = 0.745, P = 0.003), the intratumoral radiomics model (AUC = 0.791, P = 0.027), the peritumoral 3 mm radiomics model (AUC = 0.704, P = 0.001), and the combined intratumoral and peritumoral radiomic model (AUC = 0.830, P = 0.004). In the test set, the combined model also showed the highest diagnostic performance (AUC = 0.851). The combined model constructed by integrating the intratumoral and peritumoral radiomics features with the clinical and imaging features had the best diagnostic performance, with the sensitivity, specificity, and accuracy of 79.4%, 82.7%, and 81.8% in the training set, and 72.7%, 85.7%, and 83.0% in the test set, respectively.ConclusionThe combined predictive model, which integrates intratumoral and peritumoral radiomic features with clinical and imaging data, exhibited strong diagnostic performance and a clinically applicable nomogram was constructed to stratify individualized upgrade risk, assisting clinicians in making more precise decisions.