Prognostic value of radiomic features in patients with esophageal cancer treated with chemoradiotherapy.

BACKGROUND: The aim of the present study was to evaluate the prognostic value of radiomic features in patients who underwent chemoradiotherapy for esophageal cancer. METHODS: In this retrospective study, two independent cohorts of esophageal cancer patients treated with chemoradiotherapy were included. Radiomics features of each patient were extracted from pre-treatment computed tomography (CT) images. Radiomic features were selected by employing univariate and multivariate analyses in the test cohort. Selected radiomic features were verified in the validation cohort. The endpoint of the present study was overall survival. RESULTS: A total of 101 esophageal cancer patients were included in our study, with 71 patients in the test cohort and 30 patients in the validation cohort. Univariate analysis identified 158 radiomic features as prognostic factors for overall survival in the test cohort. A multivariate analysis revealed that root mean squared and Low-High-High (LHH) median were prognostic factors for overall survival with a hazard ratio of 2.23 (95% confidence interval [CI]: 1.16-4.70, P = 0.017) and 0.26 (95% CI: 0.13-0.54, P < 0.001), respectively. In the validation cohort, root mean squared high/LHH median low group had the most preferable prognosis with a median overall survival of 73.30 months (95% CI: 32.13-NA), whereas root mean squared low/LHH median low group had the poorest prognosis with a median overall survival of 9.72 months (95% CI: 2.50-NA), with a P value of < 0.001. CONCLUSIONS: We identified two radiomic features that might be independent prognostic factors of overall survival of esophageal cancer patients treated with chemoradiotherapy.

Re-salvage focal low-dose rate brachytherapy for local recurrence of prostate cancer after salvage focal low-dose rate brachytherapy.

Introduction: Salvage brachytherapy represents an effective treatment for local recurrence of prostate cancer after prior external beam radiotherapy. However, the optimal therapeutic strategies for local recurrence after salvage brachytherapy have not yet been determined. Case presentation: We describe the case of a 77-year-old man who underwent re-salvage focal low-dose rate brachytherapy for local recurrence after carbon ion radiotherapy and salvage focal low-dose rate brachytherapy. We performed re-salvage focal low-dose rate brachytherapy for the recurrence with a different type of seed, which resulted in a significant reduction in the prostate-specific antigen level. During the 35-month follow-up after re-salvage focal low-dose rate brachytherapy, no recurrence of prostate cancer and no severe radiation-related toxicities were observed. Conclusion: Our patient was successfully treated with re-salvage focal low-dose rate brachytherapy for local recurrence after salvage focal low-dose rate brachytherapy. This treatment strategy might be effective for such patients and is not associated with sexual dysfunction or severe adverse events.

Cone-beam computed-tomography-based delta-radiomic analysis for investigating prognostic power for esophageal squamous cell cancer patients undergoing concurrent chemoradiotherapy.

PURPOSE: To investigate the prognostic power of cone-beam computed-tomography (CBCT)-based delta-radiomics in esophageal squamous cell cancer (ESCC) patients treated with concurrent chemoradiotherapy (CCRT). METHODS: We collected data from 26 ESCC patients treated with CCRT. CBCT images acquired at five time points (1st-5th week) per patient during CCRT were used in this study. Radiomic features were extracted from the five CBCT images on the gross tumor volumes. Then, 17 delta-radiomic feature sets derived from five types of calculations were obtained for all the cases. Leave-one-out cross-validation was applied to investigate the prognostic power of CBCT-based delta-radiomic features. Feature selection and construction of a prediction model using Coxnet were performed using training samples. Then, the test sample was classified into high or low risk in each cross-validation fold. Survival analysis for the two groups were performed to evaluate the prognostic power of the extracted CBCT-based delta-radiomic features. RESULTS: Four delta-radiomic feature sets indicated significant differences between the high- and low-risk groups (p < 0.05). The highest C-index in the 17 delta-radiomic feature sets was 0.821 (95 % confidence interval, 0.735-0.907). That feature set had p-value of the log-rank test and hazard ratio of 0.003 and 4.940 (95 % confidence interval, 1.391-17.544), respectively. CONCLUSIONS: We investigated the potential of using CBCT-based delta-radiomics for prognosis of ESCC patients treated with CCRT. It was demonstrated that delta-radiomic feature sets based on the absolute value of relative difference obtained from the early to the middle treatment stages have high prognostic power for ESCC.

Development and Validation of a Preoperative Nomogram for Endoscopic Management Decision Making in Upper Urinary Tract Urothelial Carcinoma.

OBJECTIVE: We aimed to develop and validate a preoperative nomogram that predicts low-grade, non-muscle invasive upper urinary tract urothelial carcinoma (LG-NMI UTUC), thereby aiding in the accurate selection of endoscopic management (EM) candidates. METHODS: This was a retrospective study that included 454 patients who underwent radical surgery (Cohort 1 and Cohort 2), and 26 patients who received EM (Cohort 3). Utilizing a multivariate logistic regression model, a nomogram predicting LG-NMI UTUC was developed based on data from Cohort 1. The nomogram's accuracy was compared with conventional European Association of Urology (EAU) and National Comprehensive Cancer Network (NCCN) models. External validation was performed using Cohort 2 data, and the nomogram's prognostic value was evaluated via disease progression metrics in Cohort 3. RESULTS: In Cohort 1, multivariate analyses highlighted the absence of invasive disease on imaging (odds ratio [OR] 7.04; p = 0.011), absence of hydronephrosis (OR 2.06; p = 0.027), papillary architecture (OR 24.9; p < 0.001), and lack of high-grade urine cytology (OR 0.22; p < 0.001) as independent predictive factors for LG-NMI disease. The nomogram outperformed the two conventional models in predictive accuracy (0.869 vs. 0.759-0.821) and exhibited a higher net benefit in decision curve analysis. The model's clinical efficacy was corroborated in Cohort 2. Moreover, the nomogram stratified disease progression-free survival rates in Cohort 3. CONCLUSION: Our nomogram ( https://kmur.shinyapps.io/UTUC_URS/ ) accurately predicts LG-NMI UTUC, thereby identifying suitable candidates for EM. Additionally, the model serves as a useful tool for prognostic stratification in patients undergoing EM.

Hypoxia-inducible factor 2α protein and mRNA expression correlate with histomorphological features in clear cell renal cell carcinoma.

Hypoxia-inducible factor 2α (HIF2α) has been identified as a potential biomarker and novel target for systemic therapy in clear cell renal cell carcinoma (ccRCC). The present study aims to evaluate the association of HIF2α protein and HIF2A mRNA expression with clinicopathological factors and histomorphological features related to vasculature and inflammation of ccRCC using a localized ccRCC cohort (n = 428) and The Cancer Genome Atlas (TCGA)-KIRC cohort (n = 433). HIF2α protein expression was immunohistochemically assessed using tissue microarrays and HIF2A mRNA expression was assessed using the TCGA RNA-sequencing data. Positive HIF2α protein and high HIF2A mRNA expression were observed in 145 (33.9 %) and 142 (32.8 %) patients, respectively. Positive nuclear HIF2α protein expression was significantly associated with the clear histological phenotype and architectural patterns related to rich vascular networks (p < 0.001), and no tumor-associated immune cells status (p < 0.05) in addition to favorable prognostic factors such as lower TNM stage, lower WHO/ISUP grade, or the absence of necrosis (p < 0.001). The HIF2A mRNA expression profile by the TCGA cohort showed similar trends as the HIF2α protein profile. In addition, positive HIF2α protein and high HIF2A mRNA expression were associated with higher recurrence-free survival and overall survival, respectively (both p < 0.001). In conclusion, we comprehensively demonstrated the association of HIF2α profiles with clinicopathological factors and histomorphological features related to vasculature and inflammation at both protein and mRNA levels. Histomorphological features expressing HIF2α may provide information on HIF2α targeted therapeutic response as well as prognosis in ccRCC patients.

Effect of finerenone on diabetic kidney disease outcomes with estimated glomerular filtration rate below 25 mL/min/1.73 m2.

Background: In the Finerenone in Reducing Cardiovascular Mortality and Morbidity in Diabetic Kidney Disease trial, finerenone reduced the risk of cardiovascular events in patients with chronic kidney disease (CKD) and type 2 diabetes, while in the Finerenone in Reducing Kidney Failure and Disease Progression in Diabetic Kidney Disease trial, it improved renal and cardiovascular outcomes in patients with advanced CKD. However, no previous studies have assessed patients with CKD and type 2 diabetes with an estimated glomerular filtration rate (eGFR) below 25 mL/min/1.73 m2. Methods: Nine patients with CKD and type 2 diabetes who received finerenone 10 mg/day were analyzed retrospectively. Changes in eGFR, urinary protein, and serum potassium levels were studied from 1 year before administration of finerenone until 6 months after administration. Results: The mean baseline eGFR slope was -7.63 ± 9.84 (mL/min/1.73 m2/year). After finerenone treatment, the mean eGFR slope significantly improved -1.44 ± 3.17 (mL/min/1.73 m2/6 months, P=0.038). However, finerenone treatment did not significantly reduce proteinuria. Furthermore, finerenone did not increase serum potassium levels. Conclusions: Patients treated with finerenone showed a significantly slower decline in eGFR. Furthermore, aside from the present study, no reports have indicated the effectiveness of finerenone in patients with advanced CKD with an eGFR below 25 mL/min/1.73 m2. As confirmed in our clinical trials, the finding that finerenone is effective in a wide range of renal functions can be generalized to clinical practice. However, sample size in this study was small. Thus, further large-scale investigations will be needed.

Learning from the past and present to change the future: Endoscopic management of upper urinary tract urothelial carcinoma.

Current guidelines recommend endoscopic management (EM) for patients with low-risk upper urinary tract urothelial carcinoma, as well as those with an imperative indication. However, regardless of the tumor risk, radical nephroureterectomy is still mainly performed worldwide despite the benefits of EM, such as renal function maintenance, no hemodialysis requirement, and treatment cost reduction. This might be explained by the association of EM with a high risk of local recurrence and progression. Furthermore, the need for rigorous patient selection and close surveillance following EM may be relevant. Nevertheless, recent developments in diagnostic modalities, pathological evaluation, surgical devices and techniques, and intracavitary regimens have been reported, which may contribute to improved risk stratification and treatments with superior oncological outcomes. In this review, considering recent advances in endourology and oncology, we propose novel treatment strategies for optimal EM.

Deep learning-based predictions of clear and eosinophilic phenotypes in clear cell renal cell carcinoma.

We have recently shown that histological phenotypes focusing on clear and eosinophilic cytoplasm in clear cell renal cell carcinoma (ccRCC) correlated with prognosis and the response to angiogenesis inhibition and checkpoint blockade. This study aims to objectively show the diagnostic utility of clear or eosinophilic phenotypes of ccRCC by developing an artificial intelligence (AI) model using the TCGA-ccRCC dataset and to demonstrate if the clear or eosinophilic predicted phenotypes correlate with pathological factors and gene signatures associated with angiogenesis and cancer immunity. Before the development of the AI model, histological evaluation using hematoxylin and eosin whole-slide images of the TCGA-ccRCC cohort (n = 435) was performed by a urologic pathologist. The AI model was developed as follows. First, the highest-grade area on each whole slide image was captured for image processing. Second, the selected regions were cropped into tiles. Third, the AI model was trained using transfer learning on a deep convolutional neural network, and clear or eosinophilic predictions were scaled as AI scores. Next, we verified the AI model using a validation cohort (n = 95). Finally, we evaluated the accuracy of the prognostic predictions of the AI model and revealed that the AI model detected clear and eosinophilic phenotypes with high accuracy. The AI model stratified the patients' outcomes, and the predicted eosinophilic phenotypes correlated with adverse clinicopathological characteristics and high immune-related gene signatures. In conclusion, the AI-based histologic subclassification accurately predicted clear or eosinophilic phenotypes of ccRCC, allowing for consistently reproducible stratification for prognostic and therapeutic stratification.

MRI-Based Radiomics Differentiates Skull Base Chordoma and Chondrosarcoma: A Preliminary Study.

Chordoma and chondrosarcoma share common radiographic characteristics yet are distinct clinically. A radiomic machine learning model differentiating these tumors preoperatively would help plan surgery. MR images were acquired from 57 consecutive patients with chordoma (N = 32) or chondrosarcoma (N = 25) treated at the University of Tokyo Hospital between September 2012 and February 2020. Preoperative T1-weighted images with gadolinium enhancement (GdT1) and T2-weighted images were analyzed. Datasets from the first 47 cases were used for model creation, and those from the subsequent 10 cases were used for validation. Feature extraction was performed semi-automatically, and 2438 features were obtained per image sequence. Machine learning models with logistic regression and a support vector machine were created. The model with the highest accuracy incorporated seven features extracted from GdT1 in the logistic regression. The average area under the curve was 0.93 ± 0.06, and accuracy was 0.90 (9/10) in the validation dataset. The same validation dataset was assessed by 20 board-certified neurosurgeons. Diagnostic accuracy ranged from 0.50 to 0.80 (median 0.60, 95% confidence interval 0.60 ± 0.06%), which was inferior to that of the machine learning model (p = 0.03), although there are some limitations, such as the risk of overfitting and the lack of an extramural cohort for truly independent final validation. In summary, we created a novel MRI-based machine learning model to differentiate skull base chordoma and chondrosarcoma from multiparametric signatures.

Training of deep cross-modality conversion models with a small data set, and their application in megavoltage CT to kilovoltage CT conversion.

PURPOSE: In recent years, deep learning-based image processing has emerged as a valuable tool for medical imaging owing to its high performance. However, the quality of deep learning-based methods heavily relies on the amount of training data; the high cost of acquiring a large data set is a limitation to their utilization in medical fields. Herein, based on deep learning, we developed a computed tomography (CT) modality conversion method requiring only a few unsupervised images. METHODS: The proposed method is based on cycle-consistency generative adversarial network (CycleGAN) with several extensions tailored for CT images, which aims at preserving the structure in the processed images and reducing the amount of training data. This method was applied to realize the conversion of megavoltage computed tomography (MVCT) to kilovoltage computed tomography (kVCT) images. Training was conducted using several data sets acquired from patients with head and neck cancer. The size of the data sets ranged from 16 slices (two patients) to 2745 slices (137 patients) for MVCT and 2824 slices (98 patients) for kVCT. RESULTS: The required size of the training data was found to be as small as a few hundred slices. By statistical and visual evaluations, the quality improvement and structure preservation of the MVCT images converted by the proposed model were investigated. As a clinical benefit, it was observed by medical doctors that the converted images enhanced the precision of contouring. CONCLUSIONS: We developed an MVCT to kVCT conversion model based on deep learning, which can be trained using only a few hundred unpaired images. The stability of the model against changes in data size was demonstrated. This study promotes the reliable use of deep learning in clinical medicine by partially answering commonly asked questions, such as "Is our data sufficient?" and "How much data should we acquire?"

[Improvement in Image Quality of CBCT during Treatment by Cycle Generative Adversarial Network].

PURPOSE: Volumetric modulated arc therapy (VMAT) can acquire projection images during rotational irradiation, and cone-beam computed tomography (CBCT) images during VMAT delivery can be reconstructed. The poor quality of CBCT images prevents accurate recognition of organ position during the treatment. The purpose of this study was to improve the image quality of CBCT during the treatment by cycle generative adversarial network (CycleGAN). METHOD: Twenty patients with clinically localized prostate cancer were treated with VMAT, and projection images for intra-treatment CBCT (iCBCT) were acquired. Synthesis of PCT (SynPCT) with improved image quality by CycleGAN requires only unpaired and unaligned iCBCT and planning CT (PCT) images for training. We performed visual and quantitative evaluation to compare iCBCT, SynPCT and PCT deformable image registration (DIR) to confirm the clinical usefulness. RESULT: We demonstrated suitable CycleGAN networks and hyperparameters for SynPCT. The image quality of SynPCT improved visually and quantitatively while preserving anatomical structures of the original iCBCT. The undesirable deformation of PCT was reduced when SynPCT was used as its reference instead of iCBCT. CONCLUSION: We have performed image synthesis with preservation of organ position by CycleGAN for iCBCT and confirmed the clinical usefulness.

Publisher Correction: Prediction of malignant glioma grades using contrast-enhanced T1-weighted and T2-weighted magnetic resonance images based on a radiomic analysis.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Prediction of malignant glioma grades using contrast-enhanced T1-weighted and T2-weighted magnetic resonance images based on a radiomic analysis.

We conducted a feasibility study to predict malignant glioma grades via radiomic analysis using contrast-enhanced T1-weighted magnetic resonance images (CE-T1WIs) and T2-weighted magnetic resonance images (T2WIs). We proposed a framework and applied it to CE-T1WIs and T2WIs (with tumor region data) acquired preoperatively from 157 patients with malignant glioma (grade III: 55, grade IV: 102) as the primary dataset and 67 patients with malignant glioma (grade III: 22, grade IV: 45) as the validation dataset. Radiomic features such as size/shape, intensity, histogram, and texture features were extracted from the tumor regions on the CE-T1WIs and T2WIs. The Wilcoxon-Mann-Whitney (WMW) test and least absolute shrinkage and selection operator logistic regression (LASSO-LR) were employed to select the radiomic features. Various machine learning (ML) algorithms were used to construct prediction models for the malignant glioma grades using the selected radiomic features. Leave-one-out cross-validation (LOOCV) was implemented to evaluate the performance of the prediction models in the primary dataset. The selected radiomic features for all folds in the LOOCV of the primary dataset were used to perform an independent validation. As evaluation indices, accuracies, sensitivities, specificities, and values for the area under receiver operating characteristic curve (or simply the area under the curve (AUC)) for all prediction models were calculated. The mean AUC value for all prediction models constructed by the ML algorithms in the LOOCV of the primary dataset was 0.902 ± 0.024 (95% CI (confidence interval), 0.873-0.932). In the independent validation, the mean AUC value for all prediction models was 0.747 ± 0.034 (95% CI, 0.705-0.790). The results of this study suggest that the malignant glioma grades could be sufficiently and easily predicted by preparing the CE-T1WIs, T2WIs, and tumor delineations for each patient. Our proposed framework may be an effective tool for preoperatively grading malignant gliomas.

Radiomics Analysis for Glioma Malignancy Evaluation Using Diffusion Kurtosis and Tensor Imaging.

PURPOSE: A noninvasive diagnostic method to predict the degree of malignancy accurately would be of great help in glioma management. This study aimed to create a highly accurate machine learning model to perform glioma grading. METHODS AND MATERIALS: Preoperative magnetic resonance imaging acquired for cases of glioma operated on at our institution from October 2014 through January 2018 were obtained retrospectively. Six types of magnetic resonance imaging sequences (T2-weighted image, diffusion-weighted image, apparent diffusion coefficient [ADC], fractional anisotropy, and mean kurtosis [MK]) were chosen for analysis; 476 features were extracted semiautomatically for each sequence (2856 features in total). Recursive feature elimination was used to select significant features for a machine learning model that distinguishes glioblastoma from lower-grade glioma (grades 2 and 3). RESULTS: Fifty-five data sets from 54 cases were obtained (14 grade 2 gliomas, 12 grade 3 gliomas, and 29 glioblastomas), of which 44 and 11 data sets were used for machine learning and independent testing, respectively. We detected 504 features with significant differences (false discovery rate <0.05) between glioblastoma and lower-grade glioma. The most accurate machine learning model was created using 6 features extracted from the ADC and MK images. In the logistic regression, the area under the curve was 0.90 ± 0.05, and the accuracy of the test data set was 0.91 (10 out of 11); using a support vector machine, they were 0.93 ± 0.03 and 0.91 (10 out of 11), respectively (kernel, radial basis function; c = 1.0). CONCLUSIONS: Our machine learning model accurately predicted glioma tumor grade. The ADC and MK sequences produced particularly useful features.

[An Introduction to Radiomics: Toward a New Era of Precision Medicine].

Recently, in a medical field, quantitative data mining is a hot topic for performing a precision (or personalized) medicine. Although a molecular biological data has been mainly utilized for data mining in this field, medical images are also important minable data. Radiomics is a comprehensive analysis methodology for describing tumor phenotypes or molecular biological expressions (e.g. genotypes) using minable feature extracted from a large number of medical images. In this review paper, we introduce to a framework of the radiomics.

Cone Beam Computed Tomography Image Quality Improvement Using a Deep Convolutional Neural Network.

Introduction Cone beam computed tomography (CBCT) plays an important role in image-guided radiation therapy (IGRT), while having disadvantages of severe shading artifact caused by the reconstruction using scatter contaminated and truncated projections. The purpose of this study is to develop a deep convolutional neural network (DCNN) method for improving CBCT image quality. Methods CBCT and planning computed tomography (pCT) image pairs from 20 prostate cancer patients were selected. Subsequently, each pCT volume was pre-aligned to the corresponding CBCT volume by image registration, thereby leading to registered pCT data (pCTr). Next, a 39-layer DCNN model was trained to learn a direct mapping from the CBCT to the corresponding pCTr images. The trained model was applied to a new CBCT data set to obtain improved CBCT (i-CBCT) images. The resulting i-CBCT images were compared to pCTr using the spatial non-uniformity (SNU), the peak-signal-to-noise ratio (PSNR) and the structural similarity index measure (SSIM). Results The image quality of the i-CBCT has shown a substantial improvement on spatial uniformity compared to that of the original CBCT, and a significant improvement on the PSNR and the SSIM compared to that of the original CBCT and the enhanced CBCT by the existing pCT-based correction method. Conclusion We have developed a DCNN method for improving CBCT image quality. The proposed method may be directly applicable to CBCT images acquired by any commercial CBCT scanner.