Artificial intelligence aided precise detection of local recurrence on MRI for nasopharyngeal carcinoma: a multicenter cohort study.

Background: MRI is the routine examination to surveil the recurrence of nasopharyngeal carcinoma, but it has relatively lower sensitivity than PET/CT. We aimed to find if artificial intelligence (AI) could be competent pre-inspector for MRI radiologists and whether AI-aided MRI could perform better or even equal to PET/CT. Methods: This multicenter study enrolled 6916 patients from five hospitals between September 2009 and October 2020. A 2.5D convolutional neural network diagnostic model and a nnU-Net contouring model were developed in the training and test cohorts and used to independently predict and visualize the recurrence of patients in the internal and external validation cohorts. We evaluated the area under the ROC curve (AUC) of AI and compared AI with MRI and PET/CT in sensitivity and specificity using the McNemar test. The prospective cohort was randomized into the AI and non-AI groups, and their sensitivity and specificity were compared using the Chi-square test. Findings: The AI model achieved AUCs of 0.92 and 0.88 in the internal and external validation cohorts, corresponding to the sensitivity of 79.5% and 74.3% and specificity of 91.0% and 92.8%. It had comparable sensitivity to MRI (e.g., 74.3% vs. 74.7%, P = 0.89) but lower sensitivity than PET/CT (77.9% vs. 92.0%, P < 0.0001) at the same individual-specificities. The AI model achieved moderate precision with a median dice similarity coefficient of 0.67. AI-aided MRI improved specificity (92.5% vs. 85.0%, P = 0.034), equaled PET/CT in the internal validation subcohort, and increased sensitivity (81.9% vs. 70.8%, P = 0.021) in the external validation subcohort. In the prospective cohort of 1248 patients, the AI group had higher sensitivity than the non-AI group (78.6% vs. 67.3%, P = 0.23), albeit nonsignificant. In future randomized controlled trials, a sample size of 3943 patients in each arm would be required to demonstrate the statistically significant difference. Interpretation: The AI model equaled MRI by expert radiologists, and AI-aided MRI by expert radiologists equaled PET/CT. A larger randomized controlled trial is warranted to demonstrate the AI's benefit sufficiently. Funding: The Sun Yat-sen University Clinical Research 5010 Program (2015020), Guangdong Basic and Applied Basic Research Foundation (2022A1515110356), and Guangzhou Science and Technology Program (2023A04J1788).

Deep learning-based precise prediction and early detection of radiation-induced temporal lobe injury for nasopharyngeal carcinoma.

Background: Radiotherapy is the mainstay of treatment for nasopharyngeal carcinoma. Radiation-induced temporal lobe injury (TLI) can regress or resolve in the early phase, but it is irreversible at a later stage. However, no study has proposed a risk-based follow-up schedule for its early detection. Planning evaluation is difficult when dose-volume histogram (DVH) parameters are similar and optimization is terminated. Methods: This multicenter retrospective study included 6065 patients between 2014 and 2018. A 3D ResNet-based deep learning model was developed in training and validation cohorts and independently tested using concordance index in internal and external test cohorts. Accordingly, the patients were stratified into risk groups, and the model-predicted risks were used to develop risk-based follow-up schedules. The schedule was compared with the Radiation Therapy Oncology Group (RTOG) recommendation (every 3 months during the first 2 years and every 6 months in 3-5 years). Additionally, the model was used to evaluate plans with similar DVH parameters. Findings: Our model achieved concordance indexes of 0.831, 0.818, and 0.804, respectively, which outperformed conventional prediction models (all P < 0.001). The temporal lobes in all the cohorts were stratified into three groups with discrepant TLI-free survival. Personalized follow-up schedules developed for each risk group could detect TLI 1.9 months earlier than the RTOG recommendation. According to a higher median predicted 3-year TLI-free survival (99.25% vs. 99.15%, P < 0.001), the model identified a better plan than previous models. Interpretation: The deep learning model predicted TLI more precisely. The model-determined risk-based follow-up schedule detected the TLI earlier. The planning evaluation was refined because the model identified a better plan with a lower risk of TLI. Funding: The Sun Yat-sen University Clinical Research 5010 Program (2015020), Guangdong Basic and Applied Basic Research Foundation (2022A1515110356), Medical Scientific Research Foundation of Guangdong Province (A2022367), and Guangzhou Science and Technology Program (2023A04J1788).

Dosiomics Risk Model for Predicting Radiation Induced Temporal Lobe Injury and Guiding Individual Intensity-Modulated Radiation Therapy.

PURPOSE: We aimed to assess the value of dose distribution-based dosiomics and planning computed tomography-based radiomics to predict radiation-induced temporal lobe injury (TLI) and guide individualized intensity modulated radiation therapy. METHODS AND MATERIALS: A total of 5599 nasopharyngeal carcinoma patients were enrolled, including 2503, 1072, 988, and 1036 patients in the training, validation, prospective test, and external test cohorts, respectively. The concordance index (C-index) was used to compare the performance of the radiomics and dosiomics models with that of the quantitative analyses of normal tissue effects in the clinic and Wen's models. The predicted TLI-free survival rates of redesigned simulated plans with the same dose-volume histogram but different dose distributions for same patient in a cohort of 30 randomly selected patients were compared by the Wilcoxon matched-pairs signed-rank test. RESULTS: The radiomics and dosiomics signatures were constructed based on 30 selected computed tomography features and 10 selected dose distribution features, respectively, which were important predictors of TLI-free survival (all P <.001). However, the radiomics signature had a low C-index. The dosiomics risk model combining the dosiomics signature, D1cc, and age had favorable performance, with C-index values of 0.776, 0.811, 0.805, and 0.794 in the training, validation, prospective test, and external test cohorts, respectively, which were better than those of the quantitative analyses of normal tissue effects in the clinic model and Wen's model (all P <.001). The dosiomics risk model can further distinguish patients in a same risk category divided by other models (all P <.05). Conversely, the other models were unable to separate populations classified by the dosiomics risk model (all P > .05). Two simulated plans with the same dose-volume histogram but different dose distributions had different TLI-free survival rates predicted by dosiomics risk model (all P ≤ .002). CONCLUSIONS: The dosiomics risk model was superior to traditional models in predicting the risk of TLI. This is a promising approach to precisely predict radiation-induced toxicities and guide individualized intensity modulated radiation therapy.

Effect of Induction Chemotherapy in Nasopharyngeal Carcinoma: An Updated Meta-Analysis.

BACKGROUND: Previous meta-analysis had evaluated the effect of induction chemotherapy in nasopharyngeal carcinoma. But two trials with opposite findings were not included and the long-term result of another trial significantly differed from the preliminary report. This updated meta-analysis was thus warranted. METHODS: Literature search was conducted to identify randomized controlled trials focusing on the additional efficacy of induction chemotherapy in nasopharyngeal carcinoma. Trial-level pooled analysis of hazard ratio (HR) for progression free survival and overall survival and risk ratio (RR) for locoregional control rate and distant control rate were performed. RESULTS: Twelve trials were eligible. The addition of induction chemotherapy significantly prolonged both progression free survival (HR=0.68, 95% confidence interval [CI] 0.60-0.76, p<0.001) and overall survival (HR=0.67, 95% CI 0.54-0.80, p<0.001), with 5-year absolute benefit of 11.31% and 8.95%, respectively. Locoregional (RR=0.80, 95% CI 0.70-0.92, p=0.002) and distant control (RR=0.70, 95% CI 0.62-0.80) rates were significantly improved as well. The incidence of grade 3-4 adverse events during the concurrent chemoradiotherapy was higher in leukopenia (p=0.028), thrombocytopenia (p<0.001), and fatigue (p=0.038) in the induction chemotherapy group. CONCLUSIONS: This meta-analysis supported that induction chemotherapy could benefit patients with nasopharyngeal carcinoma in progression free survival, overall survival, locoregional, and distant control rate.

High MMP-26 expression in glioma is correlated with poor clinical outcome of patients.

To date the management of glioma remains a great challenge in cancer therapy worldwide. The identification of novel diagnostic and therapeutic methods is required. Although there is data indicating that matrix metalloproteinase (MMP)-26 serves an important role in many human cancer types, its clinical significance in glioma remains uncertain. The present study aimed to evaluate MMP-26 expression in human astrocytic glioma specimens, and investigate its role and significance in the progression of astrocytic glioma. Immunohistochemistry was performed to assess MMP-26 expression in astrocytic glioma tissues. The levels of MMP-26 expression and its relevance to the clinicopathological features and prognostic factors in patients with astrocytic glioma patients were then investigated. The results demonstrated that MMP-26 expression was significantly assocaited with the World Health Organization grade (P<0.05). Additionally, it was identified that MMP-26 expression was an effective predictor of the overall survival of patients with astrocytic glioma (P<0.05). Analyses of univariate and multivariate Cox regression confirmed that MMP-26 expression was an independent factor for evaluating the prognosis of astrocytic glioma patients (P<0.05). The current results support that MMP-26 may be a novel indicator of diagnosis and an independent factor for evaluating prognosis in patients with glioma.

Meta-analysis of the clinical characteristics and prognostic relevance of NOTCH1 and FBXW7 mutation in T-cell acute lymphoblastic leukemia.

The NOTCH1 signaling pathway is crucial for T-cell development, and NOTCH1 and/or FBXW7 mutations are frequently detected in T-cell acute lymphoblastic leukemia (T-ALL). We performed a systematic review and meta-analysis of 18 randomized controlled trials (RCTs) to assess the prognostic impact of mutations in the NOTCH1 pathway. After retrieving relevant articles from PubMed, EMBASE, and the Cochrane Library, we investigated overall survival (OS) and event-free survival (EFS) with hazard ratios (HRs) using fixed-effects or random-effects models and conducted subgroup analyses based on population and mutation status. NOTCH1/FBXW7 mutations correlated significantly with better prognosis (5-year EFS: HR, 0.57; 95% confidence interval [CI], 0.46 to 0.68; P < 0.001 and 5-year OS: HR, 0.61; 95% CI, 0.51 to 0.74; P < 0.001). The HR for 5-year EFS and OS with NOTCH1 mutations were 0.63 (95% CI, 0.53 to 0.75) and 0.76 (95% CI, 0.60 to 0.95), respectively; with FBXW7 mutations, they were 0.82 (95% CI, 0.60 to 1.11) and 0.79 (95% CI, 0.55 to 1.12), respectively. However, differences between children and adults showed no significance. We conclude that the presence of NOTCH1/FBXW7 mutations is an independent prognostic factor for 5-year EFS and 5-year OS.

Clinical significance of B7-H6 protein expression in astrocytoma.

Currently, immunotherapy by blocking the immune checkpoint inhibitors, such as anti-PD-1, has been carried out in many clinical studies on recurrent glioma, and the preliminary results are satisfactory, which provides a rationale for the exploration of immune checkpoint inhibitors in glioma. B7-H6 is a newly discovered member of the B7 family, which triggers antitumor of natural killer cell cytotoxicity and cytokine secretion by binding the NKp30 receptor. B7-H6 mRNA and protein expressions, which are not detected in normal tissues, are expressed mainly on the cell surface of various primary tumors and cell lines. However, up until now, there is no data about the clinical significance of B7-H6 expression in astrocytoma patients. The present study provides an investigation on the relationship between prognostic and clinical value of B7-H6 protein in astrocytoma tissues. All the astrocytic glioma tissues were stained for B7-H6. Immunohistochemistry stain of 122 astrocytoma samples showed that immunoreactivity of B7-H6 was seen predominantly in the cytoplasm. The B7-H6 expression did not show significant relevance with patient age, sex distribution, Karnofsky performance status score, extent of resection, and tumor location in astrocytoma patients, but B7-H6 positive expression is significantly associated with World Health Organization grade (P=0.046). However, the survival rate after operation presented no significant difference of B7-H6 expression in astrocytoma patients. Kaplan-Meier analysis and the log-rank test revealed that B7-H6 expression cannot predict the overall survival. In all, it seems that the B7-H6 expression might be a marker to differentiate the World Health Organization grade level of astrocytoma, but the prognosis value of B7-H6 in astrocytoma should be studied in detail.

Neurosurgical interventions for patients with nasopharyngeal carcinoma: a single institution experience.

BACKGROUND: Nasopharyngeal carcinoma (NPC) is a frequent head and neck cancer in southern China and Southeast Asia. The majority of NPC patients are managed by radiation oncologists, medical oncologists and head and neck surgeons. Actually, neurosurgical interventions are warranted under specific circumstances. In this article, we described our experience as neurosurgeons in the management of NPC patients. METHODS: Medical records of NPC patients who received neurosurgical procedure at Sun Yat-sen University Cancer Center were reviewed. RESULTS: Twenty-seven patients were identified. Among 27 cases, neurosurgical procedures were performed in 18 (66.7%) with radiation-induced temporal necrosis, 2 (7.4%) with radiation-induced sarcoma, 4 (14.8%) with synchronous NPC with primary brain tumors, 2 (7.4%) with recurrent NPC involving skull base, and 1 (3.7%) with metachronous skull eosinophilic granuloma, respectively. The diagnosis is challenging in specific cases and initial misdiagnoses were found in 6 (22.2%) patients. CONCLUSIONS: For NPC patients with intracranial or skull lesions, the initial diagnosis can be occasionally difficult because of the presence or a history of NPC and related treatment. Unawareness of these entities can result in misdiagnosis and subsequent improper treatment. Neurosurgical interventions are necessary for the diagnosis and treatment for these patients.

Clinical significance of B7-H4 expression in matched non-small cell lung cancer brain metastases and primary tumors.

BACKGROUND: B7-H4, a member of the inhibitory B7 family, is shown to have a profound inhibitory effect on the proliferation, activation, cytokine secretion, and development of cytotoxicity of T cells and may be involved in immune evasion in cancer patients. Although B7-H4 expression has been detected in non-small cell lung cancer (NSCLC), there are no published reports on the expression of B7-H4 in brain metastases from NSCLC. METHODS: We examined the expression of B7-H4 by immunohistochemistry in 49 cases of brain metastatic NSCLC, 18 cases of matched primary NSCLC, and 20 cases of NSCLC patients who had neither brain metastases nor other distant metastases. RESULTS: B7-H4 was highly expressed in 20 (40.8%) out of 49 brain metastases and two (11.1%) out of 18 matched primary tumors. The expression of B7-H4 in brain metastases appeared to be significantly higher than their matched primary tumors (P = 0.016). We also found that patients with high B7-H4 expression in their primary NSCLC have a higher risk of developing brain metastases (P = 0.022). Univariate analyses showed that median overall survival was significantly shorter in patients with high B7-H4 expression in brain metastases (P = 0.002). Multivariate analyses showed that B7-H4 was a significant independent prognostic indicator (P = 0.003). CONCLUSION: NSCLC patients with high B7-H4 expression may benefit from aggressive treatment and close surveillance. Furthermore, our study suggests that B7-H4 may play an important role in the metastatic process of NSCLC and is promising to be a new immune checkpoint molecule for future antitumoral immunotherapy.