Development of End-to-End AI-Based MRI Image Analysis System for Predicting IDH Mutation Status of Patients with Gliomas: Multicentric Validation.

Radiogenomics has shown potential to predict genomic phenotypes from medical images. The development of models using standard-of-care pre-operative MRI images, as opposed to advanced MRI images, enables a broader reach of such models. In this work, a radiogenomics model for IDH mutation status prediction from standard-of-care MRIs in patients with glioma was developed and validated using multicentric data. A cohort of 142 (wild-type: 32.4%) patients with glioma retrieved from the TCIA/TCGA was used to train a logistic regression model to predict the IDH mutation status. The model was evaluated using retrospective data collected in two distinct hospitals, comprising 36 (wild-type: 63.9%) and 53 (wild-type: 75.5%) patients. Model development utilized ROC analysis. Model discrimination and calibration were used for validation. The model yielded an AUC of 0.741 vs. 0.716 vs. 0.938, a sensitivity of 0.784 vs. 0.739 vs. 0.875, and a specificity of 0.657 vs. 0.692 vs. 1.000 on the training, test cohort 1, and test cohort 2, respectively. The assessment of model fairness suggested an unbiased model for age and sex, and calibration tests showed a p < 0.05. These results indicate that the developed model allows the prediction of the IDH mutation status in gliomas using standard-of-care MRI images and does not appear to hold sex and age biases.

Two fractions staged Gammaknife radiosurgery for "large" cerebral metastases.

BACKGROUND: Gammaknife radiosurgery (GKRS) is a valuable option to control cerebral metastases. However, the risk (adverse radiation effect (ARE))-benefit (local control (LC)) ratio switches when the target is too large. OBJECTIVE: In order to balance this ratio, two fractions staged GKRS protocol was conducted for "large" cerebral metastases. The aim of this study is to evaluate the outcome (LC, ARE). METHODS: A total of 39 large cerebral metastases in 35 patients were treated. The initial mean tumor volume was 14.6 cc [6.1; 35.8]. The prescription margin dose was 12 Gy on the 50% isodose line, with 2 weeks between them. A majority of primary cancer were from lung (43%), melanoma (20%) or breast (17%) origin. The mean age was 63 years old (31-89). Mean Graded Prognostic Assessment (GPA) was 2. RESULTS: At the second fraction, mean tumor volume was 10.3 cc [1.9-27.4]. The mean percentage of volume variation for decreasing lesions was 29%. At last follow-up, mean tumor volume was 7.4 cc [0-25.2]; 34 lesions decreased volume (mean 35%). A decreased volume of more than 45% after first stage GKRS was able to predict a long-term local response to staged GKRS treatment. Local control rate at 6 months and 1 year was 87.3% and 75% respectively. The rate of ARE was 7.7%. No predictive factor of local control or ARE was found in a univariate analysis. CONCLUSION: The new 2-fractions-dose-staged GKRS concept seems to be a well-tolerated and effective treatment option for large cerebral metastases.

Texture analysis- and support vector machine-assisted diffusional kurtosis imaging may allow in vivo gliomas grading and IDH-mutation status prediction: a preliminary study.

We sought to investigate, whether texture analysis of diffusional kurtosis imaging (DKI) enhanced by support vector machine (SVM) analysis may provide biomarkers for gliomas staging and detection of the IDH mutation. First-order statistics and texture feature extraction were performed in 37 patients on both conventional (FLAIR) and mean diffusional kurtosis (MDK) images and recursive feature elimination (RFE) methodology based on SVM was employed to select the most discriminative diagnostic biomarkers. The first-order statistics demonstrated significantly lower MDK values in the IDH-mutant tumors. This resulted in 81.1% accuracy (sensitivity = 0.96, specificity = 0.45, AUC 0.59) for IDH mutation diagnosis. There were non-significant differences in average MDK and skewness among the different tumour grades. When texture analysis and SVM were utilized, the grading accuracy achieved by DKI biomarkers was 78.1% (sensitivity 0.77, specificity 0.79, AUC 0.79); the prediction accuracy for IDH mutation reached 83.8% (sensitivity 0.96, specificity 0.55, AUC 0.87). For the IDH mutation task, DKI outperformed significantly the FLAIR imaging. When using selected biomarkers after RFE, the prediction accuracy achieved 83.8% (sensitivity 0.92, specificity 0.64, AUC 0.88). These findings demonstrate the superiority of DKI enhanced by texture analysis and SVM, compared to conventional imaging, for gliomas staging and prediction of IDH mutational status.

The Role of Concomitant Radiation Boost in Neoadjuvant Chemoradiotherapy for Locally Advanced Rectal Cancer.

BACKGROUND/AIM: This study analyzed the impact of concomitant boost on long-term clinical outcomes in locally advanced rectal cancer. PATIENTS AND METHODS: A total of 141 patients (median age=61 years) were treated with neoadjuvant chemoradiotherapy. Median total dose was 50.4 Gy. Forty-three patients received a concomitant boost. Concurrent chemotherapy consisted of 5-fluorouracil (5-FU), given as a 24-h continuous infusion. Mean follow-up was 83.7 months. RESULTS: The 3, 5-, and 10-year overall survival (OS) rates were 91.9%, 84.6%, and 52.9%, respectively. Recurrence-free survival (RFS) rates at 3, 5, and 10 years were 91.4%, 88.9%, and 79.3%, respectively. Metastasis-free survival (MFS) rates at 3, 5, and 10 years were 84.6%, 75.4%, and 49.9%, respectively. Overall, 9.9% of all patients achieved pathological complete response. Down-staging of T- or N-stage was achieved in 55.1% and 41.5% of patients. Multivariate analysis revealed that female sex (p=0.011), concomitant boost-radiotherapy (p=0.014), and the presence of fewer than five positive lymph nodes (p<0.001) were positive predictors of OS. Fewer than five positive lymph nodes was also a positive predictor for RFS (p=0.019). Female gender (p=0.018) and fewer than five positive lymph nodes (p<0.001) were significant predictors for MFS. CONCLUSION: Our data support the efficacy of preoperative treatment for rectal cancer in terms of local outcomes. Intensified radiotherapy using a concomitant boost has a positive effect on OS.

Prognostic value of preoperative dynamic contrast-enhanced MRI perfusion parameters for high-grade glioma patients.

INTRODUCTION: The prognostic value of the dynamic contrast-enhanced (DCE) MRI perfusion and its histogram analysis-derived metrics is not well established for high-grade glioma (HGG) patients. The aim of this prospective study was to investigate DCE perfusion transfer coefficient (Ktrans), vascular plasma volume fraction (vp), extracellular volume fraction (ve), reverse transfer constant (kep), and initial area under gadolinium concentration time curve (IAUGC) as predictors of progression-free (PFS) and overall survival (OS) in HGG patients. METHODS: Sixty-nine patients with suspected anaplastic astrocytoma or glioblastoma underwent preoperative DCE-MRI scans. DCE perfusion whole tumor region histogram parameters, clinical details, and PFS and OS data were obtained. Univariate, multivariate, and Kaplan-Meier survival analyses were conducted. Receiver operating characteristic (ROC) curve analysis was employed to identify perfusion parameters with the best differentiation performance. RESULTS: On univariate analysis, ve and skewness of vp had significant negative impacts, while kep had significant positive impact on OS (P < 0.05). ve was also a negative predictor of PFS (P < 0.05). Patients with lower ve and IAUGC had longer median PFS and OS on Kaplan-Meier analysis (P < 0.05). Ktrans and ve could also differentiate grade III from IV gliomas (area under the curve 0.819 and 0.791, respectively). CONCLUSIONS: High ve is a consistent predictor of worse PFS and OS in HGG glioma patients. vp skewness and kep are also predictive for OS. Ktrans and ve demonstrated the best diagnostic performance for differentiating grade III from IV gliomas.

Minimized doses for linear accelerator radiosurgery of brainstem metastasis.

PURPOSE: Treatment of cerebral metastases located inside the brainstem remains a challenge, as the brainstem is considered to be a neurological organ at risk, whatever the treatment strategy. We report a retrospective study of 30 consecutive patients treated in our institution between 2005 and 2007 with micromultileaf linear accelerator (LINAC)-radiosurgery for brainstem metastases, with reduced doses compared to those usually reported in the literature. METHODS AND MATERIALS: Mean follow-up was 311 days (range, 41-1351). Median age was 57 years (range, 37-82), Mean Karnofsky Index (KI) was 80. Primary tumor site was lung (n = 13), breast (n = 4), kidney (n = 4), skin (melanoma; n = 3), and others (n = 6). Primary tumor was controlled in 17 cases; extracranial metastases were controlled in 12 cases. Mean number of metastases was 1.46 (one to three); median volume was 2.82 cc (0.06-18). Dose was delivered by a micromultileaf collimator 6-MV LINAC . RESULTS: Dose administered at the 70% isodose was 13.4 Gy (range, 8.2-15). Median survival was 10 months. Local control rates at 3, 6, and 12 months were 100%, 100%, and 79% respectively. Median neurological control duration was 5 months. Neurological control rates at 3, 6, and 12 months were 73%, 42%, and 25%, respectively. No parameter was found to significantly correlate with survival, local, or cerebral control. No patients had severe side effects (Grade III-IV), according to the Radiation Therapy Oncology Group (RTOG) scale. CONCLUSION: Lower doses than previously reported can achieve the same local control and survival rates in brain metastases, with minimal side effects.

Combined proton and photon conformal radiotherapy for intracranial atypical and malignant meningioma.

PURPOSE: To evaluate retrospectively the efficacy of conformal fractionated radiotherapy combining proton and photon beams after primary surgery for treatment of atypical and malignant meningiomas. PATIENTS AND METHODS: Between September 1999 and October 2006, 24 patients (12 male, 12 female) with histopathologically proven meningioma (atypical 19, malignant 5) received postoperative combined radiotherapy with a 201-MeV proton beam at the Centre Protontherapie d'Orsay and a high-energy photon beam. Six patients underwent gross total resection and 18 a subtotal resection. Median gross tumor volume and clinical target volume were 44.7 cm(3) and 153.3 cm(3), respectively. Mean total irradiation dose was 65.01 CGE (cobalt gray equivalent), with a mean proton total dose of 34.05 CGE and a mean photon total dose 30.96 CGE. RESULTS: The median (range) follow-up interval was 32.2 (1-72) months. The overall mean local relapse-free interval was 27.2 (10-50) months, 28.3 (10-50) months for atypical meningioma and 23 (13-33) months for malignant meningioma. Ten tumors recurred locally. One-, 2-, 3-, 4-, 5-, and 8- year local control rates for the entire group of patients were 82.9% +/- 7.8%, 82.9% +/- 7.8%, 61.3% +/- 11%, 61.3% +/- 11%, 46.7% +/- 12.3%, and 46.7% +/- 12.3%, respectively. One-, 2-, 3-, 4-, 5-, and 8- year overall survival rates were 100%, 95.5% +/- 4.4%, 80.4% +/- 8.8%, 65.3% +/- 10.6%, 53.2% +/- 11.6%, and 42.6% +/- 13%, respectively. Survival was significantly associated with total dose. There was no acute morbidity of radiotherapy. One patient developed radiation necrosis 16 months after treatment. CONCLUSIONS: Postoperative combination of conformal radiotherapy with protons and photons for atypical and malignant meningiomas is a well-tolerated treatment producing long-term tumor stabilization.

Pre-operative chemoradiotherapy with oral capecitabine in locally advanced, resectable rectal cancer.

PURPOSE: The aim of the study was to evaluate the efficacy and tolerance of pre-operative chemoradiotherapy with oral capecitabine in Greek patients with locally advanced, resectable rectal cancer. MATERIALS AND METHODS: Thirty patients, 16 men and 14 women with a median age of 58 years (range, 21-75 years), with potentially resectable T3NO (30%), T3N1 (53%) and T4NO-1 (17%) rectal cancer, were treated with capecitabine (825 mg/m(2), twice daily for 7 days/week) and concomitant radiotherapy (50.4 Gy/28 fractions) for 5.5 weeks. Patients underwent surgery with total mesorectal excision 4-6 weeks later followed by 4-months of post-operative treatment with capecitabine. The primary end-point was to determine the clinical and pathological response, safety profile, preservation of the sphincter mechanism and rate of peri-operative complications. RESULTS: The median distance of rectal tumors from the anal verge was 7 cm. All patients had curative resection. Downstaging rate was 84% (25/30) on endorectal ultrasonography and 75% (22/30) on pathology findings. Pathological complete response rate was 23% (7/30). No patient had grade 4 toxicity. Grade 3 toxicity occurred in 3 patients (10%) and consisted mainly of leucopenia (6%) and hand-foot syndrome (4%). Mild or moderate toxicity was frequent, but always reversible. Twenty-four patients (80%) received sphincter-preserving surgical procedures. Peni-operative complications were seen in 6 (20%) patients and included mechanical ileus (3%), delayed wound healing (7%), wound infection (7%) and anastomotic leakage (3%). CONCLUSION: Pre-operative chemoradiotherapy with oral capecitabine in locally advanced, resectable rectal cancer achieves significant rates of tumor downstaging and sphincter preservation with a favorable safety profile.