Quantitative analysis of permeability for glioma grading using dynamic contrast-enhanced magnetic resonance imaging

Imaging biomarkers for clinical applications in neuro-oncology: current status and future perspectives

Biomarker discovery and development are popular for detecting the subtle diseases. However, biomarkers are needed to be validated and approved, and even fewer are ever used clinically. Imaging biomarkers have a crucial role in the treatment of cancer patients because they provide objective information on tumor biology, the tumor's habitat, and the tumor's signature in the environment. Tumor changes in response to an intervention complement molecular and genomic translational diagnosis as well as quantitative information. Neuro-oncology has become more prominent in diagnostics and targeted therapies. The classification of tumors has been actively updated, and drug discovery, and delivery in nanoimmunotherapies are advancing in the field of target therapy research. It is important that biomarkers and diagnostic implements be developed and used to assess the prognosis or late effects of long-term survivors. An improved realization of cancer biology has transformed its management with an increasing emphasis on a personalized approach in precision medicine. In the first part, we discuss the biomarker categories in relation to the courses of a disease and specific clinical contexts, including that patients and specimens should both directly reflect the target population and intended use. In the second part, we present the CT perfusion approach that provides quantitative and qualitative data that has been successfully applied to the clinical diagnosis, treatment and application. Furthermore, the novel and promising multiparametric MR imageing approach will provide deeper insights regarding the tumor microenvironment in the immune response. Additionally, we briefly remark new tactics based on MRI and PET for converging on imaging biomarkers combined with applications of bioinformatics in artificial intelligence. In the third part, we briefly address new approaches based on theranostics in precision medicine. These sophisticated techniques merge achievable standardizations into an applicatory apparatus for primarily a diagnostic implementation and tracking radioactive drugs to identify and to deliver therapies in an individualized medicine paradigm. In this article, we describe the critical principles for imaging biomarker characterization and discuss the current status of CT, MRI and PET in finiding imaging biomarkers of early disease.

Peritumor Edema Serves as an Independent Predictive Factor of Recurrence Patterns and Recurrence-Free Survival for High-Grade Glioma

Objective. This study is aimed at analyzing the factors affecting the recurrence patterns and recurrence-free survival (RFS) of high-grade gliomas (HGG). Methods. Eligible patients admitted to the Affiliated Hospital of Xuzhou Medical University were selected. Subsequently, the effects of some clinical data including age, gender, WHO pathological grades, tumor site, tumor size, clinical treatments, and peritumoral edema (PTE) area and molecular markers (Ki-67, MGMT, IDH-1, and p53) on HGG patients’ recurrence patterns and RFS were analyzed. Results. A total number of 77 patients were enrolled into this study. After analyzing all the cases, it was determined that tumor size and tumor site had a significant influence on the recurrent patterns of HGG, and PTE was an independent predict factor of recurrence patterns. Specifically, when the PTE was mild (<1 cm), the recurrence pattern tended to be local; in contrast, HGG was more likely to progress to marginal recurrence and distant recurrence. Furthermore, age and PTE were significantly associated with RFS; the median RFS of the population with $\text{PTE}<1\text{cm}$ (23.60 months) was obviously longer than the population with $\text{PTE}\ge 1\text{cm}$ (5.00 months). Conclusions. PTE is an independent predictor of recurrence patterns and RFS for HGG. Therefore, preoperative identification of PTE in HGG patients is crucially important, which is helpful to accurately estimate the recurrence pattern and RFS.

The Value of Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) in the Differentiation of Pseudoprogression and Recurrence of Intracranial Gliomas

Objective

The objective of this study was to determine the value of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in assessing postoperative changes in intracranial gliomas.

Method

A total of fifty-one patients who had new enhanced lesions after surgical resection followed by standard radiotherapy and chemotherapy were collected retrospectively from October 2014 to June 2021. The patients were divided into a pseudoprogression group (15 cases) and a recurrence group (36 cases) according to the pathological results of the second operation or a follow-up of more than six months. The follow-up data of all patients were complete, and DCE-MRI was performed. The images were processed to obtain the quantitative parameters K^trans, Ve, and Kep and the semiquantitative parameter iAUC, which were analysed with relevant statistical software.

Results

First, the difference in K^trans and iAUC values between the two groups was statistically significant (P < 0.05), and the difference in Ve and Kep values was not statistically significant (P > 0.05). Second, by comparing the area under the curve, threshold, sensitivity and specificity of K^trans, and iAUC, it was found that the iAUC threshold value was slightly higher than that of K^trans, and the specificity of K^trans was equal to that of iAUC, while the area under the curve and sensitivity of K^trans were higher than those of iAUC. Third, K^trans and iAUC had high accuracy in diagnosing recurrence and pseudoprogression, and K^trans had higher accuracy than iAUC.

Conclusion

In this study, DCE-MRI has a certain diagnostic value in the early differentiation of recurrence and pseudoprogression, offering a new method for the diagnosis and assessment of gliomas after surgery.

Dynamic Contrast-Enhanced MR Perfusion in Differentiation of Benign and Malignant Brain Lesions

Background- We aimed to differentiate Glioblastoma Multiforme (GBM) from benign lesions like Developmental Venous Anomaly (DVA) and Cavernous Malformation (CM) by Dynamic Contrast-Enhanced MR Perfusion (DCE-MRP) markers such as Ktrans, Ve, Kep, and IAUC. Method-We retrospectively evaluated 20 patients; 10 GBM as the malignant group, 5 CM and 5 DVA as the benign group. Ktrans, Kep, Ve, and IAUC parameters were measured by DCE-MRP, within the lesion, at perilesional nonenhancing white matter (PLWM) and at contralateral normal appearing white matter (CLWM). All benign and malignant lesions exhibited significantly increased Ktrans, Ve, and IAUC values compared to PLWM and CLWM (p < 0.001, p=0.006 and p<0.001). Result-Subtracted Kep values between lesion and PLWM were significantly different between the benign and malignant groups, as the malignant group exhibited higher subtracted Kep values (p 0.035). For the malignant group; Ktrans and IAUC values at the lesion were positively correlated (r 0.911), while Kep and Ve at CLWM were negatively and strongly correlated (r 0.798). For the benign group; Ktrans with Ve and Ktrans with IAUC lesion (r 0.708 and r 0.816 respectively), Ktrans and IAUC at PLWM (r 0.809), Ktrans and IAUC at CLWM(r 0.798) were strongly and positively correlated. Ktrans, Ve, and IAUC values can be used to demarcate the lesion in both groups. Conclusion- Ktrans strongly correlates with IAUC and they can be used instead of each other in both benign and malignant lesions. DCE-MRP cannot be used in the differentiation of malignant lesions from benign vascular lesions. However, subtracted Kep values can be used to differentiate GBM from benign vascular lesions.

Nomograms for predicting the overall survival of patients with cerebellar glioma: an analysis of the surveillance epidemiology and end results (SEER) database

At present, our understanding of cerebellar glioma is still insufficient. This study collected information on patients in the SEER database to identify the predictive factors for patients with cerebellar glioma. Data from patients with cerebellar glioma diagnosed from 1975 to 2018 were retrieved from the Surveillance Epidemiology and End Results Database. We randomly divided the patients into a training group and a validation group, established a nomogram based on the training group, and used the validation group data to verify the clinical value of the model. A total of 508 patients were included in this study. Multivariate analysis was performed based on the data before randomization, and the results showed that the patient's age, WHO grade, histological type, and extent were significantly correlated with the survival rate. The C-index of the OS nomograms of the training cohort was 0.909 (95% CI, (0.880–0.938)) and 0.932 (95% CI, (0.889–0.975)) in the validation group. The calibration curve of OS for 3 and 5 years showed that there was good consistency between the actual survival probability and the predicted survival probability. For patients with cerebellar glioma, the age at diagnosis, WHO grade of the glioma, histological type, and extension are the four factors that most strongly affect the overall survival outcomes. Furthermore, our model may be a useful tool for predicting OS in these patients.

Diagnostic Role of Diffusion-Weighted and Dynamic Contrast-Enhanced Perfusion MR Imaging in Paragangliomas and Schwannomas in the Head and Neck

BACKGROUND AND PURPOSE

Distinguishing schwannomas from paragangliomas in the head and neck and determining succinate dehydrogenase (SDH) mutation status in paragangliomas are clinically important. We aimed to assess the clinical usefulness of DWI and dynamic contrast-enhanced MR imaging in differentiating these 2 types of tumors, as well as the SDH mutation status of paragangliomas.

MATERIALS AND METHODS

This retrospective study from June 2016 to June 2020 included 42 patients with 15 schwannomas and 27 paragangliomas (10 SDH mutation-positive and 17 SDH mutation-negative). ADC values, dynamic contrast-enhanced MRI parameters, and tumor imaging characteristics were compared between the 2 tumors and between the mutation statuses of paragangliomas as appropriate. Multivariate stepwise logistic regression analysis was performed to identify significant differences in these parameters.

RESULTS

Fractional plasma volume (P ≤ .001), rate transfer constant (P = .038), time-to-maximum enhancement (P < .001), maximum signal-enhancement ratio (P < .001) and maximum concentration of contrast agent (P < .001), velocity of enhancement (P = .002), and tumor characteristics including the presence of flow voids (P = .001) and enhancement patterns (P = .027) showed significant differences between schwannomas and paragangliomas, though there was no significant difference in ADC values. In the multivariate logistic regression analysis, fractional plasma volume was identified as the most significant value for differentiation of the 2 tumor types (P = .014). ADC values were significantly higher in nonhereditary than in hereditary paragangliomas, while there was no difference in dynamic contrast-enhanced MR imaging parameters.

CONCLUSIONS

Dynamic contrast-enhanced MR imaging parameters show promise in differentiating head and neck schwannomas and paragangliomas, while DWI can be useful in detecting SDH mutation status in paragangliomas.

Prognostic Nomograms for Primary High-Grade Glioma Patients in Adult: A Retrospective Study Based on the SEER Database

Purpose

In our study, we aimed to screen the risk factors that affect overall survival (OS) and cancer-specific survival (CSS) in adult glioma patients and to develop and evaluate nomograms.

Methods

Primary high-grade gliomas patients being retrieved from the surveillance, epidemiology and end results (SEER) database, between 2004 and 2015, then they randomly assigned to a training group and a validation group. Univariate and multivariate Cox analysis models were used to choose the variables significantly correlated with the prognosis of high-grade glioma patients. And these variables were used to construct the nomograms. Next, concordance index (C-index), calibration plot and receiver operating characteristics (ROCs) curve were used to evaluate the accuracy of the nomogram model. In addition, the decision curve analysis (DCA) was used to analyze the benefit of nomogram and prognostic indicators commonly used in clinical practice.

Results

A total of 6395 confirmed glioma patients were selected from the SEER database, divided into training set (n =3166) and validation set (n =3229). Age at diagnosis, tumor grade, tumor size, histological type, surgical type, radiotherapy and chemotherapy were screened out by Cox analysis model. For OS nomogram, the C-index of the training set was 0.741 (95% CI: 0.751-0.731), and the validation set was 0.738 (95% CI: 0.748-0.728). For CSS nomogram, the C-index of the training set was 0.739 (95% CI: 0.749-0.729), and the validation set was 0.738 (95% CI: 0.748-0.728). The net benefit and net reduction in inverventions of nomograms in the decision curve analysis (DCA) was higher than histological type.

Conclusions

We developed nomograms to predict 3- and 5-year OS rates and 3- and 5-year CSS rates in adult high-grade glioma patients. Both the training set and the validation set showed good calibration and validation, indicating the clinical applicability of the nomogram and good predictive results.

Noninvasive Prediction of TERT Promoter Mutations in High-Grade Glioma by Radiomics Analysis Based on Multiparameter MRI

Objectives

To investigate the predictors of telomerase reverse transcriptase (TERT) promoter mutations in adults suffered from high-grade glioma (HGG) through radiomics analysis, develop a noninvasive approach to evaluate TERT promoter mutations.

Methods

126 adult patients with HGG (88 in the training cohort and 38 in the validation cohort) were retrospectively enrolled. Totally 5064 radiomics features were, respectively, extracted from three VOIs (necrosis, enhanced, and edema) in MRI. Firstly, an optimal radiomics signature (Radscore) was established based on LASSO regression. Secondly, univariate and multivariate logistic regression analyses were performed to investigate important potential variables as predictors of TERT promoter mutations. Besides, multiparameter models were established and evaluated. Eventually, an optimal model was visualized as radiomics nomogram for clinical evaluations.

Results

6 radiomics features were selected to build Radscore signature through LASSO regression. Among them, 5 were from necrotic VOIs and 1 was from enhanced ones. With univariate and multivariate analysis, necrotic volume percentages of core (CNV), Age, Cho/Cr, Lac, and Radscore were significantly higher in TERTm than in TERTw (p < 0.05). 4 models were built in our study. Compared with Model B (Age, Cho/Cr, Lac, and Radscore), Model A (Age, Cho/Cr, Lac, Radscore, and CNV) has a larger AUC in both training (0.955 vs. 0.917, p = 0.049) and validation (0.889 vs. 0.868, p = 0.039) cohorts. It also has higher performances in net reclassification improvement (NRI), integrated discrimination improvement (IDI), and decision curve analysis (DCA) evaluation. Conclusively, Model A was visualized as a radiomics nomogram. Calibration curve shows a good agreement between estimated and actual probabilities.

Conclusions

Age, Cho/Cr, Lac, CNV, and Radscore are important indicators for TERT promoter mutation predictions in HGG. Tumor necrosis seems to be closely related to TERT promoter mutations. Radiomics nomogram based on multiparameter MRI and CNV has higher prediction accuracies.

Guiding the first biopsy in glioma patients using estimated Ki-67 maps derived from MRI: conventional versus advanced imaging

BACKGROUND

Undersampling of gliomas at first biopsy is a major clinical problem, as accurate grading determines all subsequent treatment. We submit a technological solution to reduce the problem of undersampling by estimating a marker of tumor proliferation (Ki-67) using MR imaging data as inputs, against a stereotactic histopathology gold standard.

METHODS

MR imaging was performed with anatomic, diffusion, permeability, and perfusion sequences, in untreated glioma patients in a prospective clinical trial. Stereotactic biopsies were harvested from each patient immediately prior to surgical resection. For each biopsy, an imaging description (23 parameters) was developed, and the Ki-67 index was recorded. Machine learning models were built to estimate Ki-67 from imaging inputs, and cross validation was undertaken to determine the error in estimates. The best model was used to generate graphical maps of Ki-67 estimates across the whole brain.

RESULTS

Fifty-two image-guided biopsies were collected from 23 evaluable patients. The random forest algorithm best modeled Ki-67 with 4 imaging inputs (T2-weighted, fractional anisotropy, cerebral blood flow, Ktrans). It predicted the Ki-67 expression levels with a root mean square (RMS) error of 3.5% (R2 = 0.75). A less accurate predictive result (RMS error 5.4%, R2 = 0.50) was found using conventional imaging only.

CONCLUSION

Ki-67 can be predicted to clinically useful accuracies using clinical imaging data. Advanced imaging (diffusion, perfusion, and permeability) improves predictive accuracy over conventional imaging alone. Ki-67 predictions, displayed as graphical maps, could be used to guide biopsy, resection, and/or radiation in the care of glioma patients.

Quantitative pharmacokinetic analysis of high-temporal-resolution dynamic contrast-enhanced MRI to differentiate the normal-appearing pituitary gland from pituitary macroadenoma

PURPOSE

To evaluate the usefulness of high-temporal-resolution dynamic contrast-enhanced (DCE) MRI and quantitative pharmacokinetic analysis to differentiate the normal-appearing pituitary gland from a pituitary macroadenoma.

MATERIALS AND METHODS

Twenty-seven patients with macroadenomas underwent preoperative DCE-MRI with a temporal resolution of 5 s using compressed sensing to obtain pharmacokinetic parameters. Two independent observers localized the normal-appearing pituitary gland on post-contrast T1-weighted images before and after referring to the corresponding K^trans maps. Agreements between the localizations and intraoperative findings were evaluated using the kappa statistics. The Mann-Whitney U test was used to compare the pharmacokinetic parameters of the normal-appearing pituitary gland and adenoma.

RESULTS

For both observers, the agreement between the MRI-based localization and the intraoperative findings increased after referring to the K^trans maps (observer 1, 0.930-1; observer 2, 0.636-0.855). The normal-appearing pituitary gland had significantly higher K^trans [/min] (1.50 ± 0.80 vs 0.58 ± 0.49, P < 0.0001), k_ep [/min] (3.19 ± 1.29 vs 2.15 ± 1.18, P = 0.0049), and v_e (0.43 ± 0.15 vs 0.25 ± 0.17, P = 0.0003) than adenoma.

CONCLUSION

High-temporal-resolution DCE-MRI and quantitative pharmacokinetic analysis help accurately localize the normal-appearing pituitary gland in patients with macroadenomas. The normal-appearing pituitary gland was characterized by higher K^trans, k_ep, and v_e than macroadenoma. Dynamic contrast-enhanced MRI with high-temporal-resolution using compressed sensing was used for quantitative pharmacokinetic analysis of pituitary macroadenomas. An observer study, the use of K^trans maps improved accuracy in localizing the normal-appearing pituitary gland. As compared to an adenoma, the normal-appearing pituitary gland had significantly higher K^trans, k_ep, and v_e values.

Imaging-Based Algorithm for the Local Grading of Glioma

BACKGROUND AND PURPOSE

Gliomas are highly heterogeneous tumors, and optimal treatment depends on identifying and locating the highest grade disease present. Imaging techniques for doing so are generally not validated against the histopathologic criterion standard. The purpose of this work was to estimate the local glioma grade using a machine learning model trained on preoperative image data and spatially specific tumor samples. The value of imaging in patients with brain tumor can be enhanced if pathologic data can be estimated from imaging input using predictive models.

MATERIALS AND METHODS

Patients with gliomas were enrolled in a prospective clinical imaging trial between 2013 and 2016. MR imaging was performed with anatomic, diffusion, permeability, and perfusion sequences, followed by image-guided stereotactic biopsy before resection. An imaging description was developed for each biopsy, and multiclass machine learning models were built to predict the World Health Organization grade. Models were assessed on classification accuracy, Cohen κ, precision, and recall.

RESULTS

Twenty-three patients (with 7/9/7 grade II/III/IV gliomas) had analyzable imaging-pathologic pairs, yielding 52 biopsy sites. The random forest method was the best algorithm tested. Tumor grade was predicted at 96% accuracy (κ = 0.93) using 4 inputs (T2, ADC, CBV, and transfer constant from dynamic contrast-enhanced imaging). By means of the conventional imaging only, the overall accuracy decreased (89% overall, κ = 0.79) and 43% of high-grade samples were misclassified as lower-grade disease.

CONCLUSIONS

We found that local pathologic grade can be predicted with a high accuracy using clinical imaging data. Advanced imaging data improved this accuracy, adding value to conventional imaging. Confirmatory imaging trials are justified.

Glioma Grading on Conventional MR Images: A Deep Learning Study With Transfer Learning

Background: Accurate glioma grading before surgery is of the utmost importance in treatment planning and prognosis prediction. But previous studies on magnetic resonance imaging (MRI) images were not effective enough. According to the remarkable performance of convolutional neural network (CNN) in medical domain, we hypothesized that a deep learning algorithm can achieve high accuracy in distinguishing the World Health Organization (WHO) low grade and high grade gliomas.

Methods: One hundred and thirteen glioma patients were retrospectively included. Tumor images were segmented with a rectangular region of interest (ROI), which contained about 80% of the tumor. Then, 20% data were randomly selected and leaved out at patient-level as test dataset. AlexNet and GoogLeNet were both trained from scratch and fine-tuned from models that pre-trained on the large scale natural image database, ImageNet, to magnetic resonance images. The classification task was evaluated with five-fold cross-validation (CV) on patient-level split.

Results: The performance measures, including validation accuracy, test accuracy and test area under curve (AUC), averaged from five-fold CV of GoogLeNet which trained from scratch were 0.867, 0.909, and 0.939, respectively. With transfer learning and fine-tuning, better performances were obtained for both AlexNet and GoogLeNet, especially for AlexNet. Meanwhile, GoogLeNet performed better than AlexNet no matter trained from scratch or learned from pre-trained model.

Conclusion: In conclusion, we demonstrated that the application of CNN, especially trained with transfer learning and fine-tuning, to preoperative glioma grading improves the performance, compared with either the performance of traditional machine learning method based on hand-crafted features, or even the CNNs trained from scratch.

MRI analysis to map interstitial flow in the brain tumor microenvironment

Glioblastoma (GBM), a highly aggressive form of brain tumor, is a disease marked by extensive invasion into the surrounding brain. Interstitial fluid flow (IFF), or the movement of fluid within the spaces between cells, has been linked to increased invasion of GBM cells. Better characterization of IFF could elucidate underlying mechanisms driving this invasion in vivo. Here, we develop a technique to noninvasively measure interstitial flow velocities in the glioma microenvironment of mice using dynamic contrast-enhanced magnetic resonance imaging (MRI), a common clinical technique. Using our in vitro model as a phantom "tumor" system and in silico models of velocity vector fields, we show we can measure average velocities and accurately reconstruct velocity directions. With our combined MR and analysis method, we show that velocity magnitudes are similar across four human GBM cell line xenograft models and the direction of fluid flow is heterogeneous within and around the tumors, and not always in the outward direction. These values were not linked to the tumor size. Finally, we compare our flow velocity magnitudes and the direction of flow to a classical marker of vessel leakage and bulk fluid drainage, Evans blue. With these data, we validate its use as a marker of high and low IFF rates and IFF in the outward direction from the tumor border in implanted glioma models. These methods show, for the first time, the nature of interstitial fluid flow in models of glioma using a technique that is translatable to clinical and preclinical models currently using contrast-enhanced MRI.