Radio-Pathomic Maps of Cell Density Identify Brain Tumor Invasion beyond Traditional MRI-Defined Margins

Radiomics in glioma: emerging trends and challenges

Radiomics is a promising neuroimaging technique for extracting and analyzing quantitative glioma features. This review discusses the application, emerging trends, and challenges associated with using radiomics in glioma. Integrating deep learning algorithms enhances various radiomics components, including image normalization, region of interest segmentation, feature extraction, feature selection, and model construction and can potentially improve model accuracy and performance. Moreover, investigating specific tumor habitats of glioblastomas aids in a better understanding of glioblastoma aggressiveness and the development of effective treatment strategies. Additionally, advanced imaging techniques, such as diffusion-weighted imaging, perfusion-weighted imaging, magnetic resonance spectroscopy, magnetic resonance fingerprinting, functional MRI, and positron emission tomography, can provide supplementary information for tumor characterization and classification. Furthermore, radiomics analysis helps understand the glioma immune microenvironment by predicting immune-related biomarkers and characterizing immune responses within tumors. Integrating multi-omics data, such as genomics, transcriptomics, proteomics, and pathomics, with radiomics, aids the understanding of the biological significance of the underlying radiomics features and improves the prediction of genetic mutations, prognosis, and treatment response in patients with glioma. Addressing challenges, such as model reproducibility, model generalizability, model interpretability, and multi-omics data integration, is crucial for the clinical translation of radiomics in glioma.

Supramarginal Resection of Glioblastoma: A Review

This article discusses the evidence supporting the resection of glioblastoma beyond the borders of contrast-enhancing tumor. While several techniques for this have been described, including a so-called FLAIRectomy, lobectomy, or via the use of adjuncts such as fluorescence or intraoperative MRI, the optimal extent of additional resection has yet to be established. Many authors have noted a survival benefit with supramarginal resection without significant additional morbidity.

Radio-pathomic estimates of cellular growth kinetics predict survival in recurrent glioblastoma

Aim: A radio-pathomic machine learning (ML) model has been developed to estimate tumor cell density, cytoplasm density (Cyt) and extracellular fluid density (ECF) from multimodal MR images and autopsy pathology. In this multicenter study, we implemented this model to test its ability to predict survival in patients with recurrent glioblastoma (rGBM) treated with chemotherapy.Methods: Pre- and post-contrast T1-weighted, FLAIR and ADC images were used to generate radio-pathomic maps for 51 patients with longitudinal pre- and post-treatment scans. Univariate and multivariate Cox regression analyses were used to test the influence of contrast-enhancing tumor volume, total cellularity, mean Cyt and mean ECF at baseline, immediately post-treatment and the pre- and post-treatment rate of change in volume and cellularity on overall survival (OS).Results: Smaller Cyt and larger ECF after treatment were significant predictors of OS, independent of tumor volume and other clinical prognostic factors (HR = 3.23 × 10-6, p < 0.001 and HR = 2.39 × 105, p < 0.001, respectively). Both post-treatment volumetric growth rate and the rate of change in cellularity were significantly correlated with OS (HR = 1.17, p = 0.003 and HR = 1.14, p = 0.01, respectively).Conclusion: Changes in histological characteristics estimated from a radio-pathomic ML model are a promising tool for evaluating treatment response and predicting outcome in rGBM.

Exploratory Analysis of Radiomics and Pathomics in Uterine Corpus Endometrial Carcinoma

Uterine corpus endometrial carcinoma (EC) is one of the most common malignancies in the female reproductive system, characterized by tumor heterogeneity at both radiological and pathological scales. Both radiomics and pathomics have the potential to assess this heterogeneity and support EC diagnosis. This study examines the correlation between radiomics features from Apparent Diffusion Coefficient (ADC) maps and post-contrast T1 (T1C) images with pathomic features from pathology images in 32 patients from the CPTAC-UCEC database. 91 radiomics features were extracted from ADC maps and T1C images, and 566 pathomic features from cell detections and cell density maps at four different resolutions. Spearman's correlation and Bayes Factor analysis were used to evaluate radio-pathomic correlations. Significant cross-scale correlations were found, with strengths ranging from 0.57 to 0.89 in absolute value (9.47 × 104 < BF < 4.77 × 1014) for the ADC task, and from 0.64 and 0.70 (1.80 × 104 < BF < 5.69 × 105) for the T1C task. Most significant and high cross-scale associations were observed between ADC textural features and features from cell density maps. Correlations involving morphometric features and ADC and T1C first-order features were also observed, reflecting variations in tumor aggressiveness and tissue composition. These findings suggest that correlating radiomic features from ADC and T1C features with histopathological features can enhance understanding of EC intratumoral heterogeneity.

Non-invasive staging of endometrial fibrosis using diffusion-weighted imaging: a feasibility study

RESEARCH QUESTION

Can diffusion-weighted imaging (DWI) be used to stage endometrial fibrosis?

DESIGN

This prospective study included 41 healthy women, 30 patients with mild to moderate endometrial fibrosis (MMEF) and 102 patients with severe endometrial fibrosis (SEF). Endometrial thickness and DWI-related parameters, specifically the mean intensity (ADCROI and ADCVOI) and standard deviation (ADC-SDROI and ADC-SDVOI) of the endometrial apparent diffusion coefficient (ADC) values within a region of interest (ROI) on a mid-sagittal ADC map, and volume of interest (VOI) of the entire endometrium in the corpus uteri, were measured and analysed.

RESULTS

Endometrial thickness, endometrial ADCVOI and ADCROI were significantly higher in healthy women (11.7 mm, 1.31  ×  10-3 mm2/s and 1.36  ×  10-3 mm2/s) than MMEF patients (7.5 mm, P < 0.001; 1.23  ×  10-3 mm2/s, P = 0.001; and 1.26  ×  10-3 mm2/s, P = 0.003) and SEF patients (6.2 mm, 1.15  ×  10-3 mm2/s and 1.23  ×  10-3 mm2/s; all P < 0.001). Endometrial ADC-SDVOI and ADC-SDROI were significantly higher in SEF patients (0.24 and 0.24) than healthy women (0.16 and 0.13) and MMEF patients (0.18 and 0.16) (all P < 0.001). The degree of endometrial fibrosis was negatively correlated with endometrial thickness, endometrial ADCVOI and ADCROI (Spearman's rho = -0.662, -0.526 and -0.349) and positively correlated with endometrial ADC-SDVOI and ADC-SDROI (Spearman's rho 0.729 and 0.713) (all P < 0.001). Compared with endometrial thickness measurements, endometrial magnetic resonance imaging parameters, particularly DWI-related parameters, demonstrated excellent accuracy in distinguishing normal endometrium, MMEF and SEF (areas under the curve >0.800).

CONCLUSIONS

DWI is particularly effective for accurately staging the microstructural changes associated with endometrial fibrosis.

Noninvasive Autopsy-Validated Tumor Probability Maps Identify Glioma Invasion Beyond Contrast Enhancement

BACKGROUND AND OBJECTIVES

This study identified a clinically significant subset of patients with glioma with tumor outside of contrast enhancement present at autopsy and subsequently developed a method for detecting nonenhancing tumor using radio-pathomic mapping. We tested the hypothesis that autopsy-based radio-pathomic tumor probability maps would be able to noninvasively identify areas of infiltrative tumor beyond traditional imaging signatures.

METHODS

A total of 159 tissue samples from 65 subjects were aligned to MRI acquired nearest to death for this retrospective study. Demographic and survival characteristics for patients with and without tumor beyond the contrast-enhancing margin were computed. An ensemble algorithm was used to predict pixelwise tumor presence from pathological annotations using segmented cellularity (Cell), extracellular fluid, and cytoplasm density as input (6 train/3 test subjects). A second level of ensemble algorithms was used to predict voxelwise Cell, extracellular fluid, and cytoplasm on the full data set (43 train/22 test subjects) using 5-by-5 voxel tiles from T1, T1 + C, fluid-attenuated inversion recovery, and apparent diffusion coefficient as input. The models were then combined to generate noninvasive whole brain maps of tumor probability.

RESULTS

Tumor outside of contrast was identified in 41.5% of patients, who showed worse survival outcomes (hazard ratio = 3.90, P < .001). Tumor probability maps reliably tracked nonenhancing tumor on a range of local and external unseen data, identifying tumor outside of contrast in 69% of presurgical cases that also showed reduced survival outcomes (hazard ratio = 1.67, P = .027).

CONCLUSION

This study developed a multistage model for mapping gliomas using autopsy tissue samples as ground truth, which was able to identify regions of tumor beyond traditional imaging signatures.

Simultaneous quantification of perfusion, permeability, and leakage effects in brain gliomas using dynamic spin-and-gradient-echo echoplanar imaging MRI

OBJECTIVE

To determine the feasibility and biologic correlations of dynamic susceptibility contrast (DSC), dynamic contrast enhanced (DCE), and quantitative maps derived from contrast leakage effects obtained simultaneously in gliomas using dynamic spin-and-gradient-echo echoplanar imaging (dynamic SAGE-EPI) during a single contrast injection.

MATERIALS AND METHODS

Thirty-eight patients with enhancing brain gliomas were prospectively imaged with dynamic SAGE-EPI, which was processed to compute traditional DSC metrics (normalized relative cerebral blood flow [nrCBV], percentage of signal recovery [PSR]), DCE metrics (volume transfer constant [Ktrans], extravascular compartment [ve]), and leakage effect metrics: ΔR2,ss* (reflecting T2*-leakage effects), ΔR1,ss (reflecting T1-leakage effects), and the transverse relaxivity at tracer equilibrium (TRATE, reflecting the balance between ΔR2,ss* and ΔR1,ss). These metrics were compared between patient subgroups (treatment-naïve [TN] vs recurrent [R]) and biological features (IDH status, Ki67 expression).

RESULTS

In IDH wild-type gliomas (IDHwt-i.e., glioblastomas), previous exposure to treatment determined lower TRATE (p = 0.002), as well as higher PSR (p = 0.006), Ktrans (p = 0.17), ΔR1,ss (p = 0.035), ve (p = 0.006), and ADC (p = 0.016). In IDH-mutant gliomas (IDHm), previous treatment determined higher Ktrans and ΔR1,ss (p = 0.026). In TN-gliomas, dynamic SAGE-EPI metrics tended to be influenced by IDH status (p ranging 0.09-0.14). TRATE values above 142 mM-1s-1 were exclusively seen in TN-IDHwt, and, in TN-gliomas, this cutoff had 89% sensitivity and 80% specificity as a predictor of Ki67 > 10%.

CONCLUSIONS

Dynamic SAGE-EPI enables simultaneous quantification of brain tumor perfusion and permeability, as well as mapping of novel metrics related to cytoarchitecture (TRATE) and blood-brain barrier disruption (ΔR1,ss), with a single contrast injection.

CLINICAL RELEVANCE STATEMENT

Simultaneous DSC and DCE analysis with dynamic SAGE-EPI reduces scanning time and contrast dose, respectively alleviating concerns about imaging protocol length and gadolinium adverse effects and accumulation, while providing novel leakage effect metrics reflecting blood-brain barrier disruption and tumor tissue cytoarchitecture.

KEY POINTS

• Traditionally, perfusion and permeability imaging for brain tumors requires two separate contrast injections and acquisitions. • Dynamic spin-and-gradient-echo echoplanar imaging enables simultaneous perfusion and permeability imaging. • Dynamic spin-and-gradient-echo echoplanar imaging provides new image contrasts reflecting blood-brain barrier disruption and cytoarchitecture characteristics.

Radio-pathomic maps of glioblastoma identify phenotypes of non-enhancing tumor infiltration associated with bevacizumab treatment response

BACKGROUND

Autopsy-based radio-pathomic maps of glioma pathology have shown substantial promise inidentifying areas of non-enhancing tumor presence, which may be able to differentiate subsets of patients that respond favorably to treatments such as bevacizumab that have shown mixed efficacy evidence. We tested the hypthesis that phenotypes of non-enhancing tumor fronts can distinguish between glioblastoma patients that will respond favorably to bevacizumab and will visually capture treatment response.

METHODS

T1, T1C, FLAIR, and ADC images were used to generate radio-pathomic maps of tumor characteristics for 79 pre-treatment patients with a primary GBM or high-grade IDH1-mutant astrocytoma for this study. Novel phenotyping (hypercellular, hypocellular, hybrid, or well-circumscribed front) of the non-enhancing tumor front was performed on each case. Kaplan Meier analyses were then used to assess differences in survival and bevacizumab efficacy between phenotypes. Phenotype compartment segmentations generated longitudinally for a subset of 26 patients over the course of bevacizumab treatment, where a mixed effect model was used to detect longitudinal changes.

RESULTS

Well-Circumscribed patients showed significant/trending increases in survival compared to Hypercellular Front (HR = 2.0, p = 0.05), Hypocellular Front (HR = 2.02, p = 0.03), and Hybrid Front tumors (HR = 1.75, p = 0.09). Only patients with hypocellular or hybrid fronts showed significant survival benefits from bevacizumab treatment (HR = 2.35, p = 0.02; and HR = 2.45, p = 0.03, respectively). Hypocellular volumes decreased by an average 50.52 mm3 per day of bevacizumab treatment (p = 0.002).

CONCLUSION

Patients with a hypocellular tumor front identified by radio-pathomic maps showed improved treatment efficacy when treated with bevacizumab, and reducing hypocellular volumes over the course of treatment may indicate treatment response.

Advances in Neuro-Oncological Imaging: An Update on Diagnostic Approach to Brain Tumors

This study delineates the pivotal role of imaging within the field of neurology, emphasizing its significance in the diagnosis, prognostication, and evaluation of treatment responses for central nervous system (CNS) tumors. A comprehensive understanding of both the capabilities and limitations inherent in emerging imaging technologies is imperative for delivering a heightened level of personalized care to individuals with neuro-oncological conditions. Ongoing research in neuro-oncological imaging endeavors to rectify some limitations of radiological modalities, aiming to augment accuracy and efficacy in the management of brain tumors. This review is dedicated to the comparison and critical examination of the latest advancements in diverse imaging modalities employed in neuro-oncology. The objective is to investigate their respective impacts on diagnosis, cancer staging, prognosis, and post-treatment monitoring. By providing a comprehensive analysis of these modalities, this review aims to contribute to the collective knowledge in the field, fostering an informed approach to neuro-oncological care. In conclusion, the outlook for neuro-oncological imaging appears promising, and sustained exploration in this domain is anticipated to yield further breakthroughs, ultimately enhancing outcomes for individuals grappling with CNS tumors.

Radio-pathomic maps of glioblastoma identify phenotypes of non-enhancing tumor infiltration associated with bevacizumab treatment response

Background

Autopsy-based radio-pathomic maps of glioma pathology have shown substantial promise inidentifying areas of non-enhancing tumor presence, which may be able to differentiate subsets of patients that respond favorably to treatments such as bevacizumab that have shown mixed efficacy evidence. We tested the hypthesis that phenotypes of non-enhancing tumor fronts can distinguish between glioblastoma patients that will respond favorably to bevacizumab and will visually capture treatment response.

Methods

T1, T1C, FLAIR, and ADC images were used to generate radio-pathomic maps of tumor characteristics for 79 pre-treatment patients with a primary GBM or high-grade IDH1-mutant astrocytoma for this study. Novel phenotyping (hypercellular, hypocellular, hybrid, or well-circumscribed front) of the non-enhancing tumor front was performed on each case. Kaplan Meier analyses were then used to assess differences in survival and bevacizumab efficacy between phenotypes. Phenotype compartment segmentations generated longitudinally for a subset of 26 patients over the course of bevacizumab treatment, where a mixed effect model was used to detect longitudinal changes.

Results

Well-Circumscribed patients showed significant/trending increases in survival compared to Hypercellular Front (HR = 2.0, p = 0.05), Hypocellular Front (HR = 2.02, p = 0.03), and Hybrid Front tumors (HR = 1.75, p = 0.09). Only patients with hypocellular or hybrid fronts showed significant survival benefits from bevacizumab treatment (HR = 2.35, p = 0.02; and HR = 2.45, p = 0.03, respectively). Hypocellular volumes decreased by an average 50.52 mm3 per day of bevacizumab treatment (p = 0.002).

Conclusion

Patients with a hypocellular tumor front identified by radio-pathomic maps showed improved treatment efficacy when treated with bevacizumab, and reducing hypocellular volumes over the course of treatment may indicate treatment response.

The Brainbox-a tool to facilitate correlation of brain magnetic resonance imaging features to histopathology

Magnetic resonance imaging (MRI) has limitations in identifying underlying tissue pathology, which is relevant for neurological diseases such as multiple sclerosis, stroke or brain tumours. However, there are no standardized methods for correlating MRI features with histopathology. Thus, here we aimed to develop and validate a tool that can facilitate the correlation of brain MRI features to corresponding histopathology. For this, we designed the Brainbox, a waterproof and MRI-compatible 3D printed container with an integrated 3D coordinate system. We used the Brainbox to acquire post-mortem ex vivo MRI of eight human brains, fresh and formalin-fixed, and correlated focal imaging features to histopathology using the built-in 3D coordinate system. With its built-in 3D coordinate system, the Brainbox allowed correlation of MRI features to corresponding tissue substrates. The Brainbox was used to correlate different MR image features of interest to the respective tissue substrate, including normal anatomical structures such as the hippocampus or perivascular spaces, as well as a lacunar stroke. Brain volume decreased upon fixation by 7% (P = 0.01). The Brainbox enabled degassing of specimens before scanning, reducing susceptibility artefacts and minimizing bulk motion during scanning. In conclusion, our proof-of-principle experiments demonstrate the usability of the Brainbox, which can contribute to improving the specificity of MRI and the standardization of the correlation between post-mortem ex vivo human brain MRI and histopathology. Brainboxes are available upon request from our institution.

Radio-pathomic approaches in pediatric neuro-oncology: Opportunities and challenges

With medical software platforms moving to cloud environments with scalable storage and computing, the translation of predictive artificial intelligence (AI) models to aid in clinical decision-making and facilitate personalized medicine for cancer patients is becoming a reality. Medical imaging, namely radiologic and histologic images, has immense analytical potential in neuro-oncology, and models utilizing integrated radiomic and pathomic data may yield a synergistic effect and provide a new modality for precision medicine. At the same time, the ability to harness multi-modal data is met with challenges in aggregating data across medical departments and institutions, as well as significant complexity in modeling the phenotypic and genotypic heterogeneity of pediatric brain tumors. In this paper, we review recent pathomic and integrated pathomic, radiomic, and genomic studies with clinical applications. We discuss current challenges limiting translational research on pediatric brain tumors and outline technical and analytical solutions. Overall, we propose that to empower the potential residing in radio-pathomics, systemic changes in cross-discipline data management and end-to-end software platforms to handle multi-modal data sets are needed, in addition to embracing modern AI-powered approaches. These changes can improve the performance of predictive models, and ultimately the ability to advance brain cancer treatments and patient outcomes through the development of such models.

Radiomics in clinical trials: perspectives on standardization

The term biomarker is used to describe a biological measure of the disease behavior. The existing imaging biomarkers are associated with the known tissue biological characteristics and follow a well-established roadmap to be implemented in routine clinical practice. Recently, a new quantitative imaging analysis approach named radiomics has emerged. It refers to the extraction of a large number of advanced imaging features with high-throughput computing. Extensive research has demonstrated its value in predicting disease behavior, progression, and response to therapeutic options. However, there are numerous challenges to establishing it as a clinically viable solution, including lack of reproducibility and transparency. The data-driven nature also does not offer insights into the underpinning biology of the observed relationships. As such, additional effort is needed to establish it as a qualified biomarker to inform clinical decisions. Here we review the technical difficulties encountered in the clinical applications of radiomics and current effort in addressing some of these challenges in clinical trial designs. By addressing these challenges, the true potential of radiomics can be unleashed.

Radiomics in neuro-oncological clinical trials

The development of clinical trials has led to substantial improvements in the prevention and treatment of many diseases, including brain cancer. Advances in medicine, such as improved surgical techniques, the development of new drugs and devices, the use of statistical methods in research, and the development of codes of ethics, have considerably influenced the way clinical trials are conducted today. In addition, methods from the broad field of artificial intelligence, such as radiomics, have the potential to considerably affect clinical trials and clinical practice in the future. Radiomics is a method to extract undiscovered features from routinely acquired imaging data that can neither be captured by means of human perception nor conventional image analysis. In patients with brain cancer, radiomics has shown its potential for the non-invasive identification of prognostic biomarkers, automated response assessment, and differentiation between treatment-related changes from tumour progression. Despite promising results, radiomics is not yet established in routine clinical practice nor in clinical trials. In this Viewpoint, the European Organization for Research and Treatment of Cancer Brain Tumour Group summarises the current status of radiomics, discusses its potential and limitations, envisions its future role in clinical trials in neuro-oncology, and provides guidance on how to address the challenges in radiomics.

The relationship between radiomics and pathomics in Glioblastoma patients: Preliminary results from a cross-scale association study

Glioblastoma multiforme (GBM) typically exhibits substantial intratumoral heterogeneity at both microscopic and radiological resolution scales. Diffusion Weighted Imaging (DWI) and dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) are two functional MRI techniques that are commonly employed in clinic for the assessment of GBM tumor characteristics. This work presents initial results aiming at determining if radiomics features extracted from preoperative ADC maps and post-contrast T1 (T1C) images are associated with pathomic features arising from H&E digitized pathology images. 48 patients from the public available CPTAC-GBM database, for which both radiology and pathology images were available, were involved in the study. 91 radiomics features were extracted from ADC maps and post-contrast T1 images using PyRadiomics. 65 pathomic features were extracted from cell detection measurements from H&E images. Moreover, 91 features were extracted from cell density maps of H&E images at four different resolutions. Radiopathomic associations were evaluated by means of Spearman's correlation (ρ) and factor analysis. p values were adjusted for multiple correlations by using a false discovery rate adjustment. Significant cross-scale associations were identified between pathomics and ADC, both considering features (n = 186, 0.45 < ρ < 0.74 in absolute value) and factors (n = 5, 0.48 < ρ < 0.54 in absolute value). Significant but fewer ρ values were found concerning the association between pathomics and radiomics features (n = 53, 0.5 < ρ < 0.65 in absolute value) and factors (n = 2, ρ = 0.63 and ρ = 0.53 in absolute value). The results of this study suggest that cross-scale associations may exist between digital pathology and ADC and T1C imaging. This can be useful not only to improve the knowledge concerning GBM intratumoral heterogeneity, but also to strengthen the role of radiomics approach and its validation in clinical practice as "virtual biopsy", introducing new insights for omics integration toward a personalized medicine approach.

Cellular Density in Adult Glioma, Estimated with MR Imaging Data and a Machine Learning Algorithm, Has Prognostic Power Approaching World Health Organization Histologic Grading in a Cohort of 1181 Patients

BACKGROUND AND PURPOSE

Recent advances in machine learning have enabled image-based prediction of local tissue pathology in gliomas, but the clinical usefulness of these predictions is unknown. We aimed to evaluate the prognostic ability of imaging-based estimates of cellular density for patients with gliomas, with comparison to the gold standard reference of World Health Organization grading.

MATERIALS AND METHODS

Data from 1181 (207 grade II, 246 grade III, 728 grade IV) previously untreated patients with gliomas from a single institution were analyzed. A pretrained random forest model estimated voxelwise tumor cellularity using MR imaging data. Maximum cellular density was correlated with the World Health Organization grade and actual survival, correcting for covariates of age and performance status.

RESULTS

A maximum estimated cellular density of >7681 nuclei/mm2 was associated with a worse prognosis and a univariate hazard ratio of 4.21 (P < .001); the multivariate hazard ratio after adjusting for covariates of age and performance status was 2.91 (P < .001). The concordance index between maximum cellular density (adjusted for covariates) and survival was 0.734. The hazard ratio for a high World Health Organization grade (IV) was 7.57 univariate (P < .001) and 5.25 multivariate (P < .001). The concordance index for World Health Organization grading (adjusted for covariates) was 0.761. The maximum cellular density was an independent predictor of overall survival, and a Cox model using World Health Organization grade, maximum cellular density, age, and Karnofsky performance status had a higher concordance (C = 0.764; range 0.748-0.781) than the component predictors.

CONCLUSIONS

Image-based estimation of glioma cellularity is a promising biomarker for predicting survival, approaching the prognostic power of World Health Organization grading, with added values of early availability, low risk, and low cost.