Integrated models incorporating radiologic and radiomic features predict meningioma grade, local failure, and overall survival

Abstract

Background We investigated prognostic models based on clinical, radiologic, and radiomic feature to preoperatively identify meningiomas at risk for poor outcomes. Methods Retrospective review was performed for 303 patients who underwent resection of 314 meningiomas (57% World Health Organization grade I, 35% grade II, and 8% grade III) at two independent institutions, which comprised primary and external datasets. For each patient in the primary dataset, 16 radiologic and 172 radiomic features were extracted from preoperative magnetic resonance images, and prognostic features for grade, local failure (LF) or overall survival (OS) were identified using the Kaplan–Meier method, log-rank tests and recursive partitioning analysis. Regressions and random forests were used to generate and test prognostic models, which were validated using the external dataset. Results Multivariate analysis revealed that apparent diffusion coefficient hypointensity (HR 5.56, 95% CI 2.01–16.7, P = .002) was associated with high grade meningioma, and low sphericity was associated both with increased LF (HR 2.0, 95% CI 1.1–3.5, P = .02) and worse OS (HR 2.94, 95% CI 1.47–5.56, P = .002). Both radiologic and radiomic predictors of adverse meningioma outcomes were significantly associated with molecular markers of aggressive meningioma biology, such as somatic mutation burden, DNA methylation status, and FOXM1 expression. Integrated prognostic models combining clinical, radiologic, and radiomic features demonstrated improved accuracy for meningioma grade, LF, and OS (area under the curve 0.78, 0.75, and 0.78, respectively) compared to models based on clinical features alone. Conclusions Preoperative radiologic and radiomic features such as apparent diffusion coefficient and sphericity can predict tumor grade, LF, and OS in patients with meningioma.

Multiplatform genomic profiling and magnetic resonance imaging identify mechanisms underlying intratumor heterogeneity in meningioma

“…Thus, although high grade meningiomas can have globally low ADC values reflective of high cellularity ( Supplementary Fig. 4b) 45 , our data suggest that regions of elevated ADC in meningiomas may harbor aggressive meningioma clones. To address the possibility that spatially-distinct samples from M13, the most clinically and histopathologically aggressive tumor in our study, may have overly biased our protein expression analyses from a limited number of samples ( Fig.…”

“…In summary, our findings support the large body of literature lending credence to existing histopathologic criteria for meningioma prognostication, and indicate that existing, robust prognostic models should not be discounted when incorporating new technologies into clinical decision making. Rather, our results suggest that new and old models may be integrated to optimize clinical stratification of patients, as we have previously done using clinical, demographic, and imaging characteristics of meningioma patients alongside tumor grade 45,65 . In particular, we suggest that neurosurgeons consider image-guided sampling of regions with elevated ADC for histopathologic evaluation to determine whether they contain features that may predict aggressive clinical behavior that would otherwise go undetected.…”

“…Due to the intrinsic morbidity of salvage surgery and radiotherapy for recurrent meningioma [53][54][55] , there is an urgent, unmet need for new therapies to treat patients with high grade or recurrent tumors 4,5 . We have shown that meningiomas with elevated ADC are at risk for high grade and recurrence 45,46 , and have identified molecular signatures associated with high ADC values in vivo ( Fig. 4) and in vitro (Fig.…”

“…Within these gene sets, notable genes that were enriched in samples with high ADC values included CDH2, which encodes N-cadherin and is a regulator of noncanonical Wnt signaling and neuronal development 39,40 , and PTPRZ1, which is necessary for neuronal development 41 , and we and others have shown is involved in glioblastoma cell invasion and self-renewal [42][43][44] . Further, we previously identified high ADC as a marker of high grade meningioma and poor clinical outcomes 45,46 , suggesting that molecular correlates in regions with elevated ADC may drive meningioma tumorigenesis. In summary, these data demonstrate that radiographic intratumor heterogeneity in meningioma is associated with regional transcriptomic differences, and that high ADC distinguishes regions that are enriched in developmental gene expression programs that may underlie meningioma cell proliferation and tumor recurrence.…”

Identification of High-Risk Atypical Meningiomas According to Semantic and Radiomic Features

“…Neuro-oncologists have successfully applied radiomics in the survival assessment of glioblastoma [ 11 ]. Recent studies on the use of radiomic analysis for meningiomas suggest that they can be used to estimate WHO grades [ 12 , 13 , 14 ], with close to 90% accuracy [ 15 ]. Radiomics could also be applied for the preoperative differentiation of meningioma subtypes [ 16 ].…”