Role of dynamic contrast-enhanced and dynamic susceptibility contrast imaging in evaluating the biological features of glioma

Introduction

Brain gliomas in adults, including astrocytoma, oligodendroglioma, and glioblastoma (GBM), account for about 25.5% of all central nervous system (CNS) tumors and 80% of malignant CNS tumors (1-3). Even low-grade gliomas (LGG) often exhibit diffuse infiltrative growth of gliomas (4), so it is difficult to completely remove the tumor by surgical treatment (5). Furthermore, considering the recovery of postoperative brain function, intracranial tumors cannot be eradicated through arbitrary expansion resection like other tumors (6). Therefore, there is an urgent need for imaging modalities that can accurately evaluate gliomas prior to surgery.

The traditional grading for gliomas relies solely on histological features, which may not consistently capture the underlying biological behavior and subsequently result in varied prognostic outcomes. Nevertheless, incorporating the 2016 version of glioma classification from the World Health Organization (WHO) into molecular pathological diagnosis, along with additional improvements introduced in WHO 2021 edition (WHO-CNS 5), has substantially improved its diagnostic accuracy by leveraging advancements in molecular biology. The WHO-CNS 5 (7) simplifies the classification and grading criteria for adult diffuse gliomas into three categories: astrocytomas (with isocitrate dehydrogenase, IDH-mutant), oligodendrogliomas (with both IDH-mutant and 1p/19q codeleted), and GBMs (IDH-wildtype). The status of IDH mutation, along with 1p/19q codeletion, plays a crucial role in classifying glioma types and significantly influences prognosis. The results of numerous studies have consistently demonstrated that patients harboring IDH-mutant and 1p/19q codeleted (8-10) exhibit a more favorable prognosis. However, these assessments not only necessitate an invasive biopsy or surgical resection for acquiring tumor samples but also carry the risk of sampling errors (11). The advantage of imaging lies in its non-invasive nature and ability to provide continuous comprehensive information about gliomas, enabling evaluation of the entire tumor.

Perfusion magnetic resonance imaging (MRI) has become an indispensable tool in neuro-oncology, primarily through two principal techniques: dynamic susceptibility contrast (DSC) MRI and dynamic contrast enhanced (DCE) MRI. Both modalities provide critical insights into tumor microvascular characteristics, though distinct technical mechanisms. DSC-MRI quantifies cerebral hemodynamics via T2*-weighted signal attenuation caused by contrast agent susceptibility, with relative cerebral blood volume (rCBV) and relative cerebral blood flow (rCBF) emerging as validated biomarkers for tumor grading, prognosis assessment, and differentiation of tumor recurrence from treatment effects (12-14). Multicenter studies further demonstrate its utility in predicting IDH mutation status (15). In contrast, DCE-MRI employs T1-weighted imaging to monitor contrast agent extravasation kinetics, offering unique capabilities for preoperative evaluation and longitudinal monitoring (16). Its signal intensity-time curve enables comprehensive characterization of tumor microvasculature through three key parameters: tissue perfusion, endothelial permeability, and extravascular-extracellular space properties (17,18). The volume transfer constant (Ktrans), representing the composite effect of blood flow and vascular permeability, serves as a sensitive marker of blood-brain barrier (BBB) disruption and malignancy grade (18). Its interpretation requires careful contextualization within distinct hemodynamic environments: in non-enhancing tumors, Ktrans values are predominantly governed by perfusion deficits under low-flow conditions, whereas in enhancing lesions, they primarily reflect permeability-surface area product elevations associated with BBB breakdown (19). The volume fraction of the extravascular-extracellular space (Ve) shows additional promise as a mitotic activity indicator (20). The complementary strengths of these techniques in glioma genotypic profiling stem from their distinct technical attributes. Specifically, DCE-MRI demonstrates unique sensitivity to angiogenesis-driven genotypes (e.g., IDH1-mutant gliomas) through two technical merits: (I) pharmacokinetic modeling precision of endothelial leakage; and (II) reduced susceptibility artifacts confounding DSC measurements (12,21,22). This enables detection of hypoxia-inducible factor 1-alpha (HIF-1α)-mediated vascular permeability alterations characteristic of specific molecular subtypes (23,24). Conversely, DSC-MRI provides superior temporal resolution for mapping hypoxia-related hemodynamic shifts via CBV quantification (21), albeit limited in permeability assessment by T2* weighting constraints (22).

Current limitations in glioma characterization stem from fragmented methodological approaches that inadequately address the biological duality of tumor vascular features. While DSC-MRI and DCE-MRI respectively capture hemodynamic and permeability dimensions, prior single-modality studies (25,26) have failed to systematically compare their differential diagnostic sensitivities within identical tumor ecosystems, particularly regarding molecular subtype discrimination (IDH mutation, 1p/19q status). This methodological siloing obscures critical interactions between vascular phenotype and genotype, leaving unresolved whether these perfusion techniques offer complementary or redundant value for tumor subtyping. To address this gap, our investigation implemented a prospective dual-modality comparative design, acquiring concurrent DSC-MRI and DCE-MRI datasets within a unified glioma cohort to resolve three pivotal questions: (I) which technique (DCE or DSC) better distinguishes genotype-driven vascular phenotypes, such as IDH-mutant versus wild-type? (II) What are the complementary strengths of each modality in capturing haemodynamic (DSC) versus permeability (DCE) biomarkers? We hypothesize that DCE will outperform DSC in predicting permeability-dominant genotypes, while DSC will excel in blood flow-dominant classifications. (III) Can these perfusion parameters serve as biomarkers for predicting prognosis? Furthermore, we established an intrinsic correlation between the findings obtained from both imaging methods (DSC-MRI/DCE-MRI) with those observed through immunohistochemical staining/histopathological examination to validate their biological interpretability. This represents the first study to systematically decode the synergistic diagnostic hierarchy of DSC and DCE through genotype-anchored vascular phenotyping, thereby establishing a paradigm shift from modality-agnostic to molecularly guided perfusion imaging in gliomas. These results will challenge the current “one-size-fits-all” MRI protocols and provide actionable criteria for modality selection based on molecular profiles. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2794/rc).

Methods

Participants

A total of 46 glioma patients were prospectively enrolled from the Department of Neurosurgery at the First Hospital of China Medical University, Shenyang, China. The inclusion criteria were as follows: (I) the patient’s tumor was the first to develop, and the patient had no history of trauma, preoperative radiotherapy, or chemotherapy; (II) the patient had undergone routine MRI, DCE, and DSC before the operation; (III) all patients were pathologically diagnosed with gliomas through the detection of IDH and 1p/19q; (IV) the interval between MRI and operation was less than 1 week in all patients. The exclusion criteria were as follows: (I) patients with postoperative recurrence; (II) patients with brain tumors other than gliomas; (III) MRI images with obvious motion artifacts; (IV) patients with claustrophobia. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. All patients provided written informed consent after receiving a detailed description of the study. The study was approved by the Institutional Review Board of the First Hospital of China Medical University (No. KLS [2023]-21).

MRI protocol

All MR data were acquired in a GE pioneer 3.0T scanner (GE, SIGNA Pioneer, Waukesha, WI, USA) with a standard 16-channel head coil at the First Hospital of China Medical University. The DCE and DSC sequences were collected simultaneously with routine sequences in all patients. To address BBB leakage artifacts in DSC-MRI, a dual-bolus gadolinium protocol was implemented (27). The total contrast dose (0.1 mmol/kg gadodiamide, GE Healthcare, Ireland) was divided into two sequential boluses (0.05 mmol/kg each) to concurrently optimize diagnostic precision and renal safety (28). The first bolus (2 mL/s injection rate) was administered 50 seconds after DCE-MRI initiation, enabling baseline permeability quantification (Ktrans) and identification of BBB-compromised regions. Following a 3–5-minute interval to facilitate contrast clearance and tissue saturation, conventional post-contrast T1-weighted imaging was performed. Subsequently, the second bolus (4 mL/s injection rate) was delivered 16 seconds into the DSC-MRI acquisition, with leakage correction applied using the Boxerman-Weisskoff model (27). This algorithm integrates pre-bolus permeability data to simultaneously mitigate T1- and T2*-weighted leakage effects, thereby suppressing spurious rCBV overestimation in BBB-defective areas while preserving hemodynamic fidelity in intact regions. Each bolus was followed by a 20 mL saline flush. Prior validation studies confirm that this reduced-dose dual-bolus approach maintains perfusion parameter reliability (28), resolving the historical trade-off between contrast-induced nephropathy risk and DSC-MRI diagnostic accuracy. Detailed acquisition parameters are provided in Appendix 1.

Image post-processing

The DCE data was analyzed by image processing software (Omni-kinetics Version V2.1.1R; General Electric, Wauwatosa, WI, USA) to calculate the parameters of Pharmacokinetic Model, including Ktrans, Ve, rate constant (Kep) and fractional volume of the intravascular compartment (Vp) in accordance with the Extended Tofts model (Figure 1A-1C). Preprocessing included motion correction and baseline signal normalization. For DSC analysis, the arterial input function (AIF) was manually selected from the middle cerebral artery using semi-automated tools on the GE ADW 4.7 workstation. Leakage correction was performed via the Boxerman-Weisskoff model (27) to account for contrast extravasation. Relative CBV (rCBV) and rCBF were computed as the area under the concentration-time curve normalized to contralateral normal-appearing white matter, ensuring comparability across patients (Figure 1A-1C). Whole-tumor volume-of-interest (VOI) delineation was performed on contrast-enhanced T1-weighted and T2-weighted fluid-attenuated inversion recovery (FLAIR) sequences by consensus between a neuroradiologist and an imaging specialist, with a two-step registration protocol ensuring spatial alignment between perfusion maps. Histogram features were extracted via PyRadiomics to quantify perfusion heterogeneity (Table S1). Inter-observer agreement was validated through intraclass correlation coefficients (ICC >0.85) and Cohen’s kappa (κ>0.75) (Table S2). Detailed segmentation workflows, and full histogram feature specifications are provided in Appendix 1.

Figure 1 Representative DCE/DSC parametric maps and pathological images in glioma with different genotypes. (A) Images in 30-year-old woman with WHO grade 3 astrocytoma with IDH-mutant and 1p/19q codeleted; (B) images in 69-year-old woman with WHO grade 4 astrocytoma with IDH-mutant and 1p/19q non-codeleted; (C) images in 53-year-old woman with WHO grade 4 GBM with IDH-wildtype. Representative immunohistochemical staining for Ki-67 and MVD using the streptavidin-peroxidase method is shown. Images were captured at an original magnification of ×100. Ktrans and Kep are expressed in min⁻¹. DCE, dynamic contrast-enhanced; DSC, dynamic susceptibility contrast; GBM, glioblastoma; IDH, isocitrate dehydrogenase; Kep, rate constant; Ktrans, volume transfer constant; MVD, micro-vessel density; rCBF, relative cerebral blood flow; rCBV, relative cerebral blood volume; Ve, fractional volume of the extravascular-extracellular space; WHO, world health organization.

Pathological examination

The grades and genotypes were documented from pathological and genetic test reports, while immunohistochemistry was conducted in the laboratory. Formalinfixed, paraffin-embedded tumor sections were subjected to standardized immunohistochemical analysis using CD34 and Ki-67 antibodies (Maixin Biotechnology, Fuzhou, China). MVD quantification followed Weidner’s hotspot methodology, while Ki-67 proliferation index assessment adhered to Winyard criteria (Figure 1A-1C). All histopathological evaluations were independently performed by two blinded neuropathologists with third-party adjudication for discordant cases. Detailed protocols on specimen processing, staining parameters, and quantitative measurement procedures are provided in Appendix 1.

Statistical analysis

All statistical analyses were conducted using SPSS software (Version 22.0.0, IBM, Armonk, NY, USA). A two-tailed P value less than 0.05 was considered statistically significant. The Chi-squared test was employed for analyzing the categorical variable of demographic characteristics across different grades and genotypes. The Shapiro-Wilk test was conducted to evaluate the normality of the data distribution. Two independent samples t-tests and Mann-Whitney U tests were utilized for the continuous variable of demographic data and all histogram features of perfusion parameters (Ktrans, Ve, Kep, Vp, and rCBV, rCBF) among different grades (LGG and high-grade glioma, HGG) and genotypes (IDH-mutant and 1p/19q codeleted, IDH-mutant and 1p/19q non-codeleted, IDH-wildtype). To account for potential confounding effects of contrast enhancement on perfusion parameters in the IDH mutant type, we performed stratified analyses by segregating tumors into enhancing and non-enhancing subgroups based on post-gadolinium T1-weighted imaging. Corresponding histogram features of perfusion parameters were independently calculated for each subgroup, and intergroup differences were statistically evaluated using the Mann-Whitney U test with Benjamini-Hochberg correction for multiple comparisons.

Pearson’s correlation analysis was conducted to assess the correlations among the histogram features of DCE and DSC, Ki-67 labeling index, and MVD. Outliers in perfusion metrics, defined as values beyond ±2 standard deviations from the cohort mean, were excluded from correlation analyses. This protocol ensured robustness against extreme-value biases while maintaining pathophysiological relevance. Receiver operating characteristic (ROC) curve analysis was conducted to explore the performance of DCE and DSC histogram features in distinguishing different molecular statuses and grades of gliomas. To statistically compare the discriminative performance between DCE- and DSC-derived biomarkers, we employed DeLong’s test—a non-parametric method that quantifies differences in the areas under the curves (AUCs) by analyzing the covariance structures of paired empirical ROC curves. Optimal diagnostic thresholds for perfusion parameters were derived using Youden index, which identifies the cutoff value that maximizes the sum of sensitivity and specificity while minimizing misclassification errors.

Survival analysis

Overall survival (OS) was defined as the time from initial diagnosis to death from any cause, with patients alive at the last follow-up being right-censored. Regular follow-up assessments were conducted every 3–6 months until death or the study cutoff date [2023.12], yielding a median follow-up duration of 25 months (range, 15–35 months). To evaluate the association between imaging/clinical variables and survival outcomes, Cox proportional hazards regression models were employed. This method was selected for its ability to handle right-censored data, incorporate time-dependent covariates, and estimate hazard ratios (HRs) without restrictive assumptions about the baseline hazard function. In the univariate analysis, all candidate variables—including perfusion parameters (Ktrans, Ve, Vp, Kep, rCBV and rCBF), age, gender, chemoradiotherapy, IDH status, 1p/19q, O6-methylguanine-DNA methyltransferase (MGMT), Ki-67 and MVD—were individually assessed for their prognostic significance, with variables showing P<0.1 retained for further multivariate modeling. The multivariate analysis utilized a backward stepwise selection approach to construct the final model. To ensure model stability, multicollinearity was evaluated using variance inflation factors (VIFs), with a threshold of VIF <5 indicating acceptable collinearity. The proportional hazards assumption was rigorously verified through Schoenfeld residuals testing and log-minus-log survival plots. Variables violating this assumption were addressed via stratification or time-dependent covariate adjustments. Model performance was quantified using Harrell’s concordance index (C-index) for predictive accuracy, while calibration curves assessed goodness-of-fit.

Results

Participant demographics

A total of 46 patients (27 men, 19 women; mean age, 54±13 years; age range, 20–78 years) with gliomas were enrolled in this study. An overview of the participant stratification according to genotype and grade were shown in Table 1. Statistical evaluation using the Shapiro-Wilk test demonstrated normal distribution characteristics for age (P>0.05), therefore parametric analysis was conducted with two independent samples t-test. The results demonstrated a significant age disparity between HGGs and LGGs, with HGG patients being notably older than those with LGG. Additionally, IDH-wildtype individuals exhibited a significantly lower age compared to IDH-mutant. No notable gender differences were observed in relation to glioma WHO grade or IDH mutation status. Peritumoral edema was commonly observed in patients diagnosed with HGG and IDH-wildtype. In contrast, MGMT methylation predominated in LGG and IDH-mutant patients, while no statistically significant difference was found regarding MGMT methylation status among the patients of HGG and IDH-wildtype. The detailed results were presented in Table 1.

Statistical analysis of DCE and DSC histograms

According to the Shapiro-Wilk test results, the histogram features based on DCE and DSC perfusion parameters conformed to a non-normal distribution (P<0.05), so Mann-Whitney U tests were used. The results demonstrated that DCE-derived parameters were effective in distinguishing IDH mutation status, particularly the mean values of Ktrans, 90^th percentile of Ve and median of Kep, which also provided indications of 1p/19q co-deletion status (Figure 2). Conversely, there was no statistically significant difference for molecular typing distinction when utilizing DSC-derived parameters (Figure 2). However, both DCE and DSC-derived parameters exhibited favorable efficacy for histological classification as well (Figure 2). For a detailed comparison of histogram features related to specific parameters obtained from both DCE and DSC parameters across various molecular types and histological grades, please refer to Tables S1,S3,S4.

Figure 2 The boxplots of DCE/DSC-derived parameters across different genotypes and WHO grades. *, P<0.05; **, P<0.01; ***, P<0.001. Ktrans and Kep are expressed in min⁻¹. 1p/19q codel, 1p/19q codeleted; 1p/19q non-codel, 1p/19q non-codeleted; DCE, dynamic contrast-enhanced; DSC, dynamic susceptibility contrast; IDH, isocitrate dehydrogenase; IDH-mut, IDH-mutant; IDH-wt, IDH-wildtype; IQR, inter-quartile range; Kep, rate constant; Ktrans, volume transfer constant; P90, 90^th percentile; rCBF, relative cerebral blood flow; rCBV, relative cerebral blood volume; Ve, fractional volume of the extravascular-extracellular space; Vp, fractional volume of the intravascular compartment; WHO, World Health Organization.

Stratified analysis of IDH-mutant gliomas demonstrated distinct perfusion profiles: enhancing tumors showed significantly elevated perfusion derived parameters compared to non-enhancing cases, highlighting the necessity of enhancement-specific analyses (Table S5).

Correlation analyses between perfusion parameters and pathological indices

The Pearson correlation analysis was employed to examine the association between histogram characteristics of perfusion parameters and Ki-67, MVD. Significant correlations were observed among the Ki-67, MVD, and DCE parameter maps containing Ktrans, Ve, Kep, and Vp (P<0.05, r=0.35–0.56) (Figure 3, Figure S1). However, within the parameters of DSC, significant correlations were found only with MVD (rCBV: r=0.52, P<0.05; rCBF: r=0.36, P<0.05) rather than with Ki-67 (Figure 3, Figure S1).

Figure 3 The Pearson’s correlations matrix of perfusion parametric maps and pathological indexes. *, P<0.05; **, P<0.01; ***, P<0.001. Ktrans and Kep are expressed in min⁻¹. The color bar represents r value. Kep, rate constant; Ktrans, volume transfer constant; MVD, micro-vessel density; rCBF, relative cerebral blood flow; rCBV, relative cerebral blood volume; Ve, fractional volume of the extravascular-extracellular space; Vp, fractional volume of the intravascular compartment.

Accuracy of DCE and DSC parameters for distinguishing glioma genotypes and grades

Table 2 and Figure 4A,4B show AUC values, proportional to diagnostic accuracy, and the respective cutoff values of each DCE and DSC parameters for distinguishing glioma genotypes (IDH-mutant vs. IDH-wildtype) and grades (LGG vs. HGG). For DSC, the rCBV and rCBF measurements did not demonstrate statistical significance in distinguishing genotyping with IDH mutation status. However, DCE exhibited superior discriminatory properties for both glioma genotypes and grades.

Figure 4 Comparative diagnostic performance of DCE/DSC-derived parameters. (A,C) ROC curves and DeLong’s test for IDH-mutant versus wildtype discrimination; (B,D) ROC curves and DeLong’s test for LGG versus HGG discrimination. AUC, area under the curve; DCE, dynamic contrast enhanced; DSC, dynamic susceptibility contrast; HGG, high-grade glioma; IDH, isocitrate dehydrogenase; Kep, rate constant; Ktrans, volume transfer constant; LGG, low-grade glioma; rCBF, relative cerebral blood flow; rCBV, relative cerebral blood volume; ROC, receiver operating characteristic; Ve, fractional volume of the extravascular-extracellular space; Vp, fractional volume of the intravascular compartment.

Comparative diagnostic performance of DCE/DSC-derived parameters

The DeLong’s test revealed a statistically significant difference in AUC between DCE- and DSC-derived parameters for glioma genotyping (Figure 4C), with DCE parameters demonstrating superior discriminative power (Figure 4A). In contrast, a marginal statistical difference was observed in histological grading (Figure 4D), though both DCE and DSC parameters maintained clinically comparable AUC values (Figure 4B). These results further validate the distinct diagnostic advantages of DCE in molecular characterization and the complementary role of DSC in gliomas grading.

Survival analysis

Among the 46 glioma patients, 19 deaths (41%) occurred during the follow-up period (3 years). The median OS was 32 months [95% confidence interval (CI): 29–33]. Univariate Cox regression identified several significant prognostic factors for reduced OS: advancing age (HR =1.08/year, 95% CI: 1.03–1.13, P=0.001), IDH wildtype status (HR =1.70, 95% CI: 1.01–2.85, P=0.04), MGMT promoter methylation (HR =2.58, 95% CI: 1.01–6.60, P=0.04), and elevated values of MVD (HR =1.07, 95% CI: 1.03–1.13, P=0.001), Ktrans (HR =1.07, 95% CI: 1.03–1.13, P=0.002), Ve (HR =1.01, 95% CI: 1.00–1.01, P=0.002), Kep (HR =1.01, 95% CI: 1.00–1.01, P=0.004), rCBV (HR =1.51, 95% CI: 1.17–1.95, P=0.002), and rCBF (HR =1.03, 95% CI: 1.00–1.05, P=0.03) (Table 3).

These candidate predictors were subsequently entered into multivariate analysis following verification of model assumptions. After addressing multicollinearity (VIF <3 for all variables) and confirming proportional hazards assumptions through Schoenfeld residual testing (global test P>0.05), three independent prognostic factors emerged: MGMT promoter methylation demonstrated the strongest association with mortality risk (HR =4.15, 95% CI: 1.10–15.68, P=0.04), followed by elevated Ktrans (HR =1.08, 95% CI: 1.01–1.17, P=0.03) and increased rCBV (HR =1.87, 95% CI: 1.17–2.97, P=0.009) (Table 3). The final multivariate model exhibited good discriminative ability with a concordance index of 0.81 (95% CI: 0.43–0.96) and appropriate calibration as indicated by non-significant Hosmer-Lemeshow test results (P=0.25).

Discussion

In this prospective study, we conducted DCE and DSC simultaneously in 46 glioma patients to investigate the value of these two techniques in assessing the biological characteristics of glioma. The histogram was employed to compare glioma grades and genotypes within the same patient using DCE and DSC. The findings revealed that DSC demonstrated superior performance in determining the grades of glioma, whereas no statistical difference was observed for genotyping. This suggests that DSC primarily reflects histological tumor characteristics, whereas DCE technology exhibits notable distinctions in both genotyping and WHO grading, with better predictive capabilities than DSC. Furthermore, correlation analysis results indicated that DCE may provide a more comprehensive representation of tumor biological characteristics through its diverse features, while DSC solely correlates with MVD without reflecting tumor proliferation characteristics. Survival analysis established MGMT promoter methylation status, elevated values of Ktrans and rCBV as independent prognostic predictors, demonstrating the combined influence of epigenetic regulation and vascular pathobiology on glioma survival.

DSC-MRI in tumor grading: technical merits and genotyping limitations

DSC-MRI, widely adopted for quantifying rCBV and rCBF (29,30), has been established as a reliable biomarker for glioma grading, reflecting histopathological angiogenesis—a conclusion corroborated by our cohort where rCBV achieved an AUC of 0.742 (95% CI: 0.596–0.888) in distinguishing HGGs from LGGs (P=0.008). However, its utility in molecular genotyping remains contentious. While prior studies reported DSC’s potential to differentiate IDH mutation status (AUC =0.82 for rCBV) (31) and 1p/19q codeletion via normalized CBV (nCBV) skewness (AUC =0.690) (32), our analysis revealed no statistically significant association between DSC parameters and IDH status (AUC =0.577, P=0.374 for rCBV; AUC =0.537, P=0.668 for rCBF).

We attribute this discrepancy to two interrelated mechanisms: hemodynamic sensitivity disparities and genotype-specific vascular pathophysiology. DSC-derived metrics (rCBV, rCBF) predominantly reflect macrovascular blood volume and flow dynamics, whereas DCE-MRI captures microvascular permeability (Ktrans, Ve) sensitive to IDH mutation-driven angiogenic dysregulation (22-24,33). DSC’s diagnostic fidelity is further confounded by susceptibility artifacts (e.g., intratumoral hemorrhage, calcifications, motion), which may obscure subtle genotype-correlated perfusion signatures (34,35). This aligns with the suboptimal diagnostic accuracy (67%) of DSC-based radiomic models in predicting endothelial growth factor receptor (EGFR) amplification (36), underscoring DSC’s inadequacy in resolving genotype-specific microenvironmental changes. Thus, while DSC-MRI excels in gliomas grading by reflecting microvascular proliferation, it lacks the resolution to decode molecular subtyping.

DCE-MRI’s superior sensitivity: bridging perfusion metrics with tumor microenvironment

DCE-MRI exhibits superior reliability in glioma genotyping and grading, attributable to its higher spatial resolution, reduced magnetic susceptibility artifacts (stemming from T1-weighted sequence stability), and capacity to quantify endothelial permeability and vascular leakage (16,37,38). Pharmacokinetic parameters such as Ktrans, Ve, Kep, and Vp demonstrate robust correlations with WHO grade, microvascular density (MVD), and vascular endothelial growth factor (VEGF) expression (24,39,40), as evidenced by our cohort where Ktrans differentiated IDH-mutant HGGs from wildtype GBM (AUC =0.847, P<0.001). These findings align with Hilario et al. (41), who reported significantly lower Ktrans in IDH-mutant HGGs versus wildtype counterparts, and Lee et al. (42), linking elevated Ktrans, Ve, and Kep to 1p/19q codeletion.

Biologically, IDH mutation drives metabolic reprogramming via D-2-hydroxyglutarate (2-HG) accumulation, which stabilizes HIF-1α and upregulates VEGF signaling, fostering immature neovascularization with BBB disruption (23,24,33,42). DCE-MRI’s permeability metrics (Ktrans, Ve) capture these pathophysiological features—hyperpermeable, disorganized vasculature—that are imperceptible to DSC’s macrovascular-focused parameters. Zhang et al. (33) further validated this by associating IDH-wildtype gliomas with vascular gene expression profiles linked to Notch1-mediated remodeling, a pathway critical for angiogenesis. Future studies integrating DCE with transcriptomic or proteomic data could decode microenvironmental heterogeneity, particularly in LGGs where angiogenesis-related molecular changes remain understudied (22,33,42).

Prognostic integration of molecular and hemodynamic biomarkers

Our multivariate survival analysis delineates a tripartite prognostic paradigm encompassing molecular genetics (MGMT promoter methylation), vascular permeability (Ktrans), and hemodynamic alterations (rCBV), collectively accounting for 81% of survival variability (C-index =0.81). The dominance of MGMT promoter methylation (HR =4.15) aligns with its established role in alkylating agent resistance (43). The retained prognostic independence of Ktrans (HR =1.08) and rCBV (HR =1.87) underscores their complementary biological significance. Elevated Ktrans reflects hypoxia-driven vascular hyperpermeability through VEGF/matrix metalloproteinase-9 (MMP-9) cascades (44,45), creating niches for immune evasion via programmed death-ligand 1 (PD-L1) upregulation and macrophage polarization (46). Concurrently, increased rCBV indicates neoangiogenic “hotspots” with immature vasculature prone to microhemorrhages—a hallmark of perivascular niche expansion that facilitates tumor stemness (22,33). Clinically, the MGMT-Ktrans-rCBV triad may guide personalized intervention thresholds. For instance, bevacizumab efficacy correlates with baseline Ktrans in recurrent GBM (47), while MGMT-methylated tumors with low rCBV might benefit from de-escalated radiotherapy. Future studies should validate these biomarkers in therapeutic response cohorts and explore their integration with emerging metabolic markers.

Limitations

However, it is important to acknowledge the limitations of our study. First, the glioma sample size is relatively small, which may hinder accuracy and therefore necessitate a larger number of samples to validate these findings. Second, the present study discusses gliomas as a unified entity, while acknowledging the dynamic spatial and temporal heterogeneity induced by the intricate internal microenvironment. Subsequent investigations will employ DCE-MRI to scrutinize the internal microenvironment of gliomas. Third, the lack of a healthy control group precluded the establishment of baseline perfusion parameter values, thereby hindering the differentiation of glioma-associated microvascular alterations from normal physiological vascular heterogeneity. This limitation also impacts pathophysiological interpretations of tumor-associated angiogenesis. Fourth, the absence of longitudinal post-treatment imaging data precluded the evaluation of dynamic changes in DSC/DCE biomarkers following surgery, chemotherapy, or radiotherapy. Future studies tracking temporal variations in perfusion parameters could clarify their correlation with treatment response or tumor recurrence, thereby enhancing the clinical utility of perfusion MRI in personalized therapeutic monitoring.

Conclusions

In summary, PWI, containing DCE and DSC techniques, has served as a research hotspot in the non-invasive identification of glioma genotyping and grading (22,48-50). It has been proposed as a non-invasive approach for characterizing glioma (51). The derived parameters obtained from DCE and DSC possess inherent evaluative value. While we acknowledge the growing emphasis on multimodal approaches in neuroimaging research, the present study was specifically designed to isolate the individual contributions of these two distinct perfusion techniques. This methodological choice was motivated by our aim to avoid potential confounding effects arising from cross-modality interactions and to systematically elucidate the intrinsic value of each perfusion method in molecular subtyping and histopathological grading of gliomas. A comprehensive understanding and proficient application of these two technologies will significantly contribute to our future advancements in the diagnosis, treatment, and prognosis assessment of glioma.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2794/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2794/dss

Funding: This study was supported by research grant from the Natural Science Foundation Program of Liaoning Province (No. 2023-MSLH-407 to M.C.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2794/coif). Y.J. serves as a consultant for Siemens Healthineers in the field of MRI scientific support. Siemens Healthineers did not provide direct funding for this study. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of the First Hospital of China Medical University (No. KLS [2023]-21). All patients provided written informed consent after receiving a detailed description of the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Ostrom QT, Cioffi G, Gittleman H, Patil N, Waite K, Kruchko C, Barnholtz-Sloan JS. CBTRUS Statistical Report: Primary Brain and Other Central Nervous System Tumors Diagnosed in the United States in 2012-2016. Neuro Oncol 2019;21:v1-v100. [Crossref] [PubMed]
2. Jiang T, Nam DH, Ram Z, Poon WS, Wang J, Boldbaatar D, et al. Clinical practice guidelines for the management of adult diffuse gliomas. Cancer Lett 2021;499:60-72. [Crossref] [PubMed]
3. Sun Y, Yang Z, Deng K, Geng Y, Hu X, Song Y, Jiang R. Histogram analysis of quantitative susceptibility mapping and apparent diffusion coefficient for identifying isocitrate dehydrogenase genotypes and tumor subtypes of adult-type diffuse gliomas. Quant Imaging Med Surg 2023;13:8681-93. [Crossref] [PubMed]
4. Gusyatiner O, Hegi ME. Glioma epigenetics: From subclassification to novel treatment options. Semin Cancer Biol 2018;51:50-8. [Crossref] [PubMed]
5. Zhou Y, Xiao D, Jiang X. LncRNA RP3-525N10.2-NFKB1-PROS1 triplet-mediated low PROS1 expression is an onco-immunological biomarker in low-grade gliomas: a pan-cancer analysis with experimental verification. J Transl Med 2022;20:335. [Crossref] [PubMed]
6. Hawkins-Daarud A, Rockne RC, Anderson AR, Swanson KR. Modeling Tumor-Associated Edema in Gliomas during Anti-Angiogenic Therapy and Its Impact on Imageable Tumor. Front Oncol 2013;3:66. [Crossref] [PubMed]
7. Louis DN, Perry A, Wesseling P, Brat DJ, Cree IA, Figarella-Branger D, Hawkins C, Ng HK, Pfister SM, Reifenberger G, Soffietti R, von Deimling A, Ellison DW. The 2021 WHO Classification of Tumors of the Central Nervous System: a summary. Neuro Oncol 2021;23:1231-51. [Crossref] [PubMed]
8. Eckel-Passow JE, Lachance DH, Molinaro AM, Walsh KM, Decker PA, Sicotte H, et al. Glioma Groups Based on 1p/19q, IDH, and TERT Promoter Mutations in Tumors. N Engl J Med 2015;372:2499-508. [Crossref] [PubMed]
9. Sun H, Yin L, Li S, Han S, Song G, Liu N, Yan C. Prognostic significance of IDH mutation in adult low-grade gliomas: a meta-analysis. J Neurooncol 2013;113:277-84. [Crossref] [PubMed]
10. Izquierdo C, Barritault M, Poncet D, Cartalat S, Joubert B, Bruna J, Jouanneau E, Guyotat J, Vasiljevic A, Fenouil T, Berthezène Y, Honnorat J, Meyronet D, Ducray F. Radiological Characteristics and Natural History of Adult IDH-Wildtype Astrocytomas with TERT Promoter Mutations. Neurosurgery 2019;85:E448-56. [Crossref] [PubMed]
11. Lasocki A, Tsui A, Tacey MA, Drummond KJ, Field KM, Gaillard F. MRI grading versus histology: predicting survival of World Health Organization grade II-IV astrocytomas. AJNR Am J Neuroradiol 2015;36:77-83. [Crossref] [PubMed]
12. Arzanforoosh F, Croal PL, van Garderen KA, Smits M, Chappell MA, Warnert EAH. Effect of Applying Leakage Correction on rCBV Measurement Derived From DSC-MRI in Enhancing and Nonenhancing Glioma. Front Oncol 2021;11:648528. [Crossref] [PubMed]
13. Liu J, Li C, Chen Y, Lv X, Lv Y, Zhou J, Xi S, Dou W, Qian L, Zheng H, Wu Y, Chen Z. Diagnostic performance of multiparametric MRI in the evaluation of treatment response in glioma patients at 3T. J Magn Reson Imaging 2020;51:1154-61. [Crossref] [PubMed]
14. Steidl E, Müller M, Müller A, Herrlinger U, Hattingen E. Longitudinal, leakage corrected and uncorrected rCBV during the first-line treatment of glioblastoma: a prospective study. J Neurooncol 2019;144:409-17. [Crossref] [PubMed]
15. Manikis GC, Ioannidis GS, Siakallis L, Nikiforaki K, Iv M, Vozlic D, Surlan-Popovic K, Wintermark M, Bisdas S, Marias K. Multicenter DSC-MRI-Based Radiomics Predict IDH Mutation in Gliomas. Cancers (Basel) 2021.
16. Xue W, Du X, Wu H, Liu H, Xie T, Tong H, Chen X, Guo Y, Zhang W. Aberrant glioblastoma neovascularization patterns and their correlation with DCE-MRI-derived parameters following temozolomide and bevacizumab treatment. Sci Rep 2017;7:13894. [Crossref] [PubMed]
17. Essig M, Nguyen TB, Shiroishi MS, Saake M, Provenzale JM, Enterline DS, Anzalone N, Dörfler A, Rovira À, Wintermark M, Law M. Perfusion MRI: the five most frequently asked clinical questions. AJR Am J Roentgenol 2013;201:W495-510. [Crossref] [PubMed]
18. Heye AK, Culling RD, Valdés Hernández Mdel C, Thrippleton MJ, Wardlaw JM. Assessment of blood-brain barrier disruption using dynamic contrast-enhanced MRI. A systematic review. Neuroimage Clin 2014;6:262-74. [Crossref] [PubMed]
19. Sourbron SP, Buckley DL. Tracer kinetic modelling in MRI: estimating perfusion and capillary permeability. Phys Med Biol 2012;57:R1-33. [Crossref] [PubMed]
20. Mills SJ, du Plessis D, Pal P, Thompson G, Buonacorrsi G, Soh C, Parker GJ, Jackson A. Mitotic Activity in Glioblastoma Correlates with Estimated Extravascular Extracellular Space Derived from Dynamic Contrast-Enhanced MR Imaging. AJNR Am J Neuroradiol 2016;37:811-7. [Crossref] [PubMed]
21. Sourbron SP, Buckley DL. On the scope and interpretation of the Tofts models for DCE-MRI. Magn Reson Med 2011;66:735-45. [Crossref] [PubMed]
22. van Santwijk L, Kouwenberg V, Meijer F, Smits M, Henssen D. A systematic review and meta-analysis on the differentiation of glioma grade and mutational status by use of perfusion-based magnetic resonance imaging. Insights Imaging 2022;13:102. [Crossref] [PubMed]
23. Xie Q, Wu J, Du Z, Di N, Yan R, Pang H, Jin T, Zhang H, Wu Y, Zhang Y, Yao Z, Feng X. DCE-MRI in Human Gliomas: A Surrogate for Assessment of Invasive Hypoxia Marker HIF-1A Based on MRI-Neuronavigation Stereotactic Biopsies. Acad Radiol 2019;26:179-87. [Crossref] [PubMed]
24. Jensen RL, Mumert ML, Gillespie DL, Kinney AY, Schabel MC, Salzman KL. Preoperative dynamic contrast-enhanced MRI correlates with molecular markers of hypoxia and vascularity in specific areas of intratumoral microenvironment and is predictive of patient outcome. Neuro Oncol 2014;16:280-91. [Crossref] [PubMed]
25. Xing Z, Zhang H, She D, Lin Y, Zhou X, Zeng Z, Cao D. IDH genotypes differentiation in glioblastomas using DWI and DSC-PWI in the enhancing and peri-enhancing region. Acta Radiol 2019;60:1663-72. [Crossref] [PubMed]
26. Zhang HW, Lyu GW, He WJ, Lei Y, Lin F, Wang MZ, Zhang H, Liang LH, Feng YN, Yang JH. DSC and DCE Histogram Analyses of Glioma Biomarkers, Including IDH, MGMT, and TERT, on Differentiation and Survival. Acad Radiol 2020;27:e263-71. [Crossref] [PubMed]
27. Paulson ES, Schmainda KM. Comparison of dynamic susceptibility-weighted contrast-enhanced MR methods: recommendations for measuring relative cerebral blood volume in brain tumors. Radiology 2008;249:601-13. [Crossref] [PubMed]
28. Lewis D, Li KL, Waqar M, Coope DJ, Pathmanaban ON, King AT, Djoukhadar I, Zhao S, Cootes TF, Jackson A, Zhu X. Low-dose GBCA administration for brain tumour dynamic contrast enhanced MRI: a feasibility study. Sci Rep 2024;14:4905. [Crossref] [PubMed]
29. Moon HH, Park JE, Kim N, Kim YH, Song SW, Hong CK, Kim JH, Kim HS. Prospective longitudinal analysis of imaging-based spatiotemporal tumor habitats in glioblastoma, IDH-wild type: implication in patient outcome using multiparametric physiologic MRI. BMC Cancer 2024;24:1197. [Crossref] [PubMed]
30. Wang H, Zeng L, Wu H, Tian J, Xie H, Zhang L, Ran Q, Zhong P, Chen L, Yi L, Wang S. Preoperative vascular heterogeneity based on dynamic susceptibility contrast MRI in predicting spatial pattern of locally recurrent high-grade gliomas. Eur Radiol 2024;34:1982-93. [Crossref] [PubMed]
31. Xing Z, Yang X, She D, Lin Y, Zhang Y, Cao D. Noninvasive Assessment of Mutational Status in World Health Organization Grade II and III Astrocytomas Using DWI and DSC-PWI Combined with Conventional MR Imaging. AJNR Am J Neuroradiol 2017;38:1138-1144. [Crossref] [PubMed]
32. Lee MK, Park JE, Jo Y, Park SY, Kim SJ, Kim HS. Advanced imaging parameters improve the prediction of diffuse lower-grade gliomas subtype, IDH mutant with no 1p19q codeletion: added value to the T2/FLAIR mismatch sign. Eur Radiol 2020;30:844-54. [Crossref] [PubMed]
33. Zhang L, He L, Lugano R, Roodakker K, Bergqvist M, Smits A, Dimberg A. IDH mutation status is associated with distinct vascular gene expression signatures in lower-grade gliomas. Neuro Oncol 2018;20:1505-16. [Crossref] [PubMed]
34. Liang J, Liu D, Gao P, Zhang D, Chen H, Shi C, Luo L. Diagnostic Values of DCE-MRI and DSC-MRI for Differentiation Between High-grade and Low-grade Gliomas: A Comprehensive Meta-analysis. Acad Radiol 2018;25:338-48. [Crossref] [PubMed]
35. Wang C, Liu F, Zhang L, Song Y, Pan Z, Li G, Bian H, Yuan X. Comparison of normalized cerebral blood flow between different post-processing methods of dynamic susceptibility contrast perfusion-weighted imaging and arterial spin labeling in gliomas with different grading. Quant Imaging Med Surg 2024;14:8720-33. [Crossref] [PubMed]
36. Hu LS, Wang L, Hawkins-Daarud A, Eschbacher JM, Singleton KW, Jackson PR, et al. Uncertainty quantification in the radiogenomics modeling of EGFR amplification in glioblastoma. Sci Rep 2021;11:3932. [Crossref] [PubMed]
37. Pak E, Choi KS, Choi SH, Park CK, Kim TM, Park SH, Lee JH, Lee ST, Hwang I, Yoo RE, Kang KM, Yun TJ, Kim JH, Sohn CH. Prediction of Prognosis in Glioblastoma Using Radiomics Features of Dynamic Contrast-Enhanced MRI. Korean J Radiol 2021;22:1514-24. [Crossref] [PubMed]
38. Zhao K, Huang H, Gao E, Qi J, Chen T, Zhao G, Zhao G, Zhang Y, Wang P, Bai J, Zhang Y, Hou Z, Cheng J, Ma X. Distributed parameter model of dynamic contrast-enhanced MRI in the identification of IDH mutation, 1p19q codeletion, and tumor cell proliferation in glioma patients. Front Oncol 2024;14:1333798. [Crossref] [PubMed]
39. Piludu F, Marzi S, Pace A, Villani V, Fabi A, Carapella CM, Terrenato I, Antenucci A, Vidiri A. Early biomarkers from dynamic contrast-enhanced magnetic resonance imaging to predict the response to antiangiogenic therapy in high-grade gliomas. Neuroradiology 2015;57:1269-80. [Crossref] [PubMed]
40. Dominietto M, Dobosz M, Bürgi S, Renner A, Zahlmann G, Scheuer W, Rudin M. Quantification of antiangiogenic treatment effects on tissue heterogeneity in glioma tumour xenograft model using a combination of DCE-MRI and 3D-ultramicroscopy. Eur Radiol 2017;27:2894-902. [Crossref] [PubMed]
41. Hilario A, Hernandez-Lain A, Sepulveda JM, Lagares A, Perez-Nuñez A, Ramos A. Perfusion MRI grading diffuse gliomas: Impact of permeability parameters on molecular biomarkers and survival. Neurocirugia 2019;30:11-8. (Engl Ed). [Crossref] [PubMed]
42. Lee JY, Ahn KJ, Lee YS, Jang JH, Jung SL, Kim BS. Differentiation of grade II and III oligodendrogliomas from grade II and III astrocytomas: a histogram analysis of perfusion parameters derived from dynamic contrast-enhanced (DCE) and dynamic susceptibility contrast (DSC) MRI. Acta Radiol 2018;59:723-31. [Crossref] [PubMed]
43. Tomar MS, Kumar A, Srivastava C, Shrivastava A. Elucidating the mechanisms of Temozolomide resistance in gliomas and the strategies to overcome the resistance. Biochim Biophys Acta Rev Cancer 2021;1876:188616. [Crossref] [PubMed]
44. Awasthi R, Rathore RK, Soni P, Sahoo P, Awasthi A, Husain N, Behari S, Singh RK, Pandey CM, Gupta RK. Discriminant analysis to classify glioma grading using dynamic contrast-enhanced MRI and immunohistochemical markers. Neuroradiology 2012;54:205-13. [Crossref] [PubMed]
45. Zhang J, Xue W, Xu K, Yi L, Guo Y, Xie T, Tong H, Zhou B, Wang S, Li Q, Liu H, Chen X, Fang J, Zhang W. Dual inhibition of PFKFB3 and VEGF normalizes tumor vasculature, reduces lactate production, and improves chemotherapy in glioblastoma: insights from protein expression profiling and MRI. Theranostics 2020;10:7245-59. [Crossref] [PubMed]
46. Wei H, Fu B. Make killers sweeter: targeting metabolic checkpoints of NK cells. Nat Immunol 2020;21:970-1. [Crossref] [PubMed]
47. Battaglin F, Ou FS, Qu X, Hochster HS, Niedzwiecki D, Goldberg RM, Mayer RJ, Ashouri K, Zemla TJ, Blanke CD, Venook AP, Kabbarah O, Lenz HJ, Innocenti F. HER2 Gene Expression Levels Are Predictive and Prognostic in Patients With Metastatic Colorectal Cancer Enrolled in CALGB/SWOG 80405. J Clin Oncol 2024;42:1890-902. [Crossref] [PubMed]
48. Buz-Yalug B, Turhan G, Cetin AI, Dindar SS, Danyeli AE, Yakicier C, Pamir MN, Özduman K, Dincer A, Ozturk-Isik E. Identification of IDH and TERTp mutations using dynamic susceptibility contrast MRI with deep learning in 162 gliomas. Eur J Radiol 2024;170:111257. [Crossref] [PubMed]
49. Hu LS, D'Angelo F, Weiskittel TM, Caruso FP, Fortin Ensign SP, Blomquist MR, et al. Integrated molecular and multiparametric MRI mapping of high-grade glioma identifies regional biologic signatures. Nat Commun 2023;14:6066. [Crossref] [PubMed]
50. Moltoni G, Romano A, Capriotti G, Campagna G, Ascolese AM, Romano A, Dellepiane F, Minniti G, Signore A, Bozzao A. ASL, DSC, DCE perfusion MRI and 18F-DOPA PET/CT in differentiating glioma recurrence from post-treatment changes. Radiol Med 2024;129:1382-93. [Crossref] [PubMed]
51. Kouwenberg V, van Santwijk L, Meijer FJA, Henssen D. Reliability of dynamic susceptibility contrast perfusion metrics in pre- and post-treatment glioma. Cancer Imaging 2022;22:28. [Crossref] [PubMed]