Grading of glioma and prediction of IDH mutation status via longitudinal relaxation time in rotating frame mapping on 5-T magnetic resonance imaging: added value to amide proton transfer-weighted imaging

Introduction

Glioma is the most common primary brain tumor in adults (1), and the annual incidence is estimated to range between 5 and 7 cases per 100,000 population (2-4). The therapeutic approach and prognosis of glioma differ substantially according to glioma grade and isocitrate dehydrogenase (IDH) mutational status (5). However, preoperative assessment through noninvasive measures remains challenging. The criteria for tumor grading are based on histopathological assessment with biopsy or surgical resection, which is invasive and subject to sampling error. Inappropriate sampling from sites with a lower histologic tumor grade can lead to underestimation of the true grade (6,7). Molecular markers, particularly IDH mutations, are strongly associated with the prognosis of patients with glioma and are considered critical prognostic factors (8). IDH mutation status has been incorporated into the 2016 and 2021 World Health Organization (WHO) classifications of central nervous system (CNS) tumors. Accurate preoperative imaging analysis facilitates sample selection, surgical planning, and prognostication (9). In addition, noninvasive imaging analysis is particularly important for patients who are unable or unwilling to undergo surgery or biopsy (10).

Gadolinium-enhanced magnetic resonance imaging (MRI) is the standard imaging protocol for preoperative glioma grading. However, the accuracy of conventional contrast-enhanced MRI in grading gliomas is unsatisfactory (11). Currently, noncontrast MRI sequences, such as advanced MRI techniques, provide several advantages for grading gliomas prior to surgery. For example, the longitudinal relaxation time in the rotating frame (T1rho) mapping and amide proton transfer (APT)-weighted chemical exchange saturation transfer (CEST) imaging may reflect differences in histopathological features that may inform glioma grading (8,12,13).

APT imaging is a novel contrast-free technique that detects amide protons of proteins in tissues (14). Numerous studies have confirmed the value of APT in the clinical management of glioma, meningioma, cerebral ischemic diseases, and neurodegenerative diseases (15-19). However, challenges in APT imaging remain, including the long scanning time and inhomogeneity of the radiofrequency transmit field (B1) and main static magnetic field (B0), which limit its clinical application.

T1rho provides unique insights into the functional variations associated with microstructure and metabolism changes in tissues (20). T1rho is sensitive to the low-frequency interactions between macromolecules and the water pool and thus can be used to assess the macromolecular environment inside tissues (21). Previous studies have focused on the application of T1rho in examining the musculoskeletal system, neurodegenerative diseases, and neuropsychiatric disorders (22-25); however, few have investigated its utility in glioma. In Cao et al.’s study (26), T1rho mapping demonstrated potential value for the noninvasive prediction of molecular subtypes and WHO grades of gliomas. Moreover, Villanueva-Meyer et al. (13) reported that T1rho imaging may help differentiate tumor-related edema. However, these studies generally evaluated T1rho as a standalone technique and did not integrate it with other advanced molecular imaging approaches, such as the APT-weighted imaging described above. In addition, the bulk of the research on T1rho has been performed in typical 1.5-T or 3.0-T fields, and human studies on T1rho quantification in ultrahigh fields are rare (27).

A higher magnetic field strength provides a superior signal-to-noise ratio (SNR) (28), which is critical for the robust quantification of subtle T1rho relaxation time differences and APT-weighted magnetization transfer effects. Therefore, 5-T MRI can enable higher spatial resolution to image the microstructure (29). In addition, high-field environments facilitate optimized pulse sequence designs (e.g., longer spin-lock durations for T1rho mapping and controlled radiofrequency pulse amplitudes for APT imaging) that may be compromised at lower fields due to SNR limitations. Thus, 5-T MRI offers better sensitivity for the probing of metabolic tissue properties (30).

To the best of our knowledge, there are no reports on the combined use of T1rho and APT techniques for examining patients with glioma. We hypothesized that T1rho and APT in 5-T ultrahigh-field MRI, either alone or in combination, can be used to grade gliomas and predict IDH mutation status. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2294/rc).

Methods

Patient population

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments, and was approved by Institutional Ethics Board of The First Affiliated Hospital of University of Science and Technology of China (USTC) (approval No. 2024-RE-333). Informed consent was obtained from all patients. Patients were consecutively recruited from the Department of Neurosurgery at the First Affiliated Hospital of USTC. In October 2023, the T1rho and APT imaging sequences were incorporated into our standard MRI protocols for patients confirmed with or suspected to have a brain tumor. The inclusion criteria were as follows: (I) pathologically confirmed adult-type diffuse gliomas; (II) ≥18 years of age; (III) no prior treatment related to gliomas; and (IV) preoperatively obtained T1rho and APT images. Meanwhile, the exclusion criteria were as follows: (I) diagnosis of tumors other than adult-type diffuse gliomas; (II) unavailable IDH mutation status; (III) contraindications to MRI; (IV) images with severe motion artifacts; and (V) degraded quality of T1rho or APT images. Between January 2024 and April 2025, 50 consecutive adult patients were enrolled. The interval between MRI and surgery was less than 1 week for all patients. Pathology was performed by neuropathologists according to the 2021 WHO classification of tumors of the CNS (31). IDH mutation status was determined through either next-generation sequencing or immunohistochemistry (32,33).

Image acquisition

All patients were examined on a 5-T scanner (uMR Jupiter, United Imaging Healthcare, Shanghai, China) with a 48-channel head coil. The conventional MRI sequences included axial T1-weighted fluid-attenuated inversion recovery imaging [repetition time (TR)/echo time (TE), 1,600/7.44 ms; inversion time, 705 ms; field of view (FOV), 200×230 mm²; slice thickness, 5 mm; acquired matrix, 376×432; and reconstructed matrix, 752×864], axial T2-weighted imaging (TR/TE, 4,500/108.24 ms; FOV, 200×230 mm²; slice thickness, 5 mm; acquired matrix, 501×576; and reconstructed matrix, 752×864), and axial three-dimensional (3D) contrast-enhanced T1-weighted imaging (TR/TE, 7.0/2.3 ms; FOV, 220×256 mm²; acquired slice thickness, 0.5 mm; reconstructed slice thickness, 0.3 mm; acquired matrix, 275×320; and reconstructed matrix, 413×480).

T1rho mapping and APT-weighted imaging were performed before contrast agent administration. T1rho mapping was performed via the 3D segmented radiofrequency gradient echo approach under the following parameters: TR/TE, 9.20/4.27 ms; slice thickness, 3 mm; FOV, 220×220 mm²; acquired matrix, 256×256; reconstructed matrix, 512×512; spin-lock frequency, 500 Hz; and spin-lock time, 0, 25, and 55 ms. An adiabatically prepared constant-amplitude on-resonant spin-lock preparation pulse was used for T1rho preparation, which is robust to B0 and B1 inhomogeneities (27).

APT-weighted imaging was acquired through use of a 2D single-shot sequence with a fast spin-echo readout under the following parameters: TR/TE, 6,000/7.62 ms; slice thickness, 8 mm; FOV, 220×220 mm²; acquired matrix, 128×128; and presaturation pulses with frequency offsets at ±4, ±3.5, and ±3 ppm. The 8-mm slice thickness for APT-CEST imaging was used as a deliberate tradeoff to maximize the coverage of the tumor parenchyma within a single slice. Moreover, to prioritize the SNR for robust APT quantification, a relatively high thickness was applied to this research-specific sequence.

For B0 shimming, prescan B0 mapping was applied to produce a B0 field distribution across the FOV. Subsequently, the active shim coils generated the supplemental magnetic field, which was superimposed on the uncorrected field to calibrate the B0 field. The B1 map was generated via a rapid B1 mapping method described in the literature (34). The continuous-wave rectangular pulse was used for presaturation. CEST data with saturation intensities of 1.5, 2, and 3 µT were collected for B1 correction. The actual applied B1 was calculated based on the B1 map, and the saturated image with an actual saturation intensity of 2 µT was obtained through linear interpolation.

Image analysis

The imaging data were analyzed by postprocessing workstation of the 5-T MRI system (uWS-MR, United Imaging Healthcare), which automatically generated T1rho and magnetization transfer ratio asymmetry (MTRasym) values fitted by the following formulas for further analysis:

S(TSL)=S0(T1rhosequence)e−TSL/T1rho[1]

MTRasym=[Ssat(−Δω)−Ssat(+Δω)]/S0(APTsequence)[2]

where S(TSL) is the signal intensity with spin-lock time, S0(T1rhosequence) is the signal intensity of the T1rho sequence without application of the spin-lock module, Ssat is the signal intensity with presaturation, S0(APTsequence) is the signal intensity of APT imaging without presaturation, and Δω is the frequency offset from water. The targeted chemical exchange of interest in this study was APT; thus, Δω was set to 3.5 ppm.

Briefly, 3D contrast-enhanced T1-weighted imaging and T1rho mapping or APT-weighted imaging were first coregistered and fused via a rigid registration method on the postprocessing workstation, which helped provide precise anatomical information. On the basis of these images, regions of interest (ROIs) were marked accurately, with large vessels and cystic, necrotic, or hemorrhagic components being avoided (26).

Circular ROIs were carefully placed in the solid component and peritumoral edema of the tumor as determined by visual inspection of the 3D contrast-enhanced T1-weighted imaging. The ROIs were then mapped to the corresponding coregistered T1rho and APT MTRasym maps to obtain the corresponding T1rho and APT MTRasym values for tumor and edema, which were recorded as T1rho-tumor, T1rho-edema, APT-tumor, and APT-edema, respectively. Circular ROIs were then placed in the contralateral normal-appearing white matter (WM), with T1rho and APT parameters being recorded as T1rho-WM and APT-WM, respectively. To minimize confounding factors in the analysis, ROIs for the lesion and the contralateral normal-appearing WM were placed at the same slice level, and a consistent ROI size was used (1). The ROI size at each lesion ranged from 29.5 to 47.3 mm².

The relative T1rho of the tumor (rT1rho-tumor) was defined as the difference between T1rho-tumor and T1rho-WM. The relative T1rho of edema (rT1rho-edema) was defined as the difference between T1rho-edema and T1rho-WM. The relative APT MTRasym of tumor (rAPT-tumor) and edema (rAPT-edema) were calculated in the same manner as that described above.

All the ROIs were manually placed by two experienced neuroradiologists (with 10 and 15 years of experience in neuroradiology, respectively) who were blinded to the histopathological data. The average values of the two neuroradiologists were used for statistical analysis.

Statistical analysis

Statistical analyses were performed with SPSS software version 26.0 (IBM Corp., Armonk, NY, USA) and R software version 4.5.2 (The R Foundation for Statistical Computing, Vienna, Austria). The normality of the variables was formally assessed via the Shapiro-Wilk test. Because many of the continuous measures evaluated in this study exhibited skewed distributions, they are summarized as the median and interquartile range (IQR), with comparisons performed via the nonparametric test. The interobserver agreement for each measured value from the two neuroradiologists was analyzed according to the intraclass correlation coefficient (ICC), with ICC values greater than 0.75 indicating a correlation. Differences in T1rho and APT parameters between the low-grade glioma (LGG) and high-grade glioma (HGG) groups and between the IDH mutant-type and IDH wild-type groups were evaluated via the Mann-Whitney U test. LGG was defined as CNS WHO grade 2 and HGG as CNS WHO grade 3–4 (35,36). Furthermore, the parameters of grade 2, 3, and 4 gliomas were compared via the Kruskal-Wallis H test. Receiver operating characteristic (ROC) curves of each parameter and the curve of the combination of the T1rho and APT parameters were used to evaluate the ability of each parameter to grade gliomas and predict IDH mutation status. The optimal thresholds were determined by maximizing the Youden index, and the sensitivity and specificity at the threshold values were then recorded. The DeLong test was applied to assess statistical differences between AUCs derived from single parameters and the combined model. Values of P<0.05 were considered to indicate a significant difference.

Results

Patient characteristics

Among the 50 patients included in the study, 35 were males and 15 were females, with a mean age of 55.2±13.4 years. Histopathologic examinations revealed that 12 patients were diagnosed with LGG (WHO grade 2) and 38 with HGG (6 with WHO grade 3 and 32 with WHO grade 4). Among them, 31 were IDH wild type and 19 were IDH mutant type. The patient characteristics are listed in Table 1.

According to the 2021 WHO classification, the tumor subtypes in our cohort were further stratified as follows: oligodendroglioma, IDH-mutant and 1p/19q-codeleted (WHO grade 2: n=4; WHO grade 3: n=3); astrocytoma, IDH-mutant (WHO grade 2: n=7; WHO grade 3: n=4; WHO grade 4: n=1); and glioblastoma, IDH-wildtype (WHO grade 4: n=31).

Interobserver agreement

The interobserver agreement was almost perfect for all measured values in the solid component of the tumor, peritumoral edema area, and contralateral normal-appearing WM: the ICC values for the T1rho-based measurements were 0.955, 0.968, and 0.913, respectively, while those for the APT-based measurements were 0.853, 0.820, and 0.924, respectively.

LGG versus HGG

Among the whole patient population, T1rho-tumor, T1rho-edema, APT-tumor, and APT-edema were significantly higher than T1rho-WM and APT-WM (P<0.001). Compared with patients with LGG, those with HGG had significantly higher T1rho-tumor, rT1rho-tumor, T1rho-edema, rT1rho-edema, APT-tumor, rAPT-tumor, APT-edema, and rAPT-edema values (P≤0.036) (Table 2).

Parameters with significant differences between LGGs and HGGs and a combined model were subjected to ROC analysis (Figure 1). The cutoff value, sensitivity, specificity, and area under the curve (AUC) of each parameter were calculated and are provided in Table 3. T1rho-edema achieved the highest AUC (0.974) in differentiating HGGs from LGGs. The cutoff value with the highest Youden index for T1rho-edema, indicating ideal discriminative power, was 103.5 ms (sensitivity, 97.4%; specificity, 91.7%). Combining T1rho-edema with rAPT-tumor increased the diagnostic ability from 0.974 to 0.978 (sensitivity, 92.1%; specificity, 100.0%). Typical images of patients with LGG and HGG are shown in Figure 2. The DeLong test demonstrated that for differentiating HGG from LGG, the combined model achieved a significantly higher AUC than did rT1rho-tumor (P=0.023), APT-tumor (P=0.027), APT-edema (P<0.001), and rAPT-edema (P<0.001), whereas the difference compared with T1rho-tumor (P=0.081), rAPT-tumor (P=0.066), T1rho-edema (P=0.483), and rT1rho-edema (P=0.112) was not statistically significant.

Figure 1 Receiver operating characteristic curves for single parameters and the combined model in differentiating high-grade glioma from low-grade glioma. APT, amide proton transfer; rAPT, relative amide proton transfer; rT1rho, relative longitudinal relaxation time in rotating frame; T1rho, longitudinal relaxation time in rotating frame.

Figure 2 Examples of imaging maps acquired from two representative patients. (A-E) Glioblastoma (WHO 4) in a 46-year-old female. (F-J) Oligodendroglioma (WHO 2) in a 64-year-old female. Compared with those in the low-grade case, the T1rho values in the high-grade case were higher in both the solid portion and peritumoral edema areas, and APT-weighted imaging revealed a mild increase APT value in the solid portion and peritumoral edema areas. (A,F) T1-weighted imaging. (B,G) T2-weighted imaging. (C,H) Contrast-enhanced T1-weighted imaging. (D,I) T1rho mapping. (E,J) APT-weighted imaging. APT, amide proton transfer; T1rho, longitudinal relaxation time in rotating frame; WHO, World Health Organization.

IDH wild type versus IDH mutant type

Among the 19 IDH-mutant tumors, 17 were IDH1-mutant and 2 were IDH2-mutant. Compared with the IDH mutant-type group, the IDH wild-type group had significantly higher T1rho-tumor, rT1rho-tumor, T1rho-edema, rT1rho-edema, APT-tumor, and rAPT-tumor values (P≤0.003) (Table 2). Parameters with significant differences between the IDH wild-type and IDH mutant-type groups and the combined model were subjected to ROC analysis (Figure 3). The cutoff value, sensitivity, specificity, and AUC of each parameter were calculated (Table 4). T1rho-edema achieved the highest AUC (0.883) in differentiating IDH wild-type from IDH mutant-type glioma. The cutoff value with the highest Youden index for T1rho-edema, indicating ideal discriminative power, was 123.6 ms (sensitivity, 90.3%; specificity, 89.5%). However, by combining T1rho-edema with rAPT-tumor, the AUC slightly increased from 0.883 to 0.890, while the sensitivity slightly decreased (sensitivity, 87.1%; specificity, 89.5%). The DeLong test indicated no statistically significant differences between the combined model and any of the individual parameters (P≥0.114).

Figure 3 Receiver operating characteristic curves for T1rho-edema and rT1rho-edema in differentiating IDH wild-type and IDH mutant-type. APT, amide proton transfer; IDH, isocitrate dehydrogenase; rAPT, relative amide proton transfer; rT1rho, relative longitudinal relaxation time in rotating frame; T1rho, longitudinal relaxation time in rotating frame.

Glioma grade

Generally, significant differences in T1rho-tumor, rT1rho-tumor, T1rho-edema, rT1rho-edema, APT-tumor, rAPT-tumor, APT-edema, and rAPT-edema were observed between WHO grade 2, 3, and 4 gliomas (P≤0.007) (Table 5).

The T1rho-edema, APT-edema, and rAPT-edema values were significantly higher in WHO grade 3 gliomas than in WHO grade 2 gliomas (P=0.049, P=0.005, and P=0.001, respectively). The rAPT-edema values were significantly higher in WHO grade 4 gliomas than in WHO grade 3 gliomas (P=0.024). The T1rho-tumor, rT1rho-tumor, T1rho-edema, rT1rho-edema, APT-tumor, and rAPT-tumor values of grade 4 gliomas were significantly higher than those of grade 2 gliomas (P≤0.002).

Discussion

A key finding of this study was that T1rho mapping and APT-weighted CEST imaging, alone or in combination, could be used to differentiate LGGs from HGGs and predict IDH mutation status. The combination of these two sequences may provide incremental diagnostic performance, as reflected by a numerically higher AUC, although the AUC differences were not statistically significant in several comparisons.

In general, T1rho provides additional diagnostic value in the diagnosis of glioma. In a preliminary study by Cao et al. (26), the T1rho value of the solid component of the tumor had an AUC of 0.841 and the T1rho value of the peritumoral edema an AUC of 0.772 when T1rho was used to differentiate HGG from LGG at 3 T. The T1rho in our study demonstrated greater diagnostic performance than that reported by Cao et al. According to the ROC analyses, the maximum AUC value of the T1rho-edema alone was 0.974. This finding might be attributed to chemical exchange between different chemical shift points increasing rapidly under ultrahigh magnetic fields and contributing to rotational frame relaxation (27). Studies have confirmed that the T1rho sequence has good repeatability and consistency in normal human brain and neck tissues (21,37,38). Ultrahigh field MR scanners have a higher SNR than do conventional scanners and can provide finer details of T1rho mapping of brain lesions (39).

As a CEST technique, APT imaging has been commonly used evaluating glioma (40,41). Xu et al. (42) retrospectively examined patients with 51 glioma and found that APT MTRasym had good discriminatory ability in distinguishing different grades of glioma, with AUCs of 0.90, 0.94, and 1.00 in distinguishing between grades 2 and 3, grades 3 and 4, and grades 2 and 4, respectively. In Kang et al.’s study (7), AUC was greater than 0.8 when the mean value of APT or the 90th percentile value of APT was used to discriminate HGGs from LGGs. The 5-T scanner in our study was only equipped with a two-dimensional APT sequence, which was acquired in a single slice due to the long scanning duration. The slice displaying the maximum cross-sectional area of the tumor was imaged, which might not comprehensively represent the characteristics of the whole tumor. This may be one of the reasons for the moderate diagnostic efficacy of APT-related parameters encountered in our study. For multislice CEST imaging, the manufacturer of the devices we employ (United Imaging Healthcare) has been developing new APT-CEST sequences for 3D data acquisition. Our institution will update our scanning sequence based on this method and conduct further research on APT.

In this study, the measurement of T1rho in the peritumoral edema of gliomas showed high diagnostic efficacy, not only in the differentiation of HGGs and LGGs but also in the determination of IDH status. The formation mechanism of brain parenchymal edema around LGGs and that around HGGs differ. LGGs are less invasive and are associated with minimal disruption of the blood-brain barrier and typically mild peritumoral edema. In HGGs, the mass effect is more obvious, the tumor cells secrete a large number of vasoactive substances—which may also cause inflammatory reactions—and the blood-brain barrier is more severely damaged (43,44). IDH-mutant gliomas are generally associated with less peritumoral edema than are IDH wild-type gliomas are, reflecting differences in tumor metabolism, vascular biology, and the microenvironment (45). Conventional MRI can only allow for visual and subjective assessment of the peritumoral edema, whereas T1rho can be used to quantitatively and objectively analyze the differences in edema area. Unlike APT, the T1rho scan in this study had higher resolution and may be more accurate for ROI placement and drawing in the edema area.

Although 5-T MRI remains primarily a research tool due to its limited clinical availability (compared with the more common 3- or 1.5-T scanners), it offers distinct advantages for advanced quantitative techniques, such as T1rho mapping and APT-weighted imaging, including an enhanced SNR and improved sensitivity in detecting subtle biochemical changes in the tissue (34,46). These benefits are critical for the robust quantification of T1rho and APT values. In this study, T1rho and APT-CEST magnetization transfer ratio asymmetry maps were provided inline via a 5-T scanner without reliance on third-party software. This streamlines the workflow for radiologic technologists and radiologists via the integration of quantification algorithms natively. Additionally, 5-T MRI has been approved by China’s National Medical Products Administration (NMPA), The US Food and Drug Administration (FDA), and European Union Medical Device Regulation (MDR) for whole-body clinical use and may be applied in a wider array of clinical settings (47).

Compared with either technique alone, the combined model performed better in the differential diagnosis of LGGs and HGGs. This enhancement can be attributed to the leveraging of complementary biological information, with T1rho reflecting macromolecular relaxation properties and APT measuring peptide and protein proton exchanges with water protons (39,48). This synergy is particularly valuable in heterogeneous tumors for which single metrics may lack robustness. To our knowledge, this is the first study to validate the joint use of these ultrahigh-field MRI techniques for IDH mutation prediction and glioma grading, addressing a deficiency in previous studies that often evaluated these factors in isolation. Although conventional contrast-enhanced MRI is a simple procedure with a short scanning time, T1rho and APT imaging provide molecular information without the use of contrast media. This information can be used to supplement conventional scanning series for a wide range of patients.

In our protocol, the acquisition time was approximately 6 minutes and 59 seconds for T1rho and 3 minutes 7 seconds for APT, resulting in an additional scan-time burden of about 10 minutes. This additional time may be acceptable in preoperative protocols but may present a challenge in uncooperative patients. Both techniques may support preoperative risk stratification, guide surgical planning by highlighting biologically aggressive regions for targeted sampling, and assist treatment planning when conventional MRI findings are equivocal. In particular, positive findings within peritumoral edema may help identify the boundaries of tumor infiltration. At present, the global installed base of 5-T magnetic resonance platforms remains limited, and research evaluating the repeatability of T1rho and APT sequences across different field strengths remains lacking. The studies that have been conducted primarily evaluated the repeatability and stability of APT or T1rho imaging at either 3 or 5 T individually (21,27,49,50). Encouragingly, however, these studies have consistently reported good reproducibility. The T1rho and APT-weighted sequences used in our study are not currently part of routine clinical protocols but are being increasingly assessed in research settings. In clinical neuro-oncology workflows, T1rho mapping and APT-weighted imaging are intended to serve as complementary, noninvasive techniques that can be appended to standard preoperative brain tumor MRI protocols rather than replacing histopathology. Before these biomarkers can be broadly adopted for decision-making, prospective studies with larger cohorts, external validation, and, ideally, multicenter standardization are required to establish robustness and generalizability and to determine whether incremental imaging value translates into clinically meaningful improvements in management and outcomes.

Our study involved several limitations that should be addressed. First, the sample size was relatively small, particularly the number of patients with LGG, which may limit the statistical power and generalizability of the results. Second, the number of IDH2-mutant tumors was very limited (n=2), precluding a separate comparison between IDH1- and IDH2-mutant subgroups. Our institution is continuing to enroll patients to expand the cohort, with the aim of validating these findings in a larger sample and enabling more detailed subgroup analyses in subsequent studies. Third, manual ROI marking was based on visual analysis and could have introduced bias. In this study, we used 3D contrast-enhanced T1-weighted imaging for registration and used the difference measurement of the contralateral normal WM area to minimize bias. Finally, the reproducibility of manual ROI placement was demonstrated in this study by the high interobserver agreement between two experienced neuroradiologists. However, semiautomatic or automatic segmentation methods based on artificial intelligence should be considered in future work to further improve the reliability of the quantitative data.

Conclusions

Both T1rho mapping and APT-weighted CEST imaging can be used to grade gliomas and determine IDH mutation status. The differentiation of HGG and LGG can be further improved by combining the two techniques.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2294/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2294/dss

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-aw-2294/coif). X.S. is an employee of United Imaging Healthcare. 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 Institutional Ethics Board of The First Affiliated Hospital of University of Science and Technology of China (USTC) (No. 2024-RE-333) and informed consent was taken from all the patients.

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. Hou H, Chen W, Diao Y, Wang Y, Zhang L, Wang L, Xu M, Yu J, Song T, Liu Y, Yuan Z. 3D Amide Proton Transfer-Weighted Imaging for Grading Glioma and Correlating IDH Mutation Status: Added Value to 3D Pseudocontinuous Arterial Spin Labelling Perfusion. Mol Imaging Biol 2023;25:343-52. [Crossref] [PubMed]
2. Sehmer EA, Hall GJ, Greenberg DC, O'Hara C, Wallingford SC, Wright KA, Green AC. Incidence of glioma in a northwestern region of England, 2006-2010. Neuro Oncol 2014;16:971-4. [Crossref] [PubMed]
3. Cote DJ, Ostrom QT, Gittleman H, Duncan KR. CreveCoeur TS, Kruchko C, Smith TR, Stampfer MJ, Barnholtz-Sloan JS. Glioma incidence and survival variations by county-level socioeconomic measures. Cancer 2019;125:3390-400. [Crossref] [PubMed]
4. Ostrom QT, Price M, Neff C, Cioffi G, Waite KA, Kruchko C, Barnholtz-Sloan JS. CBTRUS Statistical Report: Primary Brain and Other Central Nervous System Tumors Diagnosed in the United States in 2016-2020. Neuro Oncol 2023;25:iv1-iv99. [Crossref] [PubMed]
5. Wu M, Jiang T, Guo M, Duan Y, Zhuo Z, Weng J, Xie C, Sun J, Li J, Cheng D, Liu X, Du J, Zhang X, Zhang Y, Liu Y. Amide proton transfer-weighted imaging and derived radiomics in the classification of adult-type diffuse gliomas. Eur Radiol 2024;34:2986-96. [Crossref] [PubMed]
6. Youssef G, Miller JJ. Lower Grade Gliomas. Curr Neurol Neurosci Rep 2020;20:21. [Crossref] [PubMed]
7. Kang XW, Xi YB, Liu TT, Wang N, Zhu YQ, Wang XR, Guo F. Grading of Glioma: combined diagnostic value of amide proton transfer weighted, arterial spin labeling and diffusion weighted magnetic resonance imaging. BMC Med Imaging 2020;20:50. [Crossref] [PubMed]
8. Joo B, Han K, Ahn SS, Choi YS, Chang JH, Kang SG, Kim SH, Zhou J, Lee SK. Amide proton transfer imaging might predict survival and IDH mutation status in high-grade glioma. Eur Radiol 2019;29:6643-52. [Crossref] [PubMed]
9. Togao O, Yoshiura T, Keupp J, Hiwatashi A, Yamashita K, Kikuchi K, Suzuki Y, Suzuki SO, Iwaki T, Hata N, Mizoguchi M, Yoshimoto K, Sagiyama K, Takahashi M, Honda H. Amide proton transfer imaging of adult diffuse gliomas: correlation with histopathological grades. Neuro Oncol 2014;16:441-8. [Crossref] [PubMed]
10. Wang D, Chen J, Ying Y, Zhao X, Mei N, Li X, Zhu Y, Cui J, Wu PY, Lu Y, Yin B. Assessment of hypoxia and its dynamic evolution in glioblastoma via qBOLD MRI: a comparative study with metformin treatment. Eur Radiol Exp 2024;8:134. [Crossref] [PubMed]
11. Zhang L, Yang LQ, Wen L, Lv SQ, Hu JH, Li QR, Xu JP, Xu RF, Zhang D. Noninvasively Evaluating the Grading of Glioma by Multiparametric Magnetic Resonance Imaging. Acad Radiol 2021;28:e137-46. [Crossref] [PubMed]
12. Barajas RF Jr, Villanueva-Meyer J, Perry A, Berger M, Cha S. Biologically aggressive regions within glioblastoma identified by spin-lock contrast T1 relaxation in the rotating frame (T1ρ) MRI. Radiol Case Rep 2017;12:827-32. [Crossref] [PubMed]
13. Villanueva-Meyer JE, Barajas RF Jr, Mabray MC, Chen W, Shankaranarayanan A, Koon P, Barani IJ, Tihan T, Cha S. Differentiation of brain tumor-related edema based on 3D T1rho imaging. Eur J Radiol 2017;91:88-92. [Crossref] [PubMed]
14. Yu H, Zhu L, Wang Y, Yue X, Wang W, Sun Z, Jiang S, Chen Y, Wen Z. Amide Proton Transfer Weighted MR Imaging for Predicting Meningioma Stiffness: A Feasibility Study. J Magn Reson Imaging 2023;57:1071-8. [Crossref] [PubMed]
15. Tang PLY, Romero AM, Nout RA, van Rij C, Slagter C, Swaak-Kragten AT, Smits M, Warnert EAH. Amide proton transfer-weighted CEST MRI for radiotherapy target delineation of glioblastoma: a prospective pilot study. Eur Radiol Exp 2024;8:123. [Crossref] [PubMed]
16. Joo B, Han K, Choi YS, Lee SK, Ahn SS, Chang JH, Kang SG, Kim SH, Zhou J. Amide proton transfer imaging for differentiation of benign and atypical meningiomas. Eur Radiol 2018;28:331-9. [Crossref] [PubMed]
17. Koike H, Morikawa M, Ideguchi R, Uetani M, Hiu T, Matsuo T. Amide proton transfer and chemical exchange saturation transfer MRI differentiates between growing and non-growing intracranial meningiomas: a pilot study. Clin Radiol 2022;77:e295-301. [Crossref] [PubMed]
18. Zhou L, Pan W, Huang R, Wang T, Wei Z, Wang H, Zhang Y, Li Y. Amide Proton Transfer-Weighted MRI, Associations with Clinical Severity and Prognosis in Ischemic Strokes. J Magn Reson Imaging 2024;60:2509-17. [Crossref] [PubMed]
19. Koike H, Morikawa M, Ishimaru H, Ideguchi R, Uetani M, Hiu T, Matsuo T, Miyoshi M. Quantitative Chemical Exchange Saturation Transfer Imaging of Amide Proton Transfer Differentiates between Cerebellopontine Angle Schwannoma and Meningioma: Preliminary Results. Int J Mol Sci 2022;23:10187. [Crossref] [PubMed]
20. Johnson CP, Follmer RL, Oguz I, Warren LA, Christensen GE, Fiedorowicz JG, Magnotta VA, Wemmie JA. Brain abnormalities in bipolar disorder detected by quantitative T1ρ mapping. Mol Psychiatry 2015;20:201-6. [Crossref] [PubMed]
21. Wang L, Chen W, Qian Y, So TY. Repeatability of quantitative T1rho magnetic resonance imaging in normal brain tissues at 3.0T. Phys Med 2023;112:102641. [Crossref] [PubMed]
22. Li Y, Hu M, Chen J, Ling Z, Zou X, Cao W, Wei F. Deep Learning Assisted Classification of T1ρ-MR Based Intervertebral Disc Degeneration Phases. J Magn Reson Imaging 2025;61:1492-500. [Crossref] [PubMed]
23. Boles Ponto LL, Magnotta VA, Menda Y, Moser DJ, Oleson JJ, Harlynn EL, DeVries SD, Wemmie JA, Schultz SK. Comparison of T(1)Rho MRI, Glucose Metabolism, and Amyloid Burden Across the Cognitive Spectrum: A Pilot Study. J Neuropsychiatry Clin Neurosci 2020;32:352-61. [Crossref] [PubMed]
24. Haris M, Yadav SK, Rizwan A, Singh A, Cai K, Kaura D, Wang E, Davatzikos C, Trojanowski JQ, Melhem ER, Marincola FM, Borthakur A. T1rho MRI and CSF biomarkers in diagnosis of Alzheimer’s disease. Neuroimage Clin 2015;7:598-604. [Crossref] [PubMed]
25. Johnson CP, Christensen GE, Fiedorowicz JG, Mani M, Shaffer JJ Jr, Magnotta VA, Wemmie JA. Alterations of the cerebellum and basal ganglia in bipolar disorder mood states detected by quantitative T1ρ mapping. Bipolar Disord 2018;20:381-90. [Crossref] [PubMed]
26. Cao M, Ding W, Han X, Suo S, Sun Y, Wang Y, Qu J, Zhang X, Zhou Y. Brain T1ρ mapping for grading and IDH1 gene mutation detection of gliomas: a preliminary study. J Neurooncol 2019;141:245-52. [Crossref] [PubMed]
27. Liu Y, Wang W, Zheng Y, Wang H, Zheng H, Liang D, Zhu Y. Magnetic resonance T1ρ quantification of human brain at 5.0 T: A pilot study. Front Phys 2022;10:1016932.
28. Liu S, Li J, Hua R, Xing Y, Wu J, Lin J, Wang J, Shan Y, Xu L, Shi F, Zeng M. Visualization of perivascular spaces in the human brain with 5-T magnetic resonance imaging. BMC Neurosci 2025;26:18. [Crossref] [PubMed]
29. Wei Z, Chen Q, Han S, Zhang S, Zhang N, Zhang L, Wang H, He Q, Cao P, Zhang X, Liang D, Liu X, Li Y, Zheng H. 5T magnetic resonance imaging: radio frequency hardware and initial brain imaging. Quant Imaging Med Surg 2023;13:3222-40. [Crossref] [PubMed]
30. Liu J, Wang Z, Yu D, Yang Y, Li Z, Wang X, Yang Y, Cheng C, Zou C, Gan J. Comparative analysis of hepatic fat quantification across 5 T, 3 T and 1.5 T: A study on consistency and feasibility. Eur J Radiol 2024;180:111709. [Crossref] [PubMed]
31. 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]
32. Choate KA, Pratt EPS, Jennings MJ, Winn RJ, Mann PB. IDH Mutations in Glioma: Molecular, Cellular, Diagnostic, and Clinical Implications. Biology (Basel) 2024;13:885. [Crossref] [PubMed]
33. Capper D, Reifenberger G, French PJ, Schweizer L, Weller M, Touat M, Niclou SP, Euskirchen P, Haberler C, Hegi ME, Brandner S, Le Rhun E, Rudà R, Sanson M, Tabatabai G, Sahm F, Wen PY, Wesseling P, Preusser M, van den Bent MJ. EANO guideline on rational molecular testing of gliomas, glioneuronal, and neuronal tumors in adults for targeted therapy selection. Neuro Oncol 2023;25:813-26. [Crossref] [PubMed]
34. Zhou J, Sun W, Li H, Song X, Xu D, Xu H. Application of 5T glutamate chemical exchange saturation transfer imaging in brain tumors: preliminary results. J Neurooncol 2024;169:581-9. [Crossref] [PubMed]
35. De Maria L, Ponzio F, Cho HH, Skogen K, Tsougos I, Gasparini M, Zeppieri M, Ius T, Ugga L, Panciani PP, Fontanella MM, Brinjikji W, Agosti E. The Current Diagnostic Performance of MRI-Based Radiomics for Glioma Grading: A Meta-Analysis. J Integr Neurosci 2024;23:100. [Crossref] [PubMed]
36. Saijo A, Ogino H, Butowski NA, Tedesco MR, Gibson D, Watchmaker PB, et al. A combinatory vaccine with IMA950 plus varlilumab promotes effector memory T-cell differentiation in the peripheral blood of patients with low-grade gliomas. Neuro Oncol 2024;26:335-47. [Crossref] [PubMed]
37. Menon RG, Zibetti MVW, Jain R, Ge Y, Regatte RR. Performance Comparison of Compressed Sensing Algorithms for Accelerating T(1ρ) Mapping of Human Brain. J Magn Reson Imaging 2021;53:1130-9. [Crossref] [PubMed]
38. Ai QYH, Zhang H, Jiang B, So TY, Mo FKF, Qamar S, Chen W, King AD. Test-retest repeatability of T1rho (T1ρ) MR imaging in the head and neck. Eur J Radiol 2021;135:109489. [Crossref] [PubMed]
39. Ma S, Wang N, Fan Z, Kaisey M, Sicotte NL, Christodoulou AG, Li D. Three-dimensional whole-brain simultaneous T1, T2, and T1ρ quantification using MR Multitasking: Method and initial clinical experience in tissue characterization of multiple sclerosis. Magn Reson Med 2021;85:1938-52. [Crossref] [PubMed]
40. Suh CH, Park JE, Jung SC, Choi CG, Kim SJ, Kim HS. Amide proton transfer-weighted MRI in distinguishing high- and low-grade gliomas: a systematic review and meta-analysis. Neuroradiology 2019;61:525-34. [Crossref] [PubMed]
41. Chen L, Huang L, Zhang J, Li S, Li Y, Tang L, Chen W, Wu M, Li T. Amide proton transfer-weighted and arterial spin labeling imaging may improve differentiation between high-grade glioma recurrence and radiation-induced brain injury. Heliyon 2024;10:e32699. [Crossref] [PubMed]
42. Xu Z, Ke C, Liu J, Xu S, Han L, Yang Y, Qian L, Liu X, Zheng H, Lv X, Wu Y. Diagnostic performance between MR amide proton transfer (APT) and diffusion kurtosis imaging (DKI) in glioma grading and IDH mutation status prediction at 3 T. Eur J Radiol 2021;134:109466. [Crossref] [PubMed]
43. Groblewska M, Mroczko B. Pro- and Antiangiogenic Factors in Gliomas: Implications for Novel Therapeutic Possibilities. Int J Mol Sci 2021;22:6126. [Crossref] [PubMed]
44. Dubois LG, Campanati L, Righy C, D'Andrea-Meira I, Spohr TC, Porto-Carreiro I, Pereira CM, Balça-Silva J, Kahn SA, DosSantos MF, Oliveira Mde A, Ximenes-da-Silva A, Lopes MC, Faveret E, Gasparetto EL, Moura-Neto V. Gliomas and the vascular fragility of the blood brain barrier. Front Cell Neurosci 2014;8:418. [Crossref] [PubMed]
45. Dasgupta A, Geraghty B, Maralani PJ, Malik N, Sandhu M, Detsky J, Tseng CL, Soliman H, Myrehaug S, Husain Z, Perry J, Lau A, Sahgal A, Czarnota GJ. Quantitative mapping of individual voxels in the peritumoral region of IDH-wildtype glioblastoma to distinguish between tumor infiltration and edema. J Neurooncol 2021;153:251-61. [Crossref] [PubMed]
46. Spear JT, Gore JC. New insights into rotating frame relaxation at high field. NMR Biomed 2016;29:1258-73. [Crossref] [PubMed]
47. Liu S, Feng L, Li Y, Ni M, Jiang C, Yuan H. Participant Discomfort During 5 T MRI Examinations and Its Contributing Factors. J Magn Reson Imaging 2025;62:588-97. [Crossref] [PubMed]
48. Essed RA, Prysiazhniuk Y, Wamelink IJ, Azizova A, Keil VC. Performance of amide proton transfer imaging to differentiate true progression from therapy-related changes in gliomas and metastases. Eur Radiol 2025;35:580-91. [Crossref] [PubMed]
49. Pflüger I, Rastogi A, Casagranda S, Papageorgakis C, Behnisch R, Liebig P, Prager M, Ippen FM, Paech D, Wick W, Bendszus M, Brugnara G, Vollmuth P. Amide proton transfer weighted MRI measurements yield consistent and repeatable results in patients with gliomas: a prospective test-retest study. Eur Radiol 2025;35:3367-79. [Crossref] [PubMed]
50. Wamelink IJHG, Kuijer JPA, Padrela BE, Zhang Y, Barkhof F, Mutsaerts HJMM, Petr J, van de Giessen E, Keil VC. Reproducibility of 3 T APT-CEST in Healthy Volunteers and Patients With Brain Glioma. J Magn Reson Imaging 2023;57:206-15. [Crossref] [PubMed]