Utilizing the Amide Proton Transfer Technique to Characterize Diffuse Gliomas Based on the WHO 2021 Classification of CNS Tumors

Introduction

Diffuse glioma is the most common primary brain tumor, accounting for approximately 80% of all primary malignant brain tumors in adults [1]. Overall survival and prognosis for relapse, as well as the provision of adjuvant therapeutic regimens, depend dramatically on both tumor grade and morphological subtype [2]. In addition, it is believed that the molecular profile of the tumor can predict the response to treatment [3], [4]. Thus, an accurate evaluation of the malignant potential of diffuse gliomas is absolutely necessary.

The last release of the WHO Classification of CNS Tumors (2021) is focused primarily on molecular diagnostics, where the main factors for grading and tumor type are the combination of IDH mutation and 1p19q codeletion statuses [5]. However, tumor grading is still carried out through visual analysis [6], which is subjective and not always accurate due to tumor heterogeneity [7]. The evaluation of the Ki-67 proliferation index has been recognized as highly beneficial in achieving this objective [8]. However, it necessitates additional immunohistochemical staining. Consequently, despite morphological analysis being regarded as the definitive method for diagnosing diffuse gliomas, there is variability among observers, including experts [9]. Furthermore, detailed histological analysis with tumor molecular profile evaluation is expensive and requires advanced laboratory techniques, which could be a significant problem in middle- and low-income countries [10].

Brain magnetic resonance imaging (MRI) plays a key role in the diagnosis, presurgical planning, surveillance, and treatment monitoring of gliomas [11]. In recent decades, many attempts have been made to accurately determine the type of diffuse brain glioma based on MRI features [11]–[13] or a combination of MRI and demographic characteristics [14]. In several studies, conventional MRI has been shown to predict both the pathological subtype [15] and even the molecular profile [16] of diffuse gliomas. To unify radiological reports and create a common vocabulary, the VASARI scoring system has been developed for a detailed description of brain gliomas (https://wiki.cancerimagingarchive.net/display/Public/VASARI+Research+Project), and its use has already shown promising results [13], [17]. On the other hand, in recent years, there have been significant advancements in image postprocessing techniques that offer a wealth of additional information extracted from conventional T2-weighted, FLAIR, pre- and postcontrast T1-weighted sequences [18]. These radiomic-based methods showed high precision for the prediction of tumor grade, IDH mutation, and 1p19q codeletion status in adults [12], [13] as well as tumor grade [19] and BRAF mutation status [20] in children. However, the signal changes on most conventional MRI sequences lack biological specificity, which limits the accuracy of noninvasive glioma characterization. Therefore, although conventional MRI is readily accessible and provides crucial anatomical details, accurately distinguishing the type and grade of a tumor solely based on conventional techniques appears to be challenging [11], [12].

The apparent diffusion coefficient (ADC) map is an essential conventional quantitative magnetic resonance sequence that is particularly valuable for analysing diffuse gliomas. It is believed that the restriction of diffusion in this map reflects the level of cellularity within the tumor [21]–[23], and a negative correlation between ADC values and glioma grade has been shown in many previous studies [21], [24]–[26].

The advancements and widespread adoption of advanced MRI sequences in recent decades have enabled clinicians and researchers to gather extensive information about tumor structure and physiology, facilitating noninvasive glioma diagnosis and evaluation of treatment effectiveness. Specifically, perfusion techniques can offer valuable noninvasive insights into the microvasculature of tumors. Several perfusion methods are available and utilized in clinical practice, showing promising results [27], [28]. One of the most popular techniques among them is dynamic susceptibility contrast imaging (DSC), which involves measuring the relative cerebral blood volume (rCBV) [12]. A recently published meta-analysis revealed that this technique can accurately predict tumor grade [28]. Another meta-analysis demonstrated the high accuracy of DSC perfusion in determining both the IDH mutation and 1p19q codeletion statuses in patients with diffuse brain glioma [29]. However, it is worth noting that data acquisition protocols for DSC can be relatively complex and vary significantly across different institutions [30]. Furthermore, this technique is unable to provide definitive values for perfusion parameters, which limits the practical applicability of the acquired data. For example, it becomes challenging to establish a universal threshold in such circumstances. Additionally, this procedure is invasive and necessitates the administration of a contrast agent.

The arterial spin labeling (ASL) technique is another promising alternative to invasive perfusion methods, and its utility has been extensively studied in different fields, including neurooncology [31]. ASL has been shown to be useful in the prediction of glioma grade [32] but much less useful in defining the molecular profiles of gliomas [33]. Moreover, studies suggest a strong correlation between the data obtained from ASL and DSC perfusion techniques [34], [35]. However, it is important to note that ASL can only provide measurements of cerebral blood flow (CBF), whereas CBV values are known to better reflect the microvasculature of the tumor [12].

An additional encouraging MRI technique used for imaging diffuse brain gliomas is amide proton transfer (APT), known for its remarkable biological specificity. It is well known that gliomas have higher protein/peptide contents than normal brain tissue [36]. Therefore, information at the protein level could potentially contribute to earlier diagnosis, more accurate delineation of boundaries, and enhanced tumor characterization [37]. The biophysical basis of APT imaging is the ability to detect mobile proteins [38]. The chemical exchange rate of amide protons with water hydrogens is a crucial parameter in this context, as it enhances the clinical utility of APT [39]. In addition, it is a noninvasive procedure that does not require the administration of any contrast agents.

APT showed better diagnostic performance than conventional MRI [40]. Moreover, it has been discovered that it is just as effective as DSC perfusion [41]. A recently published meta-analysis confirmed the effectiveness of the APT technique in distinguishing between low- and high-grade gliomas, as well as its potential for predicting histopathology noninvasively [38]. However, some authors have reported improved accuracy in grade prediction when combining APT with ADC and ASL values [42]. In addition, several studies [43], [44] have found a correlation between APT values within the glioma and tumor cellularity.

Furthermore, these studies have also shown a relationship between APT values and the extent of diffusion restriction as measured by both ADC and DKI techniques [43], [45].

Beyond that, APT seems to be a predictive factor for IDH mutation status as well [39], [46], [47]. It is important to note that the application of APT is not limited to presurgical glioma diagnosis and can also have implications for predicting overall survival, prognosis of recurrence, and assessing treatment outcomes. This has been demonstrated in multiple recently published articles [47]–[49].

As a reflection of its advantages and benefits for clinical applications, numerous articles focusing on the radiological characteristics of diffuse gliomas have been published over the past years. However, it is worth noting that most of these studies were conducted using the WHO 2016 Classification of CNS Tumors, whereas the recently released WHO 2021 update introduces significant revisions to glioma grading. Interestingly, only a few authors have made an effort to determine the subtype of glioma, despite its clear correlation with patient prognosis. Furthermore, there have been only a limited number of research papers published on the practicality of combining conventional MRI with APT data, as well as on the effectiveness of utilizing different quantitative techniques (such as APT, ADC, and ASL) for assessing diffuse gliomas [42], [50]. Moreover, many studies examining the effectiveness of advanced brain MRI techniques, specifically APT and ASL, in patients with diffuse gliomas have been limited by small sample sizes.

Thus, the main objective of this study was to investigate the usefulness of the APT technique, both alone and in combination with other quantitative MRI sequences (specifically ADC and ASL), in predicting the grade and tumor type according to the WHO 2021 Classification of CNS Tumors for diffuse gliomas. Additionally, we aimed to determine whether incorporating descriptive tumor characteristics could enhance the predictive model based on APT. Furthermore, we sought to assess the ability of APT (alone or combined with other variables) to determine the status of IDH mutations, which greatly influences prognosis. Finally, we were interested in examining the relationship between quantitative MRI data and tumor cellularity measured by Ki-67 levels.

Methods

MRI data processing

The quantitative maps of APT, CBF (ASL), and ADC were calculated automatically by the MRI system. In each case, the entire tumor was segmented with ITK-Snap software (version 4.0.0, http://www.itksnap.org) using a semiautomatic classification algorithm. Segmentation was performed based on FLAIR and T2-WI (referring to T1-WI and CE-T1WI) for contrast-negative cases and postcontrast T1-WI (referring to FLAIR, T2-WI, and T1-WI) for contrast-positive cases. The segmentation results were saved as binary masks in NifTI format. The ADC, CBF and APT maps were registered and resampled referring to T1-WI using SPM12 with a normalized mutual information cost function and 4th Degree B-Spline interpolation (http://www.fil.ion.ucl.ac.uk/spm/). Subsequently, the tumor binary mask was moved to the APT, CBF, and CBV maps, and quantitative analysis was performed using Pyradiomics tool (https://aim.hms.harvard.edu/pyradiomics).

The following signal intensity characteristics were extracted from the tumor on the APT, CBF and ADC maps: (1) mean, (2) median, (3) 10th percentile (for ADC), and (4) 90th percentile (for APT and CBF). Absolute values of all quantitative MR parameters were used. An illustration of data processing is shown in Figure 1.

• Download figure
• Open in new tab

Figure 1.

An example of post-processing of MRI data in a patient with glioblastoma (WHO grade 4, IDH1 wild type). A – Axial CE-T1WI shows vivid inhomogeneous enhancement; B – same slice after semiautomatic tumor segmentation; C-axial T2WI; D – APT map demonstrates marked elevation of metabolites in the tumor’s center; E – CBF map (ASL) with slight hyperperfusion from the tumor; F – ADC map shows moderate diffusion restriction.

Two neuroradiologists qualitatively evaluated the MRI data independently (with 5 and 2 years of neuroradiology experience, respectively). Magnetic resonance features were defined according to Visually Accessible Rembrandt Images (VASARI) imaging criteria (https://wiki.cancerimagingarchive.net/display/Public/VASARI+Research+Project), and full results are provided in Supplementary Table 2. Subsequently, qualitative features such as enhancement quality (f4), necrosis proportion (f7), and hemorrhage presence (f16) were evaluated between groups.

Results

Quantitative MRI features

A Kruskal□Wallis test was performed on the median APT values of the three groups (grades 2, 3 and 4) and revealed that there was a statistically significant difference in these values among the tested groups (χ² = 11.25, df = 2, p = 0.0036). A post hoc Dunn test was further conducted to determine the differences in measured values between each pair of groups. We found the abovementioned differences in grade 2-grade 4 and grade 3-grade 4 pairs (p = 0.03 and p = 0.01, respectively; Figure 2A).

• Download figure
• Open in new tab

Figure 2.

The results of nonparametric statistical comparison of APT median values between groups with Kruskal□Wallis (panels A and B) and Mann□Whitney tests (panels C and D). APT levels are compared among groups using boxplots, where individual data values are represented by black dots and the medians are shown by the thick red horizontal lines. The boxes represent the middle 50% of the data, ranging from the 25th to the 75th percentile. The whiskers extend to 1.5 times the interquartile range. In this figure, NS indicates p > 0.05; * represents p < 0.05; ** p < 0.01; *** p < 0.001.

• Download figure
• Open in new tab

Figure 3.

The plot represents areas under the curves (ROC curves) for two multinomial logistic regression models (one-vs-the-rest multiclass strategy) used to classify patients according to their tumor grade (panel A) or tumor type (panel B). Different grades or types are depicted in different colors. The dotted line corresponds to AUC = 0.5 or the random classifier’s performance.

In a similar vein, median APT values were compared between patients belonging to groups with different tumor types, resulting in a rejection of the null hypothesis (χ² = 17.33, df = 2, p = 0.0002). Patients with glioblastoma showed increased median APT values in comparison to those with astrocytoma and oligodendroglioma (p = 0.002 and p = 0.0004, respectively; Figure 2B).

When comparing APT median levels between patients with IDH1 mutant and wild types, the latter group demonstrated increased median values, which were statistically significant (U = 51, p = 7.37*10⁻⁵; Figure 2C). Similarly, when applied to groups with and without 1p19q codeletion, the Mann□Whitney test indicated the presence of noticeable differences in median APT levels (U = 236, p = 0.0077; Figure 2D). Differences in the mean and 90th percentile APT values were also statistically significant between groups, but these differences were less pronounced than those of median APT. We report these results in detail in Supplementary Table 3.

We did not find any differences in median CBF values between patients with different tumor grades (χ² = 2.6, p = 0.27) or types (χ² = 3.96, p = 0.14). Similarly, these values were not significantly different for patients within the IDH1 mutation (mutant/wild types) and 1p19q codeletion (yes/no) groups (U = 127, p = 0.09 and U = 134.5, p = 0.78, respectively). Comparison of the mean and 90th percentile CBF values showed the same trend and were not reported in the main text of the article.

Likewise, we did not observe any significant differences in median or mean and 10th percentile ADC levels among different grades, tumor types and IDH1 status. Here, we briefly presented only results regarding median ADC values with corresponding data for mean and 10th percentile values provided in Supplementary Table 3. Notably, the Kruskal□Wallis test returned statistically significant differences between tumor grade groups, showing that the null hypothesis about equality of group’s median values can be rejected (χ² = 6.86, df = 2, p = 0.03). However, post hoc Dunn test results revealed no significant differences upon pairwise comparison (p = 0.47 for the grade 2-grade 3 pair; p = 0.09 for the grade 2-grade 4 pair and p = 0.07 for the grade 3-grade 4 pair). Median values for tumor type groups did not differ from each other (χ² = 2.28, df = 2, p = 0.32). The Mann□Whitney test applied to ADC data also did not result in any discrepancies between mutant and wild types (U = 260, p = 0.135).

IDH1 status prediction

We further sought to predict the patient’s IDH1 class, that is, whether the patient had mutant or wild-type IDH1. For this end, several binary logistic regression models were developed. As were the cases with tumor grade and type classification, we started with APT median values being the only predictor in the model and found that this model was significantly different from the null model (χ² = 17.62, df = 1, p = 2.7*10⁻⁵). The inclusion of the age variable resulted in a substantial enhancement in the performance of the model (χ²= 12.56, df = 1, p = 0.0004). In total, these two variables provided a model with an accuracy of 0.79 [0.4, 0.97] and an area under the curve of 0.95; however, this model exhibited a low specificity rate of 30%. The likelihood ratio test indicated that further addition of variables to the model did not result in statistically significant differences compared to the model with only two variables. However, the inclusion of the predictor “necrosis” led to an increase in model accuracy up to 1 [0.66, 1], while there was minimal change in the area under the curve, which reached a value of 0.96 (χ² = 5.16, df = 3, p = 0.16). In Figure 4, we present the ROC curves for each of the aforementioned models. In addition, an increase in the median APT and age values by one unit led to the elevation of the odds ratio of membership in the wild-type group (see Table 4 for specific values).

• Download figure
• Open in new tab

Figure 4.

The plot displays areas under the curves for three binary logistic models (IDH1 group prediction) with different numbers of predictors, depicted in different colors. The dotted line corresponds to AUC = 0.5 or the random classifier’s performance.

Correlations among APT, ADC and CBF values within tumor

We observed a negative correlation between the 90^th percentile APT and 10^th percentile ADC values (r = - 0.44 [−0.66, −0.15], adjusted p = 0.008; Figure 5). Similarly, there was a weak negative correlation between the mean/median APT values and the 10^th percentile ADC values (r = - 0.39 [−0.63, −0.09], adjusted p = 0.016; not shown). No associations were revealed between CBF values (mean, median, 90th percentile) and APT values (mean, median, 90th percentile), as well as ADC values (mean, median, 10th percentile) - p > 0.1 in all comparisons, not shown.

• Download figure
• Open in new tab

Figure 5.

Spearman correlation between 90th percentile APT and 10th percentile ADC values. The gray area around the red line represents the 95% confidence interval. Individual data points are displayed as black dots.

Correlation between Ki-67 levels and MRI parameters

We found a moderate positive correlation between Ki-67 and the 90th percentile APT values (r = 0.53 [0.26, 0.72], p = 0.003; Figure 6A) and a moderate negative association between Ki-67 and the 10th percentile ADC values (r = - 0.5 [−0.70, −0.22], p = 0.004; Figure 6B). Additionally, mean and median APT values also exhibited significant correlations with Ki-67 levels, although to a lesser extent than the 90th percentile values (r = 0.47 [0.18, 0.68], p = 0.004 for both variables, not shown). Likewise, the mean and median ADC values showed a significant correlation with Ki-67 levels (r = - 0.33 [−0.59, −0.02], p = 0.046 and r = - 0.39 [−0.63, −0.09], p = 0.019, respectively; not shown). No significant correlations were observed between Ki-67 levels and mean, median, or 90th percentile CBF values (p > 0.1, not shown).

• Download figure
• Open in new tab

Figure 6.

Panel A displays the Spearman correlation between the 90th percentile APT and Ki-67 values. Panel B shows the Spearman correlation between the 10th percentile ADC values and Ki-67 levels in the patients. The gray area around the red line represents the 95% confidence interval. Individual data points are displayed as black dots.

Discussion

Diffuse gliomas present a major challenge to the modern healthcare system worldwide, leading to a substantial reduction in patients’ quality of life, posing an immediate threat to their lives, and requiring significant direct and indirect financial expenses associated with the prevention, diagnosis, and treatment of such tumors. Brain MRI is essential for diagnosing, planning surgery, monitoring progress, and treating gliomas. However, accurately distinguishing the type and grade of a tumor solely based on conventional MRI techniques seems challenging due to the lack of biological specificity in signal changes, limiting noninvasive glioma characterization. In this study, we explored the potential of using the APT technique, both on its own and in conjunction with other quantitative MRI sequences (specifically ADC and ASL), to predict the grade and tumor type based on the WHO 2021 Classification of CNS Tumors for diffuse gliomas. We discovered that incorporating APT values into the models resulted in a substantial enhancement of their performance. Grade prediction metrics based on median APT values and patient age were the highest for grade 4, with grade 3 and grade 2 demonstrating moderate classification accuracy (all AUC values exceeded 0.75). These findings align well with the literature, which demonstrates the predictive potential of APT levels in distinguishing between low- and high-grade gliomas [38]. Furthermore, we expanded upon these findings by demonstrating the capability of APT to distinguish patients within the high-grade class, specifically those categorized as grade 3 and grade 4. It is crucial to consider this aspect, as there are significant disparities in both overall and median survival rates observed among patients diagnosed with grade 3 and grade 4 diffuse brain gliomas [51]–[53]. Similar results have been reported previously by Guo and colleagues in their study involving 62 patients, further validating our findings [46]. However, the most notable contrast between their findings and ours was the existence of statistically significant differences between grade 2 and grade 3 patients in their study, a pattern that was absent in our data. This discrepancy could be attributed to the relatively small number of patients included in the grade 2 group within our study. We were also able to identify the differences in APT levels among distinct tumor type groups. The most pronounced differences were found between the glioblastoma and oligodendroglioma groups, followed by astrocytoma and glioblastoma. In contrast, there were no statistically significant differences observed between the astrocytoma and oligodendroglioma groups. To our knowledge, these results have not been showcased in earlier studies and provide definitive clinical relevance.

IDH mutation and 1p19q codeletion statuses in patients with diffuse gliomas are well-established prognostic markers [5]. We discovered that patients with wild-type IDH1 had higher APT levels than those with mutant IDH1. Consistent with our current findings, several previously published studies have also shown a similar trend, indicating that elevated APT values can serve as a significant predictor of poor overall survival [47], [54]. In the same vein, it has also been revealed that APT imaging exhibited superior performance over DKI in IDH mutation status prediction [45].

The two 1p19q codeletion groups in our study were also different from each other in terms of APT levels. Specifically, patients without codeletion were characterized by higher APT median values in comparison to patients with codeletion. However, contrary to what we have observed, Su and colleagues in their recent paper failed to identify differences between the groups in a sample of 113 patients with diffuse glioma [55]. Further research is required to address this question, as the presence of 1p19q codeletion classifies the tumor as an oligodendroglioma [5], which significantly improves prognosis for the patient.

Taken together, our findings clearly demonstrate the significant utility of the APT technique in predicting tumor grade and type, as well as identifying IDH mutation and 1p19q codeletion statuses, in patients with diffuse brain gliomas.

Hereafter, a number of multinomial logistic regression models were developed to predict patient tumor types and grades based on APT values complemented by additional parameters used as predictors. The model with the best classification performance for tumor grade, in addition to APT values, included the variable of patient age. This model demonstrated an accuracy of 0.71 and an area under the curve of 0.82, which is considered good classifier performance. The highest sensitivity values (88%) were obtained for Grade 4, while specificity was highest for Grade 3 (79%). Age is a known predictive factor for tumor malignancy and clearly impacts prognosis in patients with diffuse gliomas [56]–[58]. Therefore, entering the age of the patients in the APT-based prediction model looks logical, especially considering that this information is easy to obtain. However, we hypothesize that classification accuracy could be further improved by incorporating parameters not considered in this study.

In the case of tumor type classification, we settled on a model that includes four predictors. In addition to APT values, factors such as necrosis and the presence of hemorrhage were added. This allowed us to achieve an accuracy of 0.74 and an area under the curve of 0.93. Sensitivity and specificity reached their highest values for the glioblastoma group (87%). The balanced accuracy was lowest for the astrocytoma group, which collectively demonstrates the model’s ability to correctly detect patients with glioblastoma most effectively, while performing less accurately for astrocytomas. Conventional MRI is the diagnostic standard for patients with diffuse glioma and provides a detailed anatomical picture, as well as important qualitative tumor characteristics, such as contrast enhancement quality and the presence of necrosis or hemorrhage. Conventional brain MRI has been shown to be useful in predicting the pathological subtype [15] and even the molecular profile [16] of diffuse gliomas. The VASARI scoring system enables the quantification of tumor characteristics, making it easier to compare them between different groups [13].

In general, we found that including additional parameters (specifically the patient’s age, proportion of tumor necrosis, and presence of hemorrhage) into the regression model significantly enhanced the accuracy of predicting type, grade, and IDH mutation status beyond just APT values.

Surprisingly, we did not find differences in ADC values between tumor grades and histological subtypes. Furthermore, there were no significant differences between the IDH mutant and the IDH wild-type groups. These findings are generally inconsistent with previously published articles that revealed the high utility of ADC in the evaluation of gliomas [21], [24]–[26]. The reason could be that the primary purpose of applying ADC is to distinguish between low-grade and high-grade gliomas [24], [25], whereas our sample consisted mainly of high-grade tumors (grades 3 and 4). Another potential explanation lies in our chosen analysis method, wherein we employed an entire tumor binary mask for value extraction. This approach may have inadvertently obscured differences attributable to the inherent tissue heterogeneity of gliomas. Furthermore, it is widely acknowledged that ADC maps derived from higher b-factors offer enhanced diagnostic utility [21], whereas our study utilized the conventional acquisition method with a b-value of 1000 s/mm². It should be emphasized that certain published studies have failed to identify discernible discrepancies in ADC values within tumors when comparing low-grade and high-grade gliomas [59].

Similarly, our study found no significant differences in ASL values among different tumor grades, types, and IDH mutation statuses. These findings are also inconsistent with previous reports in the literature. Previous studies have demonstrated the utility of ASL in predicting glioma grade [34] and assessing IDH status [60]. However, Kang et al. did not observe significant differences in the 90th percentile regional cerebral blood flow (rCBF) values between low- and high-grade glioma groups [42]. A recent meta-analysis indicated that ASL shows high accuracy in glioma grading, particularly when considering maximal relative CBF values [32]. It is worth noting that our study employed absolute CBF values, which may explain the discrepancy observed compared to previously published data. Another potential explanation could be attributed to changes made by the WHO’s 2021 Classification of CNS Tumors in grading gliomas, whereas most prior research was based on the WHO 2016 version.

Another noteworthy observation was that the inclusion of ADC or ASL variables in our APT-based regression model did not yield any statistically significant enhancement to the model. Nevertheless, prior investigations have demonstrated that incorporating ADC values into APT may enhance glioma grading [42], [50]. Consequently, further studies are warranted to elucidate this aspect.

We found positive associations between tumor proliferation molecular marker (Ki-67) levels and APT values, as well as negative associations between Ki-67 and ADC values. These results are in agreement with previously published data. In a study by Yin et al. (2012), Ki-67 expression was inversely associated with ADC values in tumor parenchyma [61]. Furthermore, Yao et al. (2023) demonstrated that minimal ADC values were diagnostically comparable to the Ki-67 proliferation index in assessing pediatric glioma grade [23]. Several prior reports have also shown associations between APT values and Ki-67 levels [62]–[64]. These results support the notion that active tumor cell proliferation is linked to elevated concentrations of mobile proteins within the tumor, leading to an increased APT signal, as well as high cellularity resulting in decreased ADC values. Consequently, both ADC and APT measurements can indirectly reflect the proliferation index and subsequent malignancy of gliomas.

Moreover, our study unveiled a negative correlation between APT and ADC values within the tumor, thus corroborating earlier findings reported in the literature. A negative association has recently been shown between ADC and APT values within the tumor [43]. However, our results indicate that APT exhibits superior diagnostic accuracy when compared to ADC. Therefore, while both APT and ADC values provide insights into tumor cellularity, only APT demonstrates precise prognostic capabilities for determining tumor grade and type.

A previous study highlighted the APT technique as a viable alternative to contrast injection-dependent sequences such as DSC perfusion [41]. However, our findings indicate that APT alone lacks the necessary precision to accurately predict tumor type, and the inclusion of necrosis proportion, which can only be identified through CE-T1WI, is needed. Exploring the potential utility of combining APT and DSC values in presurgical glioma definition represents a promising research direction warranting further investigation. Our study results reaffirm the significant potential of the APT technique for the noninvasive diagnosis of diffuse gliomas. Therefore, it should be routinely incorporated alongside morphological analysis. Another valuable application of APT lies in selecting optimal sites for presurgical stereotactic biopsies. It is well established that targeted biopsy of the most malignant region of a tumor enhances diagnostic accuracy, and this region can potentially be identified using APT maps. It is important to note that beyond its role in presurgical glioma evaluation, APT holds considerable utility in assessing treatment effects and overall prognosis [47]–[49], although these aspects extend beyond the scope of our current study.

This study has several limitations. First, the sample size was small, consisting of only 42 patients, with only five cases presenting low-grade tumors. Second, our brain MRI protocol lacked a DSC perfusion sequence, thus impeding any comparison between the effectiveness of APT and rCBV values and hindering an assessment of their combined utility. Furthermore, there is a lack of follow-up data for the patients who participated in this study.

In addition, within the scope of this study, we also decided to utilize multinomial logistic regression models as classifiers. In doing so, we employed a hierarchical regression approach for variable selection and inclusion in the model. We are fully aware of alternative classifier options that could have been used in this study, such as support vector machines or decision tree classifiers. Similarly, forced entry, stepwise or all-subsets methods could have served as alternatives to hierarchical regression. Each of these methods has its own merits and limitations and cannot be deemed superior a priori over others. Therefore, the choice of methods and approaches was determined by specific research objectives.

In conclusion, the amide proton transfer technique shows significant potential for noninvasive evaluation of diffuse gliomas. To the best of our knowledge, this study represents the initial endeavor to evaluate the effectiveness of APT in the preoperative assessment of gliomas using the WHO 2021 Classification of CNS Tumors. Furthermore, it is a pioneering attempt to integrate APT data with qualitative tumor characteristics and patient demographics to enhance diagnostic accuracy rates. However, given the pilot nature of the study, further studies with larger sample sizes and comprehensive follow-up are clearly necessary to strengthen these findings.

Code availability

Cannot be applied.

Funding

No funding.

Author information

Ethics declaration