Previous studies have demonstrated that both genotype-based and histology-based classification of gliomas can predict patients’ prognosis[11, 12]. Therefore, seeking a noninvasive method to assess histology and genetic status before surgery is of great importance. In this case, the discriminative abilities of ADC in histologic subtypes, IDH, MGMT, and TERT status were assessed, respectively.
In the current study, ADC values generated from DWI (b=0 and 1000 s/mm2) decreased significantly with the WHO glioma grade, which contradicts with previous studies[5, 13]. The results indicated that the ADC values can differentiate glioblastomas from LGGs. Cell density, mitotic activity and vascularity play an important role in the pathological grading of gliomas[14]. For example, the increment of cell density can remarkably restrict the movement of water molecules, which can be reflected by ADC[14]. Therefore, glioblastomas were more prone to exhibit lower ADC values than LGGs. Louis et al. [15]discovered that glioblastomas also had lesser normal brain cells and higher tumor cells than LGGs, which may also partly explain the lower ADC values in glioblastomas. Unlike the definition of WHO I-II grade gliomas as low grade gliomas and WHO III-IV grade gliomas as high grade gliomas[5, 16], this study classified WHO II-III and WHO IV gliomas as LGGs and glioblastomas, respectively. Since IDH-mut was more commonly observed in WHO II-III grade gliomas than glioblastomas, and it could predict the prognosis of gliomas well, whatever the gliomas grading[17], the grouping method in this paper was closely relevant to patients’ prognosis.
The IDH-mut rates were 61.76% in LGGs and 7.14% in glioblastomas, respectively, both of which were lower than the reported indices (75% for LGGs and 12% for glioblastomas)[17]. The ADC values for IDH-mutated gliomas were found significantly higher than those for IDH-wildtype gliomas. Liu et al.[5] also reported that in grade II-III astrocytoma, mutant IDH showed higher ADCmean value than wild-type IDH. IDH may inhibit tumor growth through decreasing the level of nicotinamide adenine dinucleotide phosphate production[17] and hypoxia-inducible factor 1α[18]. This mechanism could decrease the cell density and partially explain the phenomenon that IDH-mutated gliomas displayed higher ADC values. In addition, IDH-mut having a direct and greater impact on ADC values than tumor grade helped to explain why IDH status could predict prognosis better than the histologic classification[11, 12].
Besides IDH, MGMT and TERT are also important genetic hallmarks of gliomas in guiding clinical treatment and evaluating prognosis[19, 20]. Recently, several studies[21-23] have supported the feasibility of ADC to evaluate the MGMT status in glioblastoma. In the present study, we found that ADC values had low accuracy and reliability in discriminating MGMT and TERT status in WHO II-III gliomas and thus had limited application in the prediction of these two genotypes. Multiple linear regression analysis also revealed that MGMT and TERT status were not independent parameters for ADC values. It was hypothesized that the increment of ADC in TERT-w and MGMT-m gliomas may be induced by coexisting factors or interactions between variables. For example, in this study, glioblastomas were more likely to have MGMT-um and IDH-wt than LGGs (P<0.0001), and consequently, the ADC values in MGMT-unmethylated gliomas might be affected by the tumor grading and concurrent IDH-wt. However, Cui et al.[16] analyzed 82 gliomas and revealed that ADC values are directly associated with MGMT immunoactivity, which is inconsistent with our results.
The strength of this study was that the abilities of ADC values were evaluated to predict WHO glioma grade and various molecular status in the same study. The overall assessment of the predictive power of DWI metrics can be available. Accessing various molecular features in one study also helps us identify the valuable genotypes, which can directly affect ADC values. As higher ADC values are associated with more favorable prognosis[16, 21], it is very important to find out meaningful genotypes for further investigation in patients’ outcomes.
Besides the intrinsic limitations of retrospective researches, the other two limitations of this study should be noted. Firstly, biopsy samples used in this study were not acquired by ADC-guided biopsy. Because the ROI-based method cannot assess the direct correlation between histopathology and ADC values, some bias can be produced, especially in more heterogeneous gliomas like glioblastomas. Secondly, the sample size is small. Thus, a larger cohort of patients are needed to verify our conclusions. Thirdly, the genetic status evaluated in this study is limited.