Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Association of dynamic susceptibility magnetic resonance imaging at initial tumor diagnosis with the prognosis of different molecular glioma subtypes

Association of dynamic susceptibility magnetic resonance imaging at initial tumor diagnosis with... Purpose The updated 2016 CNS World Health Organization classification differentiates three main groups of diffuse glioma according to their molecular characteristics: astrocytic tumors with and without isocitrate dehydrogenase (IDH) mutation and 1p/ 19q co-deleted oligodendrogliomas. The present study aimed to determine whether dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) is an independent prognostic marker within the molecular subgroups of diffuse glioma. Methods Fifty-six patients with treatment-naive gliomas and advanced preoperative MRI examination were assessed retrospec- tively. The mean and maximal normalized cerebral blood volume values from DSC-MRI within the tumors were measured. Optimal cutoff values for the 1-year progression-free survival (PFS) were defined, and Kaplan-Meier analyses were performed separately for the three glioma subgroups. Results IDH wild-type astrocytic tumors had a higher mean and maximal perfusion than IDH-mutant astrocytic tumors and oligodendrogliomas. Patients with IDH wild-type astrocytic tumors and a low mean or maximal perfusion had a significantly shorter PFS than patients of the same group with high perfusion (p = 0.0159/0.0112). Furthermore, they had a significantly higher risk for early progression (hazard ratio = 5.6/5.1). This finding was independent of the methylation status of O6-methylguanin- DNA-methyltransferase and variations of the therapy. Within the groups of IDH-mutant astrocytic tumors and oligodendrogliomas, the PFS of low and highly perfused tumors did not differ. Conclusion High perfusion upon initial diagnosis is not compellingly associated with worse short-term prognosis within the different molecular subgroups of diffuse glioma. Particularly, the overall highly perfused group of IDH wild-type astrocytic tumors contains tumors with low perfusion but unfavorable prognosis. . . . . . Keywords Glioma DSC-MRI Perfusion prognosis IDH mutation Molecular Introduction The 2016 update of the World Health Organization classifica- tion of brain tumors (2016 CNS WHO) now combines specif- * Cornelia Brendle ic molecular characteristics of gliomas beyond histological [email protected] characteristics for assignment to different subgroups in an integrated approach. Consequently, diffuse gliomas are clas- Diagnostic and Interventional Neuroradiology, Department of sified into three main groups. The first two are astrocytic tu- Radiology, Eberhard Karls University, Hoppe-Seyler-Straße 3, mors, which are divided into tumors with isocitrate dehydro- 72076 Tuebingen, Germany genase (IDH) mutation and without IDH mutation (IDH wild Neuropathology, Department of Pathology and Neuropathology, type). The third group, oligodendrogliomas, is defined by IDH Eberhard Karls University, Calwerstr. 3, 72076 Tuebingen, Germany 3 mutation, codeletion of chromosomes 1p and 19q, and reten- University Hospital for Neurosurgery, Eberhard Karls University, tion of the ATRX gene [1]. Hoppe-Seyler-Straße 3, 72076 Tuebingen, Germany 4 Magnetic resonance imaging (MRI) is the modality of Interdisciplinary Section of Neurooncology, Eberhard Karls choice for diagnosing brain tumors, as well as assessing the University, Hoppe-Seyler-Straße 3, 72076 Tuebingen, Germany 3626 Neurol Sci (2020) 41:3625–3632 tumor characteristics, the extent at the initial diagnosis, and The inclusion criteria comprised adult patient age, initial the therapy response in the course of the disease. Standard tumor diagnosis, advanced MRI with DSC perfusion before MRI is limited to visualizing morphological changes and the any treatment, pathologic examination including molecular disruption of the blood-brain barrier after injection of contrast profiling, and follow-up by clinical examination and imaging. agent. Beyond that, advanced MRI techniques give insights We assessed the age, sex, extent of tumor resection (complete into the functional properties and tissue composition of glio- versus incomplete), and the application of adjuvant therapy. mas. MR spectroscopy measures biochemical metabolites and We gathered follow-up data until progressive disease occurred can provide direct proof of IDH mutations in gliomas [2]. according to the Response Assessment in Neuro-Oncology Diffusion-weighted imaging displays the variable cellular- (RANO) criteria or until the patients were lost to follow-up ity and tissue structure in different tumor entities and the grade [1]. of malignancy [3–5]. Perfusion MRI techniques visualize the We calculated the PFS from the date of the advanced MR vascularization by quantifying the blood flow through the tis- examination to the date of tumor progression. We chose tumor sue, which is increased in malignant tissue due to neo- progression and not overall survival as the endpoint because angiogenesis [6]. Dynamic susceptibility contrast (DSC)- patients in specific glioma groups partially survive for a long MRI is the most common perfusion MRI technique. The con- time, and the progression-free interval is crucial for quality of trast agent inflow in the T2*-weighted images causes life. Furthermore, the overall survival of patients is influenced susceptibility-induced signal changes, which are used for cal- by variable therapy management in the case of progression. culating the cerebral blood volume (CBV). The absolute per- The IDH mutation status of the tumors was identified using fusion values of DSC-MRI are prone to multiple sources of an IDH1 R132H antibody, and rare IDH1/2 mutations in error, so they are usually corrected using an internal reference, IDH1 R132H-negative cases were identified by pyrosequenc- resulting in relative values (rCBV). ing [15, 16]. Nuclear ATRX status was assessed using immu- DSC-MRI is widely used in the diagnosis of gliomas and nohistochemistry and 1p/19q codeletion by a synthetic high- has the potential to predict the prognosis of glioma subtypes resolution microsatellite PCR gel [17, 18]. Possible according to the 2007 WHO classification, particularly in methylation of the O6-methylguanin-DNA- high-grade astrocytomas [7–9]. Previous studies reported that methyltransferase (MGMT) promoter of IDH wild-type lower tumor perfusion is associated with prolonged survival astrocytic tumors was identified using methylation- [7, 8, 10, 11]. DSC-MRI can also help to identify the molec- specific pyrosequencing [19]. For further analyses, we ular subgroups of diffuse glioma according to the updated differentiated three glioma subgroups according to their 2016 WHO classification, which differ substantially in their molecular characteristics: IDH wild-type astrocytic tu- prognosis [12, 13]. However, the genetic profile of a tumor is mors, IDH-mutant astrocytic tumors (ATRX loss), and usually confirmed histologically, and there is only limited oligodendrogliomas (IDH mutation, 1p19q codeletion, knowledge about the prognostic value of DSC-MRI as an ATRX retention). independent prognostic factor within a specific glioma group [14]. Therefore, the present study aims to evaluate the 1-year MRI examination progression-free survival (PFS) of treatment-naïve patients with IDH wild-type astrocytic tumors, IDH-mutant astrocytic All patients were examined on a 3-T MRI scanner (Biograph tumors, and oligodendrogliomas according to the DSC-MRI mMR MR-PET, Siemens Healthcare, Erlangen, Germany) perfusion values. using a head-neck coil. The MR sequences comprised an axial T2-weighted fluid-attenuated inversion recovery (FLAIR) se- quence as an anatomical reference (repetition time 9000 ms, echo time 94 ms, inversion time 2500 ms, slice thickness 3 Materials and methods mm, matrix size 207 × 320, field of view 189 × 220 mm ). For DSC-MRI, a pre-bolus of contrast agent was applied (0.025 Patients mmol/kg gadobutrol (Gadovist, Bayer Healthcare, Leverkusen, Germany)). Subsequently, a single-shot T2*- We retrospectively evaluated all patients with an initial diag- weighted echo-planar imaging sequence was performed dur- nosis of diffuse glioma and clinically indicated advanced MR ing the first pass of a contrast agent bolus of 0.1 mmol/kg of examination between 11/2012 and 9/2016. All patients pro- gadobutrol (Gadovist, Bayer Healthcare, Leverkusen, vided informed written consent for the scientific use of their Germany; injection rate 3 ml/s). The repetition time was data. The local ethics committee approved the study, which 1130 ms, the echo time was 31 ms, the flip angle was 60°, was performed in accordance with the Declaration of the slice thickness was 4 mm, the matrix size was 128 × 128, Helsinki. and the field of view was 230 × 230 mm . Neurol Sci (2020) 41:3625–3632 3627 Image analysis astrocytic tumors than IDH wild-type astrocytic tumors. Oligodendrogliomas had the best prognosis, but the differ- We calculated CBV parametric maps using raw data from ences between the glioma groups were not significant (p = DSC-MRI with the software Syngo® MR Perfusion 0.09). (Siemens Healthineers, Erlangen, Germany) with model- Table 1 shows detailed characteristics of the glioma sub- based leakage correction. We manually identified the middle groups. Within the groups, PFS did not correlate with age (p = cerebral artery or anterior cerebral artery for the arterial input 0.15, p = 0.30, and p = 0.65, respectively). PFS did not differ function. The matrix of the FLAIR sequence was adjusted to between sexes (p =0.46, p = 0.90, and p = 0.14, respectively) the perfusion maps using an in-house Matlab-based applica- or different extents of resection (p =0.15, p = 0.20, and p = tion (Matlab 2014b, MathWorks Natick, Massachusetts, 0.71, respectively). We could not assess the MR perfusion USA). We drew a volume of interest (VOI) covering the data in five patients due to incomplete coverage of the tumor whole tumor with signal alteration in the FLAIR sequence by imaging or a lack of arterial input function by insufficient and excluded surrounding edema, vessels, and necrotic areas inflow of the contrast agent. Two of these patients had IDH- [20, 21]. mutant astrocytic tumors, and three had IDH wild-type astro- For the reference VOI, we chose a slice in contralateral cytic tumors. normal-appearing white matter. We determined the tumor vol- ume in the T2-weighted FLAIR sequence because some tu- mors did not show contrast enhancement. The VOIs were Association of progression-free survival with mean transferred to the CBV maps automatically. We noted the tumor perfusion mean and maximal CBV inside the tumor VOI and the mean values inside the reference VOI. For further analyses, we cal- The mean rCBV was 1.2 ± 0.6 in IDH-mutant astrocytic tu- culated the relative CBV (rCBV) as the ratio of the perfusion mors, 2.7 ± 1.4 in IDH wild-type astrocytic tumors, and 1.9 ± values inside the tumor to values in the reference region. 0.8 in oligodendrogliomas. It differed significantly between the astrocytic tumors (p < 0.001), oligodendrogliomas and Statistical analysis IDH-mutant astrocytic tumors (p = 0.01), and oligodendrogliomas and IDH wild-type astrocytic tumors (p We performed Spearman’s correlation for the association of 1- = 0.049). The duration of the progression-free interval corre- year PFS with age and rCBV. We excluded patients who were lated with the mean rCBV in IDH wild-type astrocytic tumors lost to follow-up within the first year from these analyses. We (r =0.48, p = 0.04), but not in IDH-mutant astrocytic tumors defined the best rCBV cutoff values for predicting PFS in each (r = − 0.05, p = 0.89) and oligodendrogliomas (r = − 0.16, p = separate glioma group by a receiver operating characteristic 0.63). (ROC) analysis. We performed a Kaplan-Meier survival anal- In predicting the 1-year PFS of patients in the group of IDH ysis with a log-rank test for rCBV values, molecular patho- wild-type astrocytic tumors, the mean rCBV reached a sensi- logical tumor groups, sex, the extent of resection, WHO tivity of 0.78, a specificity of 0.83, and an area under the curve grades, and MGMT methylation status. We calculated the (AUC) of 0.81 (cutoff value = 2.0). Patients in this group with hazard ratios for the perfusion values. Using the Wilcoxon a mean rCBV below the cutoff value (n = 9) had a significant- test, we estimated differences in rCBV between tumor groups ly shorter 1-year PFS than those with a mean rCBV above the and MGMT promoter statuses and set the significance level at cutoff value (n =12, p = 0.02). A low mean rCBV was asso- α = 0.05. We used the software JMP (JMP 11.2.0, SAS ciated with a significantly higher risk for progression (hazard Institute, Cary, NC, USA) for the statistical calculations. ratio = 5.6; see Fig. 1 for an example). The mean rCBV did not differ between IDH wild-type astrocytic tumors with and without MGMT methylation (p = Results 0.44), and the PFS did not differ according to the MGMT methylation status. The percentage of patients receiving adju- Patients vant therapy was comparable in cases of IDH wild-type astro- cytic tumors with high and low perfusion. Only one patient Fifty-six patients were included in the present study (mean age with a highly perfused wild-type astrocytic tumor and two 48 ± 16years,33males,23females). The medianfollow-up patients with low-perfused wild-type astrocytic tumors did time was 350 days (range 11–1380 days), so we used the 1- not undergo adjuvant treatment. The WHO grades of IDH year PFS for further analyses. Tumor progression occurred in wild-type astrocytic tumors with low and high perfusion 31% of IDH-mutant astrocytic tumors, 46% of IDH wild-type showed similar distributions (WHO grade II n = 2 and n =2, astrocytic tumors, and 13% of oligodendrogliomas. The dura- WHO grade III n = 5 and n = 7, WHO grade IV n = 2 and n = tion of PFS tended to be higher in cases of IDH-mutant 3, respectively). 3628 Neurol Sci (2020) 41:3625–3632 Table 1 Patient and tumor Parameter IDH-mutant astrocytoma IDH wild-type astrocytoma Oligodendroglioma characteristics Total number 16 24 16 Progress 5 11 2 PFS rate 0.69 0.54 0.88 Mean patient age 40.3 ± 14.7 53.8 ± 17.3 46.2 ± 14.1 Patient sex (male/female) 10 / 6 15 / 9 8 / 8 Complete resection 4 2 7 Radiotherapy 10 20 10 Chemotherapy 7 19 10 WHO grade II 11 5 13 WHO grade III 3 14 3 WHO grade IV 2 5 - IDH isocitrate dehydrogenase, n number, PFS rate progression-free survival rate after 1 year, WHO World Health Organization In oligodendrogliomas, the mean rCBV could predict the rCBV differed significantly between with IDH wild- 1-year PFS with a sensitivity of 1.0, a specificity of 0.43, and type and IDH-mutant astrocytic tumors (p <0.003) an AUC of 0.64 (cutoff value = 1.7). The 1-year PFS did not and between oligodendrogliomas and IDH wild-type as- differ significantly in this glioma group, depending on the trocytic tumors (p = 0.005). However, there was no level of the mean rCBV (p = 0.87). In IDH-mutant astrocytic significant difference between oligodendrogliomas and tumors, ROC analysis with the mean rCBV did not yield a IDH-mutant astrocytic tumors (p = 0.48). The duration suitable differentiation of prognostic groups as the AUC of the progression-free interval did not correlate with was highest when all rCBV values were above any cut- the maximal rCBV in any glioma group (IDH-mutant astro- off value. Figure 2 and Table 2 show the details of the cytic tumors r = − 0.8, p = 0.83, IDH wild-type astrocytic survival analyses. tumors r =0.14, p = 0.59, oligodendrogliomas r = − 0.18, p = 0.59, respectively). In the IDH wild-type astrocytic tumor group, the maximal Association of progression-free survival with maximal rCBV reached a sensitivity of 0.33, specificity of 1.0, and tumor perfusion AUC of 0.61 in predicting the 1-year PFS (cutoff value = 4.4). Patients with a maximal rCBV below the cutoff value The maximal rCBV was 5.2 ± 3.3 in IDH-mutant astrocytic within these tumor group had a significantly shorter 1-year tumors, 7.5 ± 3.1 in IDH wild-type astrocytic tumors, PFS (p = 0.01) and a higher risk for progression (hazard ratio and 5.0 ± 1.7 in oligodendrogliomas. The maximal Fig. 1 Imaging of a 74-year-old patient with IDH wild-type astrocytomas WHO grade III (FLAIR sequence, a) with only local contrast agent uptake (post-contrast T1-weighted sequence, b), and low tumor perfusion (c, CBV map of DSC-MRI), but tumor progression within 130 days Neurol Sci (2020) 41:3625–3632 3629 ab c 1.0 1.0 1.0 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 100 150 200 250 300 350 100 150 200 250 300 350 100 150 200 250 300 350 Progression-free survival (days) Progression-free survival (days) Progression-free survival (days) Fig. 2 Progression-free survival of patients with mean tumor perfusion (c). In IDH-mutant astrocytic tumors (a), the best cutoff value was below above (gray) and below (black) the cutoff value in IDH-mutant astrocytic all measured perfusion values and limited the separation into two groups tumors (a), IDH wild-type astrocytic tumors (b), and oligodendrogliomas 5.1, p = 0.04) than patients with a maximal rCBV above the Discussion cutoff value. The maximal rCBV did not differ between IDH wild-type astrocytic tumors with and without MGMT methyl- The integrated 2016 CNS WHO of gliomas considers molec- ation (p =0.16). ular tumor characteristics as critical important prognostic fac- The percentage of patients receiving adjuvant therapy was tors beyond histological WHO grades [1, 22, 23]. According comparable in wild-type astrocytic tumors with high and low to previous reports, patients with oligodendrogliomas tended perfusion. 15 of 18 patients with high maximal rCBV and all to have longer PFS than patients with IDH-mutant astrocytic patients with low maximal rCBV underwent adjuvant treat- tumors in our cohort, and IDH wild-type astrocytic tumors had ment. The WHO grades in IDH wild-type astrocytic tumors the most unfavorable prognosis [24, 25]. The potential con- with low and high maximal rCBV showed a similar distribu- founders age, sex, and the extent of surgical resection were not tion. WHO grade III was dominant in both groups. associated with PFS in the present study [10, 11, 26]. IDH In oligodendrogliomas, the maximal rCBV predicted the 1- wild-type astrocytic tumors displayed the highest perfusion, year PFS with a sensitivity of 1.0, specificity of 0.43, and followed by oligodendrogliomas and IDH-mutant astrocytic AUC of 0.55 (cutoff value = 4.8). In IDH-mutant astrocytic tumors. This confirms previous reports that perfusion is a tumors, the maximal rCBV reached a sensitivity of 0.67, spec- marker for malignancy and the molecular characteristics of ificity of 0.81, and AUC of 0.60 (cutoff value = 5.6). The 1- diffuse gliomas [5, 13, 25, 27]. year PFS within the oligodendroglioma and IDH-mutant as- However, DSC-MRI may be dispensable for the identifi- trocytic tumor groups did not differ significantly depending cation of the molecular tumor characteristics because they are on the level of the maximal rCBV (p =0.87 and p =0.63, assessed by pathological examination in daily routine. respectively). Figure 3 and Table 2 show the details of the Therefore, we investigated the potential of DSC-MRI in survival analysis. predicting the PFS as an independent factor within the single molecular glioma groups. Knowledge about DSC-MRI in this Table 2 Survival analysis for patients with mean and maximal tumor perfusion below and above the cutoff value Parameter IDH-mutant astrocytoma IDH wild-type astrocytoma Oligodendroglioma Mean tumor perfusion Total number, </> cutoff value 14/– 9/12 7/9 PFS rate, </> cutoff value 0.79/– 0.22/0.83 0.86/0.89 p value – 0.02 0.87 Hazard ratio (95% CI) – 5.6 (1.3–37.9) 1.3 (0–32.0) Maximal tumor perfusion Total number, </> cutoff value 11/3 3/18 8/8 PFS rate, </> cutoff value 0.82/0.67 0/0.67 0.88/0.88 p value 0.63 0.01 0.87 Hazard ratio (95% CI) 0.6 (0.1–12.0) 5.1 (1.1–19.7) 10.8 (0–20.0) IDH isocitrate dehydrogenase, PFS rate progression-free survival rate after 1 year, CI confidence interval Survival probability Survival probability Survival probability 3630 Neurol Sci (2020) 41:3625–3632 ab c 1.0 1.0 1.0 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 100 150 200 250 300 350 100 150 200 250 300 350 100 150 200 250 300 350 Progression-free survival (days) Progression-free survival (days) Progression-free survival (days) Fig. 3 Progression-free survival of patients with maximal tumor perfusion above (gray) and below (black) the cutoff value in IDH-mutant astrocytic tumors (a), IDH wild-type astrocytic tumors (b), and oligodendrogliomas (c) context is limited, and previous studies concentrated on the The mean perfusion values of the whole tumor volume old WHO classification 2007 or the subgroup of WHO grade showed better performance and higher discriminatory power IV glioblastomas [9, 11]. Generally, lower tumor perfusion is than maximal perfusion values. Measuring the perfusion in a related to longer survival [7, 8, 28]. hotspot is a usual way to assess DSC-MRI in brain tumors. In oligodendrogliomas, the extent of tumor perfusion did However, our cohort comprised low-grade tumors without not predict the short-term prognosis. This finding has been contrast enhancement and occasionally lower perfusion than previously reported and might be due to the overall high per- the healthy tissue. The hotspot method does not work in this fusion and homogeneity of oligodendrogliomas [7, 10, case, and including the whole tumor volume in the measure- 24]. Likewise, the extent of tumor perfusion did not ment might represent the behavior of the tumor more thor- have any impact on the progression-free survival within oughly [20, 21, 31]. the group of IDH-mutant astrocytic tumors. The favor- DSC-MRI is a widely available, robust, and fast MR per- able prognosis and thus the low progression rate of both fusion technique [32]. It is the best established and investigat- glioma groups may hamper our results [24]. Future ed perfusion technique for gliomas. However, DSC-MRI is studies on the long-term prognosis might give new prone to bias due to contrast agent extravasation from leakage insights. through the blood-brain barrier, and appropriate correction is In IDH wild-type astrocytic tumors, the extent of tumor needed. Additionally, its evaluation is user-dependent, and the perfusion had an impact on the short-term prognosis. diagnostic quality is limited in regions with susceptibility ar- Interestingly, low perfusion was associated with a three to tifacts [32]. As an alternative, dynamic contrast-enhanced im- fourfold lower 1-year PFS rate than higher perfusion and sig- aging is based on T1-weighted images and can quantify the nificantly shorter survival. The main proportion in this group microvascular permeability [32]. Arterial spin labeling mea- was WHO grade III astrocytic tumors and not typical WHO sures the perfusion without contrast agent application and is grade IV glioblastomas, so our results might apply to them in independent of bias from leakage through the blood-brain bar- particular. rier [21]. It has shown promising results in the initial diagnosis The MGMT methylation status, which is an important and detection of recurrent disease in gliomas [33, 34]. prognostic factor in glioblastomas and might be a potential One limitation of the present study was the small number of confounder, did not differ between the low- and highly per- patients in the different glioma groups since advanced imag- fused IDH wild-type astrocytic tumors [29]. Furthermore, the ing, including DSC-MRI, is not performed on all patients therapy regimen was similar in both groups, which excludes preoperatively in our institution. Usually, only complex cases relevant bias by therapy on our results. Cimino et al. recently undergo advanced MRI for biopsy planning, tumor character- reported that IDH wild-type glioblastomas comprise several ization, and assessment of the tumor extent. This preselection further molecular subgroups with different prognosis might have biased our results. Tumor progression was not [24]. These subgroups might display different vascular- confirmed by histopathology but only by imaging according ization patterns, as seen in the present study. to the clinical standard. We defined tumor progress by the Additionally, low-perfused tumors might have a worse RANO criteria in order to avoid the inclusion of cases with response to adjuvant therapy due to diminished delivery pseudoprogression. of chemotherapeutic agents and higher resistance to ra- In conclusion, high tumor perfusion upon initial diagnosis diotherapy in hypoxic and low-perfused areas [30]. is not compellingly associated with worse short-term progno- Overall, our results suggest that higher vascularization sis within different molecular glioma subgroups. In the overall does not always result in a worse prognosis in diffuse highly perfused IDH wild-type astrocytic tumors, there are gliomas. tumors with low perfusion but unfavorable prognosis. Survival probability Survival probability Survival probability Neurol Sci (2020) 41:3625–3632 3631 Authors’ contributions All authors contributed to the study conception 4. Lu X, Xu W, Wei Y, Li T, Gao L, Fu X, Yao Y, Wang L (2019) and design. Material preparation, data collection, and analysis were per- Diagnostic performance of DWI for differentiating primary central formed by Cornelia Brendle, Uwe Klose, and Benjamin Bender. The first nervous system lymphoma from glioblastoma: a systematic review draft of the manuscript was written by Cornelia Brendle and all authors and meta-analysis. Neurol Sci 40(5):947–956. https://doi.org/10. commented on previous versions of the manuscript. All authors read and 1007/s10072-019-03732-7 approved the final manuscript. 5. Razek A, El-Serougy L, Abdelsalam M, Gaballa G, Talaat M (2020) Multi-parametric arterial spin labelling and diffusion- weighted magnetic resonance imaging in differentiation of grade Funding information Open Access funding enabled and organized by II and grade III gliomas. Pol J Radiol 84:e110–e117 Projekt DEAL. Cornelia Brendle has been supported by a grant from 6. Abdel Razek AA, Gaballa G (2011) Role of perfusion magnetic the TÜFF program of the Faculty of Medicine, Eberhard Karls resonance imaging in cervical lymphadenopathy. J Comput Assist University Tuebingen (application number 2395-0-0). Tomogr 35(1):21 –25. https://doi.org/10.1097/RCT. 0b013e3181ff9143 Compliance with ethical standards 7. Mangla R, Ginat DT, Kamalian S, Milano MT, Korones DN, Walter KA, Ekholm S (2014) Correlation between progression free survival and dynamic susceptibility contrast MRI perfusion in Conflict of interest The Department of Diagnostic and Interventional Radiology has a collaboration contract with Siemens concerning the tech- WHO grade III glioma subtypes. J Neuro-Oncol 116(2):325–331. https://doi.org/10.1007/s11060-013-1298-9 nical development of PET/MRI Biograph mMR. Otherwise, the authors declare no conflict of interest. 8. Rau MK, Braun C, Skardelly M, Schittenhelm J, Paulsen F, Bender B, Ernemann U, Bisdas S (2014) Prognostic value of blood flow estimated by arterial spin labeling and dynamic susceptibility Ethical approval All procedures performed in studies involving human contrast-enhanced MR imaging in high-grade gliomas. J Neuro- participants were in accordance with the ethical standards of the institu- Oncol 120(3):557–566. https://doi.org/10.1007/s11060-014-1586- tional research committee of the Medical Faculty of the University of Tuebingen and with the 1964 Helsinki declaration and its later amend- 9. Romano A, Pasquini L, Di Napoli A, Tavanti F, Boellis A, Rossi ments or comparable ethical standards. For this type of study formal Espagnet MC, Minniti G, Bozzao A (2018) Prediction of survival in consent is not required. patients affected by glioblastoma: histogram analysis of perfusion MRI. J Neuro-Oncol 139(2):455–460. https://doi.org/10.1007/ Informed consent Informed consent was obtained from all individual s11060-018-2887-4 participants included in the study. 10. Coban G, Mohan S, Kural F, Wang S, O’Rourke DM, Poptani H (2015) Prognostic value of dynamic susceptibility contrast- Open Access This article is licensed under a Creative Commons enhanced and diffusion-weighted mr imaging in patients with glio- Attribution 4.0 International License, which permits use, sharing, adap- blastomas. AJNR Am J Neuroradiol 36(7):1247–1252. https://doi. tation, distribution and reproduction in any medium or format, as long as org/10.3174/ajnr.A4284 you give appropriate credit to the original author(s) and the source, pro- 11. Bonekamp D, Deike K, Wiestler B, Wick W, Bendszus M, vide a link to the Creative Commons licence, and indicate if changes were Radbruch A, Heiland S (2015) Association of overall survival in made. The images or other third party material in this article are included patients with newly diagnosed glioblastoma with contrast- in the article's Creative Commons licence, unless indicated otherwise in a enhanced perfusion MRI: comparison of intraindividually matched credit line to the material. If material is not included in the article's T1 - and T2 (*) -based bolus techniques. J Magn Reson Imaging Creative Commons licence and your intended use is not permitted by 42(1):87–96. https://doi.org/10.1002/jmri.24756 statutory regulation or exceeds the permitted use, you will need to obtain 12. Brendle C, Hempel JM, Schittenhelm J, Skardelly M, Tabatabai G, permission directly from the copyright holder. To view a copy of this Bender B, Ernemann U, Klose U (2017) Glioma grading and de- licence, visit http://creativecommons.org/licenses/by/4.0/. termination of IDH mutation status and ATRX loss by DCE and asl perfusion. Clin Neuroradiol. 28:421–428. https://doi.org/10.1007/ s00062-017-0590-z 13. Kickingereder P, Sahm F, Radbruch A, Wick W, Heiland S, Deimling A, Bendszus M, Wiestler B (2015) IDH mutation status References is associated with a distinct hypoxia/angiogenesis transcriptome signature which is non-invasively predictable with rCBV imaging 1. Louis DN, Perry A, Reifenberger G, von Deimling A, Figarella- in human glioma. Sci Rep 5:16238. https://doi.org/10.1038/ Branger D, Cavenee WK, Ohgaki H, Wiestler OD, Kleihues P, srep16238 Ellison DW (2016) The 2016 World Health Organization classifi- 14. Zhang HW, Lyu GW, He WJ, Lei Y, Lin F, Wang MZ, Zhang H, cation of tumors of the central nervous system: a summary. Acta Liang LH, Feng YN, Yang JH (2020) DSC and DCE Histogram neuropathologica 131(6):803–820. https://doi.org/10.1007/s00401- analyses of glioma biomarkers, including IDH, MGMT, and TERT, 016-1545-1 on differentiation and survival. Acad Radiol. https://doi.org/10. 2. Cuccarini V, Antelmi L, Pollo B, Paterra R, Calatozzolo C, Nigri A, 1016/j.acra.2019.12.010 DiMeco F, Eoli M, Finocchiaro G, Brenna G, Tramacere I, 15. Schittenhelm J, Mittelbronn M, Meyermann R, Melms A, Tatagiba Bruzzone MG, Anghileri E (2020) In vivo 2-hydroxyglutarate- M, Capper D (2011) Confirmation of R132H mutation of isocitrate proton magnetic resonance spectroscopy (3T, PRESS technique) dehydrogenase 1 as an independent prognostic factor in anaplastic in treatment-naive suspect lower-grade gliomas: feasibility and ac- astrocytoma. Acta neuropathologica 122(5):651–652. https://doi. curacy in a clinical setting. Neurol Sci 41(2):347–355. https://doi. org/10.1007/s00401-011-0885-0 org/10.1007/s10072-019-04087-9 16. Hartmann C, Meyer J, Balss J, Capper D, Mueller W, Christians A, 3. El-Serougy L, Abdel Razek AA, Ezzat A, Eldawoody H, El-Morsy Felsberg J, Wolter M, Mawrin C, Wick W, Weller M, Herold- A (2016) Assessment of diffusion tensor imaging metrics in differ- Mende C, Unterberg A, Jeuken JW, Wesseling P, Reifenberger entiating low-grade from high-grade gliomas. Neuroradiol J 29(5): G, von Deimling A (2009) Type and frequency of IDH1 and 400–407. https://doi.org/10.1177/1971400916665382 IDH2 mutations are related to astrocytic and oligodendroglial 3632 Neurol Sci (2020) 41:3625–3632 differentiation and age: a study of 1,010 diffuse gliomas. Acta defined by the 2016 WHO classification for brain tumors: histo- neuropathol 118(4):469–474. https://doi.org/10.1007/s00401-009- gram analysis approach. Neuroradiology 61(5):545–555. https:// 0561-9 doi.org/10.1007/s00234-019-02173-5 17. Reuss DE, Sahm F, Schrimpf D, Wiestler B, Capper D, Koelsche C, 26. Jain R, Poisson LM, Gutman D, Scarpace L, Hwang SN, Holder Schweizer L, Korshunov A, Jones DT, Hovestadt V, Mittelbronn CA, Wintermark M, Rao A, Colen RR, Kirby J, Freymann J, Jaffe M, Schittenhelm J, Herold-Mende C, Unterberg A, Platten M, CC, Mikkelsen T, Flanders A (2014) Outcome prediction in pa- Weller M, Wick W, Pfister SM, von Deimling A (2015) ATRX tients with glioblastoma by using imaging, clinical, and genomic and IDH1-R132H immunohistochemistry with subsequent copy biomarkers: focus on the nonenhancing component of the tumor. number analysis and IDH sequencing as a basis for an “integrated” Radiology 272(2):484–493. https://doi.org/10.1148/radiol. diagnostic approach for adult astrocytoma, oligodendroglioma and glioblastoma. Acta Neuropathol 129(1):133–146. https://doi.org/ 27. Hempel JM, Schittenhelm J, Klose U, Bender B, Bier G, Skardelly 10.1007/s00401-014-1370-3 M, Tabatabai G, Castaneda Vega S, Ernemann U, Brendle C (2019) 18. Thon N, Eigenbrod S, Grasbon-Frodl EM, Ruiter M, Mehrkens JH, In vivo molecular profiling of human glioma : cross-sectional ob- Kreth S, Tonn JC, Kretzschmar HA, Kreth FW (2009) Novel mo- servational study using dynamic susceptibility contrast magnetic lecular stereotactic biopsy procedures reveal intratumoral homoge- resonance perfusion imaging. Clin Neuroradiol 29(3):479–491. neity of loss of heterozygosity of 1p/19q and TP53 mutations in https://doi.org/10.1007/s00062-018-0676-2 World Health Organization grade II gliomas. J Neuropathol Exp 28. Hashido T, Saito S, Ishida T (2020) A radiomics-based comparative Neurol 68(11):1219–1228. https://doi.org/10.1097/NEN. study on arterial spin labeling and dynamic susceptibility contrast 0b013e3181bee1f1 perfusion-weighted imaging in gliomas. Sci Rep 10(1):6121. 19. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, https://doi.org/10.1038/s41598-020-62658-9 Weller M, Kros JM, Hainfellner JA, Mason W, Mariani L, 29. Binabaj MM, Bahrami A, ShahidSales S, Joodi M, Joudi Mashhad Bromberg JE, Hau P, Mirimanoff RO, Cairncross JG, Janzer RC, M, Hassanian SM, Anvari K, Avan A (2017) The prognostic value Stupp R (2005) MGMT gene silencing and benefit from temozolo- of MGMT promoter methylation in glioblastoma: a meta-analysis mide in glioblastoma. N Engl J Med 352(10):997–1003. https://doi. of clinical trials. J Cell Physiol. 233:378–386. https://doi.org/10. org/10.1056/NEJMoa043331 1002/jcp.25896 20. Brendle C, Hempel JM, Schittenhelm J, Skardelly M, Reischl G, 30. Lemasson B, Chenevert TL, Lawrence TS, Tsien C, Sundgren PC, Bender B, Ernemann U, la Fougere C, Klose U (2018) Glioma Meyer CR, Junck L, Boes J, Galban S, Johnson TD, Rehemtulla A, grading by dynamic susceptibility contrast perfusion and (11)C- Ross BD, Galban CJ (2013) Impact of perfusion map analysis on methionine positron emission tomography using different regions early survival prediction accuracy in glioma patients. Transl Oncol of interest. Neuroradiology 60(4):381–389. https://doi.org/10.1007/ 6(6):766–774 s00234-018-1993-5 31. Xing Z, Zhang H, She D, Lin Y, Zhou X, Zeng Z, Cao D (2019) 21. Bonm AV, Ritterbusch R, Throckmorton P, Graber JJ (2020) IDH genotypes differentiation in glioblastomas using DWI and Clinical imaging for diagnostic challenges in the management of DSC-PWI in the enhancing and peri-enhancing region. Acta gliomas: a review. J Neuroimaging 30(2):139–145. https://doi.org/ Radiol 60(12):1663 –1672. https://doi.org/10.1177/ 10.1111/jon.12687 22. Dong C, Yuan Z, Li Q, Wang Y (2018) The clinicopathological and 32. van Dijken BRJ, van Laar PJ, Smits M, Dankbaar JW, Enting RH, prognostic significance of TP53 alteration in K27M mutated glio- van der Hoorn A (2019) Perfusion MRI in treatment evaluation of mas: an individual-participant data meta-analysis. Neurol Sci 39(7): glioblastomas: clinical relevance of current and future techniques. J 1191–1201. https://doi.org/10.1007/s10072-018-3407-1 Magn Reson Imaging 49(1):11–22. https://doi.org/10.1002/jmri. 23. Xiao Q, Yang S, Ding G, Luo M (2018) Anti-vascular endothelial growth factor in glioblastoma: a systematic review and meta-anal- 33. Abdel Razek AAK, Talaat M, El-Serougy L, Gaballa G, ysis. Neurol Sci 39(12):2021–2031. https://doi.org/10.1007/ Abdelsalam M (2019) Clinical applications of arterial spin labeling s10072-018-3568-y in brain tumors. J Comp Assist Tomogr 43(4):525–532. https://doi. 24. Cimino PJ, Zager M, McFerrin L, Wirsching HG, Bolouri H, org/10.1097/RCT.0000000000000873 Hentschel B, von Deimling A, Jones D, Reifenberger G, Weller 34. Razek A, El-Serougy L, Abdelsalam M, Gaballa G, Talaat M M, Holland EC (2017) Multidimensional scaling of diffuse glio- (2018) Differentiation of residual/recurrent gliomas from mas: application to the 2016 World Health Organization classifica- postradiation necrosis with arterial spin labeling and diffusion ten- tion system with prognostically relevant molecular subtype discov- sor magnetic resonance imaging-derived metrics. Neuroradiology ery. Acta Neuropathol Commun 5(1):39. https://doi.org/10.1186/ 60(2):169–177. https://doi.org/10.1007/s00234-017-1955-3 s40478-017-0443-7 25. Latysheva A, Emblem KE, Brandal P, Vik-Mo EO, Pahnke J, Roysland K, Hald JK, Server A (2019) Dynamic susceptibility Publisher’snote Springer Nature remains neutral with regard to jurisdic- contrast and diffusion MR imaging identify oligodendroglioma as tional claims in published maps and institutional affiliations. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Neurological Sciences Unpaywall

Association of dynamic susceptibility magnetic resonance imaging at initial tumor diagnosis with the prognosis of different molecular glioma subtypes

Download PDF
8 pages

Loading next page...
 
/lp/unpaywall/association-of-dynamic-susceptibility-magnetic-resonance-imaging-at-8iGwZvgjsO

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
1590-1874
DOI
10.1007/s10072-020-04474-7
Publisher site
See Article on Publisher Site

Abstract

Purpose The updated 2016 CNS World Health Organization classification differentiates three main groups of diffuse glioma according to their molecular characteristics: astrocytic tumors with and without isocitrate dehydrogenase (IDH) mutation and 1p/ 19q co-deleted oligodendrogliomas. The present study aimed to determine whether dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) is an independent prognostic marker within the molecular subgroups of diffuse glioma. Methods Fifty-six patients with treatment-naive gliomas and advanced preoperative MRI examination were assessed retrospec- tively. The mean and maximal normalized cerebral blood volume values from DSC-MRI within the tumors were measured. Optimal cutoff values for the 1-year progression-free survival (PFS) were defined, and Kaplan-Meier analyses were performed separately for the three glioma subgroups. Results IDH wild-type astrocytic tumors had a higher mean and maximal perfusion than IDH-mutant astrocytic tumors and oligodendrogliomas. Patients with IDH wild-type astrocytic tumors and a low mean or maximal perfusion had a significantly shorter PFS than patients of the same group with high perfusion (p = 0.0159/0.0112). Furthermore, they had a significantly higher risk for early progression (hazard ratio = 5.6/5.1). This finding was independent of the methylation status of O6-methylguanin- DNA-methyltransferase and variations of the therapy. Within the groups of IDH-mutant astrocytic tumors and oligodendrogliomas, the PFS of low and highly perfused tumors did not differ. Conclusion High perfusion upon initial diagnosis is not compellingly associated with worse short-term prognosis within the different molecular subgroups of diffuse glioma. Particularly, the overall highly perfused group of IDH wild-type astrocytic tumors contains tumors with low perfusion but unfavorable prognosis. . . . . . Keywords Glioma DSC-MRI Perfusion prognosis IDH mutation Molecular Introduction The 2016 update of the World Health Organization classifica- tion of brain tumors (2016 CNS WHO) now combines specif- * Cornelia Brendle ic molecular characteristics of gliomas beyond histological [email protected] characteristics for assignment to different subgroups in an integrated approach. Consequently, diffuse gliomas are clas- Diagnostic and Interventional Neuroradiology, Department of sified into three main groups. The first two are astrocytic tu- Radiology, Eberhard Karls University, Hoppe-Seyler-Straße 3, mors, which are divided into tumors with isocitrate dehydro- 72076 Tuebingen, Germany genase (IDH) mutation and without IDH mutation (IDH wild Neuropathology, Department of Pathology and Neuropathology, type). The third group, oligodendrogliomas, is defined by IDH Eberhard Karls University, Calwerstr. 3, 72076 Tuebingen, Germany 3 mutation, codeletion of chromosomes 1p and 19q, and reten- University Hospital for Neurosurgery, Eberhard Karls University, tion of the ATRX gene [1]. Hoppe-Seyler-Straße 3, 72076 Tuebingen, Germany 4 Magnetic resonance imaging (MRI) is the modality of Interdisciplinary Section of Neurooncology, Eberhard Karls choice for diagnosing brain tumors, as well as assessing the University, Hoppe-Seyler-Straße 3, 72076 Tuebingen, Germany 3626 Neurol Sci (2020) 41:3625–3632 tumor characteristics, the extent at the initial diagnosis, and The inclusion criteria comprised adult patient age, initial the therapy response in the course of the disease. Standard tumor diagnosis, advanced MRI with DSC perfusion before MRI is limited to visualizing morphological changes and the any treatment, pathologic examination including molecular disruption of the blood-brain barrier after injection of contrast profiling, and follow-up by clinical examination and imaging. agent. Beyond that, advanced MRI techniques give insights We assessed the age, sex, extent of tumor resection (complete into the functional properties and tissue composition of glio- versus incomplete), and the application of adjuvant therapy. mas. MR spectroscopy measures biochemical metabolites and We gathered follow-up data until progressive disease occurred can provide direct proof of IDH mutations in gliomas [2]. according to the Response Assessment in Neuro-Oncology Diffusion-weighted imaging displays the variable cellular- (RANO) criteria or until the patients were lost to follow-up ity and tissue structure in different tumor entities and the grade [1]. of malignancy [3–5]. Perfusion MRI techniques visualize the We calculated the PFS from the date of the advanced MR vascularization by quantifying the blood flow through the tis- examination to the date of tumor progression. We chose tumor sue, which is increased in malignant tissue due to neo- progression and not overall survival as the endpoint because angiogenesis [6]. Dynamic susceptibility contrast (DSC)- patients in specific glioma groups partially survive for a long MRI is the most common perfusion MRI technique. The con- time, and the progression-free interval is crucial for quality of trast agent inflow in the T2*-weighted images causes life. Furthermore, the overall survival of patients is influenced susceptibility-induced signal changes, which are used for cal- by variable therapy management in the case of progression. culating the cerebral blood volume (CBV). The absolute per- The IDH mutation status of the tumors was identified using fusion values of DSC-MRI are prone to multiple sources of an IDH1 R132H antibody, and rare IDH1/2 mutations in error, so they are usually corrected using an internal reference, IDH1 R132H-negative cases were identified by pyrosequenc- resulting in relative values (rCBV). ing [15, 16]. Nuclear ATRX status was assessed using immu- DSC-MRI is widely used in the diagnosis of gliomas and nohistochemistry and 1p/19q codeletion by a synthetic high- has the potential to predict the prognosis of glioma subtypes resolution microsatellite PCR gel [17, 18]. Possible according to the 2007 WHO classification, particularly in methylation of the O6-methylguanin-DNA- high-grade astrocytomas [7–9]. Previous studies reported that methyltransferase (MGMT) promoter of IDH wild-type lower tumor perfusion is associated with prolonged survival astrocytic tumors was identified using methylation- [7, 8, 10, 11]. DSC-MRI can also help to identify the molec- specific pyrosequencing [19]. For further analyses, we ular subgroups of diffuse glioma according to the updated differentiated three glioma subgroups according to their 2016 WHO classification, which differ substantially in their molecular characteristics: IDH wild-type astrocytic tu- prognosis [12, 13]. However, the genetic profile of a tumor is mors, IDH-mutant astrocytic tumors (ATRX loss), and usually confirmed histologically, and there is only limited oligodendrogliomas (IDH mutation, 1p19q codeletion, knowledge about the prognostic value of DSC-MRI as an ATRX retention). independent prognostic factor within a specific glioma group [14]. Therefore, the present study aims to evaluate the 1-year MRI examination progression-free survival (PFS) of treatment-naïve patients with IDH wild-type astrocytic tumors, IDH-mutant astrocytic All patients were examined on a 3-T MRI scanner (Biograph tumors, and oligodendrogliomas according to the DSC-MRI mMR MR-PET, Siemens Healthcare, Erlangen, Germany) perfusion values. using a head-neck coil. The MR sequences comprised an axial T2-weighted fluid-attenuated inversion recovery (FLAIR) se- quence as an anatomical reference (repetition time 9000 ms, echo time 94 ms, inversion time 2500 ms, slice thickness 3 Materials and methods mm, matrix size 207 × 320, field of view 189 × 220 mm ). For DSC-MRI, a pre-bolus of contrast agent was applied (0.025 Patients mmol/kg gadobutrol (Gadovist, Bayer Healthcare, Leverkusen, Germany)). Subsequently, a single-shot T2*- We retrospectively evaluated all patients with an initial diag- weighted echo-planar imaging sequence was performed dur- nosis of diffuse glioma and clinically indicated advanced MR ing the first pass of a contrast agent bolus of 0.1 mmol/kg of examination between 11/2012 and 9/2016. All patients pro- gadobutrol (Gadovist, Bayer Healthcare, Leverkusen, vided informed written consent for the scientific use of their Germany; injection rate 3 ml/s). The repetition time was data. The local ethics committee approved the study, which 1130 ms, the echo time was 31 ms, the flip angle was 60°, was performed in accordance with the Declaration of the slice thickness was 4 mm, the matrix size was 128 × 128, Helsinki. and the field of view was 230 × 230 mm . Neurol Sci (2020) 41:3625–3632 3627 Image analysis astrocytic tumors than IDH wild-type astrocytic tumors. Oligodendrogliomas had the best prognosis, but the differ- We calculated CBV parametric maps using raw data from ences between the glioma groups were not significant (p = DSC-MRI with the software Syngo® MR Perfusion 0.09). (Siemens Healthineers, Erlangen, Germany) with model- Table 1 shows detailed characteristics of the glioma sub- based leakage correction. We manually identified the middle groups. Within the groups, PFS did not correlate with age (p = cerebral artery or anterior cerebral artery for the arterial input 0.15, p = 0.30, and p = 0.65, respectively). PFS did not differ function. The matrix of the FLAIR sequence was adjusted to between sexes (p =0.46, p = 0.90, and p = 0.14, respectively) the perfusion maps using an in-house Matlab-based applica- or different extents of resection (p =0.15, p = 0.20, and p = tion (Matlab 2014b, MathWorks Natick, Massachusetts, 0.71, respectively). We could not assess the MR perfusion USA). We drew a volume of interest (VOI) covering the data in five patients due to incomplete coverage of the tumor whole tumor with signal alteration in the FLAIR sequence by imaging or a lack of arterial input function by insufficient and excluded surrounding edema, vessels, and necrotic areas inflow of the contrast agent. Two of these patients had IDH- [20, 21]. mutant astrocytic tumors, and three had IDH wild-type astro- For the reference VOI, we chose a slice in contralateral cytic tumors. normal-appearing white matter. We determined the tumor vol- ume in the T2-weighted FLAIR sequence because some tu- mors did not show contrast enhancement. The VOIs were Association of progression-free survival with mean transferred to the CBV maps automatically. We noted the tumor perfusion mean and maximal CBV inside the tumor VOI and the mean values inside the reference VOI. For further analyses, we cal- The mean rCBV was 1.2 ± 0.6 in IDH-mutant astrocytic tu- culated the relative CBV (rCBV) as the ratio of the perfusion mors, 2.7 ± 1.4 in IDH wild-type astrocytic tumors, and 1.9 ± values inside the tumor to values in the reference region. 0.8 in oligodendrogliomas. It differed significantly between the astrocytic tumors (p < 0.001), oligodendrogliomas and Statistical analysis IDH-mutant astrocytic tumors (p = 0.01), and oligodendrogliomas and IDH wild-type astrocytic tumors (p We performed Spearman’s correlation for the association of 1- = 0.049). The duration of the progression-free interval corre- year PFS with age and rCBV. We excluded patients who were lated with the mean rCBV in IDH wild-type astrocytic tumors lost to follow-up within the first year from these analyses. We (r =0.48, p = 0.04), but not in IDH-mutant astrocytic tumors defined the best rCBV cutoff values for predicting PFS in each (r = − 0.05, p = 0.89) and oligodendrogliomas (r = − 0.16, p = separate glioma group by a receiver operating characteristic 0.63). (ROC) analysis. We performed a Kaplan-Meier survival anal- In predicting the 1-year PFS of patients in the group of IDH ysis with a log-rank test for rCBV values, molecular patho- wild-type astrocytic tumors, the mean rCBV reached a sensi- logical tumor groups, sex, the extent of resection, WHO tivity of 0.78, a specificity of 0.83, and an area under the curve grades, and MGMT methylation status. We calculated the (AUC) of 0.81 (cutoff value = 2.0). Patients in this group with hazard ratios for the perfusion values. Using the Wilcoxon a mean rCBV below the cutoff value (n = 9) had a significant- test, we estimated differences in rCBV between tumor groups ly shorter 1-year PFS than those with a mean rCBV above the and MGMT promoter statuses and set the significance level at cutoff value (n =12, p = 0.02). A low mean rCBV was asso- α = 0.05. We used the software JMP (JMP 11.2.0, SAS ciated with a significantly higher risk for progression (hazard Institute, Cary, NC, USA) for the statistical calculations. ratio = 5.6; see Fig. 1 for an example). The mean rCBV did not differ between IDH wild-type astrocytic tumors with and without MGMT methylation (p = Results 0.44), and the PFS did not differ according to the MGMT methylation status. The percentage of patients receiving adju- Patients vant therapy was comparable in cases of IDH wild-type astro- cytic tumors with high and low perfusion. Only one patient Fifty-six patients were included in the present study (mean age with a highly perfused wild-type astrocytic tumor and two 48 ± 16years,33males,23females). The medianfollow-up patients with low-perfused wild-type astrocytic tumors did time was 350 days (range 11–1380 days), so we used the 1- not undergo adjuvant treatment. The WHO grades of IDH year PFS for further analyses. Tumor progression occurred in wild-type astrocytic tumors with low and high perfusion 31% of IDH-mutant astrocytic tumors, 46% of IDH wild-type showed similar distributions (WHO grade II n = 2 and n =2, astrocytic tumors, and 13% of oligodendrogliomas. The dura- WHO grade III n = 5 and n = 7, WHO grade IV n = 2 and n = tion of PFS tended to be higher in cases of IDH-mutant 3, respectively). 3628 Neurol Sci (2020) 41:3625–3632 Table 1 Patient and tumor Parameter IDH-mutant astrocytoma IDH wild-type astrocytoma Oligodendroglioma characteristics Total number 16 24 16 Progress 5 11 2 PFS rate 0.69 0.54 0.88 Mean patient age 40.3 ± 14.7 53.8 ± 17.3 46.2 ± 14.1 Patient sex (male/female) 10 / 6 15 / 9 8 / 8 Complete resection 4 2 7 Radiotherapy 10 20 10 Chemotherapy 7 19 10 WHO grade II 11 5 13 WHO grade III 3 14 3 WHO grade IV 2 5 - IDH isocitrate dehydrogenase, n number, PFS rate progression-free survival rate after 1 year, WHO World Health Organization In oligodendrogliomas, the mean rCBV could predict the rCBV differed significantly between with IDH wild- 1-year PFS with a sensitivity of 1.0, a specificity of 0.43, and type and IDH-mutant astrocytic tumors (p <0.003) an AUC of 0.64 (cutoff value = 1.7). The 1-year PFS did not and between oligodendrogliomas and IDH wild-type as- differ significantly in this glioma group, depending on the trocytic tumors (p = 0.005). However, there was no level of the mean rCBV (p = 0.87). In IDH-mutant astrocytic significant difference between oligodendrogliomas and tumors, ROC analysis with the mean rCBV did not yield a IDH-mutant astrocytic tumors (p = 0.48). The duration suitable differentiation of prognostic groups as the AUC of the progression-free interval did not correlate with was highest when all rCBV values were above any cut- the maximal rCBV in any glioma group (IDH-mutant astro- off value. Figure 2 and Table 2 show the details of the cytic tumors r = − 0.8, p = 0.83, IDH wild-type astrocytic survival analyses. tumors r =0.14, p = 0.59, oligodendrogliomas r = − 0.18, p = 0.59, respectively). In the IDH wild-type astrocytic tumor group, the maximal Association of progression-free survival with maximal rCBV reached a sensitivity of 0.33, specificity of 1.0, and tumor perfusion AUC of 0.61 in predicting the 1-year PFS (cutoff value = 4.4). Patients with a maximal rCBV below the cutoff value The maximal rCBV was 5.2 ± 3.3 in IDH-mutant astrocytic within these tumor group had a significantly shorter 1-year tumors, 7.5 ± 3.1 in IDH wild-type astrocytic tumors, PFS (p = 0.01) and a higher risk for progression (hazard ratio and 5.0 ± 1.7 in oligodendrogliomas. The maximal Fig. 1 Imaging of a 74-year-old patient with IDH wild-type astrocytomas WHO grade III (FLAIR sequence, a) with only local contrast agent uptake (post-contrast T1-weighted sequence, b), and low tumor perfusion (c, CBV map of DSC-MRI), but tumor progression within 130 days Neurol Sci (2020) 41:3625–3632 3629 ab c 1.0 1.0 1.0 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 100 150 200 250 300 350 100 150 200 250 300 350 100 150 200 250 300 350 Progression-free survival (days) Progression-free survival (days) Progression-free survival (days) Fig. 2 Progression-free survival of patients with mean tumor perfusion (c). In IDH-mutant astrocytic tumors (a), the best cutoff value was below above (gray) and below (black) the cutoff value in IDH-mutant astrocytic all measured perfusion values and limited the separation into two groups tumors (a), IDH wild-type astrocytic tumors (b), and oligodendrogliomas 5.1, p = 0.04) than patients with a maximal rCBV above the Discussion cutoff value. The maximal rCBV did not differ between IDH wild-type astrocytic tumors with and without MGMT methyl- The integrated 2016 CNS WHO of gliomas considers molec- ation (p =0.16). ular tumor characteristics as critical important prognostic fac- The percentage of patients receiving adjuvant therapy was tors beyond histological WHO grades [1, 22, 23]. According comparable in wild-type astrocytic tumors with high and low to previous reports, patients with oligodendrogliomas tended perfusion. 15 of 18 patients with high maximal rCBV and all to have longer PFS than patients with IDH-mutant astrocytic patients with low maximal rCBV underwent adjuvant treat- tumors in our cohort, and IDH wild-type astrocytic tumors had ment. The WHO grades in IDH wild-type astrocytic tumors the most unfavorable prognosis [24, 25]. The potential con- with low and high maximal rCBV showed a similar distribu- founders age, sex, and the extent of surgical resection were not tion. WHO grade III was dominant in both groups. associated with PFS in the present study [10, 11, 26]. IDH In oligodendrogliomas, the maximal rCBV predicted the 1- wild-type astrocytic tumors displayed the highest perfusion, year PFS with a sensitivity of 1.0, specificity of 0.43, and followed by oligodendrogliomas and IDH-mutant astrocytic AUC of 0.55 (cutoff value = 4.8). In IDH-mutant astrocytic tumors. This confirms previous reports that perfusion is a tumors, the maximal rCBV reached a sensitivity of 0.67, spec- marker for malignancy and the molecular characteristics of ificity of 0.81, and AUC of 0.60 (cutoff value = 5.6). The 1- diffuse gliomas [5, 13, 25, 27]. year PFS within the oligodendroglioma and IDH-mutant as- However, DSC-MRI may be dispensable for the identifi- trocytic tumor groups did not differ significantly depending cation of the molecular tumor characteristics because they are on the level of the maximal rCBV (p =0.87 and p =0.63, assessed by pathological examination in daily routine. respectively). Figure 3 and Table 2 show the details of the Therefore, we investigated the potential of DSC-MRI in survival analysis. predicting the PFS as an independent factor within the single molecular glioma groups. Knowledge about DSC-MRI in this Table 2 Survival analysis for patients with mean and maximal tumor perfusion below and above the cutoff value Parameter IDH-mutant astrocytoma IDH wild-type astrocytoma Oligodendroglioma Mean tumor perfusion Total number, </> cutoff value 14/– 9/12 7/9 PFS rate, </> cutoff value 0.79/– 0.22/0.83 0.86/0.89 p value – 0.02 0.87 Hazard ratio (95% CI) – 5.6 (1.3–37.9) 1.3 (0–32.0) Maximal tumor perfusion Total number, </> cutoff value 11/3 3/18 8/8 PFS rate, </> cutoff value 0.82/0.67 0/0.67 0.88/0.88 p value 0.63 0.01 0.87 Hazard ratio (95% CI) 0.6 (0.1–12.0) 5.1 (1.1–19.7) 10.8 (0–20.0) IDH isocitrate dehydrogenase, PFS rate progression-free survival rate after 1 year, CI confidence interval Survival probability Survival probability Survival probability 3630 Neurol Sci (2020) 41:3625–3632 ab c 1.0 1.0 1.0 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 100 150 200 250 300 350 100 150 200 250 300 350 100 150 200 250 300 350 Progression-free survival (days) Progression-free survival (days) Progression-free survival (days) Fig. 3 Progression-free survival of patients with maximal tumor perfusion above (gray) and below (black) the cutoff value in IDH-mutant astrocytic tumors (a), IDH wild-type astrocytic tumors (b), and oligodendrogliomas (c) context is limited, and previous studies concentrated on the The mean perfusion values of the whole tumor volume old WHO classification 2007 or the subgroup of WHO grade showed better performance and higher discriminatory power IV glioblastomas [9, 11]. Generally, lower tumor perfusion is than maximal perfusion values. Measuring the perfusion in a related to longer survival [7, 8, 28]. hotspot is a usual way to assess DSC-MRI in brain tumors. In oligodendrogliomas, the extent of tumor perfusion did However, our cohort comprised low-grade tumors without not predict the short-term prognosis. This finding has been contrast enhancement and occasionally lower perfusion than previously reported and might be due to the overall high per- the healthy tissue. The hotspot method does not work in this fusion and homogeneity of oligodendrogliomas [7, 10, case, and including the whole tumor volume in the measure- 24]. Likewise, the extent of tumor perfusion did not ment might represent the behavior of the tumor more thor- have any impact on the progression-free survival within oughly [20, 21, 31]. the group of IDH-mutant astrocytic tumors. The favor- DSC-MRI is a widely available, robust, and fast MR per- able prognosis and thus the low progression rate of both fusion technique [32]. It is the best established and investigat- glioma groups may hamper our results [24]. Future ed perfusion technique for gliomas. However, DSC-MRI is studies on the long-term prognosis might give new prone to bias due to contrast agent extravasation from leakage insights. through the blood-brain barrier, and appropriate correction is In IDH wild-type astrocytic tumors, the extent of tumor needed. Additionally, its evaluation is user-dependent, and the perfusion had an impact on the short-term prognosis. diagnostic quality is limited in regions with susceptibility ar- Interestingly, low perfusion was associated with a three to tifacts [32]. As an alternative, dynamic contrast-enhanced im- fourfold lower 1-year PFS rate than higher perfusion and sig- aging is based on T1-weighted images and can quantify the nificantly shorter survival. The main proportion in this group microvascular permeability [32]. Arterial spin labeling mea- was WHO grade III astrocytic tumors and not typical WHO sures the perfusion without contrast agent application and is grade IV glioblastomas, so our results might apply to them in independent of bias from leakage through the blood-brain bar- particular. rier [21]. It has shown promising results in the initial diagnosis The MGMT methylation status, which is an important and detection of recurrent disease in gliomas [33, 34]. prognostic factor in glioblastomas and might be a potential One limitation of the present study was the small number of confounder, did not differ between the low- and highly per- patients in the different glioma groups since advanced imag- fused IDH wild-type astrocytic tumors [29]. Furthermore, the ing, including DSC-MRI, is not performed on all patients therapy regimen was similar in both groups, which excludes preoperatively in our institution. Usually, only complex cases relevant bias by therapy on our results. Cimino et al. recently undergo advanced MRI for biopsy planning, tumor character- reported that IDH wild-type glioblastomas comprise several ization, and assessment of the tumor extent. This preselection further molecular subgroups with different prognosis might have biased our results. Tumor progression was not [24]. These subgroups might display different vascular- confirmed by histopathology but only by imaging according ization patterns, as seen in the present study. to the clinical standard. We defined tumor progress by the Additionally, low-perfused tumors might have a worse RANO criteria in order to avoid the inclusion of cases with response to adjuvant therapy due to diminished delivery pseudoprogression. of chemotherapeutic agents and higher resistance to ra- In conclusion, high tumor perfusion upon initial diagnosis diotherapy in hypoxic and low-perfused areas [30]. is not compellingly associated with worse short-term progno- Overall, our results suggest that higher vascularization sis within different molecular glioma subgroups. In the overall does not always result in a worse prognosis in diffuse highly perfused IDH wild-type astrocytic tumors, there are gliomas. tumors with low perfusion but unfavorable prognosis. Survival probability Survival probability Survival probability Neurol Sci (2020) 41:3625–3632 3631 Authors’ contributions All authors contributed to the study conception 4. Lu X, Xu W, Wei Y, Li T, Gao L, Fu X, Yao Y, Wang L (2019) and design. Material preparation, data collection, and analysis were per- Diagnostic performance of DWI for differentiating primary central formed by Cornelia Brendle, Uwe Klose, and Benjamin Bender. The first nervous system lymphoma from glioblastoma: a systematic review draft of the manuscript was written by Cornelia Brendle and all authors and meta-analysis. Neurol Sci 40(5):947–956. https://doi.org/10. commented on previous versions of the manuscript. All authors read and 1007/s10072-019-03732-7 approved the final manuscript. 5. Razek A, El-Serougy L, Abdelsalam M, Gaballa G, Talaat M (2020) Multi-parametric arterial spin labelling and diffusion- weighted magnetic resonance imaging in differentiation of grade Funding information Open Access funding enabled and organized by II and grade III gliomas. Pol J Radiol 84:e110–e117 Projekt DEAL. Cornelia Brendle has been supported by a grant from 6. Abdel Razek AA, Gaballa G (2011) Role of perfusion magnetic the TÜFF program of the Faculty of Medicine, Eberhard Karls resonance imaging in cervical lymphadenopathy. J Comput Assist University Tuebingen (application number 2395-0-0). Tomogr 35(1):21 –25. https://doi.org/10.1097/RCT. 0b013e3181ff9143 Compliance with ethical standards 7. Mangla R, Ginat DT, Kamalian S, Milano MT, Korones DN, Walter KA, Ekholm S (2014) Correlation between progression free survival and dynamic susceptibility contrast MRI perfusion in Conflict of interest The Department of Diagnostic and Interventional Radiology has a collaboration contract with Siemens concerning the tech- WHO grade III glioma subtypes. J Neuro-Oncol 116(2):325–331. https://doi.org/10.1007/s11060-013-1298-9 nical development of PET/MRI Biograph mMR. Otherwise, the authors declare no conflict of interest. 8. Rau MK, Braun C, Skardelly M, Schittenhelm J, Paulsen F, Bender B, Ernemann U, Bisdas S (2014) Prognostic value of blood flow estimated by arterial spin labeling and dynamic susceptibility Ethical approval All procedures performed in studies involving human contrast-enhanced MR imaging in high-grade gliomas. J Neuro- participants were in accordance with the ethical standards of the institu- Oncol 120(3):557–566. https://doi.org/10.1007/s11060-014-1586- tional research committee of the Medical Faculty of the University of Tuebingen and with the 1964 Helsinki declaration and its later amend- 9. Romano A, Pasquini L, Di Napoli A, Tavanti F, Boellis A, Rossi ments or comparable ethical standards. For this type of study formal Espagnet MC, Minniti G, Bozzao A (2018) Prediction of survival in consent is not required. patients affected by glioblastoma: histogram analysis of perfusion MRI. J Neuro-Oncol 139(2):455–460. https://doi.org/10.1007/ Informed consent Informed consent was obtained from all individual s11060-018-2887-4 participants included in the study. 10. Coban G, Mohan S, Kural F, Wang S, O’Rourke DM, Poptani H (2015) Prognostic value of dynamic susceptibility contrast- Open Access This article is licensed under a Creative Commons enhanced and diffusion-weighted mr imaging in patients with glio- Attribution 4.0 International License, which permits use, sharing, adap- blastomas. AJNR Am J Neuroradiol 36(7):1247–1252. https://doi. tation, distribution and reproduction in any medium or format, as long as org/10.3174/ajnr.A4284 you give appropriate credit to the original author(s) and the source, pro- 11. Bonekamp D, Deike K, Wiestler B, Wick W, Bendszus M, vide a link to the Creative Commons licence, and indicate if changes were Radbruch A, Heiland S (2015) Association of overall survival in made. The images or other third party material in this article are included patients with newly diagnosed glioblastoma with contrast- in the article's Creative Commons licence, unless indicated otherwise in a enhanced perfusion MRI: comparison of intraindividually matched credit line to the material. If material is not included in the article's T1 - and T2 (*) -based bolus techniques. J Magn Reson Imaging Creative Commons licence and your intended use is not permitted by 42(1):87–96. https://doi.org/10.1002/jmri.24756 statutory regulation or exceeds the permitted use, you will need to obtain 12. Brendle C, Hempel JM, Schittenhelm J, Skardelly M, Tabatabai G, permission directly from the copyright holder. To view a copy of this Bender B, Ernemann U, Klose U (2017) Glioma grading and de- licence, visit http://creativecommons.org/licenses/by/4.0/. termination of IDH mutation status and ATRX loss by DCE and asl perfusion. Clin Neuroradiol. 28:421–428. https://doi.org/10.1007/ s00062-017-0590-z 13. Kickingereder P, Sahm F, Radbruch A, Wick W, Heiland S, Deimling A, Bendszus M, Wiestler B (2015) IDH mutation status References is associated with a distinct hypoxia/angiogenesis transcriptome signature which is non-invasively predictable with rCBV imaging 1. Louis DN, Perry A, Reifenberger G, von Deimling A, Figarella- in human glioma. Sci Rep 5:16238. https://doi.org/10.1038/ Branger D, Cavenee WK, Ohgaki H, Wiestler OD, Kleihues P, srep16238 Ellison DW (2016) The 2016 World Health Organization classifi- 14. Zhang HW, Lyu GW, He WJ, Lei Y, Lin F, Wang MZ, Zhang H, cation of tumors of the central nervous system: a summary. Acta Liang LH, Feng YN, Yang JH (2020) DSC and DCE Histogram neuropathologica 131(6):803–820. https://doi.org/10.1007/s00401- analyses of glioma biomarkers, including IDH, MGMT, and TERT, 016-1545-1 on differentiation and survival. Acad Radiol. https://doi.org/10. 2. Cuccarini V, Antelmi L, Pollo B, Paterra R, Calatozzolo C, Nigri A, 1016/j.acra.2019.12.010 DiMeco F, Eoli M, Finocchiaro G, Brenna G, Tramacere I, 15. Schittenhelm J, Mittelbronn M, Meyermann R, Melms A, Tatagiba Bruzzone MG, Anghileri E (2020) In vivo 2-hydroxyglutarate- M, Capper D (2011) Confirmation of R132H mutation of isocitrate proton magnetic resonance spectroscopy (3T, PRESS technique) dehydrogenase 1 as an independent prognostic factor in anaplastic in treatment-naive suspect lower-grade gliomas: feasibility and ac- astrocytoma. Acta neuropathologica 122(5):651–652. https://doi. curacy in a clinical setting. Neurol Sci 41(2):347–355. https://doi. org/10.1007/s00401-011-0885-0 org/10.1007/s10072-019-04087-9 16. Hartmann C, Meyer J, Balss J, Capper D, Mueller W, Christians A, 3. El-Serougy L, Abdel Razek AA, Ezzat A, Eldawoody H, El-Morsy Felsberg J, Wolter M, Mawrin C, Wick W, Weller M, Herold- A (2016) Assessment of diffusion tensor imaging metrics in differ- Mende C, Unterberg A, Jeuken JW, Wesseling P, Reifenberger entiating low-grade from high-grade gliomas. Neuroradiol J 29(5): G, von Deimling A (2009) Type and frequency of IDH1 and 400–407. https://doi.org/10.1177/1971400916665382 IDH2 mutations are related to astrocytic and oligodendroglial 3632 Neurol Sci (2020) 41:3625–3632 differentiation and age: a study of 1,010 diffuse gliomas. Acta defined by the 2016 WHO classification for brain tumors: histo- neuropathol 118(4):469–474. https://doi.org/10.1007/s00401-009- gram analysis approach. Neuroradiology 61(5):545–555. https:// 0561-9 doi.org/10.1007/s00234-019-02173-5 17. Reuss DE, Sahm F, Schrimpf D, Wiestler B, Capper D, Koelsche C, 26. Jain R, Poisson LM, Gutman D, Scarpace L, Hwang SN, Holder Schweizer L, Korshunov A, Jones DT, Hovestadt V, Mittelbronn CA, Wintermark M, Rao A, Colen RR, Kirby J, Freymann J, Jaffe M, Schittenhelm J, Herold-Mende C, Unterberg A, Platten M, CC, Mikkelsen T, Flanders A (2014) Outcome prediction in pa- Weller M, Wick W, Pfister SM, von Deimling A (2015) ATRX tients with glioblastoma by using imaging, clinical, and genomic and IDH1-R132H immunohistochemistry with subsequent copy biomarkers: focus on the nonenhancing component of the tumor. number analysis and IDH sequencing as a basis for an “integrated” Radiology 272(2):484–493. https://doi.org/10.1148/radiol. diagnostic approach for adult astrocytoma, oligodendroglioma and glioblastoma. Acta Neuropathol 129(1):133–146. https://doi.org/ 27. Hempel JM, Schittenhelm J, Klose U, Bender B, Bier G, Skardelly 10.1007/s00401-014-1370-3 M, Tabatabai G, Castaneda Vega S, Ernemann U, Brendle C (2019) 18. Thon N, Eigenbrod S, Grasbon-Frodl EM, Ruiter M, Mehrkens JH, In vivo molecular profiling of human glioma : cross-sectional ob- Kreth S, Tonn JC, Kretzschmar HA, Kreth FW (2009) Novel mo- servational study using dynamic susceptibility contrast magnetic lecular stereotactic biopsy procedures reveal intratumoral homoge- resonance perfusion imaging. Clin Neuroradiol 29(3):479–491. neity of loss of heterozygosity of 1p/19q and TP53 mutations in https://doi.org/10.1007/s00062-018-0676-2 World Health Organization grade II gliomas. J Neuropathol Exp 28. Hashido T, Saito S, Ishida T (2020) A radiomics-based comparative Neurol 68(11):1219–1228. https://doi.org/10.1097/NEN. study on arterial spin labeling and dynamic susceptibility contrast 0b013e3181bee1f1 perfusion-weighted imaging in gliomas. Sci Rep 10(1):6121. 19. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, https://doi.org/10.1038/s41598-020-62658-9 Weller M, Kros JM, Hainfellner JA, Mason W, Mariani L, 29. Binabaj MM, Bahrami A, ShahidSales S, Joodi M, Joudi Mashhad Bromberg JE, Hau P, Mirimanoff RO, Cairncross JG, Janzer RC, M, Hassanian SM, Anvari K, Avan A (2017) The prognostic value Stupp R (2005) MGMT gene silencing and benefit from temozolo- of MGMT promoter methylation in glioblastoma: a meta-analysis mide in glioblastoma. N Engl J Med 352(10):997–1003. https://doi. of clinical trials. J Cell Physiol. 233:378–386. https://doi.org/10. org/10.1056/NEJMoa043331 1002/jcp.25896 20. Brendle C, Hempel JM, Schittenhelm J, Skardelly M, Reischl G, 30. Lemasson B, Chenevert TL, Lawrence TS, Tsien C, Sundgren PC, Bender B, Ernemann U, la Fougere C, Klose U (2018) Glioma Meyer CR, Junck L, Boes J, Galban S, Johnson TD, Rehemtulla A, grading by dynamic susceptibility contrast perfusion and (11)C- Ross BD, Galban CJ (2013) Impact of perfusion map analysis on methionine positron emission tomography using different regions early survival prediction accuracy in glioma patients. Transl Oncol of interest. Neuroradiology 60(4):381–389. https://doi.org/10.1007/ 6(6):766–774 s00234-018-1993-5 31. Xing Z, Zhang H, She D, Lin Y, Zhou X, Zeng Z, Cao D (2019) 21. Bonm AV, Ritterbusch R, Throckmorton P, Graber JJ (2020) IDH genotypes differentiation in glioblastomas using DWI and Clinical imaging for diagnostic challenges in the management of DSC-PWI in the enhancing and peri-enhancing region. Acta gliomas: a review. J Neuroimaging 30(2):139–145. https://doi.org/ Radiol 60(12):1663 –1672. https://doi.org/10.1177/ 10.1111/jon.12687 22. Dong C, Yuan Z, Li Q, Wang Y (2018) The clinicopathological and 32. van Dijken BRJ, van Laar PJ, Smits M, Dankbaar JW, Enting RH, prognostic significance of TP53 alteration in K27M mutated glio- van der Hoorn A (2019) Perfusion MRI in treatment evaluation of mas: an individual-participant data meta-analysis. Neurol Sci 39(7): glioblastomas: clinical relevance of current and future techniques. J 1191–1201. https://doi.org/10.1007/s10072-018-3407-1 Magn Reson Imaging 49(1):11–22. https://doi.org/10.1002/jmri. 23. Xiao Q, Yang S, Ding G, Luo M (2018) Anti-vascular endothelial growth factor in glioblastoma: a systematic review and meta-anal- 33. Abdel Razek AAK, Talaat M, El-Serougy L, Gaballa G, ysis. Neurol Sci 39(12):2021–2031. https://doi.org/10.1007/ Abdelsalam M (2019) Clinical applications of arterial spin labeling s10072-018-3568-y in brain tumors. J Comp Assist Tomogr 43(4):525–532. https://doi. 24. Cimino PJ, Zager M, McFerrin L, Wirsching HG, Bolouri H, org/10.1097/RCT.0000000000000873 Hentschel B, von Deimling A, Jones D, Reifenberger G, Weller 34. Razek A, El-Serougy L, Abdelsalam M, Gaballa G, Talaat M M, Holland EC (2017) Multidimensional scaling of diffuse glio- (2018) Differentiation of residual/recurrent gliomas from mas: application to the 2016 World Health Organization classifica- postradiation necrosis with arterial spin labeling and diffusion ten- tion system with prognostically relevant molecular subtype discov- sor magnetic resonance imaging-derived metrics. Neuroradiology ery. Acta Neuropathol Commun 5(1):39. https://doi.org/10.1186/ 60(2):169–177. https://doi.org/10.1007/s00234-017-1955-3 s40478-017-0443-7 25. Latysheva A, Emblem KE, Brandal P, Vik-Mo EO, Pahnke J, Roysland K, Hald JK, Server A (2019) Dynamic susceptibility Publisher’snote Springer Nature remains neutral with regard to jurisdic- contrast and diffusion MR imaging identify oligodendroglioma as tional claims in published maps and institutional affiliations.

Journal

Neurological Sciences – Unpaywall

Published: May 28, 2020

There are no references for this article.