Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Role of Dynamic Susceptibility Contrast Perfusion MRI in Glioma Progression Evaluation

Role of Dynamic Susceptibility Contrast Perfusion MRI in Glioma Progression Evaluation Hindawi Journal of Oncology Volume 2021, Article ID 1696387, 9 pages https://doi.org/10.1155/2021/1696387 Review Article Role of Dynamic Susceptibility Contrast Perfusion MRI in Glioma Progression Evaluation 1 1 2 3 4 5 Guanmin Quan, Kexin Zhang, Yawu Liu, Jia-Liang Ren, Deyou Huang, Weiwei Wang, and Tao Yuan Department of Medical Imaging, e Second Hospital of Hebei Medical University, Shijiazhuang, China Department of Clinical Radiology, Kuopio University Hospital, Kuopio, Finland GE Healthcare China, Beijing, China Department of Radiology, Affiliated Hospital of Youjiang Medical University for Nationalities, Baise, China Department of Radiology, Handan Central Hospital, Handan, China Correspondence should be addressed to Tao Yuan; 420490790@qq.com Received 23 May 2020; Revised 23 January 2021; Accepted 28 January 2021; Published 9 February 2021 Academic Editor: (omas R. Chauncey Copyright © 2021 Guanmin Quan et al. (is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Accurately and quickly differentiating true progression from pseudoprogression in glioma patients is still a challenge. (is study aimstoexploreifdynamicsusceptibilitycontrast-(DSC-)MRIcanimprovetheevaluationofgliomaprogression.Weenrolled65 gliomapatientswithsuspectedgadolinium-enhancinglesion.LongitudinalMRIfollow-up(mean590days,range:210–2670days) or re-operation (n �3) was used to confirm true progression (n �51) and pseudoprogression (n �14). We assessed the diagnostic performance of each MRI variable and the different combinations. Our results showed that the relative cerebral blood volume (rCBV) in the true progression group (1.094, 95%CI: 1.135–1.636) was significantly higher than that of the pseudoprogression group (0.541±0.154) (p<0.001). Among the 18 patients who had serial DSC-MRI, the rCBV of the progression group (0.480, 95%CI: 0.173–0.810) differed significantly from pseudoprogression (-0.083, 95%CI: −1.138–0.620) group (p � 0.015). With an rCBV threshold of 0.743, the sensitivity and specificity for discriminating true progression from pseudoprogression were 76.5% and92.9%,respectively.(eCho/CrandCho/NAAratiosofthetrueprogressiongroup(2.520,95%CI:2.331–2.773;2.414 ±0.665, respectively) were higher than those of the pseudoprogression group (1.719±0.664; 1.499±0.500, respectively) ((p � 0.001), (p<0.001), respectively). (e areas under ROC curve (AUCs) of enhancement pattern, MRS, and DSC-MRI for the differ- entiation were 0.782, 0.881, and 0.912, respectively. Interestingly, when combined enhancement pattern, MRS, and DSC-MRI variables, the AUC was 0.965 and achieved sensitivity 90.2% and specificity 100.0%. Our results suggest that DSC-MRI can significantlyimprovethediagnosticperformanceforidentifyinggliomaprogression.DSC-MRIcombinedwithconventionalMRI may promptly distinguish true gliomas progression from pseudoprogression when the suspected gadolinium-enhancing lesion was found, without the need for a long-term follow-up. preexisting enhancement on contrast-enhanced T1-weighted 1. Introduction (CE-T1WI) MR series, especially in patients with high-grade Gliomas are the most common primary brain tumors in gliomas. Pseudoprogression lesions can appear both in the adults [1]. At present, the standard treatment for gliomas firstthreemonthsandlongaftercompletionofradiationwith involves maximal safe tumor resection, followed by radio- anincidenceof9%–36%[2,3].(epseudoprogressionlesions therapy and concomitant and adjuvant chemotherapy with are caused by normal tissue response to treatment, whereas temozolomide. Recurrence and tissue response to the treat- true progression is the result of tumor recurrence and needs mentorso-calledpseudoprogressionarebothassociatedwith salvage therapy [4]. Hence, it is essential to discriminate true a high incidence of new enhancement or enlargement of progression from pseudoprogression as early as possible. 2 Journal of Oncology As a golden standard, histopathology is limited in chemoradiotherapy [6]; (6) had follow-up MRI at least 12 clinical practice for the disadvantage of invasive sampling, months after chemoradiotherapy; (7) had multi-parametric sampling error, and high cost [5]. MRI is the most common MR imaging when suspected gadolinium-enhancing lesion modality of follow-up tool in patients with posttreatment was found in the first follow-up MRI. glioma. However, the standard imaging protocol in the 2.2. Conventional and Perfusion MRI. All studies were Response Assessment in Neuro-Oncology (RANO) criteria, including fluid-attenuated inversion recovery (FLAIR) and performed on a 3T MRI scanner (Achieva; Philips Medical System, Best, (e Netherlands), using an 8-channel phased- CE-T1WI, does not specifically require to discriminate the true progression from the pseudoprogression lesions [4, 6]. array coil for acquisition. (e conventional MR protocol included precontrast and postcontrast T1WI, T2WI, and What is more, long-term follow-up should be made to FLAIR. (e multi-parametric MR imaging protocol in- discriminate true progression from pseudoprogression cludedT WI,T WI,FLAIR,DWI,MRS,andDSC-MRI.(e when new gadolinium-enhancing lesion or increase of the 1 2 parameters of DSC-MRI: perfusion-weighted gradient-echo preexisting enhancing lesion is detected. Consequently, echo-planar sequence; repetition time, 2000ms; echo time, longitudinal follow-up MRI and various advanced func- 40ms; slice thickness, 6.5mm; the data was acquired every tional MR techniques have been applied to obtain a de- finitive diagnosis. (e long waiting time for a definitive one second for a total of 1min 30sec, with Gd-DTPA (Gadovist,BayerScheringPharma,Berlin,Germany,0.2ml/ diagnosis by means of follow-up MRI is a psychological torture to patients and may delay salvage therapy. On the kg of body weight and maximum dose of 20ml) injected with a MRI-compatible power injector at a rate of 3ml/sec, other hand, in spite of promising results, the application of advanced functional MR techniques is limited for their followed by a 20-ml 0.9% saline flush using same flow rate. vagueresultsaswellasnotbeingaroutineexamintheclinic 2.3. Imaging Analysis. (e imaging analysis and post- [6–8]. processing were performed on a workstation (PHILIPS Previous studies have shown that MR perfusion- Extended MR WorkSpace 2.6.3.4). (e enhancement pat- weightedimaging,especiallydynamicsusceptibilitycontrast terns of residual cavity wall were divided into thin-linear perfusion MRI (DSC-MRI), can reflect tumor angiogenesis (partial or entire wall enhancement with thickness <3mm), which is valuable for glioma grading, estimation of prog- thick-linear (partial or entire wall enhancement of 3–5mm nosis, and differentiating tumor recurrence from treatment- in thickness), and nodular wall enhancement (with nodular related changes [5, 9, 10]. But there are only a few reports enhancement of 5–10mm in thickness) patterns [14]. Ce- that assessed the impact of combined DSC-MRI and other rebral blood volume (CBV) was calculated based on signal MRI features on identifying the true tumor progression intensity-time curves in transverse T -weighted sequence. [5, 9, 11]. Since functional MR imaging techniques can We placed 3 circular ROIs in the region with the largest reflect different pathological changes, such as diffusion- enhancement (Figure 1), and the mean CBV was calculated weightedimaging(DWI)forwatermoleculemovement,and and used in quantitative analysis. We calculated the relative magnetic resonance spectroscopy (MRS) for neoplastic CBV (rCBV) of suspected gadolinium-enhancing lesion as metabolites, we hypothesized that combination of DSC- well as ipsilateral normal tissues by dividing the CBV values MRI, DWI, and MRS can improve the accuracy for iden- oftheseregionswiththeCBVvaluesofcontralateralnormal tification of true progression lesionsamongthe gadolinium- white matter [15]. (e rCBV difference between two se- enhancing lesions. (erefore, the objective of this study was quential perfusion imaging was calculated for those patients to explore whether DSC-MRI can improve diagnostic per- with longitudinal DSC-MRI examinations [10]. (e ap- formance of multiparametric MRI in the evaluation of parentdiffusioncoefficient(ADC)wasmeasuredinthesame progression in gliomas patients after standard treatment. regions as CBV measurements. (e conventional and functional MRI characteristics were analyzed independently 2. Materials and Methods by two neuroradiologists (with 10 and 16 years of neuro- radiology experience, respectively). When a disagreement 2.1. Patients. (is study was approved by the Institutional existed, a consent was reached after consulting another Review Board. (e informed consent was waived for its neuroradiologist (with 25-year experience in neuroradiol- retrospective nature. A total of consecutive 65 glioma pa- ogy). (e outcome of the tumor was assessed according to tients with suspicious gadolinium-enhancing lesion were the updated RANO criteria for gliomas [6, 16]. True pro- recruited(36males,29females;meanage,46.5±14.3years), gressionwasdefinedwhenpatientswithnewlyenhancement according to the following eligibility criteria: (1) gliomas lesions or with increase the size of enhancement lesions were confirmed by histology; (2) gross-total resection of continuously in the follow-up period after accomplishment tumors [12]; (3) completed the standard treatment of chemoradiotherapy. If there were stable or regressing according to National Comprehensive Cancer Network enhancement lesions, the patients were defined as pseu- (NCCN)guideline,includinggross-totalresection,radiation doprogression [6]. therapy and chemotherapy (CCRT) after surgery, and six cycles of adjuvant temozolomide [13]; (4) had standard clinical MRI before and after chemoradiotherapy; (5) 2.4. Statistical Analysis. Statistical analysis utilized the gadolinium-enhancing lesion enlarged or presence of new software SPSS for windows release 25.0 (SPSS Inc., Chicago, gadolinium-enhancing lesion within the first month after IL, USA). Categorical variables were analyzed with log-rank Journal of Oncology 3 Figure 1: Example of region of interest (ROI) for measurement of CBV (the circle in red). (is 37-year-old female patient with path- ologically proven glioblastoma was ascribed to true progression group. test. (e quantitative data between true progression and dose (p � 0.615), histopathologic grade (p � 0.451), and pseudoprogression groups, including rCBV, Cho/Cr, Cho/ Karnofsky performance status (KPS) scores (p � 0.154). NAA, and ADC, were compared by using two-tailed Stu- AmongthepatientswithDWIdata(54/65),therewasno dent’s t-test when they were in non-normal distribution, or significant difference in ADC between two groups using Mann–WhitneyU testwhen they werein non-normal (p � 0.067)inspiteoflowerADCvaluefortrueprogression distribution. (e interobserver consistency between the two group (Table 1). (e Cho/Cr and Cho/NAA of true pro- neuroradiologists was evaluated with intraclass correlation gression group (2.520, 95%CI: 2.331–2.773; 2.414±0.665, coefficient(ICC).(esurvivaltimeswereestimatedwiththe respectively) were higher (p≤0.002) than those of pseu- Kaplan–Meier methods. Receiver operating characteristic doprogression group (1.719±0.664; 1.499±0.500, respec- (ROC) curve analysis was employed in determining the best tively) (Table 2), whereas there was no significant difference cutoffvaluesofrCBVandmetaboliteratiosindifferentiating in NAA/Cr ratio between true progression (1.030±1.100) true progression and pseudoprogression by maximizing the and pseudoprogression groups (1.300±0.750) (p � 0.485). sum of sensitivity and specificity. (e diagnostic perfor- Usingacutoff Cho/Crvalueof2.475, asensitivity51.0%and mance of all variables was measured as area under ROC specificity 92.9% were achieved; using a cutoff Cho/NAA curve (AUC). (e level of significance was set at p<0.05. value of 2.155, a sensitivity 64.7% and specificity 100.0% were achieved for separating the two groups. (e rCBV of normal brain tissue was 0.993 ±0.106 (0.854–1.209). (ere was significant difference between the 3. Results rCBV of normal brain tissue and that of true progression (e median follow-up span was 590 days (range: lesions (p<0.001), so did between the normal brain tissue 210–2670days). (e median progression-free survival (PFS) and pseudoprogression lesions (p<0.001). (e rCBV of was 360 days [95% confidence interval (CI): 399–580 days] true progression group (1.094, 95%CI: 1.135–1.636) was and median overall survival (OS) was 590 days (95% CI: significantly higher than that of pseudoprogression group 603–790 days) (Figure 2). 22 patients (33.846%) were dead (0.541±0.154) (p<0.001) (Table 2). Among 65 patients, 18 during the follow-up period. (e demographic and MR patients (34.0%, 14 patients with true progression and 4 imaging characteristics of the 65 patients are summarized in patients with pseudoprogression lesions) had serial DSC- Table 1. Fifty-one patients (28 males, 47.290±14.380 years) MRI. Changes in rCBV at subsequent follow-up differed werediagnosedastrueprogression(histologicallyconfirmed significantly (p � 0.015) between true progression (0.480, in 3 cases) and fourteen patients (8 males, 43.640±14.370 95%CI: 0.173–0.810) and pseudoprogression (−0.083, 95% years) were diagnosed as pseudoprogression. (ere was no CI: −1.138–0.620) groups. (e between-group comparison significant difference in patient age (p � 0.403), radiation revealed a significant difference between the 4 Journal of Oncology 100 100 80 80 60 60 40 40 20 20 0 0 0 500 1000 1500 2000 2500 0 1000 2000 3000 PFS (d) OS (d) Pseudoprogression Pseudoprogression True progression True progression (a) (b) Figure 2: Kaplan–Meier survival curve for PFS and OS according to true progression and pseudoprogression. PFS, progression-free survival; OS, overall survival. Table 1: Clinical characteristics of 65 glioma patients. Characteristics Pseudoprogression (n �14) True progression (n �51) t/X p value Mean age±SD(years) 43.640±14.372 47.290±14.375 −0.842 0.403 Gender 0.022 0.881 Male 8 28 Female 6 23 1p19q codeletion(+) (n �23) 1(33.33%) 4(21.05%) — 0.814 Promoter of MGMT methylation (n �12) 1(33.33%) 5(55.56%) — 0.574 IDH mutation (n �13) 1(50.00%) 3(27.27%) — 0.178 WHO grading — 0.451 II 1 3 III 9 24 IV 4 24 KPS score −1.427 0.154 Median 90.000 90.000 95%CI 90.000–90.000 86.890–89.580 T2-FLAIR mismatch — 0.676 Yes 1 9 No 13 42 Enhancement of residual cavity wall — 0.001 (in-linear 11 12 (ick-linear 1 10 Nodular 2 29 Total dose (GyRBE) −0.503 0.615 Median 60 59.92 95%CI 55.761–60.502 57.669–59.462 SVZ involvement — 0.204 Yes 7 36 No 7 15 ADC mean (mm /s) (n �54) 903.142±491.652 523.000 (484.950–668.610) −1.842 0.067 MGMT, O -methyl-guanine methyl transferase; KPS, Karnofsky Performance Score; SVZ, subventricular zone; ADC mean, apparent diffusion coefficient mean; T2-FLAIR mismatch, the presence of a complete/near-complete hyperintense signal on T2-weighted (T2W) MRI sequences, in combination with a relative hypointense signal on fluid attenuation inversion recovery (FLAIR) MR sequences except for a hyperintense peripheral rim; Bold p value indicates statistically significant association, - indicates Fisher’s exact test. Percent survival Percent survival Journal of Oncology 5 100.0 Table 2: Comparison of DSC and MRS variables. Pseudoprogression True progression p value rCBV <0.001 80.0 Median/mean 0.541±0.154 1.094 95%CI (range) 0.360–0.892 1.135–1.636 Cho/Cr ratio 0.001 60.0 Median/mean 1.719±0.664 2.520 95%CI (range) 0.680–3.160 2.331–2.773 Cho/NAA ratio 1.499±0.495 2.414±0.665 <0.001 40.0 rCBV, relative cerebral blood volume; Cho, choline; NAA, N-acetylas- partate; Cr, creatin; CI, confidence interval, Bold p value indicates statis- tically significant association. 20.0 pseudoprogression and the true progression groups corre- 0.0 sponding to a large effect size (Cohen’s d �1.392). (e ROC 0.0 20.0 40.0 60.0 80.0 100.0 analysis showed that, using a cutoff rCBV value of 0.045, it 1 - specificity achieved AUC 0.904, sensitivity 100.0%, and specificity CE CE + DSC 75.0% to separate the two groups. MRS DSC + MRS Figure 3 shows the ROC analyses of the diagnostic DSC CE + DSC + MRS performance using different variables and their combina- CE + MRS Reference line tions for distinguishing true progression and pseudoprog- Figure 3: Comparison of ROC curve analyses for different MR ression. (e AUCs, sensitivity, and specificity of metabolites variables and their combinations. (e combination of ratios, and rCBV were significantly larger than those of CE+DSC+MRS presents the high diagnostic performance, and enhancement pattern of residual cavity wall (Figure 3, Ta- the AUC was 0.965 (95%CI 0.920–1.000), with significant level of ble 3). With the cutoff rCBV value of 0.743, the AUC, p<0.05 to any other variables and other combinations. CE, sensitivity, and specificity of DSC-MRI for distinguishing contrastenhancementpatternofresidualcavitywall;MRS,Cho/Cr true progression and pseudoprogression were 0.912, 76.5%, and Cho/NAA ratios; DSC, dynamic susceptibility contrast and92.9%,respectively.(eAUC,sensitivity,andspecificity perfusion. of the combination of Cho/Cr and Cho/NAA were 0.881, 88.2%, and 78.6%, respectively. (e AUC and specificity of Accurately and quickly distinguishing true progression MRS+DSC were similar to those of CE+MRS+DSC from pseudoprogression in glioma patients after standard combination model. However, the sensitivity and Youden chemoradiotherapy remains a major clinical challenge [6]. index of MRS+DSC were relatively lower (Figure 3, Ta- Newly enhancement lesions or increase of the preexisting ble 3). Interestingly, when we combined all MR variables, enhancement lesions, accompanied with mass-effect, as well including enhancement patterns of residual cavity wall, asvasogeniccerebraledema,canbedetectedbothinpatients rCBV, and metabolites ratios, the diagnostic performance with true progression and in patients with pseudoprog- was significantly improved (AUC 0.965, sensitivity 90.2%, ression lesions [17]. Generally, longitudinal follow-up MRI and specificity 100.0%). over several months or stereotactic biopsy is needed for (e agreement was excellent between the two neuro- definitive diagnosis. In perfusion-weighted imaging, CBV radiologists for evaluation of the functional MR variables, could be used as an important biomarker for neoplastic including rCBV (ICC 0.979, 95% CI: 0.966–0.987), ADC vasculature [18, 19]. Our results support the hypothesis that (ICC 0.995, 95% CI: 0.991–0.997), Cho/Cr (ICC 0.964, 95% DSC-MRI can add information for definitive diagnosis of CI: 0.941–9.978), and Cho/NAA (ICC 0.958, 95% CI: glioma patients with gadolinium-enhancing lesions after 0.932–0.974). treatment. In this study, rCBV is a valuable imaging marker Figures 4 and 5 show the classic examples of true pro- for distinguishing true progression from pseudoprogression gression and pseudoprogression lesions. in glioma patients. Increased rCBV as surrogate biomarker of active growing tumor could potentially reduce the ne- cessity for biopsy and their associated risks, thereby sim- 4. Discussion plifyingtheposttreatmentevaluationprocessanddecreasing We applied DSC-MRI, along with conventional MRI se- the care cost. On the other hand, DSC-MRI is the most quences, DWI, and MRS, to assess the impact of perfusion common and available perfusion technique in the current commercialMRequipment.(us,wesuggestthatDSC-MRI parameter on the differential diagnosis of true progression versus pseudoprogression in patients with gliomas. Our should be conventionally employed in the evaluation of posttreatment gliomas. results suggested that a combination of DSC-MRI, contrast enhanced T W imaging, and MRS can greatly improve the (e DSC-MRI has been widely used in evaluation of the diagnostic performance in distinguishing true progression suspicious lesions in posttreatment glioma patients from pseudoprogression. [5, 18–20]. (e rCBV cutoff is important for identifying true Sensitivity 6 Journal of Oncology Table 3: ROC curve analyses of diagnostic performance of various variables and their combinations. Characteristics AUC 95%CI Sensitivity Specificity Youden index CE 0.782 0.643–0.920 0.765 0.786 0.551 MRS 0.881 0.788–0.973 0.882 0.786 0.668 DSC 0.912 0.837–0.988 0.765 0.929 0.694 CE+MRS 0.930 0.867–0.993 0.882 0.857 0.739 CE+DSC 0.936 0.873–0.998 0.863 0.857 0.720 DSC+MRS 0.965 0.924–1.000 0.863 1.000 0.863 CE+MRS+DSC 0.965 0.923–1.000 0.902 1.000 0.902 CE, contrast enhancement pattern of residual cavity wall; MRS, Cho/Cr and Cho/NAA ratios; DSC, dynamic susceptibility contrast perfusion. RL (a) (b) (c) (d) Figure 4: Example of pseudoprogression. A 21-year-old man with pathologically proven glioblastoma. A new enhancing lesion (arrow) developed in his left frontal lobe three months after completion of radiotherapy (a). No hyper-perfusion was detected on his DSC-MRI (b), with rCBV value of 0.71. MRS display Cho/Cr ratio of 1.20 and Cho/NAA 0.46 (c). (e enhancing lesions in initial follow-up MRI disappeared in the MRI 23 months after completion of radiotherapy (d). progression. However, there is no consensus on the cutoff imaging techniques [10, 21]. We enrolled gliomas patients value of rCBV (0.71–5.01) for distinguishing true progression withdifferentgrades(WHOgradeII-IV)andwithgross-total frompseudoprogressionlesions[8,18,19,21].Inastudyof44 resection of tumors, and followed by standard therapy glioblastomapatients,Blaseletal.foundthecutoffvalueof2.2 according to NCCN guideline. However, in Blasel et al.’s for rCBV yielded a sensitivity of 65% and a specificity of 71% study [20], only glioblastoma patients (WHO grade IV) were [20], whereas, in the present study, with the cutoff value of recruited,including4patientswhoonlyhadbiopsyinsteadof 0.743 for rCBV, we achieved higher sensitivity (76.5%) and tumorresection.(eirtherapyregimenwasdiverse,including specificity (92.9%). (is discrepancy may be attributed to the Stupp’s method, radiotherapy only, and second- and third- diversity of inclusion criteria and treatment regimens, as well line therapy in some patients. (us, the higher rCBV cutoff as the lack of standardization in reference standard and value in Blasel et al.’s study indicted higher mitotic activity Journal of Oncology 7 RL (a) (b) (c) (d) Figure 5: Example of true progression. An 18-year-old woman with pathologically proven glioblastoma. A new irregular ring enhancing lesion (arrow) developed in her right temporal-occipital lobe three months after completion of radiotherapy (a). Hyper-perfusion (arrow) was detected on her DSC-MRI (b), with rCBV value of 1.11. (e MRS display Cho/Cr ratio of 4.15 and Cho/NAA 3.95 (c). (e enhancing lesions enlarged significantly one month later (d). Tumor recurrence was confirmed pathologically after reoperation. and higher vascularity in their patients. In spite of the dif- MRS metabolites ratios provide useful information in distinguishing true progression from pseudoprogression ferenceofcutoffvalueofrCBV,thediagnosticperformanceof our study was similar to other authors’ [8, 19]. (is indicted lesions [8, 9, 23, 24]. Among several metabolic ratios [8, 24], DSC-MRI, in spite of diverse rCBV cutoff value, can be used Cho/Crratiowasconsideredasthemostfavorablemarkerin to improve the evaluation of progression in glioma patients differentiating between true progressions and pseudoprog- after treatment [5]. ression. Cho/Cr ratio achieved excellent sensitivity (91%) Longitudinal change in rCBV may improve the dis- and specificity (95%) in a higher grade gliomas study [8]. crimination between true progression and pseudoprog- However, in our study the sensitivities of MRS were rela- ression lesions [10, 22]. (e diagnostic performance, tively lower (51.0%) with the cutoff values of Cho/Cr ratio including AUC, sensitivity, and specificity, of rCBV dif- 2.475 and Cho/NAA ratio 2.115. A possible explanation for ferencebetweentwosequentialexaminationswasnoticeably this discrepancy might be because we included lower-grade higher than that of initial rCBV in our study. Various (WHO grade II and III) gliomas, which probably had lower metabolite concentration in the enhancing regions. Inter- pathological changes representing a wide range of local cerebral blood volume, including radiation-induced vas- estingly, our results showed that MRS index ratios achieved culopathy and necrosis, could coexist shortly after treat- good diagnostic efficiency (AUC 0.881). Our findings sup- ment. (us, the rCBV value at the initial DSC-MRI may be port previous finding that MRS is a promising functional less effective in discriminating the true progression from MR technique for the evaluation of treatment response in pseudoprogression lesions [10, 22]. However, the destiny of glioma [8]. However, we must keep in mind that using MRS suspicious lesions becomes apparent over time. As the time ratios as a biomarker in the differential diagnosis has its goes on, increase in rCBV could be detected from those intrinsic limitations; i.e., it is difficult to obtain universal enhancing lesions that contain viable tumor cells, especially cutoff values due to partial volume effect resulting from in the most active and aggressive region of the lesions. relativelylargevoxelsize,andtheratiosarevulnerabletothe 8 Journal of Oncology diagnostic performance, only18 of 65 patients in our cohort local field homogeneity, such as water, lipids, and surgical clips. had the sequential DSC-MRI data. In conclusion, DSC-MRI, especially longitudinal change (e roleof DWI in differentiating true progressionfrom pseudoprogressionlesionsremainsambiguous.Moststudies in rCBV, showed satisfactory diagnostic performance and showedtheaccuracyofADCvalueinthedifferentiationwas canbeausefultoolindistinguishingtruegliomaprogression the lowest among all functional MR techniques [5, 8, 23]. from pseudoprogression. Moreover, the combination of (e treatment effects, such as gliosis, coagulation necrosis, DSC-MRIandconventionalmorphologicalimagingfeatures macrophages invasion, and demyelination, can reduce ADC and other advanced MR techniques would further improve value. Necrosis could be found in many recurrent tumor identifying the progression, which may greatly facilitate the individualized management of posttreatment glioma pa- lesions [23]. (us, the diagnostic value of DWI is weakened due to similar changes of diffusivity but quite different tients. (is combination may even eliminate a long-term follow-up when a suspected gadolinium-enhancing lesion is pathologic processes. In literatures, the sensitivity and specificity of ADC value in the differentiation varied widely found. (95% CI: 60–80%, and 95% CI: 77–93%, respectively) [8]. In our study, there was no significant difference in ADC be- Data Availability tweenthetwotypesoflesions.Similarresultswerepresented (e data used to support the findings of this study are in previous studies [23, 25]. However, in another previous available from the corresponding author upon request. study, the ADC value of recurrence lesions was significantly lower than that of non-recurrence lesions [5]. (is dis- crepancymaybeexplainedasbiasfromtherelativelysmaller Disclosure sample size and inclusion of relatively lower-grade gliomas Guanmin Quan and Kexin Zhang are the co-first authors. patients. Moreover, there was no consensus on the cutoff ADC value for this discrimination [5, 26, 27]. Conflicts of Interest Since conventional morphologic imaging findings are often limited in differentiation of true progression from (e authors declare that they have no conflicts of interest. pseudoprogression lesionsingliomas,variousadvancedMR imaging modalities, including DWI, MRS, and DSC-MRI, Authors’ Contributions canprovideadditionalphysiologicormetabolicinformation about posttreatment gliomas. However, the diagnostic ac- TY, KZ, DH, J-LR, and QG made a substantial contribution curacy of each technique is still limited and different to the concept and design, acquisition of data or analysis, [5, 8, 15, 28]. A multiparametric MR imaging has the po- andinterpretationofdata;GQ,YL,WW,andTYdraftedthe tential to improve the overall diagnostic performance. In article and revised it critically for relevant intellectual Matsusue et al.’s study [28], which enrolled 15 subjects, the content;GQ,KZ,WW,andDHperformedMRexamination diagnostic accuracy of combined multiparametric MRI and follow-up of patients; all the authors approved the final (93.3%) was higher than those of ADC ratio (86.7%), rCBV versionofthemanuscript.GuanminQuanandKexinZhang (86.7%), and MRS (Cho/Cr and Cho/NAA) (84.6%). Our contributed equally to this study. results confirmed the superiority of multiparametric MRI, because advanced imaging could provide more compre- Acknowledgments hensiveinformationwithrespecttotumorpathophysiology. In the present study with relatively larger cohort with DSC- (is study was supported by the Specialist Leadership MRI, we further combined conventional morphologic Project of Hebei Province (No. 361004, GQ), Technology imaging findings and advanced MRI data, especially the Tracking Program for Medical Application of Hebei Prov- addition of DSC-MRI, which can be more reliable in ince (No. G201725, GQ), and the Health and Family differentiating true tumor progression from Planning Commission of Guangxi Zhuang Autonomous pseudoprogression. (e combination of CE-T1WI, MRS, Region (Z2016425, DH). and DSC achieved the highest diagnostic performance (AUC, 0.965). Based on these results, we recommend that References patients with suspicious enhancing lesions in conventional follow-up MRI should additionally receive advanced MR [1] Q. T. Ostrom, H. Gittleman, G. Truitt et al., “CBTRUS sta- tisticalreport:primarybrainandothercentralnervoussystem examinations, such as MRS and DSC-MRI. tumors diagnosed in the United States in 2011–2015,” Neuro Several limitations of this study should be mentioned. Oncology, vol. 20, no. 4, pp. iv1–iv86, 2018. First, the number of the patients in this retrospective study [2] S. C. (ust, M. J. van den Bent, and M. Smits, “Pseudo- was relatively small because we only recruited those patients progression of brain tumors,” Journal of Magnetic Resonance with DSC-MRI and also with sufficient MRI follow-up data. Imaging, vol. 48, no. 3, pp. 571–589, 2018. Second, the final diagnosis of most patients (95.4%) in the [3] A. W. Abbasi, H. E. Westerlaan, G. A. Holtman, K. M. Aden, current study was based on the follow-up data. (ird, we P. J. van Laar, and A. van der Hoorn, “Incidence of tumour measured CBV and DWI in a 2D mode, which could not progression and pseudoprogression in high-grade gliomas: a reflect local pathologic changes completely. Fourth, al- systematic review and meta-analysis,” Clinical Neuroradiol- though the longitudinal change in rCBV could improve the ogy, vol. 28, no. 3, pp. 401–411, 2018. Journal of Oncology 9 [4] K. Parvez, A. Parvez, and G. Zadeh, “(e diagnosis and MRI,” American Journal of Roentgenology, vol. 198, no. 1, pp. 19–26, 2012. treatment of pseudoprogression, radiation necrosis and brain tumor recurrence,” International Journal of Molecular Sci- [19] A. Romano, M. C. Rossi Espagnet, L. F. Calabria et al., “Clinical applications of dynamic susceptibility contrast ences, vol. 15, no. 7, pp. 11832–11846, 2014. [5] Y.Yang,Y.Yang,X.Wuetal.,“AddingDSCPWIandDWIto perfusion-weighted MR imaging in brain tumours,” La Radiologia Medica, vol. 117, no. 3, pp. 445–460, 2012. BT-RADS can help identify postoperative recurrence in pa- [20] S.Blasel,A.Zagorcic,A.Jurcoaneetal.,“PerfusionMRIinthe tients with high-grade gliomas,” Journal of Neurooncology, evaluation of suspected glioblastoma recurrence,” Journal of vol. 46, no. 2, pp. 363–371, 2010. Neuroimaging, vol. 26, no. 1, pp. 116–123, 2016. [6] B. Ma, J. O. Blakeley, X. Hong et al., “Applying amide proton [21] J. Dongas, A. T. Asahina, S. Bacchi, and S. Patel, “Magnetic transfer-weighted MRI to distinguish pseudoprogression resonance perfusion imaging in the diagnosis of high-grade from true progression in malignant gliomas,” Journal of glioma progression and treatment-related changes: a sys- Magnetic Resonance Imaging,vol.44,no.2,pp.456–462,2016. tematic review,” Open Journal of Modern Neurosurgery, [7] H. S. Nguyen, N. Milbach, S. L. Hurrell et al., “Progressing vol. 08, no. 03, pp. 282–305, 2018. bevacizumab-induced diffusion restriction is associated with [22] J. L. Boxerman, B. M. Ellingson, S. Jeyapalan et al., “Longi- coagulative necrosis surrounded by viable tumor and de- tudinal DSC-MRI for distinguishing tumor recurrence from creased overall survival in patients with recurrent glioblas- pseudoprogression in patients with a high-grade glioma,” toma,” American Journal of Neuroradiology, vol. 37, no. 12, American Journal of Clinical Oncology, vol. 40, no. 3, pp. 2201–2208, 2016. pp. 228–234, 2017. [8] B. R. J. van Dijken, P. J. van Laar, G. A. Holtman, and [23] B. Bobek-Billewicz, G. Stasik-Pres, H. Majchrzak et al., A. van der Hoorn, “Diagnostic accuracy of magnetic reso- “Differentiation between brain tumor recurrence and radia- nance imaging techniques for treatment response evaluation tion injury using perfusion, diffusion-weighted imaging and in patients with high-grade glioma, a systematic review and MR spectroscopy,” Folia Neuropathology, vol. 48, no. 2, meta-analysis,” European Radiology, vol. 27, no. 10, pp. 81–92, 2010. pp. 4129–4144, 2017. [24] H. Zhang, L. Ma, Q. Wang, X. Zheng, C. Wu, and B.-n. Xu, [9] J. Liu, C. Li, Y. Chen et al., “Diagnostic performance of “Role of magnetic resonance spectroscopy for the differen- multiparametric MRI in the evaluation of treatment response tiation of recurrent glioma from radiation necrosis: a sys- in glioma patients at 3T,” Journal of Magnetic Resonance tematic review and meta-analysis,” European Journal of Imaging, vol. 51, no. 4, pp. 1154–1161, 2020. Radiology, vol. 83, no. 12, pp. 2181–2189, 2014. [10] E. Steidl, M. Muller, ¨ A. Muller, ¨ U. Herrlinger, and [25] C. Asao, Y. Korogi, M. Kitajima et al., “Diffusion-weighted E. Hattingen, “Longitudinal, leakage corrected and uncor- imaging of radiation-induced brain injury for differentiation rected rCBV during the first-line treatment of glioblastoma: a from tumor recurrence,” AJNR American Journal of Neuro- prospectivestudy,” Journal of Neuro-Oncology,vol.144,no.2, radiology, vol. 26, no. 6, pp. 1455–1460, 2005. pp. 409–417, 2019. [26] P. A. Hein, C. J. Eskey, J. F. Dunn et al., “Diffusion-weighted [11] M. Law, “Advanced imaging techniques in brain tumors,” imaginginthefollow-upoftreatedhigh-gradegliomas:tumor Cancer Imaging, vol. 9, no. Special Issue A, pp. S4–S9, 2009. recurrence versus radiation injury,” AJNR American Journal [12] G. Ekinci, I. N. Akpınar, F. Baltacıoglu ˘ et al., “Early-post- of Neuroradiology, vol. 25, no. 2, pp. 201–209, 2004. operative magnetic resonance imaging in glial tumors: pre- [27] P.C.Sundgren,X.Fan,P.Weybrightetal.,“Differentiationof diction oftumor regrowthand recurrence,” European Journal recurrent brain tumor versus radiation injury using diffusion of Radiology, vol. 45, no. 2, pp. 99–107, 2003. tensor imaging in patients with new contrast-enhancing le- [13] L. B. Nabors, J. Portnow, M. Ammirati et al., “NCCN sions,” Magnetic Resonance Imaging, vol. 24, no. 9, guidelines insights: central nervous system cancers, version pp. 1131–1142, 2006. 1.2017,” Journal of the National Comprehensive Cancer Net- [28] E. Matsusue, J. R. Fink, J. K. Rockhill, T. Ogawa, and work, vol. 15, no. 11, pp. 1331–1345, 2017. K. R. Maravilla, “Distinction between glioma progression and [14] B.R.Kim,S.H.Choi,T.J.Yunetal.,“MRimaginganalysisof post-radiationchangebycombinedphysiologicMRimaging,” non-measurable enhancing lesions newly appearing after Neuroradiology, vol. 52, no. 4, pp. 297–306, 2010. concom-itantchemoradiotherapyinglioblastomapatientsfor prognosis prediction,” PLoS One, vol.11, no.11, p. e0166096, [15] M. L. White, Y. Zhang, F. Yu et al., “Post-operative perfusion and diffusion MR imaging and tumor progression in high- grade gliomas,” PLoS One, vol. 4, no. 3, p. e0213905, 2019. [16] P. Y. Wen, D. R. Macdonald, D. A. Reardon et al., “Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group,” Journal of Clinical Oncology, vol. 28, no. 11, pp. 1963–1972, 2010. [17] A.Siu,J.J.Wind,J.B.Iorgulescu,T.A.Chan,Y.Yamada,and J. H. Sherman, “Radiation necrosis following treatment of high grade glioma-a review of the literature and current understanding,” Acta Neurochirurgica, vol. 154, no. 2, pp. 191–201, 2012. [18] G.M.Fatterpekar,D.Galheigo,A.Narayana,G.Johnson,and E. Knopp, “Treatment-related change versus tumor recur- rence in high-grade gliomas: a diagnostic conundrum-use of dynamic susceptibility contrast-enhanced (DSC) perfusion http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Oncology Hindawi Publishing Corporation

Role of Dynamic Susceptibility Contrast Perfusion MRI in Glioma Progression Evaluation

Download PDF
9 pages

Loading next page...
 
/lp/hindawi-publishing-corporation/role-of-dynamic-susceptibility-contrast-perfusion-mri-in-glioma-YO0k5eOENn

References (29)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2021 Guanmin Quan et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1687-8450
eISSN
1687-8469
DOI
10.1155/2021/1696387
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Oncology Volume 2021, Article ID 1696387, 9 pages https://doi.org/10.1155/2021/1696387 Review Article Role of Dynamic Susceptibility Contrast Perfusion MRI in Glioma Progression Evaluation 1 1 2 3 4 5 Guanmin Quan, Kexin Zhang, Yawu Liu, Jia-Liang Ren, Deyou Huang, Weiwei Wang, and Tao Yuan Department of Medical Imaging, e Second Hospital of Hebei Medical University, Shijiazhuang, China Department of Clinical Radiology, Kuopio University Hospital, Kuopio, Finland GE Healthcare China, Beijing, China Department of Radiology, Affiliated Hospital of Youjiang Medical University for Nationalities, Baise, China Department of Radiology, Handan Central Hospital, Handan, China Correspondence should be addressed to Tao Yuan; 420490790@qq.com Received 23 May 2020; Revised 23 January 2021; Accepted 28 January 2021; Published 9 February 2021 Academic Editor: (omas R. Chauncey Copyright © 2021 Guanmin Quan et al. (is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Accurately and quickly differentiating true progression from pseudoprogression in glioma patients is still a challenge. (is study aimstoexploreifdynamicsusceptibilitycontrast-(DSC-)MRIcanimprovetheevaluationofgliomaprogression.Weenrolled65 gliomapatientswithsuspectedgadolinium-enhancinglesion.LongitudinalMRIfollow-up(mean590days,range:210–2670days) or re-operation (n �3) was used to confirm true progression (n �51) and pseudoprogression (n �14). We assessed the diagnostic performance of each MRI variable and the different combinations. Our results showed that the relative cerebral blood volume (rCBV) in the true progression group (1.094, 95%CI: 1.135–1.636) was significantly higher than that of the pseudoprogression group (0.541±0.154) (p<0.001). Among the 18 patients who had serial DSC-MRI, the rCBV of the progression group (0.480, 95%CI: 0.173–0.810) differed significantly from pseudoprogression (-0.083, 95%CI: −1.138–0.620) group (p � 0.015). With an rCBV threshold of 0.743, the sensitivity and specificity for discriminating true progression from pseudoprogression were 76.5% and92.9%,respectively.(eCho/CrandCho/NAAratiosofthetrueprogressiongroup(2.520,95%CI:2.331–2.773;2.414 ±0.665, respectively) were higher than those of the pseudoprogression group (1.719±0.664; 1.499±0.500, respectively) ((p � 0.001), (p<0.001), respectively). (e areas under ROC curve (AUCs) of enhancement pattern, MRS, and DSC-MRI for the differ- entiation were 0.782, 0.881, and 0.912, respectively. Interestingly, when combined enhancement pattern, MRS, and DSC-MRI variables, the AUC was 0.965 and achieved sensitivity 90.2% and specificity 100.0%. Our results suggest that DSC-MRI can significantlyimprovethediagnosticperformanceforidentifyinggliomaprogression.DSC-MRIcombinedwithconventionalMRI may promptly distinguish true gliomas progression from pseudoprogression when the suspected gadolinium-enhancing lesion was found, without the need for a long-term follow-up. preexisting enhancement on contrast-enhanced T1-weighted 1. Introduction (CE-T1WI) MR series, especially in patients with high-grade Gliomas are the most common primary brain tumors in gliomas. Pseudoprogression lesions can appear both in the adults [1]. At present, the standard treatment for gliomas firstthreemonthsandlongaftercompletionofradiationwith involves maximal safe tumor resection, followed by radio- anincidenceof9%–36%[2,3].(epseudoprogressionlesions therapy and concomitant and adjuvant chemotherapy with are caused by normal tissue response to treatment, whereas temozolomide. Recurrence and tissue response to the treat- true progression is the result of tumor recurrence and needs mentorso-calledpseudoprogressionarebothassociatedwith salvage therapy [4]. Hence, it is essential to discriminate true a high incidence of new enhancement or enlargement of progression from pseudoprogression as early as possible. 2 Journal of Oncology As a golden standard, histopathology is limited in chemoradiotherapy [6]; (6) had follow-up MRI at least 12 clinical practice for the disadvantage of invasive sampling, months after chemoradiotherapy; (7) had multi-parametric sampling error, and high cost [5]. MRI is the most common MR imaging when suspected gadolinium-enhancing lesion modality of follow-up tool in patients with posttreatment was found in the first follow-up MRI. glioma. However, the standard imaging protocol in the 2.2. Conventional and Perfusion MRI. All studies were Response Assessment in Neuro-Oncology (RANO) criteria, including fluid-attenuated inversion recovery (FLAIR) and performed on a 3T MRI scanner (Achieva; Philips Medical System, Best, (e Netherlands), using an 8-channel phased- CE-T1WI, does not specifically require to discriminate the true progression from the pseudoprogression lesions [4, 6]. array coil for acquisition. (e conventional MR protocol included precontrast and postcontrast T1WI, T2WI, and What is more, long-term follow-up should be made to FLAIR. (e multi-parametric MR imaging protocol in- discriminate true progression from pseudoprogression cludedT WI,T WI,FLAIR,DWI,MRS,andDSC-MRI.(e when new gadolinium-enhancing lesion or increase of the 1 2 parameters of DSC-MRI: perfusion-weighted gradient-echo preexisting enhancing lesion is detected. Consequently, echo-planar sequence; repetition time, 2000ms; echo time, longitudinal follow-up MRI and various advanced func- 40ms; slice thickness, 6.5mm; the data was acquired every tional MR techniques have been applied to obtain a de- finitive diagnosis. (e long waiting time for a definitive one second for a total of 1min 30sec, with Gd-DTPA (Gadovist,BayerScheringPharma,Berlin,Germany,0.2ml/ diagnosis by means of follow-up MRI is a psychological torture to patients and may delay salvage therapy. On the kg of body weight and maximum dose of 20ml) injected with a MRI-compatible power injector at a rate of 3ml/sec, other hand, in spite of promising results, the application of advanced functional MR techniques is limited for their followed by a 20-ml 0.9% saline flush using same flow rate. vagueresultsaswellasnotbeingaroutineexamintheclinic 2.3. Imaging Analysis. (e imaging analysis and post- [6–8]. processing were performed on a workstation (PHILIPS Previous studies have shown that MR perfusion- Extended MR WorkSpace 2.6.3.4). (e enhancement pat- weightedimaging,especiallydynamicsusceptibilitycontrast terns of residual cavity wall were divided into thin-linear perfusion MRI (DSC-MRI), can reflect tumor angiogenesis (partial or entire wall enhancement with thickness <3mm), which is valuable for glioma grading, estimation of prog- thick-linear (partial or entire wall enhancement of 3–5mm nosis, and differentiating tumor recurrence from treatment- in thickness), and nodular wall enhancement (with nodular related changes [5, 9, 10]. But there are only a few reports enhancement of 5–10mm in thickness) patterns [14]. Ce- that assessed the impact of combined DSC-MRI and other rebral blood volume (CBV) was calculated based on signal MRI features on identifying the true tumor progression intensity-time curves in transverse T -weighted sequence. [5, 9, 11]. Since functional MR imaging techniques can We placed 3 circular ROIs in the region with the largest reflect different pathological changes, such as diffusion- enhancement (Figure 1), and the mean CBV was calculated weightedimaging(DWI)forwatermoleculemovement,and and used in quantitative analysis. We calculated the relative magnetic resonance spectroscopy (MRS) for neoplastic CBV (rCBV) of suspected gadolinium-enhancing lesion as metabolites, we hypothesized that combination of DSC- well as ipsilateral normal tissues by dividing the CBV values MRI, DWI, and MRS can improve the accuracy for iden- oftheseregionswiththeCBVvaluesofcontralateralnormal tification of true progression lesionsamongthe gadolinium- white matter [15]. (e rCBV difference between two se- enhancing lesions. (erefore, the objective of this study was quential perfusion imaging was calculated for those patients to explore whether DSC-MRI can improve diagnostic per- with longitudinal DSC-MRI examinations [10]. (e ap- formance of multiparametric MRI in the evaluation of parentdiffusioncoefficient(ADC)wasmeasuredinthesame progression in gliomas patients after standard treatment. regions as CBV measurements. (e conventional and functional MRI characteristics were analyzed independently 2. Materials and Methods by two neuroradiologists (with 10 and 16 years of neuro- radiology experience, respectively). When a disagreement 2.1. Patients. (is study was approved by the Institutional existed, a consent was reached after consulting another Review Board. (e informed consent was waived for its neuroradiologist (with 25-year experience in neuroradiol- retrospective nature. A total of consecutive 65 glioma pa- ogy). (e outcome of the tumor was assessed according to tients with suspicious gadolinium-enhancing lesion were the updated RANO criteria for gliomas [6, 16]. True pro- recruited(36males,29females;meanage,46.5±14.3years), gressionwasdefinedwhenpatientswithnewlyenhancement according to the following eligibility criteria: (1) gliomas lesions or with increase the size of enhancement lesions were confirmed by histology; (2) gross-total resection of continuously in the follow-up period after accomplishment tumors [12]; (3) completed the standard treatment of chemoradiotherapy. If there were stable or regressing according to National Comprehensive Cancer Network enhancement lesions, the patients were defined as pseu- (NCCN)guideline,includinggross-totalresection,radiation doprogression [6]. therapy and chemotherapy (CCRT) after surgery, and six cycles of adjuvant temozolomide [13]; (4) had standard clinical MRI before and after chemoradiotherapy; (5) 2.4. Statistical Analysis. Statistical analysis utilized the gadolinium-enhancing lesion enlarged or presence of new software SPSS for windows release 25.0 (SPSS Inc., Chicago, gadolinium-enhancing lesion within the first month after IL, USA). Categorical variables were analyzed with log-rank Journal of Oncology 3 Figure 1: Example of region of interest (ROI) for measurement of CBV (the circle in red). (is 37-year-old female patient with path- ologically proven glioblastoma was ascribed to true progression group. test. (e quantitative data between true progression and dose (p � 0.615), histopathologic grade (p � 0.451), and pseudoprogression groups, including rCBV, Cho/Cr, Cho/ Karnofsky performance status (KPS) scores (p � 0.154). NAA, and ADC, were compared by using two-tailed Stu- AmongthepatientswithDWIdata(54/65),therewasno dent’s t-test when they were in non-normal distribution, or significant difference in ADC between two groups using Mann–WhitneyU testwhen they werein non-normal (p � 0.067)inspiteoflowerADCvaluefortrueprogression distribution. (e interobserver consistency between the two group (Table 1). (e Cho/Cr and Cho/NAA of true pro- neuroradiologists was evaluated with intraclass correlation gression group (2.520, 95%CI: 2.331–2.773; 2.414±0.665, coefficient(ICC).(esurvivaltimeswereestimatedwiththe respectively) were higher (p≤0.002) than those of pseu- Kaplan–Meier methods. Receiver operating characteristic doprogression group (1.719±0.664; 1.499±0.500, respec- (ROC) curve analysis was employed in determining the best tively) (Table 2), whereas there was no significant difference cutoffvaluesofrCBVandmetaboliteratiosindifferentiating in NAA/Cr ratio between true progression (1.030±1.100) true progression and pseudoprogression by maximizing the and pseudoprogression groups (1.300±0.750) (p � 0.485). sum of sensitivity and specificity. (e diagnostic perfor- Usingacutoff Cho/Crvalueof2.475, asensitivity51.0%and mance of all variables was measured as area under ROC specificity 92.9% were achieved; using a cutoff Cho/NAA curve (AUC). (e level of significance was set at p<0.05. value of 2.155, a sensitivity 64.7% and specificity 100.0% were achieved for separating the two groups. (e rCBV of normal brain tissue was 0.993 ±0.106 (0.854–1.209). (ere was significant difference between the 3. Results rCBV of normal brain tissue and that of true progression (e median follow-up span was 590 days (range: lesions (p<0.001), so did between the normal brain tissue 210–2670days). (e median progression-free survival (PFS) and pseudoprogression lesions (p<0.001). (e rCBV of was 360 days [95% confidence interval (CI): 399–580 days] true progression group (1.094, 95%CI: 1.135–1.636) was and median overall survival (OS) was 590 days (95% CI: significantly higher than that of pseudoprogression group 603–790 days) (Figure 2). 22 patients (33.846%) were dead (0.541±0.154) (p<0.001) (Table 2). Among 65 patients, 18 during the follow-up period. (e demographic and MR patients (34.0%, 14 patients with true progression and 4 imaging characteristics of the 65 patients are summarized in patients with pseudoprogression lesions) had serial DSC- Table 1. Fifty-one patients (28 males, 47.290±14.380 years) MRI. Changes in rCBV at subsequent follow-up differed werediagnosedastrueprogression(histologicallyconfirmed significantly (p � 0.015) between true progression (0.480, in 3 cases) and fourteen patients (8 males, 43.640±14.370 95%CI: 0.173–0.810) and pseudoprogression (−0.083, 95% years) were diagnosed as pseudoprogression. (ere was no CI: −1.138–0.620) groups. (e between-group comparison significant difference in patient age (p � 0.403), radiation revealed a significant difference between the 4 Journal of Oncology 100 100 80 80 60 60 40 40 20 20 0 0 0 500 1000 1500 2000 2500 0 1000 2000 3000 PFS (d) OS (d) Pseudoprogression Pseudoprogression True progression True progression (a) (b) Figure 2: Kaplan–Meier survival curve for PFS and OS according to true progression and pseudoprogression. PFS, progression-free survival; OS, overall survival. Table 1: Clinical characteristics of 65 glioma patients. Characteristics Pseudoprogression (n �14) True progression (n �51) t/X p value Mean age±SD(years) 43.640±14.372 47.290±14.375 −0.842 0.403 Gender 0.022 0.881 Male 8 28 Female 6 23 1p19q codeletion(+) (n �23) 1(33.33%) 4(21.05%) — 0.814 Promoter of MGMT methylation (n �12) 1(33.33%) 5(55.56%) — 0.574 IDH mutation (n �13) 1(50.00%) 3(27.27%) — 0.178 WHO grading — 0.451 II 1 3 III 9 24 IV 4 24 KPS score −1.427 0.154 Median 90.000 90.000 95%CI 90.000–90.000 86.890–89.580 T2-FLAIR mismatch — 0.676 Yes 1 9 No 13 42 Enhancement of residual cavity wall — 0.001 (in-linear 11 12 (ick-linear 1 10 Nodular 2 29 Total dose (GyRBE) −0.503 0.615 Median 60 59.92 95%CI 55.761–60.502 57.669–59.462 SVZ involvement — 0.204 Yes 7 36 No 7 15 ADC mean (mm /s) (n �54) 903.142±491.652 523.000 (484.950–668.610) −1.842 0.067 MGMT, O -methyl-guanine methyl transferase; KPS, Karnofsky Performance Score; SVZ, subventricular zone; ADC mean, apparent diffusion coefficient mean; T2-FLAIR mismatch, the presence of a complete/near-complete hyperintense signal on T2-weighted (T2W) MRI sequences, in combination with a relative hypointense signal on fluid attenuation inversion recovery (FLAIR) MR sequences except for a hyperintense peripheral rim; Bold p value indicates statistically significant association, - indicates Fisher’s exact test. Percent survival Percent survival Journal of Oncology 5 100.0 Table 2: Comparison of DSC and MRS variables. Pseudoprogression True progression p value rCBV <0.001 80.0 Median/mean 0.541±0.154 1.094 95%CI (range) 0.360–0.892 1.135–1.636 Cho/Cr ratio 0.001 60.0 Median/mean 1.719±0.664 2.520 95%CI (range) 0.680–3.160 2.331–2.773 Cho/NAA ratio 1.499±0.495 2.414±0.665 <0.001 40.0 rCBV, relative cerebral blood volume; Cho, choline; NAA, N-acetylas- partate; Cr, creatin; CI, confidence interval, Bold p value indicates statis- tically significant association. 20.0 pseudoprogression and the true progression groups corre- 0.0 sponding to a large effect size (Cohen’s d �1.392). (e ROC 0.0 20.0 40.0 60.0 80.0 100.0 analysis showed that, using a cutoff rCBV value of 0.045, it 1 - specificity achieved AUC 0.904, sensitivity 100.0%, and specificity CE CE + DSC 75.0% to separate the two groups. MRS DSC + MRS Figure 3 shows the ROC analyses of the diagnostic DSC CE + DSC + MRS performance using different variables and their combina- CE + MRS Reference line tions for distinguishing true progression and pseudoprog- Figure 3: Comparison of ROC curve analyses for different MR ression. (e AUCs, sensitivity, and specificity of metabolites variables and their combinations. (e combination of ratios, and rCBV were significantly larger than those of CE+DSC+MRS presents the high diagnostic performance, and enhancement pattern of residual cavity wall (Figure 3, Ta- the AUC was 0.965 (95%CI 0.920–1.000), with significant level of ble 3). With the cutoff rCBV value of 0.743, the AUC, p<0.05 to any other variables and other combinations. CE, sensitivity, and specificity of DSC-MRI for distinguishing contrastenhancementpatternofresidualcavitywall;MRS,Cho/Cr true progression and pseudoprogression were 0.912, 76.5%, and Cho/NAA ratios; DSC, dynamic susceptibility contrast and92.9%,respectively.(eAUC,sensitivity,andspecificity perfusion. of the combination of Cho/Cr and Cho/NAA were 0.881, 88.2%, and 78.6%, respectively. (e AUC and specificity of Accurately and quickly distinguishing true progression MRS+DSC were similar to those of CE+MRS+DSC from pseudoprogression in glioma patients after standard combination model. However, the sensitivity and Youden chemoradiotherapy remains a major clinical challenge [6]. index of MRS+DSC were relatively lower (Figure 3, Ta- Newly enhancement lesions or increase of the preexisting ble 3). Interestingly, when we combined all MR variables, enhancement lesions, accompanied with mass-effect, as well including enhancement patterns of residual cavity wall, asvasogeniccerebraledema,canbedetectedbothinpatients rCBV, and metabolites ratios, the diagnostic performance with true progression and in patients with pseudoprog- was significantly improved (AUC 0.965, sensitivity 90.2%, ression lesions [17]. Generally, longitudinal follow-up MRI and specificity 100.0%). over several months or stereotactic biopsy is needed for (e agreement was excellent between the two neuro- definitive diagnosis. In perfusion-weighted imaging, CBV radiologists for evaluation of the functional MR variables, could be used as an important biomarker for neoplastic including rCBV (ICC 0.979, 95% CI: 0.966–0.987), ADC vasculature [18, 19]. Our results support the hypothesis that (ICC 0.995, 95% CI: 0.991–0.997), Cho/Cr (ICC 0.964, 95% DSC-MRI can add information for definitive diagnosis of CI: 0.941–9.978), and Cho/NAA (ICC 0.958, 95% CI: glioma patients with gadolinium-enhancing lesions after 0.932–0.974). treatment. In this study, rCBV is a valuable imaging marker Figures 4 and 5 show the classic examples of true pro- for distinguishing true progression from pseudoprogression gression and pseudoprogression lesions. in glioma patients. Increased rCBV as surrogate biomarker of active growing tumor could potentially reduce the ne- cessity for biopsy and their associated risks, thereby sim- 4. Discussion plifyingtheposttreatmentevaluationprocessanddecreasing We applied DSC-MRI, along with conventional MRI se- the care cost. On the other hand, DSC-MRI is the most quences, DWI, and MRS, to assess the impact of perfusion common and available perfusion technique in the current commercialMRequipment.(us,wesuggestthatDSC-MRI parameter on the differential diagnosis of true progression versus pseudoprogression in patients with gliomas. Our should be conventionally employed in the evaluation of posttreatment gliomas. results suggested that a combination of DSC-MRI, contrast enhanced T W imaging, and MRS can greatly improve the (e DSC-MRI has been widely used in evaluation of the diagnostic performance in distinguishing true progression suspicious lesions in posttreatment glioma patients from pseudoprogression. [5, 18–20]. (e rCBV cutoff is important for identifying true Sensitivity 6 Journal of Oncology Table 3: ROC curve analyses of diagnostic performance of various variables and their combinations. Characteristics AUC 95%CI Sensitivity Specificity Youden index CE 0.782 0.643–0.920 0.765 0.786 0.551 MRS 0.881 0.788–0.973 0.882 0.786 0.668 DSC 0.912 0.837–0.988 0.765 0.929 0.694 CE+MRS 0.930 0.867–0.993 0.882 0.857 0.739 CE+DSC 0.936 0.873–0.998 0.863 0.857 0.720 DSC+MRS 0.965 0.924–1.000 0.863 1.000 0.863 CE+MRS+DSC 0.965 0.923–1.000 0.902 1.000 0.902 CE, contrast enhancement pattern of residual cavity wall; MRS, Cho/Cr and Cho/NAA ratios; DSC, dynamic susceptibility contrast perfusion. RL (a) (b) (c) (d) Figure 4: Example of pseudoprogression. A 21-year-old man with pathologically proven glioblastoma. A new enhancing lesion (arrow) developed in his left frontal lobe three months after completion of radiotherapy (a). No hyper-perfusion was detected on his DSC-MRI (b), with rCBV value of 0.71. MRS display Cho/Cr ratio of 1.20 and Cho/NAA 0.46 (c). (e enhancing lesions in initial follow-up MRI disappeared in the MRI 23 months after completion of radiotherapy (d). progression. However, there is no consensus on the cutoff imaging techniques [10, 21]. We enrolled gliomas patients value of rCBV (0.71–5.01) for distinguishing true progression withdifferentgrades(WHOgradeII-IV)andwithgross-total frompseudoprogressionlesions[8,18,19,21].Inastudyof44 resection of tumors, and followed by standard therapy glioblastomapatients,Blaseletal.foundthecutoffvalueof2.2 according to NCCN guideline. However, in Blasel et al.’s for rCBV yielded a sensitivity of 65% and a specificity of 71% study [20], only glioblastoma patients (WHO grade IV) were [20], whereas, in the present study, with the cutoff value of recruited,including4patientswhoonlyhadbiopsyinsteadof 0.743 for rCBV, we achieved higher sensitivity (76.5%) and tumorresection.(eirtherapyregimenwasdiverse,including specificity (92.9%). (is discrepancy may be attributed to the Stupp’s method, radiotherapy only, and second- and third- diversity of inclusion criteria and treatment regimens, as well line therapy in some patients. (us, the higher rCBV cutoff as the lack of standardization in reference standard and value in Blasel et al.’s study indicted higher mitotic activity Journal of Oncology 7 RL (a) (b) (c) (d) Figure 5: Example of true progression. An 18-year-old woman with pathologically proven glioblastoma. A new irregular ring enhancing lesion (arrow) developed in her right temporal-occipital lobe three months after completion of radiotherapy (a). Hyper-perfusion (arrow) was detected on her DSC-MRI (b), with rCBV value of 1.11. (e MRS display Cho/Cr ratio of 4.15 and Cho/NAA 3.95 (c). (e enhancing lesions enlarged significantly one month later (d). Tumor recurrence was confirmed pathologically after reoperation. and higher vascularity in their patients. In spite of the dif- MRS metabolites ratios provide useful information in distinguishing true progression from pseudoprogression ferenceofcutoffvalueofrCBV,thediagnosticperformanceof our study was similar to other authors’ [8, 19]. (is indicted lesions [8, 9, 23, 24]. Among several metabolic ratios [8, 24], DSC-MRI, in spite of diverse rCBV cutoff value, can be used Cho/Crratiowasconsideredasthemostfavorablemarkerin to improve the evaluation of progression in glioma patients differentiating between true progressions and pseudoprog- after treatment [5]. ression. Cho/Cr ratio achieved excellent sensitivity (91%) Longitudinal change in rCBV may improve the dis- and specificity (95%) in a higher grade gliomas study [8]. crimination between true progression and pseudoprog- However, in our study the sensitivities of MRS were rela- ression lesions [10, 22]. (e diagnostic performance, tively lower (51.0%) with the cutoff values of Cho/Cr ratio including AUC, sensitivity, and specificity, of rCBV dif- 2.475 and Cho/NAA ratio 2.115. A possible explanation for ferencebetweentwosequentialexaminationswasnoticeably this discrepancy might be because we included lower-grade higher than that of initial rCBV in our study. Various (WHO grade II and III) gliomas, which probably had lower metabolite concentration in the enhancing regions. Inter- pathological changes representing a wide range of local cerebral blood volume, including radiation-induced vas- estingly, our results showed that MRS index ratios achieved culopathy and necrosis, could coexist shortly after treat- good diagnostic efficiency (AUC 0.881). Our findings sup- ment. (us, the rCBV value at the initial DSC-MRI may be port previous finding that MRS is a promising functional less effective in discriminating the true progression from MR technique for the evaluation of treatment response in pseudoprogression lesions [10, 22]. However, the destiny of glioma [8]. However, we must keep in mind that using MRS suspicious lesions becomes apparent over time. As the time ratios as a biomarker in the differential diagnosis has its goes on, increase in rCBV could be detected from those intrinsic limitations; i.e., it is difficult to obtain universal enhancing lesions that contain viable tumor cells, especially cutoff values due to partial volume effect resulting from in the most active and aggressive region of the lesions. relativelylargevoxelsize,andtheratiosarevulnerabletothe 8 Journal of Oncology diagnostic performance, only18 of 65 patients in our cohort local field homogeneity, such as water, lipids, and surgical clips. had the sequential DSC-MRI data. In conclusion, DSC-MRI, especially longitudinal change (e roleof DWI in differentiating true progressionfrom pseudoprogressionlesionsremainsambiguous.Moststudies in rCBV, showed satisfactory diagnostic performance and showedtheaccuracyofADCvalueinthedifferentiationwas canbeausefultoolindistinguishingtruegliomaprogression the lowest among all functional MR techniques [5, 8, 23]. from pseudoprogression. Moreover, the combination of (e treatment effects, such as gliosis, coagulation necrosis, DSC-MRIandconventionalmorphologicalimagingfeatures macrophages invasion, and demyelination, can reduce ADC and other advanced MR techniques would further improve value. Necrosis could be found in many recurrent tumor identifying the progression, which may greatly facilitate the individualized management of posttreatment glioma pa- lesions [23]. (us, the diagnostic value of DWI is weakened due to similar changes of diffusivity but quite different tients. (is combination may even eliminate a long-term follow-up when a suspected gadolinium-enhancing lesion is pathologic processes. In literatures, the sensitivity and specificity of ADC value in the differentiation varied widely found. (95% CI: 60–80%, and 95% CI: 77–93%, respectively) [8]. In our study, there was no significant difference in ADC be- Data Availability tweenthetwotypesoflesions.Similarresultswerepresented (e data used to support the findings of this study are in previous studies [23, 25]. However, in another previous available from the corresponding author upon request. study, the ADC value of recurrence lesions was significantly lower than that of non-recurrence lesions [5]. (is dis- crepancymaybeexplainedasbiasfromtherelativelysmaller Disclosure sample size and inclusion of relatively lower-grade gliomas Guanmin Quan and Kexin Zhang are the co-first authors. patients. Moreover, there was no consensus on the cutoff ADC value for this discrimination [5, 26, 27]. Conflicts of Interest Since conventional morphologic imaging findings are often limited in differentiation of true progression from (e authors declare that they have no conflicts of interest. pseudoprogression lesionsingliomas,variousadvancedMR imaging modalities, including DWI, MRS, and DSC-MRI, Authors’ Contributions canprovideadditionalphysiologicormetabolicinformation about posttreatment gliomas. However, the diagnostic ac- TY, KZ, DH, J-LR, and QG made a substantial contribution curacy of each technique is still limited and different to the concept and design, acquisition of data or analysis, [5, 8, 15, 28]. A multiparametric MR imaging has the po- andinterpretationofdata;GQ,YL,WW,andTYdraftedthe tential to improve the overall diagnostic performance. In article and revised it critically for relevant intellectual Matsusue et al.’s study [28], which enrolled 15 subjects, the content;GQ,KZ,WW,andDHperformedMRexamination diagnostic accuracy of combined multiparametric MRI and follow-up of patients; all the authors approved the final (93.3%) was higher than those of ADC ratio (86.7%), rCBV versionofthemanuscript.GuanminQuanandKexinZhang (86.7%), and MRS (Cho/Cr and Cho/NAA) (84.6%). Our contributed equally to this study. results confirmed the superiority of multiparametric MRI, because advanced imaging could provide more compre- Acknowledgments hensiveinformationwithrespecttotumorpathophysiology. In the present study with relatively larger cohort with DSC- (is study was supported by the Specialist Leadership MRI, we further combined conventional morphologic Project of Hebei Province (No. 361004, GQ), Technology imaging findings and advanced MRI data, especially the Tracking Program for Medical Application of Hebei Prov- addition of DSC-MRI, which can be more reliable in ince (No. G201725, GQ), and the Health and Family differentiating true tumor progression from Planning Commission of Guangxi Zhuang Autonomous pseudoprogression. (e combination of CE-T1WI, MRS, Region (Z2016425, DH). and DSC achieved the highest diagnostic performance (AUC, 0.965). Based on these results, we recommend that References patients with suspicious enhancing lesions in conventional follow-up MRI should additionally receive advanced MR [1] Q. T. Ostrom, H. Gittleman, G. Truitt et al., “CBTRUS sta- tisticalreport:primarybrainandothercentralnervoussystem examinations, such as MRS and DSC-MRI. tumors diagnosed in the United States in 2011–2015,” Neuro Several limitations of this study should be mentioned. Oncology, vol. 20, no. 4, pp. iv1–iv86, 2018. First, the number of the patients in this retrospective study [2] S. C. (ust, M. J. van den Bent, and M. Smits, “Pseudo- was relatively small because we only recruited those patients progression of brain tumors,” Journal of Magnetic Resonance with DSC-MRI and also with sufficient MRI follow-up data. Imaging, vol. 48, no. 3, pp. 571–589, 2018. Second, the final diagnosis of most patients (95.4%) in the [3] A. W. Abbasi, H. E. Westerlaan, G. A. Holtman, K. M. Aden, current study was based on the follow-up data. (ird, we P. J. van Laar, and A. van der Hoorn, “Incidence of tumour measured CBV and DWI in a 2D mode, which could not progression and pseudoprogression in high-grade gliomas: a reflect local pathologic changes completely. Fourth, al- systematic review and meta-analysis,” Clinical Neuroradiol- though the longitudinal change in rCBV could improve the ogy, vol. 28, no. 3, pp. 401–411, 2018. Journal of Oncology 9 [4] K. Parvez, A. Parvez, and G. Zadeh, “(e diagnosis and MRI,” American Journal of Roentgenology, vol. 198, no. 1, pp. 19–26, 2012. treatment of pseudoprogression, radiation necrosis and brain tumor recurrence,” International Journal of Molecular Sci- [19] A. Romano, M. C. Rossi Espagnet, L. F. Calabria et al., “Clinical applications of dynamic susceptibility contrast ences, vol. 15, no. 7, pp. 11832–11846, 2014. [5] Y.Yang,Y.Yang,X.Wuetal.,“AddingDSCPWIandDWIto perfusion-weighted MR imaging in brain tumours,” La Radiologia Medica, vol. 117, no. 3, pp. 445–460, 2012. BT-RADS can help identify postoperative recurrence in pa- [20] S.Blasel,A.Zagorcic,A.Jurcoaneetal.,“PerfusionMRIinthe tients with high-grade gliomas,” Journal of Neurooncology, evaluation of suspected glioblastoma recurrence,” Journal of vol. 46, no. 2, pp. 363–371, 2010. Neuroimaging, vol. 26, no. 1, pp. 116–123, 2016. [6] B. Ma, J. O. Blakeley, X. Hong et al., “Applying amide proton [21] J. Dongas, A. T. Asahina, S. Bacchi, and S. Patel, “Magnetic transfer-weighted MRI to distinguish pseudoprogression resonance perfusion imaging in the diagnosis of high-grade from true progression in malignant gliomas,” Journal of glioma progression and treatment-related changes: a sys- Magnetic Resonance Imaging,vol.44,no.2,pp.456–462,2016. tematic review,” Open Journal of Modern Neurosurgery, [7] H. S. Nguyen, N. Milbach, S. L. Hurrell et al., “Progressing vol. 08, no. 03, pp. 282–305, 2018. bevacizumab-induced diffusion restriction is associated with [22] J. L. Boxerman, B. M. Ellingson, S. Jeyapalan et al., “Longi- coagulative necrosis surrounded by viable tumor and de- tudinal DSC-MRI for distinguishing tumor recurrence from creased overall survival in patients with recurrent glioblas- pseudoprogression in patients with a high-grade glioma,” toma,” American Journal of Neuroradiology, vol. 37, no. 12, American Journal of Clinical Oncology, vol. 40, no. 3, pp. 2201–2208, 2016. pp. 228–234, 2017. [8] B. R. J. van Dijken, P. J. van Laar, G. A. Holtman, and [23] B. Bobek-Billewicz, G. Stasik-Pres, H. Majchrzak et al., A. van der Hoorn, “Diagnostic accuracy of magnetic reso- “Differentiation between brain tumor recurrence and radia- nance imaging techniques for treatment response evaluation tion injury using perfusion, diffusion-weighted imaging and in patients with high-grade glioma, a systematic review and MR spectroscopy,” Folia Neuropathology, vol. 48, no. 2, meta-analysis,” European Radiology, vol. 27, no. 10, pp. 81–92, 2010. pp. 4129–4144, 2017. [24] H. Zhang, L. Ma, Q. Wang, X. Zheng, C. Wu, and B.-n. Xu, [9] J. Liu, C. Li, Y. Chen et al., “Diagnostic performance of “Role of magnetic resonance spectroscopy for the differen- multiparametric MRI in the evaluation of treatment response tiation of recurrent glioma from radiation necrosis: a sys- in glioma patients at 3T,” Journal of Magnetic Resonance tematic review and meta-analysis,” European Journal of Imaging, vol. 51, no. 4, pp. 1154–1161, 2020. Radiology, vol. 83, no. 12, pp. 2181–2189, 2014. [10] E. Steidl, M. Muller, ¨ A. Muller, ¨ U. Herrlinger, and [25] C. Asao, Y. Korogi, M. Kitajima et al., “Diffusion-weighted E. Hattingen, “Longitudinal, leakage corrected and uncor- imaging of radiation-induced brain injury for differentiation rected rCBV during the first-line treatment of glioblastoma: a from tumor recurrence,” AJNR American Journal of Neuro- prospectivestudy,” Journal of Neuro-Oncology,vol.144,no.2, radiology, vol. 26, no. 6, pp. 1455–1460, 2005. pp. 409–417, 2019. [26] P. A. Hein, C. J. Eskey, J. F. Dunn et al., “Diffusion-weighted [11] M. Law, “Advanced imaging techniques in brain tumors,” imaginginthefollow-upoftreatedhigh-gradegliomas:tumor Cancer Imaging, vol. 9, no. Special Issue A, pp. S4–S9, 2009. recurrence versus radiation injury,” AJNR American Journal [12] G. Ekinci, I. N. Akpınar, F. Baltacıoglu ˘ et al., “Early-post- of Neuroradiology, vol. 25, no. 2, pp. 201–209, 2004. operative magnetic resonance imaging in glial tumors: pre- [27] P.C.Sundgren,X.Fan,P.Weybrightetal.,“Differentiationof diction oftumor regrowthand recurrence,” European Journal recurrent brain tumor versus radiation injury using diffusion of Radiology, vol. 45, no. 2, pp. 99–107, 2003. tensor imaging in patients with new contrast-enhancing le- [13] L. B. Nabors, J. Portnow, M. Ammirati et al., “NCCN sions,” Magnetic Resonance Imaging, vol. 24, no. 9, guidelines insights: central nervous system cancers, version pp. 1131–1142, 2006. 1.2017,” Journal of the National Comprehensive Cancer Net- [28] E. Matsusue, J. R. Fink, J. K. Rockhill, T. Ogawa, and work, vol. 15, no. 11, pp. 1331–1345, 2017. K. R. Maravilla, “Distinction between glioma progression and [14] B.R.Kim,S.H.Choi,T.J.Yunetal.,“MRimaginganalysisof post-radiationchangebycombinedphysiologicMRimaging,” non-measurable enhancing lesions newly appearing after Neuroradiology, vol. 52, no. 4, pp. 297–306, 2010. concom-itantchemoradiotherapyinglioblastomapatientsfor prognosis prediction,” PLoS One, vol.11, no.11, p. e0166096, [15] M. L. White, Y. Zhang, F. Yu et al., “Post-operative perfusion and diffusion MR imaging and tumor progression in high- grade gliomas,” PLoS One, vol. 4, no. 3, p. e0213905, 2019. [16] P. Y. Wen, D. R. Macdonald, D. A. Reardon et al., “Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group,” Journal of Clinical Oncology, vol. 28, no. 11, pp. 1963–1972, 2010. [17] A.Siu,J.J.Wind,J.B.Iorgulescu,T.A.Chan,Y.Yamada,and J. H. Sherman, “Radiation necrosis following treatment of high grade glioma-a review of the literature and current understanding,” Acta Neurochirurgica, vol. 154, no. 2, pp. 191–201, 2012. [18] G.M.Fatterpekar,D.Galheigo,A.Narayana,G.Johnson,and E. Knopp, “Treatment-related change versus tumor recur- rence in high-grade gliomas: a diagnostic conundrum-use of dynamic susceptibility contrast-enhanced (DSC) perfusion

Journal

Journal of Oncology – Hindawi Publishing Corporation

Published: Feb 9, 2021

There are no references for this article.