Prediction of Survival Outcomes in Patients with Glioma Using Magnetic Resonance Imaging (MRI): A Systematic Review and Meta-Analysis

Abstract

Background:

Glioma is the most common malignancy in the central nervous system. Even with optimal therapies, glioblastoma (the most aggressive form of glioma) is incurable, with only 26.5% of patients having a 2-year survival rate. The present meta-analysis evaluated the association of magnetic resonance imaging (MRI)-derived parameters in glioma patients with progression-free survival (PFS) and overall survival. Eligible clinical articles on glioma patients included those that contained an evaluation of the association between MRI findings, PFS, and overall length of survival.

Methods:

Review of the literature included the following databases: WHO International Clinical Trials Registry Platform; Google Scholar; Web of Science; PubMed; SIGLE; NYAM; Scopus; Randomized controlled trial (RCT); Virtual Health Library (VHL); Cochrane Collaboration; EMBASE; and Clinical Trials.

Results:

The current review included 20 studies, and covered 2097 patients with gliomas. There were 1310 patients with glioblastoma and 320 with astrocytoma. There were 161 patients with grade-2 gliomas and 111 patients with grade-3. Tumour necrosis, peritumoural oedema, and multiple lesions were associated with PFS, as well as tumour necrosis and peritumoural oedema with overall survival.

Conclusions:

The present meta-analysis highlighted the ability of MRI to predict PFS and overall survival in patients with gliomas. This is crucial to identify patients at risk for poor survival outcomes and for individualising the treatment plan for such patients.

The PROSPERO Registration:

CRD42023489535, https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=489535.

Graphical abstract

Keywords

glioblastoma / glioma / MRI / survival

1. Introduction

Glioma is the commonest malignancy in the central nervous system (CNS) [1]. Gliomas are diffuse infiltrative tumours arising from the glial cells of the CNS. In the United States, 6 cases per 100,000 population are diagnosed with glioma yearly. There are three types of gliomas based on the phenotypic cell characteristics: astrocytoma; ependymoma; and oligodendroglioma. The most malignant type of glioma is glioblastoma, and the least malignant is pilocytic astrocytoma [2, 3]. Managing gliomas necessitates adequate pre-operative imaging, surgery planning, and post-operative care. The surgical management is based mainly on tumour resection or stereotactic biopsy with the intraoperative determination of the tumour margins using accurate imaging modalities [4, 5].

Glioblastoma carries a poor prognosis even after standard management. Even with the optimal therapies, glioblastoma is incurable, with only 26.5% of patients having a 2-year overall survival rate. The median survival time of glioblastoma is 14 months, and it decreases to 30 weeks at recurrence [6, 7]. The presence of a glioma tumour is primarily evaluated using gadolinium-enhanced magnetic resonance imaging (MRI). However, this diagnostic modality has a low accuracy with some treatment modalities, resulting in a pseudoresponse phenomenon. The phenomenon is associated with a rapid decline in contrast leakage of the gadolinium, thereby producing a failure to reflect tumour size or activity accurately [8]. Diffusion imaging has gained attraction recently as an effective diagnostic tool for assessing tumour response and progression. Recently, there has been a great interest in implementing multi-parametric MRI sequences in estimating tumour infiltration and determining treatment in patients with glioma [9, 10]. MRI is a non-invasive tool that provides information on the physiological characteristics of the tumour. These data have been widely used to estimate the baseline risk and survival, which are essential to assigning glioma patients to the necessary treatment [11].

MRI parameters are routinely obtained from glioma patients, independent of the treatment received. These may provide crucial predictive information for prognosis and evaluation of treatment response. Feasible and accurate prognostic modalities for patients with gliomas may help to individualize treatment for each patient, in order to improve long-term outcomes. The prognosis of gliomas is unpredictable, and depends on many factors. The aggressiveness of the tumour is one of the most significant factors [12, 13]. Of note, grades I and II gliomas are low-grade tumours with a reasonable prognosis, with approximately 95% of patients showing a 5-year survival rate (grade I). The survivability is less predictable and poorer with more advanced diseases, with only 7% of patients with glioblastoma surviving for 5 years after diagnosis [14]. The average survival time of glioblastoma is 12–18 months, with approximately only 25% of patients surviving for more than 12 months [15].

There was abundant literature evaluating the predictive value of MRI findings in patients with gliomas. Previously published systematic reviews investigated the accuracy of MRI in diagnosing, differentiating, and assessing treatment responses among glioma patients [16, 17, 18]. However, the literature reflected differences in MRI techniques, disease differentiation, treatment modalities, patients’ characteristics, and study methods [19]. These differences highlight the need for a conclusive report to estimate the findings in different MRI modalities that may predict survival outcomes in patients with gliomas. Such knowledge is essential in order to assign patients at higher mortality risk to the necessary management plan and to decrease potential consequences of gliomas. Therefore, the present meta-analysis evaluated the association between MRI-derived parameters and progression-free survival (PFS), and overall survival, in glioma patients.

2. Methods

This systematic review and meta-analysis followed the recommendations offered by the Cochrane collaboration [20] and by Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for conducting systematic reviews and meta-analyses [21] (Supplementary Materials-PRISMA_2020_checklist). The protocol of the present meta-analysis was registered prospectively in the PROSPERO database (Number: CRD42023489535).

2.1 Literature Search

A comprehensive review of the literature was performed from the beginning of the literature on the topic until 12th December 2023. Review of the literature included the following databases: WHO International Clinical Trials Registry Platform; Google Scholar; Web of Science; PubMed; SIGLE; NYAM; Scopus; Randomized controlled trial (RCT); Virtual Health Library (VHL); Cochrane Collaboration; EMBASE; and Clinical Trials. The search string included the following keywords: MRI; Magnetic Resonance Imaging; Magnetic Resonance Image; Glioma; Gliomas; Glioblastoma; Mortality; Death; Survival; Survivability; and Prognosis. The search strategy imposed no restrictions related to patients’ demographics or to study demographics. Citation tracking and screening of the references of previous reviews was performed in addition to cross-referencing to include all potentially eligible articles.

2.2 Study Selection

All clinical articles that were used included (a) patients with gliomas; (b) evaluation of the association between MRI findings and PFS; and (c) overall survival data of patients with gliomas. The present study implemented the Cox regression model to calculate the hazard ratio (HR) for the time to relevant outcomes. We did not restrict the data as to glioma grade, type, differentiation, or diagnostic criteria. Furthermore, we excluded irrelevant articles, unextractable data, reviews, guidelines, case reports, cadaveric articles, studies that implemented artificial intelligence models, errata, case series, letters, comments, meeting abstracts, posters, and book chapters. The screening processes were carried out independently to determine the relevant articles that fulfil the inclusion criteria and to exclude irrelevant studies. A PRISMA flowchart was designed to document the search process, screening, and article-exclusion causes at each step of the literature review.

2.3 Data Extraction

The source-related data were extracted, including the title, study ID, study URL, study location, study period, and study design. Methods-related data were extracted, including the eligibility criteria, the imaging technique, the management of gliomas, study outcomes, and follow-up protocols. Baseline patients’ demographic data were retrieved, including sample size, age, weight, body mass index (BMI), and comorbid disorders. The tumour-related variables were retrieved, including the type, histopathological grade, location, initial therapy, extent of resection, and post-operative radiotherapy or chemotherapy. included: contrast enhancement; tumour necro The MRI parameters examined for association with PFS, and overall survival were sis; the extent of resection; peritumoural oedema; tumour volume, diameter, and margins; apparent diffusion coefficient; and multiple lesions.

2.4 Quality Assessment

The quality of the eligible observational studies was evaluated using the National Institute of Health quality-assessment tool [22]. The analyzed articles were sorted into ‘good’, ‘fair’, and ‘bad’.

2.5 Data Analysis

The summary of HR was computed by pooling the HR from all the relevant articles. A fixed-effect model was used when methodological and statistical homogeneity between the study variables was established. A random-effects model was used when the statistical heterogeneity was found. Statistical homogeneity was determined using Higgins I² statistic, at >50%, and the Cochrane Q (χ² test), at p \le 0.10 [23]. Publication bias was assumed in the presence of an asymmetrical funnel plot and based on Egger’s regression test (p 0.10). Subgroup analysis was performed on the type of glioma. Review Manager, v. 5.4 (The Nordic Cochrane Centre, Copenhagen, Denmark) [24], was used to analyze the data. The significant associations with PFS and overall survival were based on p \le 0.05.

3. Results

A systematic review of the literature yielded 2709 studies. Of them, 713 duplicate reports were excluded, resulting in 1996 studies eligible for title and abstract screening. A further 1968 reports were removed, resulting in 128 studies that were eligible for full-text screening. Subsequently, 106 studies were excluded, resulting in 22 articles eligible for data extraction. Three reports with unextractable data were excluded, and one study was identified through citation tracking. A total of twenty articles were finally included for systematic review and meta-analysis (see Fig. 1).

3.1 Demographic Characteristics and Quality Assessment of the Analyzed studies

The present meta-analysis included 20 articles, which included 2097 glioma patients [25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44]. There were four prospective studies and 15 retrospective designs. Six studies included patients from the United States, and four articles included patients from China. The age of the analyzed patients ranged from 36 to 63 years. There were 1159 males and 748 females. There were 1310 patients with glioblastoma and 320 with astrocytoma. There were 161 patients with grade-2 gliomas and 111 patients with grade-3. There were 104 patients with frontal lobe involvement, and 39 with parietal lobe involvement. The temporal lobe was affected in 89 patients and the insular region was affected in 10 patients. The follow-up period ranged from 12 to 65 months, with overall survival time ranging from 7.6 months to 18.4 months. All the analyzed articles showed good quality, with scores ranging from 66.66% to 75% (Table 1, Ref. [25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44]).

3.2 Factors Associated with Progression-Free Survival (PFS)

Contrast enhancement. Ten articles evaluated the association between contrast enhancement of gliomas, using MRI, and PFS [25, 26, 27, 29, 31, 32, 33, 38, 39, 41]. In the random-effects model (I² = 83%, p 0.001), there was no significant association between contrast enhancement and PFS (HR = 1.15; 95% Confidence Interval (CI): 0.84, 1.56; p = 0.39). Subgroup analysis based on the type of glioma revealed no statistically significant association between contrast enhancement and PFS among patients with glioblastoma (HR = 1.25; 95% CI: 0.94, 1.68; p = 0.13). There was no evidence of publication bias (Egger’s regression test, p > 0.10) with the symmetrical distribution of the analyzed articles within the funnel plot (Fig. 2A,B).

Tumour necrosis. The association between tumour necrosis of gliomas and PFS was evaluated in 6 articles [26, 29, 31, 38, 39, 44]. There was no significant impact of tumour necrosis finding in MRI and PFS (HR = 1.39; 95% CI: 0.94, 2.06; p = 0.10) in the random-effects model (I² = 32%; p = 0.19). Subgroup analysis revealed no statistically significant association between tumour necrosis and PFS among patients with glioblastoma (HR = 1.68; 95% CI: 0.69, 4.08; p = 0.25) (Fig. 2C).

Peritumour edema. Four articles evaluated the association between peritumour edema and PFS [25, 28, 33, 41]. In the random-effects model (I² = 0%; p = 0.78), there was a statistically significant negative relationship between peritumour edema development and PFS (HR = 2.68; 95% CI: 1.54, 4.64; p = 0.005) (Fig. 2D).

Tumour volume. The association between tumour volume and PFS was assessed in four articles [25, 34, 35, 38]. There was no statistically significant association between tumour volume and PFS (HR = 1.09; 95% CI: 0.99, 1.21; p = 0.07) in the random-effects model (I² = 0%; p = 0.44) (Fig. 2E).

Tumour diameter. Three studies evaluated the impact of tumour diameter on PFS among patients with gliomas [38, 40, 44]. In the random-effects model (I² = 81%; p = 0.005), there was no statistical association between tumour diameter and PFS (HR = 1.37; 95% CI: 0.68, 2.73; p = 0.38) (Fig. 3A).

Apparent diffusion coefficient. The impact of the apparent diffusion coefficient on PFS among patients with gliomas was reported in two studies [29, 32]. There was no statistically significant association between the apparent diffusion coefficient and PFS (HR = 0.69; 95% CI: 0.05, 9.06; p = 0.77) in the random-effects model (I² = 82%; p = 0.02) (Fig. 3B).

Multiple lesions. Three articles evaluated the association between multiple lesions, seen on MRI, and PFS [32, 43, 44]. In the random-effects model (I² = 12%; p = 0.32), there was no significant association between multiple lesions and PFS (HR = 1.57; 95% CI: 0.92, 2.69; p = 0.10) (Fig. 3C).

3.3 Factors Associated with Overall Survival

Contrast enhancement. Eight articles evaluated the impact of contrast enhancement on overall survival in patients with glioma [25, 26, 27, 32, 33, 37, 39, 41]. No significant association was observed between contrast enhancement and overall survival (HR = 1.12; 95% CI: 0.87, 1.44; p = 0.38) in the random-effects model (I² = 69%; p = 0.002). Subgroup analysis revealed no significant association between contrast enhancement and overall survival (HR = 1.37; 95% CI: 0.78, 2.40; p = 0.48) (Fig. 3D).

Tumour necrosis. The impact of tumour necrosis on the overall survival in patients with gliomas was evaluated in three studies [26, 39, 44]. There was a statistically significant negative association between tumour necrosis and overall length of survival (HR = 1.74; 95% CI: 1.04, 2.90; p = 0.03) (Fig. 3E).

Peritumour edema. Four studies showed an association between peritumoural edema and overall survival in glioma patients [25, 28, 32, 41]. Pooling the effect sizes in the random-effects model (I² = 0%; p = 0.40) revealed a statistically significant negative association between peritumoural edema development and overall survival time (HR = 2.43; 95% CI: 1.42, 4.17; p = 0.001) (Fig. 3F).

Tumour volume. The impact of the tumour volume on the overall survival in patients with gliomas was reported in three articles [25, 34, 35]. In the random-effects model (I² = 96%; p 0.001), there was no significant association between tumour volume and overall survival (HR = 1.34; 95% CI: 0.83, 2.16; p = 0.24) (Fig. 4A).

Tumour diameter. Three articles reported the association between tumour diameter and overall survival in patients with gliomas [33, 40, 44]. There was no significant association between tumour diameter and overall survival (HR = 1.71; 95% CI: 0.21, 13.98; p = 0.62) in the random-effects model (I² = 81%; p = 0.005) (Fig. 4B).

Apparent diffusion coefficient. The association between the apparent diffusion coefficient and overall survival in patients with gliomas was evaluated in three articles [30, 36, 42]. There was no statistically significant association between apparent diffusion coefficient and overall survival (HR = 1.06; 95% CI: 0.50, 2.25; p = 0.88) in the random-effects model (I² = 88%; p = 0.0002) (Fig. 4C).

Multiple lesions. Three studies evaluated the association between multiple lesions observed in MRI and overall survival among patients with gliomas [33, 43, 44]. Meta-analysis revealed no statistically significant association between multiple lesions and overall survival (HR = 1.42; 95% CI: 0.92, 2.19; p = 0.11) in the random-effects model (I² = 0%; p = 0.38) (Fig. 4D).

Tumour margins. The impact of tumour margins evaluated by MRI on the overall survival in patients with gliomas was reported in two articles [31, 41]. In the random-effects model (I² = 93%; p = 0.0002), there was no statistically significant association between tumour margins and overall survival (HR = 2.16; 95% CI: 0.05, 86.13; p = 0.68) (Fig. 4E).

4. Discussion

MRI technology has evolved as a non-invasive tool for diagnosing, evaluating, and predicting treatment responses in patients with gliomas. It overcomes the drawbacks of molecular biomarkers that can only be obtained by invasive procedures such as resection or biopsy. Post-operative MRI has a significant role for symptomatic patients, particularly in determining recurrence and pseudoresponse. MRI may change the treatment strategy and help clinical decision-making for glioma patients [45, 46]. However, conflicting results were reported regarding the value of different MRI findings in predicting PFS and overall survival in patients with gliomas [47, 48]. The present meta-analysis evaluated the association between MRI-derived parameters and PFS, and overall survival time among patients with gliomas. In particular, tumour necrosis visualized with MRI was associated with 1.39 times shorter PFS, and tumour volume was associated with 1.37 times shorter PFS. Peritumour oedema was associated with more than twofold shorter PFS, and multiple lesions revealed 1.57 times shorter PFS. In this respect, peritumoural oedema showed 2.43 times shorter overall survival. Tumour volume and tumour diameter showed a significant association with overall survival, with an HR of 1.4 and 1.37, respectively. The presence of multiple lesions showed shorter overall survival with an HR of 1.42, and tumour margins showed more than twofold shorter overall survival.

The present meta-analysis showed the predictive ability of MRI to diagnose patients with poor survival outcomes. This knowledge helps to stratify patients and potentially provide targets for individualized treatment guidelines. Consistent with these findings, Brancato et al. [49] reported that MRI metrics provided useful information to predict the survival of patients with glioblastoma, particularly if combined with multi-modality imaging properties and clinical factors. Oltra-Sastre et al. [50] reported on the importance of MRI in predicting clinical outcomes in patients with gliomas. Zhou et al. [51] showed the ability of non-invasive imaging modalities to predict treatment response and prognosis in patients with high-grade gliomas before surgery. Tumour necrosis was associated with shorter PFS and overall survival. The presence of tumour necrosis reflects the aggressiveness of the gliomas, and which carry poor prognosis. This necrosis is due to activation of the hypoxia-mediated coagulation system, which results in endothelial proliferation, intravascular thrombosis, and necrosis [52]. Peritumoural oedema was associated with worse survival outcomes, with approximately two times shorter PFS and overall survival than patients without oedema [53]. Contrary to this finding, Liu et al. [54] reported inconclusive results regarding the association between pre-operative peritumoural oedema and overall survival outcomes in patients with gliomas. The discrepancy between their findings and this meta-analysis may be attributable to the number of analyzed articles included in both systematic reviews. Tumour size and diameter were associated with poor survival outcomes, worsening the PFS, and shortening of overall survival by approximately 40%. Tumour necrosis, peritumoural oedema, and tumour size were easy to determine with routine MRI evaluations and could provide instructive information for clinical practice [55]. Another factor that may be associated with survival outcomes, but was not examined in the present review, is glioma-associated macrophages. These have a vital role in the recurrence of glioblastoma microenvironment. Macrophages reflect tumour aggressiveness and overall survival in patients with glioblastoma [56].

The present study quantified the ability of MRI to predict PFS and overall survival in patients with gliomas. However, the results of this study should be cautiously interpreted in the context of some limitations. The main limitation is that MRI technology needs more standardization with different technical modalities and diagnostic approaches for patients with gliomas. This limitation and the difference in study designs, and in type, grade, location, and differentiation of glioma, may result in significant heterogeneity among the analyzed predictors. The majority of the analyzed articles were retrospective designs, conveying a higher risk of information selection and reporting biases. As mentioned, there are other factors that may be associated with survival outcomes in patients with gliomas, but these factors have not been studied sufficiently to subject them to meta-analysis Prospective-cohort studies with strict methods are necessary to mitigate the limitations of the present meta-analysis.

5. Conclusions

The present meta-analysis highlighted the ability of MRI to predict PFS and overall survival in patients with gliomas. Tumour necrosis, peritumoural oedema, or multiple lesions seen in MRI were associated with shorter PFS. There was an association between overall survival and peritumoural oedema, tumour volume, diameter, or multiple lesions. Subgroubing based on the type of glioma revealed no significant association between contrast enhancement, tumour necrosis, and survival outcomes. Identifying such evidence is critical for identifying patients at higher risk of survival outcomes and individualising the treatment plan for the patients.

Availability of Data and Materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

References

[1] Gao X, Liu Y, Hong S, Yang H, Guan B, Ma X. PLEKHA4 is Associated with Tumour Microenvironment, Stemness, Proliferation and Poor Prognosis of Gliomas. Journal of Integrative Neuroscience. 2023; 22: 135.

[2] Mesfin FB, Al-Dhahir MA. Gliomas. StatPearls Publishing: Treasure Island (FL). 2023.

[3] Lopes MBS. The 2017 World Health Organization classification of tumors of the pituitary gland: a summary. Acta Neuropathologica. 2017; 134: 521–535.

[4] Khan MN, Sharma AM, Pitz M, Loewen SK, Quon H, Poulin A, et al. High-grade glioma management and response assessment-recent advances and current challenges. Current Oncology (Toronto, Ont.). 2016; 23: e383–e391.

[5] Kumthekar P, Raizer J, Singh S. Low-grade glioma. Cancer Treatment and Research. 2015; 163: 75–87.

[6] Iwamoto FM, Abrey LE, Beal K, Gutin PH, Rosenblum MK, Reuter VE, et al. Patterns of relapse and prognosis after bevacizumab failure in recurrent glioblastoma. Neurology. 2009; 73: 1200–1206.

[7] Lamborn KR, Yung WKA, Chang SM, Wen PY, Cloughesy TF, DeAngelis LM, et al. Progression-free survival: an important end point in evaluating therapy for recurrent high-grade gliomas. Neuro-oncology. 2008; 10: 162–170.

[8] de Groot JF, Fuller G, Kumar AJ, Piao Y, Eterovic K, Ji Y, et al. Tumor invasion after treatment of glioblastoma with bevacizumab: radiographic and pathologic correlation in humans and mice. Neuro-oncology. 2010; 12: 233–242.

[9] LaViolette PS, Cohen AD, Prah MA, Rand SD, Connelly J, Malkin MG, et al. Vascular change measured with independent component analysis of dynamic susceptibility contrast MRI predicts bevacizumab response in high-grade glioma. Neuro-oncology. 2013; 15: 442–450.

[10] Farid N, Almeida-Freitas DB, White NS, McDonald CR, Muller KA, Vandenberg SR, et al. Restriction-Spectrum Imaging of Bevacizumab-Related Necrosis in a Patient with GBM. Frontiers in Oncology. 2013; 3: 258.

[11] Wang K, Wang Y, Fan X, Wang J, Li G, Ma J, et al. Radiological features combined with IDH1 status for predicting the survival outcome of glioblastoma patients. Neuro-oncology. 2016; 18: 589–597.

[12] Friedman HS, Prados MD, Wen PY, Mikkelsen T, Schiff D, Abrey LE, et al. Bevacizumab alone and in combination with irinotecan in recurrent glioblastoma. Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology. 2009; 27: 4733–4740.

[13] Li ZC, Yan J, Zhang S, Liang C, Lv X, Zou Y, et al. Glioma survival prediction from whole-brain MRI without tumor segmentation using deep attention network: a multicenter study. European Radiology. 2022; 32: 5719–5729.

[14] Cohen-Gadol A. Overview of Survival and Glioma. 2023. Available at: https://www.aaroncohen-gadol.com/en/patients/glioma/survival/overview (Accessed: 3 June 2023).

[15] Mohammed S, Dinesan M, Ajayakumar T. Survival and quality of life analysis in glioblastoma multiforme with adjuvant chemoradiotherapy: a retrospective study. Reports of Practical Oncology and Radiotherapy. 2022; 27: 1026–1036.

[16] Wan B, Wang S, Tu M, Wu B, Han P, Xu H. The diagnostic performance of perfusion MRI for differentiating glioma recurrence from pseudoprogression: A meta-analysis. Medicine. 2017; 96: e6333.

[17] van Santwijk L, Kouwenberg V, Meijer F, Smits M, Henssen D. A systematic review and meta-analysis on the differentiation of glioma grade and mutational status by use of perfusion-based magnetic resonance imaging. Insights into Imaging. 2022; 13: 102.

[18] Dongas J, Asahina AT, Bacchi S, Patel S. Magnetic resonance perfusion imaging in the diagnosis of high-grade glioma progression and treatment-related changes: a systematic review. Open Journal of Modern Neurosurgery. 2018; 8: 282.

[19] Poon MTC, Sudlow CLM, Figueroa JD, Brennan PM. Longer-term (≥2 years) survival in patients with glioblastoma in population-based studies pre- and post-2005: a systematic review and meta-analysis. Scientific Reports. 2020; 10: 11622.

[20] Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al. Cochrane handbook for systematic reviews of interventions version 6.2 (updated February 2021). Cochrane. 2021. Available at: https://training.cochrane.org/handbook (Accessed: 3 June 2021).

[21] Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA Statement. Open Medicine: a Peer-reviewed, Independent, Open-access Journal. 2009; 3: e123–e130.

[22] National Heart, Lung, and Blood Institute. Study Quality Assessment Tools. 2021. Available at: https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools (Accessed: 3 July 2022).

[23] Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ (Clinical Research Ed.). 2003; 327: 557–560.

[24] Borenstein M. Comprehensive meta‐analysis software. Systematic reviews in health research: meta‐analysis in context (pp. 535–548). John Wiley & Sons Ltd.: New Jersey, USA. 2022.

[25] Burth S, Kickingereder P, Eidel O, Tichy D, Bonekamp D, Weberling L, et al. Clinical parameters outweigh diffusion- and perfusion-derived MRI parameters in predicting survival in newly diagnosed glioblastoma. Neuro-oncology. 2016; 18: 1673–1679.

[26] Chen C, Huang R, MacLean A, Muzikansky A, Mukundan S, Wen PY, et al. Recurrent high-grade glioma treated with bevacizumab: prognostic value of MGMT methylation, EGFR status and pretreatment MRI in determining response and survival. Journal of Neuro-oncology. 2013; 115: 267–276.

[27] Choi YS, Kim DW, Lee SK, Chang JH, Kang SG, Kim EH, et al. The Added Prognostic Value of Preoperative Dynamic Contrast-Enhanced MRI Histogram Analysis in Patients with Glioblastoma: Analysis of Overall and Progression-Free Survival. AJNR. American Journal of Neuroradiology. 2015; 36: 2235–2241.

[28] Gupta T, Tibdewal A, Mohanty S, Pietsch T, Kannan S, Juvekar S, et al. Correlation of conventional magnetic resonance imaging features with O6-methylguanine-DNA-methyltransferase gene promoter methylation status and survival outcomes in patients with newly diagnosed glioblastoma: Single-center correlative imaging substudy from a prospective clinical trial. Glioma. 2018; 1: 50–58.

[29] Hong EK, Choi SH, Shin DJ, Jo SW, Yoo RE, Kang KM, et al. Comparison of Genetic Profiles and Prognosis of High-Grade Gliomas Using Quantitative and Qualitative MRI Features: A Focus on G3 Gliomas. Korean Journal of Radiology. 2021; 22: 233–242.

[30] Hou BL, Wen S, Katsevman GA, Liu H, Urhie O, Turner RC, et al. Magnetic Resonance Imaging Parameters and Their Impact on Survival of Patients with Glioblastoma: Tumor Perfusion Predicts Survival. World Neurosurgery. 2018; S1878-S1878-8750(18)32908-5.

[31] Lasocki A, Buckland ME, Molinaro T, Xie J, Whittle JR, Wei H, et al. Correlating MRI features with additional genetic markers and patient survival in histological grade 2-3 IDH-mutant astrocytomas. Neuroradiology. 2023; 65: 1215–1223.

[32] Li C, Yan JL, Torheim T, McLean MA, Boonzaier NR, Zou J, et al. Low perfusion compartments in glioblastoma quantified by advanced magnetic resonance imaging and correlated with patient survival. Radiotherapy and Oncology. 2019; 134: 17–24.

[33] Liu H, Shen L, Huang X, Zhang G. Maximal tumor diameter in the preoperative tumor magnetic resonance imaging (MRI) T2 image is associated with prognosis of Grade II Glioma. Medicine. 2021; 100: e24850.

[34] Liu S, Wang Y, Fan X, Ma J, Qiu X, Jiang T. Association of MRI-classified subventricular regions with survival outcomes in patients with anaplastic glioma. Clinical Radiology. 2017; 72: 426.e1–426.e6.

[35] Liu S, Wang Y, Xu K, Wang Z, Fan X, Zhang C, et al. Relationship between necrotic patterns in glioblastoma and patient survival: fractal dimension and lacunarity analyses using magnetic resonance imaging. Scientific Reports. 2017; 7: 8302.

[36] Nakamura H, Murakami R, Hirai T, Kitajima M, Yamashita Y. Can MRI-derived factors predict the survival in glioblastoma patients treated with postoperative chemoradiation therapy? Acta Radiologica. 2013; 54: 214–220.

[37] Patel CB, Fazzari E, Chakhoyan A, Yao J, Raymond C, Nguyen H, et al. ¹⁸F-FDOPA PET and MRI characteristics correlate with degree of malignancy and predict survival in treatment-naïve gliomas: a cross-sectional study. Journal of Neuro-oncology. 2018; 139: 399–409.

[38] Pérez-Beteta J, Molina-García D, Martínez-González A, Henares-Molina A, Amo-Salas M, Luque B, et al. Morphological MRI-based features provide pretreatment survival prediction in glioblastoma. European Radiology. 2019; 29: 1968–1977.

[39] Sacli-Bilmez B, Danyeli AE, Yakicier MC, Aras FK, Pamir MN, Özduman K, et al. Magnetic resonance spectroscopic correlates of progression free and overall survival in “glioblastoma, IDH-wildtype, WHO grade-4”. Frontiers in Neuroscience. 2023; 17: 1149292.

[40] Schwarzenberg J, Czernin J, Cloughesy TF, Ellingson BM, Pope WB, Geist C, et al. 3’-deoxy-3’-18F-fluorothymidine PET and MRI for early survival predictions in patients with recurrent malignant glioma treated with bevacizumab. Journal of Nuclear Medicine. 2012; 53: 29–36.

[41] Shin I, Park YW, Ahn SS, Kim J, Chang JH, Kim SH, et al. Clinical factors and conventional MRI may independently predict progression-free survival and overall survival in adult pilocytic astrocytomas. Neuroradiology. 2022; 64: 1529–1537.

[42] Tatekawa H, Yao J, Oughourlian TC, Hagiwara A, Wang C, Raymond C, et al. Maximum Uptake and Hypermetabolic Volume of 18F-FDOPA PET Estimate Molecular Status and Overall Survival in Low-Grade Gliomas: A PET and MRI Study. Clinical Nuclear Medicine. 2020; 45: e505–e511.

[43] Young GS, Macklin EA, Setayesh K, Lawson JD, Wen PY, Norden AD, et al. Longitudinal MRI evidence for decreased survival among periventricular glioblastoma. Journal of Neuro-oncology. 2011; 104: 261–269.

[44] Zhao K, Liu R, Li Z, Liu M, Zhao Y, Xue Z, et al. The imaging features and prognosis of gliomas involving the subventricular zone: An MRI study. Clinical Neurology and Neurosurgery. 2022; 222: 107465.

[45] Pasqualetti F, Malfatti G, Cantarella M, Gonnelli A, Montrone S, Montemurro N, et al. Role of magnetic resonance imaging following postoperative radiotherapy in clinical decision-making of patients with high-grade glioma. La Radiologia Medica. 2022; 127: 803–808.

[46] Chaskis C, Neyns B, Michotte A, De Ridder M, Everaert H. Pseudoprogression after radiotherapy with concurrent temozolomide for high-grade glioma: clinical observations and working recommendations. Surgical Neurology. 2009; 72: 423–428.

[47] Law M, Young RJ, Babb JS, Peccerelli N, Chheang S, Gruber ML, et al. Gliomas: predicting time to progression or survival with cerebral blood volume measurements at dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging. Radiology. 2008; 247: 490–498.

[48] Tykocki T, Eltayeb M. Ten-year survival in glioblastoma. A systematic review. Journal of Clinical Neuroscience: Official Journal of the Neurosurgical Society of Australasia. 2018; 54: 7–13.

[49] Brancato V, Nuzzo S, Tramontano L, Condorelli G, Salvatore M, Cavaliere C. Predicting Survival in Glioblastoma Patients Using Diffusion MR Imaging Metrics-A Systematic Review. Cancers. 2020; 12: 2858.

[50] Oltra-Sastre M, Fuster-Garcia E, Juan-Albarracin J, Sáez C, Perez-Girbes A, Sanz-Requena R, et al. Multi-parametric MR Imaging Biomarkers Associated to Clinical Outcomes in Gliomas: A Systematic Review. Current Medical Imaging Reviews. 2019; 15: 933–947.

[51] Zhou Q, Xue C, Ke X, Zhou J. Treatment Response and Prognosis Evaluation in High-Grade Glioma: An Imaging Review Based on MRI. Journal of Magnetic Resonance Imaging. 2022; 56: 325–340.

[52] Homma T, Fukushima T, Vaccarella S, Yonekawa Y, Di Patre PL, Franceschi S, et al. Correlation among pathology, genotype, and patient outcomes in glioblastoma. Journal of Neuropathology and Experimental Neurology. 2006; 65: 846–854.

[53] Wu CX, Lin GS, Lin ZX, Zhang JD, Liu SY, Zhou CF. Peritumoral edema shown by MRI predicts poor clinical outcome in glioblastoma. World Journal of Surgical Oncology. 2015; 13: 97.

[54] Liu SY, Mei WZ, Lin ZX. Pre-operative peritumoral edema and survival rate in glioblastoma multiforme. Onkologie. 2013; 36: 679–684.

[55] Liang J, Lv X, Lu C, Ye X, Chen X, Fu J, et al. Prognostic factors of patients with Gliomas - an analysis on 335 patients with Glioblastoma and other forms of Gliomas. BMC Cancer. 2020; 20: 35.

[56] Montemurro N, Pahwa B, Tayal A, Shukla A, De Jesus Encarnacion M, Ramirez I, et al. Macrophages in Recurrent Glioblastoma as a Prognostic Factor in the Synergistic System of the Tumor Microenvironment. Neurology International. 2023; 15: 595–608.

Funding

Science and Technology Bureau Program of Huzhou(2021GBY46)