Introduction
Materials and methods
Development
Inclusion and exclusion criteria
Information sources, search strategy and study selection
Data collection process and data items
Development of grading system
Validation
Setting and participants
Variables and data sources
Study size and statistical methods
Results
Development
Study selection
Study characteristics
First author (year) | Study design | Patients characteristics | Pre-operative MRI features | Surgical outcomes | Findings |
---|---|---|---|---|---|
[1] | Case-control study |
n = 100 with GBM; 50 pts with long OS, 50 pts with short OS. | Multifocal disease Laterality Periventricular (subventricular zone) location | OS | OS reduced if both hemispheres involved (p = 0.074, univariate) OS reduced with periventricular location (p = 0.038, univariate) |
[4] | Retrospective cohort study |
n = 54 with GBM | Volume Periventricular location | EOR | EOR reduced with increased volume (p < 0.05, univariate) EOR reduced with periventricular location (p < 0.05, univariate) |
[6] | Retrospective cohort study |
n = 52 with GBM; 26 pts with periventricular location, 26 pts with other location. | Diameter Periventricular location | OS | OS reduced with periventricular location (p = 0.02, univariate) |
[7] | Retrospective cohort study to construct and validate scoring system |
n = 393 | Brain lobe Diameter Periventricular location Eloquent location | OS | OS reduced with periventricular location (p = 0.04, multivariate) |
[8] | Retrospective cohort study |
n = 129 with GBM; aged >65 years. | Brain lobe Diameter Eloquent location | OS | OS reduced with increased diameter (p = 0.002, multivariate) |
[9] | Multicentre retrospective cohort study to validate scoring system |
n = 334 with GBM | Periventricular location | OS | OS reduced with periventricular location (p = 0.04, multivariate) |
[10] | Retrospective cohort study |
n = 100 with GBM; KPS <60. | Diameter Brain lobe Periventricular location | OS | OS increased with increased diameter > 2 cm (p = 0.01, multivariate) |
[11] | Retrospective cohort study |
n = 233 with glioma; 121 with GBM. | Corpus callosum involvement | EOR PFS OS | EOR, PFS and OS reduced with corpus callosum involvement (p < 0.001, p = 0.006 and p = 0.011 respectively; multivariate) |
[16] | Retrospective cohort study |
n = 46 with GBM | Brain lobe Diameter | OS | OS reduced with temporal location (p = 0.0130, multivariate) OS reduced with increased diameter (p = 0.0395, multivariate) |
[21] | Retrospective cohort study |
n = 48 with GBM | Volume Necrosis Contrast enhancement Associated oedema | OS | OS reduced with necrosis (p < 0.001, multivariate) OS reduced with enhancement (p = 0.003, multivariate) OS reduced with oedema (p < 0.004, multivariate) |
[22] | Retrospective series |
n = 516 with GBM | Laterality | OS | OS reduced if both hemispheres involved (p < 0.01, multivariate) |
[23] | Retrospective cohort study |
n = 45 with GBM | NER:CER Area of NER Definition of margins Laterality Deep white matter involvement Associated oedema | OS PFS | NER crossing the midline (OS p = 0.0125, PFS p = 0.0661, multivariate). |
[25] | Prospective cohort study |
n = 80 with HGG; aged 60–83 years. | Volume Eloquent location | OS | No significant associations found. |
[27] | Retrospective cohort study |
n = 420 with GBM | Necrosis Contrast enhancement Deep location Eloquent location Associated oedema Midline shift | OS | OS reduced with necrosis (p = 0.01, multivariate) OS reduced with contrast enhancement (p = 0.02, multivariate) OS reduced with eloquent location (p = 0.02, univariate) OS reduced with oedema (p = 0.04, univariate) |
[28] | Retrospective cohort study |
n = 116 with GBM | Laterality Brain lobe Diameter Cysts Contrast enhancement Deep or eloquent location Associated oedema Midline shift | PFS OS | PFS was dependent upon brain lobe (p = 0.001, multivariate) OS reduced with deep or eloquent location (p = 0.009, multivariate) |
[30] | Retrospective cohort study |
n = 205 with GBM | Deep or eloquent location | OS | OS reduced with deep or eloquent location (p < 0.0001, multivariate) |
[33] | Multicentre retrospective cohort study to construct and validate scoring system |
n = 143 with recurrent GBM; 34 pts to devise scoring system and 109 pts to validate. | Brain lobe Laterality Volume Eloquent location | OS | OS reduced with greater volume (p < 0.001, univariate) OS reduced with eloquent location (p < 0.001, univariate) |
[35] | Multicentre retrospective cohort study |
n = 110 with GBM | Diameter Associated oedema | OS | OS reduced with associated oedema (p = 0.006, multivariate) |
[39] | Retrospective cohort study (using prospectively collected trial data) |
n = 243 with GBM; 121 pts received 5-ALA, 122 pts did not. | Volume Hemisphere Periventricular location Eloquent location Associated oedema Midline shift | EOR OS | EOR reduced with eloquent location (p = 0.0231, univariate) |
[41] | Retrospective cohort study |
n = 65 with GBM; received 5-ALA. | Brain lobe Volume Periventricular location | OS | OS reduced with periventricular location (p = 0.008, multivariate) |
Grading system
Pre-operative MRI feature | Number of studies that analysed | Number of studies that found feature predictive of outcome |
---|---|---|
Periventricular or deep locationa
| 12 | 8 |
Eloquent location | 8 | 5 |
Sizeb
| 13 | 5 |
Associated oedema | 6 | 3 |
Corpus callosum involvement or bilateral locationc
| 6 | 2 |
Pre-operative MRI feature | Score |
---|---|
Periventricular or deep location | |
≥10 mm from ventricle | 0 |
<10 mm from ventricle | 1 |
Corpus callosum or bilateral location | |
No corpus callosum involvement | 0 |
Corpus callosum involvement or bilateral location | 1 |
Eloquent location | |
Not eloquent location | 0 |
Eloquent location (motor or sensory cortex, language cortex, insula or basal ganglia) | 1 |
Largest diameter of tumour (mm) | |
<40 | 0 |
≥40 | 1 |
Associated oedema | |
<10 mm from contrast-enhancing tumour | 0 |
≥10 mm from contrast-enhancing tumour | 1 |
Total | 0–5 0–1 Low complexity 2–3 Moderate complexity 4–5 High complexity |
Validation
Participants and descriptive data
Sex | |
- male/female | 1.9:1 |
Age median | |
- (interquartile range) | 59 years (46–70) |
GBM | |
- primary: recurrent | 7:1 |
Outcome data and main results
Pre-operative MRI feature | Number |
---|---|
Periventricular or deep location | |
≥10 mm from ventricle | 20 (22.7%) |
<10 mm from ventricle | 68 (77.3%) |
Corpus callosum or bilateral location | |
No corpus callosum involvement | 59 (67.0%) |
Corpus callosum involvement or bilateral location | 29 (33.0%) |
Eloquent location | |
Not eloquent location | 50 (56.8%) |
Eloquent location (motor or sensory cortex, language cortex, insula or basal ganglia) | 38 (43.2%) |
Largest diameter of tumour (mm) | |
<40 | 26 (29.5%) |
≥40 | 62 (70.5%) |
Associated oedema | |
<10 mm from contrast-enhancing tumour | 27 (30.7%) |
≥10 mm from contrast-enhancing tumour | 61 (69.3%) |
Total | 88 |
Grade | Number | Complete resection of contrast-enhancing tumour | Major complications |
---|---|---|---|
Low complexity(0–1 points) | 14 | 7 (50.0%) | 0 (0%) |
Moderate complexity(2–3 points) | 45 | 7 (15.6%) | 1 (2.2%) |
High complexity(4–5 points) | 29 | 1 (3.4%) | 2 (6.9%) |
Total | 88 | 15 (17.0%) | 3 (3.4%) |