Non-prognostic features (NPF)
| NPF features | Non-prognostic evidence base | ||||||
| Feature | Population(s) or Subgroup(s) | Comments | Individual patients* | Individual studies* | Meta-analytical references | Publication years (first, last) | GRADE score |
| Clinical features | |||||||
| Sex: male vs female | Adults and children with FLE, TLE, ET, tuberous sclerosis, MRI neg TLE, repeat surgery, hemispherectomies, low grade gliomas | All were non-significant, a large proportion even on weighted univariate tests, which otherwise tend to overestimate significance. Individual unweighted effect sizes ranged from OR 0.83 (0.42, 1.64)c in repeat surgery for focal DRE14 to OR 1.44 (0.86, 2.41) in MRI negative TLE.25 | 5974 | 148 | Englot, Wang; Englot, Rolston; Zhang, Hu; Fallah, Guyatt; Ibrahim, Morgan; Wang, Zhang; Krucoff, Chan; Englot, Breshears; Hu, Zhang; Shan, Fan; Cao, Liu | 2012–2018 | +++ Moderate |
| Epilepsia partialis continua (EPC) | Children undergoing hemispherectomies | Not significant on unweighted univariate testing and result is from only one meta-analysis. | 127 | 7 | Cao, Liu | 2016 | ++ Low |
| Imaging features | |||||||
| Number of cortical tubers | Tuberous sclerosis | Numbers of tubers did not predict outcomes. See online supplemental table 5 for observations on methodology, sample sizes, adjustments and heterogeneity. | 286 | 24 | Zhang, Hu; Fallah, Guyatt; Ibrahim, Morgan | 2013–2015 | ++ Low |
| Magnetic 1H spectroscopy | TLE adults and children | Probably no more valuable than conventional MRI abnormality (online supplemental table 5). | 121 | 22 | Willmann, Wennberg | 2006 | + Very low |
| Encephalomalacia | Adults and children | Encephalomalacia was NS in the Cochrane meta-analysis, it was also not significant on subgroup analyses.22 | 317 | 5 | West, Nevitt | 2019 | + Very Low |
| Enhancement, oedema, and/or mass effect | Low grade gliomas in adults | These combined features are not clinically prognostic of low-grade glioma resection for seizure freedom. Although NS, the point estimate and CI are unavailable. | 2641 | 23 | Shan, Fan | 2018 | + Very low |
| Vascular lesions | Adults and children with TL and ET | Only one meta-analysis investigated this in 2004, comprising only three individual studies, its pathological counterpart was also NS.22 | <<3511 | 3 | Tonini, Beghi | 2004 | + Very low |
| Neurophysiological features | |||||||
| Intraoperative invasive EEG | Children and adults with FLE | Electrico-corticography did not effect outcomes | 1024 | 21 | Englot, Wang | 2012 | +++ Moderate |
| Video telemetry and long-term monitoring | Children and adults with FLE, lesional and non-lesional TLE and ET | Lesional TLE cases do well, and this was the only subgroup in which long-term monitoring had a point effect size estimate greater than 1. | 1738 | 65 | Englot, Wang; Kobulashvili, Kuchukhidze | 2012, 2018 | + Very low |
| Surgical technique or anatomic features | |||||||
| Mesial vs lateral TL focus | MRI neg TLE | Mesial or lateral TLE, as determined by sEEG, subdural grids, or ATL/SAH vs neocortectomy, are not significant. | 92 | 8 | Wang, Zhang | 2016 | + Very low |
| Side of resection (left vs right) | Children and adults, Non-lesional, TLE, FLE, ET, MRI negative TLE, repeat surgery, hemispherectomies | This feature is unlikely to be prognostic—see online supplemental table 5. | 6550 | 188 | Tonini, Beghi; West, Nevitt; Willmann, Wennberg; Ansari, Tubbs; Englot, Wang Englot; Rolston; Wang, Zhang; Krucoff, Chan; Englot, Breshears; Hu, Zhang; Ansari, Maher | 2004–2019 | +++ Moderate |
| Frontal, central, or posterior resections vs other | ET, adults, non-lesional | Not prognostic | 81 | ? | Ansari, Tubbs | 2010 | + Very low |
| Geographical location of surgery | Tuberous sclerosis in children | Only one meta-analysis investigated North America vs Elsewhere, and the GRADE score is from this meta-analysis alone. | 186 | 20 | Ibrahim, Morgan | 2015 | +++ Moderate |
| Pathological features | |||||||
| Neuro-migrational defects | TL and ET children and adults | There was a trend whereby neuromigrational deficits were negative prognostic factors, but the number of participants in this analysis was unclear. | ? | 6 | Tonini, Beghi | 2004 | + Very Low |
| Astrocytoma vs non-astrocytoma | Low grade gliomas in adults | The exact numbers of patients were not provided for this particular analysis. | <2641 | <23 | Shan, Fan | 2018 | + Very Low |
See online supplemental table 5 for more details and online supplemental materials for full list of references.
*Upper bound of estimate, not including subgroup analyses.
m, multivariate; u, univariate; c, calculated (usually unweighted) effect size; ET, extratemporal; FLE, frontal lobe epilepsy; MCD, malformations of cortical development; NS, not significant; PLS, projection to latent space; TL, temporal lobe.