Role of magnetoencephalography in predicting the epileptogenic zone and post-operative seizure outcome - A retrospective study

•One of the largest cohorts of operated cases exploring the role MEG in prediction of seizure outcome.•Concordance of MEG with resection cavity was associated with good post-operative outcomes with a sensitivity of 83 %.•The sensitivity of MEG ranged from 83 to 70 from 6 months to 5 years follow up, whereas specificity was 68 % & 40 %.•Level 1 concordance was associated with a good outcome in 78% vs 66 % in temporal vs extratemporal at 6 months follow-up.•No definite association was noted between MEG cluster characteristics and outcome except cluster strength. Study assessed the role of MSI in predicting the post-operative seizure outcome. This retrospective study included patients who underwent MEG and epilepsy surgery and had a minimum 6 months of postoperative follow-up. Concordance of MEG cluster with post-surgical resection cavity was classified as follows Class I) Concordant and region-specific, Class II) Concordant and region non-specific, Class III) Concordant lateralization only and Class IV) Discordant lateralization. The relationship between MSI concordance and post-operative seizure outcome was assessed. A total of 183 patients (M: F = 109:74) were included. The mean age at onset of seizures: 8.0 ± 6.4 years. The dipoles were frequent in 123(67.2 %). The primary cluster orientation was regular in 59 (32.2 %) and mixed in 124 (67.8 %) patients. Concordance between MEG and resection cavity: Class I - 124 (67.8 %), class II- 30 (16.4 %), class III- 23 (12.6 %), and class IV- 6 (3.3 %). The post-surgically mean duration of follow-up was 19.52 ± 11.27 months. At 6-month follow-up period, 144 (78.7 %) patients had complete seizure freedom out of which 106 (73.6 %) had class I concordance. Concordance of MEG with resection cavity was associated with a good outcome at 6 months (p = 0.001), 1 year (p = 0.001), 2 years (p = 0.0005) and 5 years (p = 0.04). MEG cluster characteristics had no association with seizure outcome except the strength of the cluster and outcome at 3 years (p = 0.02) follow-up. The study supports that the complete resection of the MEG cluster had high chance of seizure-freedom and can be used as a complementary noninvasive presurgical evaluation tool.