Epilepsia | 2021
Response: “Low statistical power in a study predicting seizure outcome”
Abstract
Dear EditorsinChief of Epilepsia, We thank Drs. Dickey and Pederson for their interest in our article and raising important points for clarification. Magnetic resonanceguided laser interstitial thermal therapy (MRgLiTT) is a novel procedure, and there is a current gap in knowledge of presurgical prognostic factors that are associated with surgical outcome. It is important to note that MRgLiTT is a more selective procedure than anterior temporal lobectomy (ATL), and our results demonstrate that prognostic factors for MRgLiTT may not be the same as for ATL. Therefore, we are cautious about extrapolating seizurefreedom rates from ATL data and applying them to MRgLiTT. We agree that findings from small studies should be interpreted cautiously, including the possibility of overestimation of effect size. Metaanalysis provides the best estimate of the true effect size, albeit with limitation that the individual studies that contribute to them are subject to low power. Despite this reduced chance of detecting a true effect, our study demonstrates that the presence of bilateral interictal epileptiform discharges (IEDs) is associated with poor outcome, with adjusted odds ratio of 0.05 with 95% confidence interval of 0.01– 0.46. This adjusted confidence interval includes the pre hoc unadjusted odds ratio of 0.4 that the authors derived from ATL studies. The presence of bilateral IEDs is not a contraindication for MRgLiTT, and we agree that their presence may prompt the clinician to be more cautious in interpreting the presurgical evaluation results. Further studies are needed and necessary to improve surgical outcomes and our findings remain to be replicated in larger cohorts.