Neurochirurgiczne leczenie epilepsji po wstępnym zabiegu termokoagulacji radioczęstościowej: Wyniki i znaczenie prognostyczne odpowiedzi na leczenie - przegląd systematyczny i metaanaliza

Stereoelectroencephalography-guided radiofrequency thermocoagulation (SEEG-RFTC) is used not only as a minimally invasive treatment for drug-resistant focal epilepsy, but also as an in vivo test of an SEEG-defined epileptogenic network. Whether response to RFTC is associated with the outcome of subsequent surgery remains uncertain. We evaluated seizure outcomes after later resective, disconnective, or ablative surgery and the association between RFTC response and later surgical outcome. PubMed/MEDLINE, Embase, Scopus, and Web of Science were searched from inception to June 1, 2026. Eligible studies included patients treated with SEEG-RFTC and reporting therapeutic, safety, downstream-treatment, or postsurgical outcomes. Single-arm outcomes were pooled using random-effects logit models. Sparse permanent-complication data were analyzed with a random-intercept logistic model, and responder-versus-nonresponder comparisons with random-effects odds ratios; Hartung-Knapp inference was examined as a small-sample sensitivity analysis. Risk of bias was assessed using Joanna Briggs Institute tools, with additional Quality In Prognosis Studies assessment of the three studies contributing to the prognostic meta-analysis. Twenty-nine original clinical studies met the eligibility criteria. Six studies comprising 194 patients reported extractable outcomes after qualifying subsequent resective, disconnective, or ablative epilepsy surgery; 141 patients achieved a favorable seizure outcome. The pooled favorable-outcome proportion was 72.4% (95% confidence interval [CI], 65.6%-78.3%; I = 0%). Three studies directly compared responders with nonresponders. Responders had numerically higher odds of favorable later outcome under the conventional random-effects model (odds ratio, 3.86; 95% CI, 0.80-18.57; I = 47.1%), but Hartung-Knapp small-sample inference substantially widened the uncertainty (odds ratio, 3.86; 95% CI, 0.19-77.91; P = .193). The pooled proportion proceeding to another epilepsy procedure was 22.0% (95% CI, 11.1%-38.8%), with substantial heterogeneity across clinical pathways (I = 84.1%), while permanent neurological complications occurred in 2.7% (95% CI, 1.0%-6.9%). Approximately seven of 10 selected patients had a favorable seizure outcome after subsequent resective, disconnective, or ablative epilepsy surgery following SEEG-RFTC. A positive RFTC response may strengthen confidence in the surgical hypothesis, but current evidence does not establish independent prognostic value. Current evidence does not support using RFTC nonresponse alone to exclude an otherwise appropriate candidate from definitive epilepsy surgery. SEEG-RFTC may therefore serve as a decision-informing component of network-based surgical evaluation rather than solely as a focal ablative treatment.