Headache Outcomes After Flow Diversion or Coiling of Unruptured Intracranial Aneurysms: A Systematic Review and Meta-analysis.
Abstract
Background
Headache often prompts the imaging that reveals an unruptured intracranial aneurysm (UIA), but how endovascular treatment affects headache burden remains poorly defined. We pooled the incidence of de-novo and post-procedural headache and the rates of improvement and worsening, with treatment modality analyzed as an exploratory, across-study subgroup.
Methods
PubMed, Scopus, and Web of Science were searched for studies reporting post-treatment headache outcomes after flow diversion or coiling of UIAs; studies that could not separate the endovascular arm from clipping, or unruptured from ruptured aneurysms, were excluded. Proportions were pooled with random-effects generalized linear mixed models on the logit scale, with 95% confidence intervals (CIs), 95% prediction intervals (PIs), I2, and modality subgroup tests. Post hoc subgroup, meta-regression, and sensitivity analyses examined ascertainment window, ascertainment method, design, and influential cohorts.
Results
Nineteen studies (2142 patients; eight flow-diverter and 14 coiling cohorts) were included. Among baseline headache-free patients, de-novo headache occurred in 30.2% (95% CI 19.4-43.6; I2=80%; k=9). New or post-procedural headache across the entire treated cohort was 14.4% (95% CI, 8.0-24.6; k=14). Among patients with pre-existing headache, 71.4% improved (95% CI, 58.8-81.4; k=11), and 5.6% worsened (95% CI, 2.2-13.9; k=6). No statistically significant between-subgroup differences were detected (de-novo p=0.42; post-procedural p=0.43; improvement p=0.29); however, this comparison was limited by few flow-diverter cohorts and predominantly indirect comparisons. De-novo headache was more frequent when ascertained within six weeks than when persistence beyond three months was required (36.7% versus 22.6%; meta-regression on follow-up duration p=0.025).
Conclusions
About one in three previously headache-free patients develops a new headache after endovascular treatment of a UIA, whereas roughly seven in ten with pre-existing headache improve, and worsening is uncommon. Estimates varied widely, partly with the timing of headache ascertainment; the available data are insufficient to establish whether flow diverters and coils differ.