For the millions of adults living with chronic knee pain unresponsive to physical therapy, injections, or anti-inflammatories, a minimally invasive vascular procedure has been gaining traction — yet clinical practice has remained fragmented. A formal expert consensus now offers the field its clearest roadmap to date, standardizing who qualifies for genicular artery embolization (GAE), how it should be performed, and how outcomes should be tracked.
Using a three-round Delphi process, 29 interventional and musculoskeletal radiologists from diverse geographic backgrounds evaluated 75 clinical statements, with consensus defined as ≥75% of panelists rating a statement ≥7 out of 10. Key agreed-upon inclusion criteria include knee pain refractory to conservative treatment for at least three months, with qualifying diagnoses spanning osteoarthritis, tendinopathies, post-surgical pain, and recurrent hemarthrosis. Pre-procedure workup should include standardized outcome measures, physical examination, and plain radiographs; contrast-enhanced MRI was endorsed as optional, primarily for synovitis grading and phenotyping. On the technical side, ipsilateral antegrade transfemoral access emerged as the preferred approach, with embolization of all visible genicular arteries showing hypervascular blush. Notably, panelists found no evidence favoring either temporary or permanent embolic agents over the other in safety or efficacy — a point that leaves room for operator discretion.
GAE targets pathological neovascularization supplying inflamed synovial tissue, a mechanism distinct from surgical debridement or joint replacement. The procedure has been evaluated in multiple small randomized controlled trials showing meaningful pain reduction in mild-to-moderate knee osteoarthritis, but study heterogeneity has complicated cross-trial comparisons. This Delphi consensus addresses that gap by harmonizing patient selection and technical parameters — a prerequisite for designing the larger, more definitive trials the field still needs. Importantly, consensus documents carry inherent limitations: they reflect expert opinion, not randomized evidence, and the 29-panelist size, while appropriate for Delphi methodology, constrains generalizability. Still, for clinicians managing treatment-refractory knee pain in patients not yet candidates for arthroplasty, this framework represents a meaningful step toward consistent, evidence-informed practice.