The purpose of this clinical trial is to determine whether genicular artery embolization (GAE) is an effective way to treat knee pain from osteoarthritis. Pain from arthritis is often due to underlying inflammation in the joint. The inflammation is associated with increased abnormal blood flow going to the specific area of pain. If the investigator can reduce the blood flow, the inflammation can be reduced and the pain can be controlled.

The GAE procedure is an experimental procedure to decrease the blood flow (embolize) to the specific region of the knee that is causing the pain. This will be done by infusing microscopic spheres into the specific blood vessel (genicular artery) supplying the area of pain in the knee. This is done through a procedure which is called an angiogram, which is done entirely through a pinhole at the creased of the thigh, using twilight (conscious) sedation.

The investigators have already completed an initial trial at UCLA, and shown that this procedure is safe and effective. The purpose of this new trial is to compare outcomes of people undergoing the GAE procedure to those who do not undergo the procedure. A total of 120 patients will be enrolled, and 2/3 of the patients will be randomly selected to undergo the GAE procedure. 1/3 will not undergo the procedure. This is known as a randomized trial. During the trial, all subjects will also get MRIs and fluid withdrawn from their knee joints at various timepoints in order to precisely figure out how the procedure works on a closer (microscopic) level.

This is a single-center, prospective, open-label randomized trial to compare outcomes of GAE versus observation, and to identify biomarkers and imaging endpoints that change in response to GAE.

Subjects will be considered enrolled in the study once they have provided informed consent and have been determined to meet all eligibility criteria. A total of 100 subjects will be enrolled in the double-arm study and will be followed for 24 months.

The study will involve a screening period in which patient eligibility is determined. Once eligibility is confirmed, subjects will then be randomized to GAE or observation in a 2:1 ratio. Group sample sizes of 80 and 40 achieve approximately 85% power to reject the null hypothesis of no difference in the clinical response defined by 50% or greater reduction in WOMAC between the treatment and observation arms with a significance level (alpha) of 0.050 using a two-sided two-proportions using Fisher's exact test. With 120 symptomatic knee osteoarthritis subjects, the power of study is approximately 85% where we assume a target response rate of 67.5% for the experimental arm and 31% for the observational arm with up to 20% drop-out rate prior to 6 months. All study subjects will undergo the initial procedures, consisting of a history and physical exam, dynamic contrast-enhanced knee MRI, blood serum and joint aspiration with biochemical analysis.

For the treatment arm, subjects will undergo GAE with Embozene microspheres (100 micron) (Varian Medical Systems). Observation group will not receive the GAE procedure at baseline, and will be counseled on conservative management with a referral or prescription for physical therapy. Follow-up visits will be at 1 month (± 2 weeks), 3 months (± 2 weeks), 6 months (± 2 weeks), 12 months (± 2 weeks), and 24 months (± 2 weeks) post-procedure. At these visits, subjects will complete the WOMAC, KOOS5, visual analog scale (VAS) pain score, undergo a directed physical examination, and report any new adverse events (AEs). At 6 and 24 months of follow-up, all subjects will also undergo dynamic contrast-enhanced knee MRI, and serum and joint aspiration with biochemical analysis.

At 6 months, subjects in the observation group will have the option to crossover to the GAE treatment arm. Subjects in the crossover group will then undergo GAE, with follow up at 1 month (± 2 weeks), 3 months (± 2 weeks), 6 months (± 2 weeks), 12 months (± 2 weeks), and 24 months (± 2 weeks) post-GAE. For the purposes of data analyses, the follow-up evaluations for crossover subjects will match those of the initial GAE subjects. For example, if a subject in the observation arm undergoes GAE crossover at 6 months, that will be considered the treatment visit, and then subject will undergo evaluation at 1, 3, 6, 12, and 24 months after GAE.

Inclusion Criteria

Subjects must meet all of the following inclusion criteria in order to be eligible for this study:

• Provided informed consent
• Age ≥ 40 years and less than 80 years
• Ineligibility for or refusal of surgical management
• Moderate-severe knee pain as determined by visual analog scale > 5 out of 10
• Osteoarthritis based on X-ray. Kellgren-Lawrence score > 2 based on radiograph completed within 3 months of procedure date.
• Resistant/failed conservative treatment (e.g. NSAIDS/physical therapy/steroid joint injection/ hyaluronic acid joint injection) for at least 3 months
• Able to comply with all treatments and protocol follow-up visits

Exclusion Criteria

Subjects that meet any of the following exclusion criteria will not be eligible for this study:

• Mild knee pain as determined by visual analog scale < 5 out of 10
• OA on knee radiograph resulting in greater than 20 degree varus or valgus angulation
• Moderate loss of kidney function, define as estimated glomerular filtration rate of less than 45 mL/min.
• Significant arterial atherosclerosis that would limit selective angiography
• History of fibromyalgia, autoimmune, or inflammatory disorder
• History of any lumbar spine surgery, spine injections, or radicular pain in the extremity believed to originate from the spine
• Allergy to iodinated contrast agents
• Active Infection or malignancy
• Recent (within 12 months) or active nicotine use
• Prior total or partial knee replacement in the subject knee
• Active pregnancy
• Uncorrectable bleeding diasthesis
• Other medical, social, or psychological/physical conditions that, in the opinion of the investigator, preclude the patient from participating in the study.

The IPD generated during this study will not be shared outside of the study team. Only de-identified IPD results will be made available in publications or presentations. This approach ensures participant confidentiality and aligns with the study design.

Unending: Beginning 1 year after publication with no end date.

A statistical analysis plan has been created. All de-identified IPD will be stored on a HIPAA compliant online database, that is only accessible by delegated study team members.