Clinically Significant Hematoma as a Complication of Cooled Radiofrequency Ablation of the Genicular Nerves; a Case Series

Dear Editor, Genicular nerve radiofrequency ablation (RFA) has gained traction as a pain management strategy for patients with recalcitrant knee pain due to osteoarthritis who are poor surgical candidates or elect for nonsurgical pain management. A multicenter study assessing the treatment effect of this procedure has demonstrated improvement in pain and function in this population [1]. The procedure is considered relatively safe, with only single cases of a third-degree skin burn, pes tendon injury, and septic arthritis being reported [2–4]. Here, we report the first published cases of large peri-articular hematomas and a case of hemarthrosis following cooled genicular nerve RFA. We discuss relevant anatomical considerations and strategies to mitigate the risk of vascular trespass and hematoma formation. A 60-year-old male with atrial fibrillation, on apixaban, was referred for evaluation of chronic right medial knee pain due to osteoarthritis (OA) that had failed to improve with physical therapy (PT), oral analgesics, and intra-articular steroid injections. He was deemed a poor candidate for total knee arthroplasty (TKA) because of poorly controlled diabetes. Apixaban was discontinued three days before prognostic genicular blocks and RFA, which were performed uneventfully. He was instructed to resume apixaban 12 hours postprocedure and discharged home. Twenty-eight days later, the patient presented to a local emergency department (ED) with exacerbation of his chronic knee pain where a computerized axial tomography (CAT) scan demonstrated a 4.6×2.3×10.8 cm hematoma within the distal vastus medialis muscle. His clinical follow-up confirmed a normal neurovascular exam with fullness of the medial quadriceps muscle to palpation. The patient was monitored for several months, with eventual resolution of knee pain and no future sequalae. A 76-year-old female with systemic lupus erythematosus, recurrent venous thromboembolism on warfarin, and multiple cardiac comorbidities was evaluated for chronic right knee pain due to OA that failed to improve with prior PT, injections, and oral analgesics. She was evaluated by an orthopedic surgeon for TKA candidacy but was deemed to be too high-risk. Warfarin was held (with enoxaparin bridge) until her preprocedure International Normalised Ratio (INR) was 1.2, and the patient underwent successful genicular nerve blocks followed by RFA. She was discharged home to restart enoxaparin 1mg/kg twice daily and warfarin 12 hours postprocedure. Two days postintervention, she presented to the ED with a swollen knee and complaints of oozing from all three needle puncture sites. Radiographs of the knee confirmed a large effusion and circumferential tissue swelling. Joint aspiration confirmed hemarthrosis, and pressure dressings successfully stopped the oozing. Knee pain and swelling persisted after discharge, and she eventually underwent right TKA. A 54-year-old female with obesity and primary biliary cirrhosis status after orthotopic liver transplantation was evaluated for treatment of recalcitrant chronic right knee pain due to OA. She was not a surgical candidate due to morbid obesity. After successful prognostic blocks were performed, genicular RFA was performed several weeks later without incident. Three days postprocedure, she developed knee swelling and stiffness and was found to have soft-tissue swelling and a palpable joint effusion. Magnetic resonance imaging of the knee confirmed a 3.1×3.2×0.7 cm hematoma within the deep anteromedial soft tissues along the fascial plane of the distal vastus medialis muscle (Figures 1 and 2) . Pain was managed with prn oxycodone and subsequently improved to her baseline pain level. Magnetic resonance image of a coronal image sequence demonstrating a hematoma within the distal vastus medialis muscle. Magnetic resonance image of an axial T2 image sequence demonstrating a hematoma within the vastus medialis muscle. Water cooled RFA (Coolief, Halyard Health, Alpharetta, GA, USA) was performed in the standard fashion to target the superior medial genicular nerve (SMGN), superior lateral genicular nerve (SLGN), and inferior medial genicular nerves (IMGNs), as described in prior study protocols [1,5] . This case series is the first to report the development of clinically significant hematomas due to the genicular RFA procedure, despite moderate adherence to American Society of Regional Anesthesia (ASRA) guidelines for anticoagulation management for interventional procedures. Given the occurrence of three hematomas within a single pain practice, this complication likely occurs more frequently than is currently recognized and is under-reported in the literature. The superior medial (SMGA), superior lateral (SLGA), and inferior medial genicular arteries (IMGA) derive their origin from the popliteal artery and form a rich anastomotic network that may be disrupted by radiofrequency needle placement when targeting the SMGN, SLGN, or IGMN with the standard technique [1,5]. In our series, two of three cases involved hematoma formation within the distal vastus medialis muscle, suggestive of superior medial genicular artery injury. It is important that the interventionalist be vigilant for this complication, as delayed intervention could lead to continued spontaneous hematomas, pseudoaneurysm, pain, and mass effect on adjacent structures, leading to neurovascular compression and/or thrombosis. Viable steps to mitigate potential bleeding risk after genicular RFA include strict adherence to ASRA guidelines for interventional spine and pain procedures in patients on antiplatelet and anticoagulant medications. Classification of water cooled genicular RFA as a deep peripheral block due to the larger 17-gauge needle size and proximity to arterial structures should be considered, despite being a peripheral joint intervention. In our cohort, preprocedural adherence to ASRA guidelines was followed, yet two anticoagulated patients experienced hematomas, and early resumption of anticoagulation could be causative; this emphasizes the importance of strict pre- and postprocedural compliance with known guidelines. Anatomically along the periosteal target, adjacent to the genicular nerves, lie the corresponding genicular arteries. Arterial injury is a distinct possibility if needles contact the periosteum and walk off; therefore, we suggest avoidance of contact with the periosteum to potentially reduce bleeding risk, though this is speculative, as potential vascular injury can occur at any juncture of needle trespass. Diagnostic ultrasound may be considered by providers who perform genicular nerve RFA to map out arterial locations before placement of fluoroscopically guided cannulas or as an adjunct/alternative in high-risk patients. Despite the relative safety of genicular nerve RFA, providers should be aware of the potential for vascular injury and subsequent hematoma formation, particularly in patients at risk of hemorrhagic complications. If patients report increased or new pain after ablation procedures, a high index of suspicion must be present to exclude this potentially under-recognized complication. Conflicts of interest: Zachary L. McCormick, MD, serves on the Board of Directors of the Spine Intervention Society. There are no other potential conflicts of interest to disclose on the part of any of the other authors.