1
Indications & Patient Selection
Indications
- Chronic knee osteoarthritis (OA) pain — Kellgren-Lawrence (KL) grade II–IV, failed conservative therapy (NSAIDs, physical therapy, injections × 3 months)
- Positive diagnostic genicular nerve block (GNB) — ≥50% pain relief on diagnostic block is required and is the strongest predictor of successful RFA
- Post-total knee arthroplasty (TKA) persistent pain — genicular nerves may remain intact and continue to mediate pain after arthroplasty
- Failed intra-articular injections — patients with ongoing pain despite corticosteroid or hyaluronic acid injections
- Pain ≥ 6 months duration with functional impairment
Contraindications & Workup
- Absolute: Active knee joint infection · Uncorrectable coagulopathy · Planned TKA within 3 months · Pacemaker without cardiology clearance (bipolar RFA mode can be used with pacemaker after clearance)
- Relative: Prior ipsilateral knee surgery altering anatomy · Prior RFA at same site without adequate washout (≥6 months)
- Workup: Weight-bearing knee X-ray (KL grading) · Baseline VAS pain score · Functional assessment (WOMAC or KOOS) · Confirm pain ≥6 months · INR / anticoagulation status
- Note: OA grade does NOT predict RFA response — even KL IV patients can have excellent relief; response is predicted by GNB result, not imaging
2
Pre-Procedure Checklist
Confirm positive diagnostic GNB. Patient must have had a prior diagnostic genicular nerve block with ≥50% pain relief. Document the degree and duration of relief. RFA without a positive GNB has significantly worse outcomes.
Fluoroscopy (C-arm) setup. Patient supine on fluoroscopy table. Feet and ankles can be taped to the table to ensure leg immobility. C-arm in AP orientation. Confirm bilateral condyles can be superimposed on lateral view.
RF equipment. Confirm RF generator availability (Halyard Coolief, Avanos, or Stryker Sievert). Set up 18G RF cannulas with 10-mm active tip × 3. Introducer needles if required by system. Confirm generator settings: 80°C × 90 sec (conventional) or 55°C × 150 sec (cooled tip).
Three-target protocol. Plan for three target nerves: Superior Medial Genicular (SMG), Superior Lateral Genicular (SLG), Inferior Medial Genicular (IMG). Mark targets on the pre-procedure fluoroscopic scout image. Review knee X-ray to appreciate bone landmarks.
Medications prepared. 1% lidocaine for skin entry (local anesthesia at 3 sites). 2% lidocaine 0.5 mL per site for pre-ablation anesthesia. Dexamethasone 4 mg + bupivacaine 0.5% for post-ablation injection at each site (reduces neuritis).
Anticoagulation hold. SIR Category 1 procedure. Heparin: hold 4–6h; LMWH: hold 24h; Warfarin: hold 5d (INR <2.0); DOACs: hold 24–48h. Aspirin may be continued. Confirm INR if applicable.
Sedation decision. Diagnostic GNB: no sedation required. RFA: mild IV sedation (midazolam 1–2 mg ± fentanyl 25–100 mcg) acceptable if patient prefers; fully awake is also feasible. Patient must be able to report sensory stimulation findings during nerve testing.
3
Relevant Anatomy
Three-Nerve Target Protocol
- Superior Medial Genicular Nerve (SMG): Runs along periosteum at the junction of the medial femoral condyle and medial femoral shaft — the "metaphyseal junction" on the medial side. Target: center of the notch between shaft and condyle on AP fluoroscopy.
- Superior Lateral Genicular Nerve (SLG): Junction of the lateral femoral condyle and lateral femoral shaft. Analogous position laterally. Note proximity to recurrent fibular (peroneal) nerve — do not target too far lateral or inferior.
- Inferior Medial Genicular Nerve (IMG): Junction of the medial tibial condyle and medial tibial shaft, just distal to the medial joint line. Most commonly under-targeted in failed RFA cases.
- Key rule: All three nerves run along periosteum at the epiphysis-diaphysis junction ("metaphyseal junction"). Periosteal contact = correct depth.
At-Risk Structures
- Saphenous nerve — runs medially near the SMG target; provides sensory innervation to the medial knee and distal leg. Injury (SMG too posterior) → medial knee and calf numbness.
- Recurrent fibular nerve (superficial peroneal branch) — near the SLG; injury → foot drop. Avoid ablating too far lateral or inferior at the SLG target.
- 5-nerve extension protocols add the Inferior Lateral Genicular nerve and Middle Genicular nerve — only recommended in experienced hands given increased peroneal nerve risk.
- Sensory stimulation (50 Hz, <0.5V) is the safety mechanism to confirm nerve location before ablation. Motor stimulation (2 Hz) rules out motor nerve proximity.
4
Technique
Default RadCall approach · share your own below
Supplies
C-arm fluoroscopy unit
18G RF cannula, 10-mm active tip × 3
RF generator (Halyard Coolief / Avanos / Stryker)
22G spinal needle × 3 (for diagnostic GNB)
1% lidocaine (skin entry)
2% lidocaine 0.5 mL per site (pre-ablation)
Dexamethasone 4 mg
Bupivacaine 0.5% 0.5 mL per site (post-ablation)
ChloraPrep
Sterile drape
Sterile dressing
Phase A — Diagnostic Genicular Nerve Block (separate visit)
Steps
1
Position and C-arm setup
Patient supine, knee slightly flexed. Tape feet/ankles to table to prevent movement. C-arm in AP orientation. Confirm condyles are superimposed on lateral view before beginning.
2
Identify three targets under fluoroscopy
AP view: SMG at medial femoral shaft-condyle junction (metaphyseal junction, medial side). SLG at lateral femoral shaft-condyle junction. IMG at medial tibial shaft-condyle junction, just distal to the medial joint line.
3
Advance 22G needle to periosteum
Local skin wheal with 1–2 mL 1% lidocaine at each entry site (avoid the target nerve). Advance 22G spinal needle to periosteum at each of the three target sites under AP fluoroscopy. Confirm periosteal contact — needle should "bounce" off bone.
▶ AP fluoroscopy — genicular nerve block needle placement
AP fluoroscopy: three-needle placement at superior medial, superior lateral, and inferior medial genicular nerve targets — needles at periosteum at junction of femoral shaft and condyle.
4
Inject lidocaine — small volume
Inject 0.5–1.0 mL 1% lidocaine per site (total ≤ 2 mL per nerve). Keep volumes small to avoid spread to adjacent structures and false-positive block. Total: 3 sites.
5
Assess pain relief at 30 min
Patient records VAS pain score pre- and 30 min post-block. ≥50% improvement = positive GNB → proceed to RFA. Document result formally before scheduling RFA.
Phase B — Radiofrequency Ablation (if GNB positive)
Steps
1
Position — same as diagnostic block
Patient supine, knee slightly flexed. C-arm AP. Same three target sites as the diagnostic block: SMG, SLG, IMG at their respective periosteal metaphyseal junctions.
2
Place 18G RF cannula to periosteum
Local skin wheal (1% lidocaine) at each of three entry sites. Advance 18G RF cannula (10-mm active tip) to periosteum at each target under AP fluoroscopy. Confirm with lateral view — tip should be at mid-depth of condyle, not anterior capsule or posterior.
▶ Lateral fluoroscopy — RF cannula depth confirmation
Lateral view confirms RF cannula tip at periosteal target — ensure cannula is flush with bone surface at each of the three targets before sensory testing.
3
Sensory stimulation — 50 Hz
Sensory stimulation at 50 Hz. Paresthesia reproduced in the knee at <0.5V = correct location (nerve confirmed). No paresthesia at >0.5V = off-target → withdraw and redirect. Do NOT ablate at off-target location.
4
Motor stimulation — 2 Hz
Motor stimulation at 2 Hz. Confirm no motor response at <2× the sensory threshold. This is especially important at the SMG (near saphenous nerve) and SLG (near recurrent fibular/peroneal nerve). If motor response present → reposition before ablating.
5
Pre-ablation anesthesia
Inject 0.5 mL 2% lidocaine through the RF cannula at each confirmed target site. Wait 60 seconds for anesthetic effect before ablation.
6
Radiofrequency ablation
Ablate each target sequentially. Conventional RF: 80°C × 90 seconds per site. Cooled-tip alternative (Coolief): 55–60°C × 150 seconds per site (larger lesion size). Treat all three targets: SMG, SLG, IMG. Total procedure time approximately 30–45 min.
7
Post-ablation injection at each site
After ablation at each site, inject 1 mL dexamethasone 4 mg + 0.5 mL bupivacaine 0.5% through the cannula before removal. This significantly reduces post-procedure neuritis (Gonzalez et al. data).
NOTE — 5-Nerve Protocol:
Some protocols extend to 5 nerves by adding the Inferior Lateral Genicular nerve and Middle Genicular nerve. Outcomes are similar to the 3-nerve protocol in published data. The 5-nerve approach is not recommended for less experienced operators given increased risk to the peroneal nerve at the inferior lateral target.
Sign in to view and create community cards
5
Fluoroscopic Landmarks
Target Localization by View
| Target Nerve | AP View Landmark | Lateral View Confirmation | Key Pitfall |
|---|---|---|---|
| Superior Medial Genicular (SMG) | Medial femur — center of the "notch" between the femoral shaft and medial femoral condyle (metaphyseal junction) | Mid-depth of condyle; not anterior capsule, not posterior | Too posterior → saphenous nerve → medial numbness |
| Superior Lateral Genicular (SLG) | Lateral femur — analogous junction of lateral femoral shaft and lateral femoral condyle | Mid-depth of condyle laterally | Too lateral/inferior → peroneal nerve → foot drop risk |
| Inferior Medial Genicular (IMG) | Medial tibia — junction of medial tibial shaft and medial tibial condyle, just distal to the medial joint line | Anterior to fibular head level, mid-depth of tibial condyle | Most commonly missed target — ensure cannula is distal enough |
AP View Essentials
- All three metaphyseal junctions are visible on a single AP view of the knee
- SMG and IMG are on the medial side; SLG on the lateral side
- The RF cannula tip should overlay the periosteum at the junction — not on the joint space, not on the shaft alone
- Confirm position at each of the three sites before moving to lateral confirmation
Lateral View Essentials
- Superimpose condyles on lateral view for accurate depth assessment
- Tip should project at the mid-depth of the condyle
- Too anterior = anterior fat pad / anterior capsule — incorrect
- Too posterior = posterior capsule, popliteal vessels, neural structures — incorrect and dangerous
6
Troubleshooting
Problem
Paresthesia down leg or foot during sensory stimulation
Likely cause: RF cannula too posterior — near saphenous nerve (medial, leg paresthesia) or near recurrent fibular / peroneal nerve (lateral, foot or dorsum paresthesia).
Next step: Stop stimulation immediately. Withdraw cannula 2–3 mm and redirect anteriorly and slightly superiorly. Repeat sensory stimulation. Do NOT ablate until knee-specific paresthesia is reproduced at <0.5V. If paresthesia persists in the foot or lower leg, reposition significantly before retesting.
Problem
Motor response at low voltage during 2 Hz stimulation
Likely cause: RF cannula near a motor nerve — most often peroneal nerve at the SLG target (ankle dorsiflexion or eversion), or saphenous-adjacent motor branch at SMG.
Next step: Reposition cannula before ablation. Acceptable motor threshold = no motor response at <2× the sensory threshold. If motor firing occurs below this margin, redirect and retest both sensory and motor stimulation before proceeding.
Problem
Incomplete or short-lived pain relief after RFA
Likely cause: Most commonly the IMG was under-targeted (too proximal on the tibia, missing the metaphyseal junction). Less commonly, inadequate sensory stimulation confirmation before ablation.
Next step: Review fluoroscopic images — confirm IMG was at the medial tibial shaft-condyle junction distal to the joint line. Consider repeat RFA targeting 5 nerves (adding inferior lateral genicular and middle genicular). Repeat RFA is safe at ≥6 months.
Problem
Post-procedure pain flare (days 1–14)
Likely cause: Neuritis from thermal ablation of genicular nerve fibers — occurs in approximately 10–15% of patients; self-limited.
Next step: Reassure patient this is expected and transient. Manage with NSAIDs and ice. Resolves within 2 weeks in most patients. Post-ablation dexamethasone injection (at time of procedure) significantly reduces the incidence. Do not interpret early pain flare as procedure failure.
7
Complications
Immediate / Periprocedural
- Neuritis / post-procedure pain flare (10–15%) — most common complication; thermal reaction around ablated nerve; self-limited, resolves within 2 weeks; mitigated by post-ablation dexamethasone + bupivacaine injection
- Saphenous nerve injury — SMG cannula placed too posterior → medial knee and calf numbness / dysesthesia; usually temporary; avoid by confirming no leg paresthesia at sensory stimulation and maintaining anterior trajectory
- Peroneal nerve injury / foot drop — SLG cannula placed too lateral or inferior → ankle dorsiflexion weakness; most serious complication; avoid by strict motor stimulation threshold testing and conservative SLG targeting
- Skin burn — rare with proper technique; RF tip must not contact dermis; ensure adequate needle depth before ablation
Delayed
- Infection — rare (<1%); sterile technique essential; present as worsening pain, warmth, fever; treat with antibiotics; joint aspiration and culture if septic arthritis suspected
- Incomplete ablation — inadequate lesion size (too low temperature or too short duration); protocol-compliant technique (80°C × 90 sec or cooled tip 55°C × 150 sec) minimizes this; repeat RFA safe at ≥6 months
- Limited duration of effect — nerves regenerate over 6–12 months; pain recurrence expected; repeat RFA is safe and effective and can be offered when pain returns
- Hematoma — rare; SIR Category 1 procedure; usually self-limited; compressive dressing and observation
8
Critical Pearls
✶
ALWAYS perform diagnostic GNB first — it is the gatekeeper to RFA. A positive GNB (≥50% pain relief) is the strongest predictor of successful RFA outcomes. Proceeding to RFA without a confirmatory block leads to significantly worse results. This is not optional — it is the standard of care.
✶
Sensory stimulation confirms target position — ≤0.5V is the threshold. Paresthesia in the knee at <0.5V = correct location (nerve confirmed). Threshold >0.5V = off-target → reposition before ablating. Never skip stimulation testing. It is both a safety and an efficacy step.
✶
Post-ablation dexamethasone + bupivacaine significantly reduces neuritis. Inject 1 mL dexamethasone 4 mg + 0.5 mL bupivacaine 0.5% at each site after ablation before removing the cannula. This is supported by Gonzalez et al. data and should be routine. It also improves immediate post-procedure comfort and reduces callbacks.
✶
Duration of effect is 6–12 months; repeat RFA is safe and effective. Genicular nerves regenerate. Patients should be counseled that pain relief is not permanent. Repeat RFA at ≥6 months can restore the same degree of relief. Frame this as a maintenance procedure rather than a cure — sets realistic expectations and improves patient satisfaction.
✶
OA grade does not predict RFA response — GNB does. Even patients with severe radiographic OA (KL grade IV) can achieve excellent pain relief with genicular nerve RFA. Do not exclude patients based on X-ray findings alone. Select based on clinical presentation, duration of pain, failed conservative therapy, and most importantly — a positive diagnostic GNB.
✶
IMG is the most commonly missed target in failed RFA cases. The inferior medial genicular nerve target on the medial tibial condyle is frequently under-targeted — the cannula is placed too proximally (at or above the joint line) rather than distal to it at the metaphyseal junction. Review lateral fluoroscopic views carefully for IMG depth and position at the time of the procedure.
9
References & Community Cards
Outcome Data Summary
| Timepoint | KOOS Score | WOMAC Score | VAS Pain |
|---|---|---|---|
| Baseline | 16.7 | 23 | 9–10 |
| 2 weeks | 81 | 65 | 4 |
| 1 month | 84 | 65 | 2 |
| 3 months | 86 | 65 | 1 |
| 6 months | 91 | 82 | 0 |
Data from Gonzalez FM, Broida SE, Kokabi N (Prologo & Ray, Thieme 2024) — cooled RFA, SIFK patient. KOOS = Knee Injury and Osteoarthritis Outcome Score; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; VAS = Visual Analog Scale.
Citations
- Prologo JD, Ray CE Jr., eds. Advanced Pain Management in Interventional Radiology: A Case-Based Approach. Thieme; 2024. Ch. 34: Gonzalez FM, Broida SE, Kokabi N — Nerve Ablations I (Genicular RF).
- Choi WJ, Hwang SJ, Song JG, et al. Radiofrequency treatment relieves chronic knee osteoarthritis pain: a double-blind randomized controlled trial. Pain. 2011;152(3):481–487.
- Jamison DE, Cohen SP. Radiofrequency techniques to treat chronic knee pain: a comprehensive review of anatomy, effectiveness, treatment parameters, and patient selection. J Pain Res. 2018;11:1879–1888.
- Sayyid S, Younan Y, Sharma G, et al. Subchondral insufficiency fracture of the knee: grading, risk factors, and outcome. Skeletal Radiology. 2019;48(12):1961–1974.
- SIR Standards of Practice Committee. Consensus Guidelines for Periprocedural Management of Coagulation Status. J Vasc Interv Radiol. 2012;23(6):727–736.
9
References & Resources
Primary sources · Key data · Related procedures
Key Guidelines
- ISIS Practice Guidelines
- ASIPP Systematic Review on Genicular Nerve RFA
- ACR Appropriateness Criteria
Primary References
- Choi WJ et al. Radiofrequency treatment relieves chronic knee osteoarthritis pain: a double-blind randomized controlled trial. Pain. 2011;152(3):481-487.
- Prologo JD, Ray CE Jr., eds. Advanced Pain Management in Interventional Radiology. Thieme; 2024. Ch. 34: Genicular Nerve Radiofrequency Ablation.
- Iannaccone F et al. A meta-analysis of clinical effectiveness of radiofrequency neurotomy for facet and sacroiliac joint pain. Pain Physician. 2012;15(4):297-304.