Comparison of radiofrequency ablation and phenol neurolysis for genicular nerve ablation in end-stage knee osteoarthritis

Authors

Abstract

Aim Knee osteoarthritis (OA) is one of the most common causes of chronic pain and disability in older adults [1]. When patients have advanced knee OA and do not respond to conservative treatments, yet are not candidates for surgery, genicular nerve interventions have emerged as minimally invasive options [2].
Materials and Methods In this retrospective cohort study, records of 54 patients with end-stage knee OA and chronic knee pain who underwent genicular nerve ablation at a pain clinic were reviewed (29 patients received fluoroscopy-guided RFA under sedation, and 25 received ultrasound-guided injections of 6% phenol under local anesthesia). Pain was assessed on an 11-point Numeric Rating Scale, and knee function was measured using the Lequesne Index. These outcomes were recorded before the intervention and at the 3-month post-procedure follow-up.
Results By 3 months, both treatments produced significant pain relief and functional improvement. In the RFA group, mean NRS pain scores fell from 8.5 ± 0.7 at baseline to 5.0 ± 1.8 at 3 months. Similarly, the phenol group’s NRS score dropped from 8.6 ± 0.5 to 4.6 ± 1.7. Lequesne Index scores improved in parallel: RFA patients went from 18.7 ± 2.3 to 12.3 ± 4.5, and phenol patients from 18.9 ± 1.7 to 13.1 ± 3.4. There were no statistically significant differences between the RFA and phenol groups.
Discussion In this study, both genicular RFA and phenol neurolysis provided significant short-term pain reduction and functional improvement.

Keywords

Introduction

Knee osteoarthritis is the most common form of arthritis and a leading cause of chronic pain and disability in older people [1]. Total knee arthroplasty is the definitive treatment for end- stage (advanced) knee OA, but many patients either cannot undergo surgery or prefer not to [3]. This leaves a treatment gap for individuals who have exhausted conservative measures (such as analgesic medications, physical therapy, and intra- articular injections) yet still suffer from severe knee pain and are not surgical candidates [2, 4]. In recent years, clinicians have explored minimally invasive nerve-targeting interventions to bridge this gap by providing analgesia through denervation of the knee’s sensory nerves [5]. One such intervention is genicular nerve radiofrequency ablation (RFA), an established technique for managing refractory knee OA pain [6]. RFA involves applying heat (typically around 80°C for 60–90 seconds at each target) to the articular branches of the genicular nerves, creating a controlled lesion that disrupts nerve function [7]. Until recently, the effectiveness and safety of genicular nerve phenol neurolysis had not been well established in large studies [8]. The present study aimed to directly compare the outcomes of conventional RFA versus phenol genicular nerve neurolysis in patients with end-stage knee OA (Kellgren–Lawrence grade 3–4).

Materials and Methods

Study Design and Patients
This research was a retrospective cohort study conducted at a single tertiary pain clinic, reviewing patient records from 2024 to 2025. Patients diagnosed with severe knee OA (Kellgren– Lawrence grade 3 or 4) who had chronic knee pain unresponsive to conservative therapy, a positive response to a genicular nerve block test, and who underwent genicular nerve ablation for pain relief were included. Patients without complete 3-month follow-up data were excluded. A total of 54 patients met the inclusion criteria, of whom 29 were treated with RFA and 25 with phenol neurolysis. The choice of RFA versus phenol was determined by the treating physician’s preference and patient- specific factors. Baseline characteristics of the two groups are summarized in Table 1. The RFA and phenol groups were similar in age, sex distribution, and baseline pain and function scores (all p-values > 0.5). Approximately 70% of patients in each group had Kellgren–Lawrence grade 4 disease, indicative of extremely severe osteoarthritis.
Interventions
For the RFA group, genicular nerve ablation was performed in a day-surgery operating room under fluoroscopic guidance. The superior medial, superior lateral, and inferior medial genicular nerve regions were targeted at the typical bony landmarks near the junction of the femur and tibia. Proper needle placement was confirmed (e.g., via sensory stimulation), and then an RFA probe was used to create a thermal lesion at each site (approximately 80°C for 90 seconds per site). Patients were under moderate intravenous sedation during these procedures to improve comfort, as creating the lesions can be painful.
In the phenol group, patients underwent chemical neurolysis of the same genicular nerve regions (superomedial, superolateral, and inferomedial) using ultrasound guidance. After a local anesthetic was applied to numb the skin, a needle was advanced under ultrasound visualization to each target area near the periosteum. At each site, about 1.5 mL of a 6% phenol- in-glycerol solution was slowly injected after careful aspiration to ensure the needle was not in a blood vessel. Phenol causes protein coagulation within nerve fibers, effectively producing a chemical ablation of the nerve. These phenol procedures were done in an outpatient clinic setting with the patient awake (no sedation needed). All patients were observed for a short period after the intervention and then discharged the same day with standard post-procedure care instructions.
Outcomes
Two primary outcomes were defined: knee pain intensity and knee functional status. Pain intensity was measured using an 11-point Numeric Rating Scale (NRS), where 0 represents no pain and 10 represents the worst imaginable pain. Functional status was measured using the Lequesne Algofunctional Index for knee OA, which ranges from 0 to 24 points (higher scores indicate more severe pain and disability). Both measures were recorded before the intervention (baseline) and at the 3-month follow-up visit. The primary endpoints for analysis were the change in NRS pain score and the change in Lequesne Index from baseline to 3 months. Any procedure-related complications observed within the 3-month follow-up period (for example, signs of infection, nerve injury, or persistent numbness/ paresthesia) were also documented.
Statistical Analysis
IBM SPSS version 26 was used for data analysis. Within- group changes were assessed with the Wilcoxon signed-rank test, and between-group changes were compared with the Mann–Whitney U test. Because the Shapiro–Wilk test indicated that some change scores were not normally distributed, non- parametric methods were chosen. A significance level of p < 0.05 was adopted. Cohen’s d was calculated for effect sizes, and 95% confidence intervals were provided where appropriate.
Ethical Approval
This study was approved by the Ethics Committee of Süleyman Demirel University (Date: 2025-08-18, No: 35).

Results

Patient Characteristics
A total of 54 patients (mean age approximately 68 years) were analyzed, with 29 in the RFA group and 25 in the phenol group. As shown in Table 1, the two groups were well matched at baseline. On average, patients in both groups reported severe knee pain before treatment (mean baseline NRS pain was about 8.5 out of 10 in each group), and they had very high disability from OA (mean Lequesne Index around 19 out of 24, indicating “extremely severe” OA-related impairment). The distribution of radiographic severity was also similar: roughly 70% of each group had Kellgren–Lawrence grade 4 disease (the highest severity), with the remainder at grade 3. The RFA group was 38% female and the phenol group 24% female, but this difference was not statistically significant given the sample size. No significant differences in any baseline pain or functional measures were observed between the RFA and phenol patients (all p > 0.6), indicating the groups were comparable before the interventions.
Pain Relief (Within-Group)
Both interventions led to significant reductions in knee pain from baseline to the 3-month follow-up. In the RFA group, the mean NRS pain score improved from 8.48 ± 0.69 at baseline to 4.97 ± 1.78 at 3 months. This is an average pain decrease of about –3.52 points (95% CI –4.08 to –2.95; p < 0.001, Wilcoxon signed-rank test), which is highly significant. The effect size for pain reduction with RFA was very large (Cohen’s d ≈ 2.38), indicating a substantial clinical improvement. The phenol group showed a similarly large improvement: their mean NRS pain score dropped from 8.56 ± 0.51 at baseline to 4.64 ± 1.68 at 3 months (an average decrease of –3.92 points, 95% CI –4.50 to –3.34; p < 0.001). This pain reduction was also highly significant and had a very large effect size (d ≈ 2.78).
Functional Improvement (Within-Group)
Knee function also improved significantly in both groups over the 3-month period, as measured by the Lequesne Algofunctional Index. In the RFA group, the mean Lequesne score declined from 18.69 ± 2.30 at baseline to 12.28 ± 4.52 at 3 months. This represents an average improvement of –6.41 points (95% CI –7.61 to –5.22; p < 0.001, Wilcoxon), which is statistically significant. The effect size for this functional improvement was large (paired Cohen’s d ≈ 2.04). Similarly, the phenol group’s mean Lequesne score went from 18.92 ± 1.71 at baseline to 13.12 ± 3.39 at 3 months, an average improvement of –5.80 points (95% CI –6.91 to –4.69; p < 0.001), with a very large effect size (d ≈ 2.16).
Comparison Between RFA and Phenol
When the outcomes of the RFA and phenol groups were compared, no significant differences were observed in the degree of pain relief or functional improvement at 3 months. The phenol group’s average pain reduction was slightly larger than the RFA group’s (–3.92 vs –3.52 NRS points), but this ~0.4-point difference was not statistically meaningful (Mann–Whitney U test, p = 0.22). The 95% confidence interval for the difference in NRS change between groups ranged from approximately –1.19 to +0.39, crossing zero and indicating no clear advantage of one treatment over the other. Similarly, the RFA group had a marginally greater mean improvement in the Lequesne Index (–6.41 vs –5.80 points), but this ~0.6-point difference was not significant either (p = 0.53). The 95% CI for the between-group difference in functional change (approximately –0.97 to +2.20) also crossed zero. In other words, within the limits of the sample, both interventions yielded essentially equivalent outcomes.
Safety and Tolerability
Both treatments were generally well tolerated, with no serious adverse events observed in either group. Three patients (5.6% of the cohort) experienced minor complications, with a similar incidence in each group: 1 patient (3.4%) in the RFA group and 2 patients (8.0%) in the phenol group (this difference was not significant, p = 0.60). All of the noted complications were transient sensory disturbances consistent with mild neuritis.

Discussion

In this retrospective series of patients with advanced knee osteoarthritis, both genicular nerve RFA and phenol neurolysis yielded meaningful pain relief and functional improvement over a 3-month period. The two treatments performed almost identically in terms of efficacy: on average, patients in each group experienced roughly a 3.5–4 point reduction in NRS pain, along with about a 6-point improvement in their Lequesne functional score. These changes were highly significant statistically and also appear to be clinically important. Approximately half of the patients in both groups achieved at least a 50% reduction in pain.
A key finding of this study is that no significant difference in efficacy was observed between the RFA and phenol groups. In this study, genicular nerve ablation provided effective pain relief within each group, but neither treatment proved superior when directly compared. Likewise, no between-group disparity in functional improvement was found. These findings suggest that chemical neurolysis with phenol (at 6% concentration) can be just as effective as thermal ablation, at least in the short term.
The magnitude of pain relief observed in this study (approximately a 3.5–3.9 point drop in NRS) is similar to that reported in many RFA studies in the literature [9, 10]. With regard to phenol neurolysis, a prospective study demonstrated that phenol genicular injections provided sustained benefits for at least 6 months [8]. The 3-month results (pain roughly 8.6 down to 4.6; Lequesne 19 down to 13) observed in the present study are in line with those findings, indicating that phenol can produce large reductions in pain and disability in the intermediate term.
Both treatments were well tolerated, with only minor, transient adverse events noted. RFA allows more precise control of lesion size and placement [11]. Phenol, a liquid neurolytic agent, may diffuse into adjacent tissues if not carefully deposited, though using small volumes and meticulous technique can mitigate this risk [12, 13].
Because the follow-up in this study was limited to 3 months, the longer-term durability of these interventions cannot be determined. Clinically, after genicular nerve ablation, pain relief often lasts 6–12 months before nerve regeneration and pain recurrence [8, 14]. A recent meta-analysis in knee osteoarthritis rated the certainty of evidence for RFA as very low and found modest, short-term, non-sustained benefits, advising against routine use pending stronger evidence [15]. The ongoing RADIOPHENOL randomized trial comparing RFA, phenol, and conservative care with ≥6-month follow-up should yield higher- level data on efficacy and durability [2].

Limitations

This study is retrospective in design, which introduces potential biases such as selection bias due to the absence of randomization. Treatment allocation (RFA vs. phenol) was based on clinical decision-making rather than random assignment, and the sample size (54 patients in total) was relatively modest. The follow-up period was limited to 3 months, so only short- term outcomes were captured. Only patient-reported outcome measures were collected, without inclusion of objective functional tests. Furthermore, outcomes like quality of life or changes in analgesic medication use after the procedures were not formally assessed, which might have provided a more comprehensive view of each treatment’s impact.

Conclusion

Both radiofrequency ablation and phenol neurolysis of the genicular nerves appear to be effective options for managing chronic knee pain in end-stage osteoarthritis. In this study, each treatment resulted in significant pain reduction and functional improvement at 3 months, and neither was found to be superior to the other in terms of outcomes. Ultrasound-guided phenol neurolysis (using a 6% phenol solution) seems to be a feasible, less resource-intensive alternative to conventional RFA for patients with severe knee OA who need pain relief but either cannot undergo surgery or wish to delay it. The safety profiles of both procedures were favorable, with only minor transient neuropathic symptoms observed (occurring slightly more frequently with phenol). Future prospective studies with larger sample sizes and longer follow-up will be needed to confirm the long-term comparative efficacy and safety of phenol versus RFA.

References

1. Hawker GA, King LK. The burden of osteoarthritis in older adults. Clin Geriatr Med. 2022;38(2):181-92.
   PubMed Google Scholar
2. de Wit PR, van Beek R, Schokker M, et al. Genicular nerve radiofrequency ablation, phenol neurolysis or conservative medical management in patients with knee osteoarthritis: Protocol for the RADIOPHENOL randomized controlled multicentre trial. BMJ Open. 2025;15(7):e094576.
   PubMed Google Scholar
3. Steinhaus ME, Christ AB, Cross MB. Total knee arthroplasty for knee osteoarthritis: Support for a foregone conclusion? HSS Journal. 2017;13(2):207- 10.
   PubMed Google Scholar
4. Chen SH, Kuan TS, Kao MJ, Wu WT, Chou LW. Clinical effectiveness in severe knee osteoarthritis after intra-articular platelet-rich plasma therapy in association with hyaluronic acid injection: Three case reports. Clin Interv Aging. 2016;11:1213-9.
   PubMed Google Scholar
5. Kanjanapanang N, Madrid R, Lin P, et al. Effectiveness of genicular nerve radiofrequency ablation in osteoarthritis and post-surgical knee pain: Systematic review. Pain Med. 2025;pnaf115.
   PubMed Google Scholar
6. Kidd VD, Strum SR, Strum DS, Shah J. Genicular nerve radiofrequency ablation for painful knee arthritis: The why and the how. JBJS Essential Surgical Techniques. 2019;9(1):e10.
   PubMed Google Scholar
7. Gupta M, Abdallah RT, Abd-Elsayed A, et al. A review of nonsurgical neurolytic procedures for neuropathic pain. J Pain Res. 2025;18:879-95.
   PubMed Google Scholar
8. Risso RC, Ferraro LHC, Frederico TN, et al. Chemical ablation of genicular nerve with phenol for pain relief in patients with knee osteoarthritis: A prospective study. Pain Pract. 2021;21(4):438-44.
   PubMed Google Scholar
9. Alomari A, Bhatia A. An update on radiofrequency denervation for arthritis- related knee joint pain: A synthesis of the current evidence. BJA Education. 2024;24(5):164-72.
   PubMed Google Scholar
10. Dias JVAMO, de Moraes VRAP, de Freitas MTB, et al.Genicular nerve ablation in knee osteoarthritis: A randomized prospective study. Acta Ortop Bras. 2025;33(3):e289218.
    PubMed Google Scholar
11. McCormick ZL, Cohen SP, Walega DR, Kohan L. Technical considerations for genicular nerve radiofrequency ablation: Optimizing outcomes. Reg Anesth Pain Med. 2021;46(6):518-23.
    PubMed Google Scholar
12. Wilkinson AJ, Chai T, Roldan CJ, Huh BK. Genicular nerve neurolysis with phenol for chronic knee pain: A case series. Interventional Pain Medicine. 2023;2(1):100182.
    PubMed Google Scholar
13. Cushman DM, Monson N, Conger A, Kendall RW, Henrie AM, McCormick ZL. Use of 0.5 mL and 1.0 mL of local anesthetic for genicular nerve blocks. Pain Med. 2019;20(5):1049-52.
    PubMed Google Scholar
14. Cedeño DL, Vallejo A, Kelley CA, Tilley DM, Kumar N. (n.d.). Comparison study: Comparisons of lesion volumes and shapes produced by a radiofrequency system with a cooled, a protruding, or a monopolar probe. Pain Physician. 2017;20(6):E915-22.
    PubMed Google Scholar
15. Almeida M, Saragiotto BT, Hunter DJ, et al. Efficacy and safety of minimally invasive interventions targeting the genicular nerves for knee osteoarthritis: A meta-analysis. Osteoarthritis Cartilage. 2025;33(5):535-47.
    PubMed Google Scholar

Declarations

Additional Information

Publisher’s Note
Bayrakol MP remains neutral with regard to jurisdictional and institutional claims.

Rights and Permissions

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). To view a copy of the license, visit https://creativecommons.org/licenses/by-nc/4.0/

View full license details →

About This Article