Cooled radiofrequency ablation provides extended clinical utility in the management of knee osteoarthritis: 12-month results from a prospective, multi-center, randomized, cross-over trial comparing cooled radiofrequency ablation to a single hyaluronic acid injection

Knee osteoarthritis (OA) is a painful and debilitating disease that often affects patients for years [1]. While total knee arthroplasty (TKA) is widely considered a definitive treatment for late stage knee OA, non-surgical options are useful for symptomatic management. Patients experiencing knee OA suffer from pain an average of 9 years before becoming candidates for surgical intervention [2].

Nonsurgical treatment options for knee OA symptoms include weight loss, activity modification and physical therapy [3]. If these do not provide adequate relief, nonsteroidal anti-inflammatory drugs (NSAIDs) and acetaminophen can be taken to mitigate pain. However, it should be noted that these pharmacologic interventions can present significant adverse events (AEs) [4]. Intra-articular steroids (IAS) injections have been utilized to manage knee OA symptoms, but studies have demonstrated they may only provide short-term pain relief [5, 6]. Other studies have cautioned that multiple steroid injections may lead to accelerated knee osteoarthritis progression [7]. Platelet rich plasma injections have also been employed to manage knee OA pain, but questions remain regarding efficacy and the lack of standardization of treatments [8‐10]. Hyaluronic acid (HA), otherwise known as “viscosupplementation” injections, are another treatment option. Some clinical trials have shown modest effects of HA injections when managing knee OA pain [11], but larger statistical analysis have concluded the benefits of HA are clinically insignificant [12].

Radiofrequency ablation (RFA) is the targeted delivery of radiofrequency energy through a probe that causes the thermal degradation of nerve structures via ionic heating. The areas of thermal degradation are referred to as lesions. Traditional radiofrequency ablation probes operate at a set temperature of 80 °C. Cooled radiofrequency ablation (CRFA) uses internally cooled radiofrequency probes which are able to delivery more energy to surrounding tissues. While internally cooled probes operate at a set temperature of 60 °C, temperatures in tissues beyond the probe tip reach 80 °C. As a result of the internal cooling of the probe, larger lesions are created which can help overcome physiological variability of nerve location and increase the likelihood of treatment success [13]. Additionally, recent research has highlighted distinct physiological differences between lesions created by CRFA probes compared to standard radiofrequency probes, that may account for the extended durability of pain relief when using CRFA [14]. Clinical studies have demonstrated that CRFA can provide 12-months of pain relief for the majority of patients undergoing this procedure [15‐17]. A subset of subjects receiving CRFA have reported pain relief extending through the 18 and 24-month timepoints [18]. A previous study demonstrated the effectiveness of CRFA versus IAS injections [15, 18, 19] and CRFA vs HA at 6-months [20], but no study has compared CRFA to HA with 12-month follow-up and a cohort of HA patients that could crossover to receive CRFA after 6-months.

Thus, the purposes of this study were to: (1) evaluate the efficacy of CRFA for the treatment of knee OA pain at 12 months, and (2) evaluate HA patients who crossed over to CRFA after 6-months of treatment.

There is currently a demand for prospective studies evaluating the methods and techniques used for nonoperative management of knee OA pain and disability [12, 28]. The number of patients experiencing symptoms associated with knee OA is increasing dramatically. While effective, TKR is not always necessary or indicated, and most patients benefit from nonoperative management of symptoms during disease progression. This study demonstrated that CRFA was effective at pain relief, reduction of stiffness, and improvement in physical function, global outcomes and quality of life at 12 months, and that patients who crossed over from HA demonstrated improvements in all the same domains.

Patients in the original CRFA cohort demonstrated pain relief that extended to the 12-month timepoint with a significant improvement from baseline. The percentage of subjects within the CRFA cohort in this trial reporting ≥50% pain relief at the 12-month timepoint (65.2%) was similar to what has been reported in previous CRFA trials (65.4%) [15]. Furthermore, the mean decrease in NRS of subjects in this CRFA cohort (4.0) matched with previously reported 12-month results of other trials (4.3) [15].

In addition to improvements in total WOMAC score, subjects within the CRFA cohort also experienced improvements in WOMAC pain score that were above the 12–18% improvement considered the minimal clinically important difference in OA [29]. Other randomized, blinded clinical trials conducted with HA have shown modest benefits in WOMAC scores. Lin et al. showed statistically significant improvements in WOMAC scores at 1 month (14% improvement) [30]. However, at the 6-month timepoint, there was a 0% improvement compared to baseline. At 12 months, there was a 6% decrease in WOMAC score, suggesting the durability of HA was attenuated over time.

CRFA was associated with an increase in the general health of subjects, as reported by self-reported measures GPE and EQ-5D-5 L. The majority (63.6%) of subjects receiving CRFA reported improved GPE at the 12-month timepoint. Those within the CRFA reported a change in EQ-5D-5 L of 0.14 points from baseline, which exceeds the minimal clinically important difference in EQ-5D-5 L of 0.074 [31].

This study also demonstrated that CRFA can be offered to patients who continue to experience pain and discomfort following viscosupplementation injections. The previously reported 6-month results from this trial indicate that HA does not provide extended pain relief when managing chronic knee pain caused by osteoarthritis [20]. Additionally, the majority (83%) of subjects within the HA cohort elected to crossover and receive CRFA after 6 months. Once these crossover subjects received CRFA, there was a significant reduction in pain relief after CRFA treatment. A higher percentage of subjects within this cohort (64.5%) reported ≥50% pain relief at the 12-month timepoint than in previous trials (48.6%) [15], although the crossover cohort in the Davis et al. trial first received IAS then CRFA. Other trials studying pain relief following HA have reported sustained decreases in WOMAC pain subscale score [32]. Subjects reported a baseline of 7.52 ± 0.58 that remained lower at 6 weeks (4.66 ± 0.47), 12 weeks (5.00 ± 0.60), 24 weeks (5.00 ± 0.50) and 52 weeks (4.00 ± 0.60).

Of note, patients who elected to crossover did experience measurable improvements in pain, as their NRS scores lowered 2.0 points at the 6-month time period following crossover to receive CRFA, representing an improvement of 36.1%. This change in NRS meets the minimal clinically important change previously reported as a decrease by 2.0 or a percent chance score of − 33.0% [33]. Additionally, during this time period, subjects within this cohort saw a 27.7% improvement in WOMAC pain score, which exceeds the minimal clinically important difference in osteoarthritis [29].

Quality of life measurements showed improvements in the crossover cohort. Subjects in this cohort reported an increase in GPE from their 6-month to 12-month timepoint. Previous studies have demonstrated that HA has an approximate 6-month durability of patient impression. Chevalier et al. showed that 33.9% of subjects reported doing ‘very well’ or ‘well’ at 26 weeks after treatment, as measured by patient global assessment [11], which closely mirror the GPE scores at the 6-month timepoint in this trial. However, GPE scores increased significantly in the crossover cohort following CRFA treatment, suggesting that the improvements in patient impression were the result of CRFA treatment. Furthermore, this cohort saw increases in EQ-5D-5 L score after receiving CRFA, suggesting that the overall impression of treatment was positive following the CRFA procedure.

Several subjects (n = 14) within the origin HA cohort that did not crossover to receive CRFA. These subjects were not deemed medically appropriate candidates for CRFA by their treating physician. Subjects that did not crossover to receive CRFA reporting 12-month outcomes (n = 11) saw long-term benefits in terms of pain and function at this timepoint. Study investigators were surprised to see long-term pain relief from this cohort, as summaries of clinical literature do not suggest long-term durability of this treatment. One clinical trial demonstrated that HA treatment had a durability of approximately 3 months [34] while other studies have demonstrated that pain relief can extend to 26 weeks [11]. Meta-analyses assessing the benefits and risks of viscosupplementation for adults with symptomatic knee OA have concluded that it is associated with a small and clinically irrelevant benefit [35]. However, the clinical trial reported herein was not powered to examine the long-term durability of HA and firm conclusions cannot be made on such a small sample size.

Overall, AE profiles were similar across cohorts and were consistent with published literature. No SAEs related to either procedure were observed. Skin burns have been reported following traditional RF procedures [36]. One instance of skin burn was previously reported following CRFA in this trial at the lower medial needle insertion site, which resolved on its own [20]. Review articles have suggested that vascular injuries may be a risk during CRFA procedures [37], and there are case reports of hemarthrosis [38] and septic arthrosis [39] following CRFA for knee OA but none have been observed in the clinical trials previously conducted [15—17,19,20], nor were observed in this study.

Since trial inception, there have been a number of anatomical studies of the knee published. Genicular nerve targets for this trial were based on available literature [22]. Tran et al demonstrated a total of 10 nerves innervating the knee [40]. Fonkue et al recently published work describing 5 potential genicular nerve targets [41]. It is possible that optimization of the procedure informed on recent anatomical work may lead to even better clinical outcomes in future studies. However, the targets used in this study resulted in consistent, favorable clinical outcomes with prior CRFA studies [15, 16, 19].

Limitations of this study included the lack of blinding as a result of pragmatic study design. The open-label nature of the trial allowed the opportunity for bias. Additionally, this trial was designed as a single-arm crossover study, so subjects receiving CRFA were not eligible to receive HA injections. Some subjects with minimal or no pain (i.e. an NRS score of 0.0) crossed over from HA to CRFA. However, the majority of the subjects within the crossover cohort did report significant improvements in pain following CRFA.

Overall, these results closely mirrored the durability of CRFA demonstrated in previous trials extending 12 months [15—17]. Furthermore, this study demonstrated that subjects who received HA prior to CRFA can still receive substantial benefit from CRFA. At the 12-month timepoint, subjects in both CRFA and crossover cohorts reported lowered NRS pain scores. These study results suggest that patients may benefit by receiving CRFA initially rather than HA, but those that receive CRFA after HA may still expect improvement in outcomes.