The effects of genicular nerves radiofrequency ablation treatment on postural stability in patients with knee osteoarthritis

Abstract

Objectives

This study aimed to investigate the effects of genicular nerves radiofrequency ablation (GNRFA) treatment on postural stability in patients with knee osteoarthritis.

Patients and Methods

Twenty-two patients were included in the study. Postural stability was assessed using the walk across test, sit-to-stand test, step-up and over test and modified clinical test of balance in sensory interaction tests with the NeuroCom Balance Master device. A numeric rating scale and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) criterion was used to determine changes pain levels and the functionality. Three evaluations were performed before treatment, 3 weeks and 3 months after treatment.

Results

After GNRFA, changes in patients’ postural stability parameters showed that the lift-up index and the movement time improved significantly in both lower extremities 3 weeks and 3 months. There was a significant reduction in pain levels [8.09±1.01 to 2.41±1.73 in the 3^rd week and to 4.18±1.79 in the 3^rd month (p<0.0001 for both)]. A significant improvement in WOMAC functionality scores was seen; they started with a mean of 58.69±11.25 at baseline and changed to 23.05±12.08 at the 3^rd week and 36.15±16.45 at the 3^rd month (p<0.001 for both).

Conclusion

After GNRFA treatment, the lift-up index and the movement time, which are the dynamic postural stability parameters has improved significantly. These results show that GNRFA treatment contributes positively to postural stability.

Osteoarthritis (OA) is the most common cause of knee joint degeneration. It is characterized by degeneration of articular cartilage and failure of the joint and surrounding structures.^{[1, 2]}

There are many conservative options for the treatment of knee OA. Genicular nerve radiofrequency (RF) ablation (GNRFA) is used in patients who have failed conservative treatment and have failed surgical intervention or are not suitable for surgical intervention. GNRFA is used to restore joint function and relieve pain by destroying the nerves innervating the painful knee or by disrupting the transmission of pain signals. GNRFA provides short-term (3 to 6 months) and sometimes longer-term pain palliation in patients. GNRFA has also been shown to be safe in most studies.^[3] Besides, pain may even persist in some patients after knee replacement.^{[4, 5]} Therefore, GNRFA treatment is an increasingly popular modality for suitable patients.^[5]

Knee OA has an impact on individuals’ postural stability and pain perception. It is essential that understanding the complex relationships between these factors for knee OA rehabilitation.^[6] In OA, the number of mechanoreceptors in the knee joint is also reduced, which impairs the proprioception provided by the knee. This situation interferes with postural stability.^[7] Pain, which is a prominent symptom in knee OA, can affect functionality. Additionally, pain can exacerbate functional limitations and affect the relationship with postural stability.^[8]

A review of the literature has not established a definite cause-and-effect relationship between reduced knee joint pain and balance control.^[9] Although decreasing the sensory inputs of the patients provides pain relief and an increase in functionality, decreased proprioceptive sense may cause postural instability.^[10] Genicular nerves receive sensory signals from the joint capsule and GNRFA targets the terminal branches of these nerves. Since the genicular nerves also travel around the periosteum before the joint capsule, they probably affect proprioception and balance control.^[5] In a review evaluating patients with persistent pain after total knee arthroplasty, an increase in the rate of falls was reported in some patients. It was emphasized that the reason for this was the decrease in proprioceptive inputs.^[11] However, another recent study found that GNRFA had a positive effect on dynamic balance and did not cause any decrease in static balance.^[5] Consequently, in the literature, there are different results regarding the effect of GNRFA on postural stability. While pain reduction is the primary outcome measure in studies evaluating GNRFA, studies on its effect on physical performance, balance and postural stability are insufficient.^{[3, 4, 5, 6, 7, 8, 9, 10, 11, 12]}

Our primary aim was to observe the changes in the postural stability parameters of patients with knee OA after GNRFA treatment. Our secondary aim was to evaluate the effect of GNRFA on pain and functionality of the patients with knee OA.

PATIENTS AND METHODS

Participants

This prospective study was conducted at the interventional pain clinic of department of physical medicine and rehabilitation of a tertiary care center between January 2019 and April 2020. Patients with knee OA planned for GNRFA due to unresponsiveness to conservative treatment include exercise, pharmacological agents and physical treatment were considered for the study.

Patients with stage 3 or 4 OA, according to Kellgren-Lawrence staging, who did not respond to conservative treatment and agreed to participate were included in the study.

Exclusion criteria were acute knee pain (less than 3 months), knee OA stage 1-2 according to Kellgren-Lawrence staging, presence of neurological disorders such as a history of polyneuropathy that reduces balance and increases the risk of falls, presence of diabetes mellitus, presence of severe neurological and psychiatric disorders that might interfere with understanding commands and speech, signs of active infection around the knee and cardiopulmonary disease that might cause hemodynamic instability, intra-articular injection into the knee joint in the last three months, total knee joint replacement or implantation of a pacemaker, taking anticoagulant drugs, taking drugs that might affect balance, being pregnant or having a history of bleeding disorders.

Experimental protocol

An anteroposterior fluoroscopic view of the knee was taken from the patient in the supine position. Three target points were set for injection: the medial and lateral diaphyseal-metaphyseal transition points of the femur, and the medial diaphyseal-metaphyseal transition point of the tibia. These three target points were determined. Then, the RF cannula was placed with fluoroscopic guidance at these three points and local analgesia was administered with 2 cc of 1% lidocaine. Each RF cannula was connected with an RF electrode to the output jack of the RF generator. First, sensory stimulation was performed from each RF cannula. It was questioned whether the patient felt pain or any sensory stimulus. Then, motor stimulation was performed from the RF cannula to prevent motor branch damage. GNRFA was then performed at 80 °C for 90 seconds. The superior lateral genicular nerve, superior medial genicular nerve, and inferior medial genicular nerve were targeted in GNRFA. After the procedure, one-third of the mixture of 40 mg of triamcinolone, 1 cc of 0.5% bupivacaine, and 1 cc of saline solution was injected separately to prevent neuritis in genicular nerves.

Patients included in the study were evaluated by a physician who was not involved in the interventions. Patients were assessed before treatment and 3 weeks and 3 months after treatment.

The pain levels of the patients were determined before the procedure and 1 hour after the procedure, while the patient was being observed. Then, pain levels were questioned at each visit at the 3^rd week and 3^rd month. A numeric rating scale (NRS), a pain scale, was used to measure and monitor pain intensity with which “0” was scored as no pain and “10” as the most severe pain possible. The patients scored their pain between 0 and 10.

The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) criteria were used to determine changes in the functionality of the patients. The WOMAC criteria have three separate sections: pain, stiffness, and physical function, with 24 questions. Each question is scored according to a Likert scale as 0=absent, 1=mild, 2=moderate, 3=severe, and 4=very severe. The score of each section is calculated on its own and a total score is obtained. This score ranges from 0 to 100. High scores indicate increased pain and stiffness and impaired physical function.

Postural stability

Postural stability of patients was assessed using the NeuroCom Balance Master device. Healthy volunteers were also assessed to understand how much the patients’ initial postural parameters were affected. This method is a valid and reliable method for assessing postural stability.^[13] Walk across (WA) test, sit-to-stand test, step-up and over test and modified clinical test of balance in sensory interaction (mCTSIB) were used. The postural stability of the patient group was assessed before the procedure and again 3 weeks and 3 months after the procedure.

The WA test is designed to evaluate the patient’s normal walking pattern on a force plate. This test assesses the average stride width and length, the symmetry between the stride lengths of the left and right legs, and the walking speed.

The sit-to-stand test involves rising from a stable surface positioned on the platform. This test focuses on two key aspects. First, it measures the initial swing rate, indicating how much the patient sways as they stand up. Second, it assesses the symmetry of weight distribution.

The step-up-and-over test involve several steps: stepping up onto a step with one leg, swinging the trailing leg over the step, and then placing the trailing leg onto the platform. During this test, three key parameters are assessed: the lift-up index, which measures the maximum lifting force exerted by both legs (expressed as percent lifting power per kilogram); the impact index, which evaluates the force used to lift the body in the vertical plane; and the movement time, which refers to the duration taken to step down.

The mCTSIB assesses static balance by measuring the speed of center of gravity oscillation (in degrees per second) while the individual stands on the platform under various conditions: with eyes open on a firm surface (firm EO), eyes closed on a firm surface (firm EC), eyes open on an unstable surface (foam EO), and eyes closed on an unstable surface (foam EC).

Ethical statement

This study was approved by the Marmara University Faculty of Medicine Local Ethics Committee and conducted in accordance with the declaration of Helsinki Ethical Principles for Medical Research Involving Human Subjects (protocol number: 09.2018.663, date: 05.10.2018). Written and verbal informed consent was obtained from each subject.

Statistical analysis

Since there were no previous studies that investigated the change of postural stability parameters before and after GNRFA, sample sizing was based on the changes in visual analog scale (VAS). According to the results a previous study,^[14] when the effect size of change in VAS is 0.6, with an alpha of 0.05 and a power of 85%, at least 21 patients was planned to be recruited. A post-hoc strength analysis showed that the according to the change in left leg lift-up index in step up and over test, the study has reached a power of 90% with 22 patients included with an alpha of 0.05. The Statistical Package for the Social Sciences version 20.0 (IBM Corp., Armonk, NY, USA) were used for data analyzes. To evaluate the changes during the study period in postural stability parameters, repeated measures analysis of variance (ANOVA) with a Bonferroni correction was used because the data were normally distributed according to the Shapiro-Wilk test, and the patients were evaluated in three consecutive occasions during the study period. Because three measurements were made, p<0.025 was considered significant in repeated analyses.

RESULTS

A total of 30 patients were included. 8 patients were dropped out of the study because they did not attend their 3 month follow up. A total of 22 patients were included in the final analysis. There were 19 (86%) female participants in both groups. The patient flowchart can be seen in Figure 1. The mean age of the patients was 57.86±4.61 years. The mean body mass index of the patients was 31.67±5.24 kg/m². Seventeen (77%) of the patients had Kellgren-Lawrence grade 3 OA. The average duration of patients’ knee symptoms was 48.4 months.

Figure 1.

Flowchart.

WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index.

There was a significant reduction in pain levels and a significant improvement in WOMAC pain, functionality, and stiffness scores in the patient group at all checkpoints. The mean pain score decreased from 8.09±1.01 to 2.41±1.73 in the 3^rd week and to 4.18±1.79 in the 3^rd month (p<0.001 for both with respective effect sizes of d=4 and d=2.69). WOMAC pain scores decreased from a mean of 50.00±12.81 at baseline to 17.95±10.76 at the 3^rd week and 28.72±14.29 at the 3^rd month (p<0.001 for both, with respective effect sizes of d=2.70 and d=1.56). WOMAC stiffness scores decreased from 40.22±25.07 to 11.93±15.65 at the 3^rd week and 19.88±20.65 at the 3^rd month (p<0.001 for both with respective effect sizes of d=1.35 and d=0.88). A significant improvement in WOMAC functionality scores was seen; they started with a mean of 58.69±11.25 at baseline and changed to 23.05±12.08 at the 3^rd week and 36.15±16.45 at the 3^rd month (p<0.001 for both with respective effect sizes of d=3.02 and d=1.59). Complications during and after the procedure were not observed.

Changes in patients’ postural stability parameters showed that the lift-up index and the movement time improved significantly in both lower extremities 3 weeks and 3 months after the intervention. Other parameters showed no significant changes. The results of these changes were summarized in Table 1.

TABLE 1. Changes in postural stability parameters of the patient group.

Parameter	Baseline mean ± SD	3rd week mean ± SD	3rd month mean ± SD	Baseline vs. 3rd week p-value (95% CI)	Baseline vs. 3rd month p-value (95% CI)
Walk across	-
Step width (cm)	17.51±3.18	16.66±2.63 -	17.76±2.43 -	p=0.09 (-0.14; – 1.83)	p=0.56 (-1.16; 0.65)
Gait speed (cm/s)	53.94±9.03	52.85±9.42	54.65±10.55	p=0.63 (-3.63; -5.81)	p=0.73 (-5.00; -3.59)
mCTSIB sway velocity (°/s)
Eyes open, soft surface	0.86±0.18 -	0.90±0.22 -	0.85±0.23 -	p=0.41 (-0.03; 0.13)	p=0.88 (-0.08; 0.06)
Eyes open, firm surface	0.51±0.30 -	0.46±0.15 -	0.51±0.20 -	p=0.34 (-0.06; 0.17)	p=0.92 (-0.12; 0.13)
Eyes closed, soft surface	1.22±0.27 -	1.13±0.29 -	1.18±0.26 -	p=0.20 (-0.12; 0.05)	p=0.33 (-0.11; -0.13)
Eyes closed, firm surface	0.52±0.18 -	0.47±0.16 -	0.53±0.17 -	p=0.22 (-0.03; -0.13)	p=0.75 (-0.02; - 0.07)
Step-up and over
Lift-up index, left leg (% body weight)	26.11±5.86 -	32.25±21.57 -	32.46±12.37	p=0.20 (-15.84; -3.57)	p=0.01 (-11.27; -1.42)
Lift-up index, right leg (% body weight)	26.72±9.01 -	31.65±7.52 -	29.71±7.77 -	p=0.01 (-8.73; -1.13)	p=0.02 (-5.52; 0.47)
Movement time, left leg (s)	2.71±0.85 -	2.25±0.76 -	2.26±0.42 -	p=0.008 (0.13; 0.78)	p=0.001 (0.08; 0.81)
Movement time, right leg (s)	2.80±1.31 -	2.18±0.81 -	2.36±0.98 -	p=0.001 (0.26; 0.96)	p=0.002 (0.07; 0.79)
Impact index, left leg (% body weight)	67.76±24.76	73.73±40.97 -	73.06±33.44 -	p=0.48 (-23.11; -11.61)	p=0.44 (-19.71; -9.12)
Impact index, right leg (% body weight)	66.05±27.58 -	70.62±30.49 -	65.53±26.36 -	p=0.52 (-19.38; -10.24)	p=0.91 (-9.53; 10.56)
Sit-to-stand
Weight transfer (s)	0.88±0.66 -	0.62±0.70 -	0.57±0.35 -	p=0.14 (-0.61; 0.41)	p=0.14 (-0.81; 0.42)
Rising index (% body weight)	16.02±12.96 -	17.72±11.89 -	18.74±14.11 -	p=0.62 (-8.75; 5.34)	p=0.48 (-10.74; -5.29)
Sway velocity (°/s)	4.42±0.92 -	4.73±1.02 -	4.78±3.99 -	p=0.12 (-0.71; 0.09)	p=0.25 (-7.35; 2.04)

SD: Standard deviation; CI: Confidence interval; mCTSIB: Modified clinical test of balance in sensory interaction.

DISCUSSION

This study aimed to observe the changes in the postural stability parameters of patients with knee OA after GNRFA treatment changes in patients’ postural stability parameters showed that the lift-up index and the movement time improved significantly in both lower extremities 3 weeks and 3 months after the intervention. Other parameters showed no significant changes.

The secondary aim was to evaluate the effect of GNRFA on pain and functionality of the patients. The NRS of pain after GNRFA treatment decreased significantly at the end of 12 weeks compared to the baseline. Similarly, the same results were observed in the pain sub-criteria of the WOMAC OA index. In the other domains of the WOMAC OA index, in which the functionality and stiffness of the patients were evaluated, a significant decrease in the values of the patients continued at the end of 12 weeks. The effect of GNRFA on pain and WOMAC scores were consistent with the literature.^{[5, 15, 16, 17]} This shows that GNRFA can be used to reduce pain and increase functionality of patients as stated in many studies.

When the literature is evaluated, to our knowledge, there is only one study on the effect of GNRFA on postural stability. Thirteen patients underwent GNRFA and then their pain and postural stability were evaluated by Chang et al.^[5] Among the tests applied to the patients, the patients gained a significant acceleration after GNRFA in the step test in which dynamic balance was evaluated, while no change was observed in the single leg stance test, 2-minute walking test, and knee joint proprioception test after GNRFA treatment. But, in here, it was highlighted that there was no significant deterioration after GNRFA application to the nerves that receive knee sensation, and it was emphasized that GNRFA is a method that does not worsen knee proprioception.^[5] Therefore, our study is novel in evaluating the effect of GNRFA on postural stability.

In another study, Hassan et al.^[18] divided the patients with knee OA into two groups and administered intra-articular saline injection to one group and intra-articular bupivacaine injection to the other group. After injections, the decrease in pain level, postural sway and change in proprioception were evaluated in both groups. It was found that a decrease in proprioception was detected in the bupivacaine group. The authors attributed this to the fact that the local anesthetic might affect the neuroreceptors in the joint. Although pain decreased in both groups, the fact that local anesthetics reduce the sense of proprioception, just as they reduce the sense of pain by acting on receptors. Postural sway worsened after application in the local anesthetic group, but it was not statistically significant. This again suggests that intra-articular local anesthetic administration also affects postural sway via intra-articular receptors.^[18] However, our study did not show a decrease in postural stability, rather, patients’ stability parameters increased after intervention. This may show that pain causes a more obvious disruption in postural stability than any intervention done to relieve it. More robust research is warranted in this regard.

In the present study, a significant improvement was found in the movement time and lift-up index parameters, subtests of the step up and over test, which evaluate the patients’ dynamic balance. An increase in the lift-up index suggests that there is an increase in the vertical stepping power of the patients. Similarly, the significant shortening after GNRFA in the movement time, in which the entire stepping swing and then the whole process of putting the foot on the ground is evaluated, proves that dynamic postural balance does not worsen in GNRFA patients. Besides, there was no worsening after GNRFA treatment in the other postural balance parameters. These results were consistent with the study of Chang et al.^[5] that showed that no deterioration in the postural stability of the patients after GNRFA application.

Balance control is very important in both static and dynamic postures. Postural sway is often used as an indicator of static standing balance. However, falls and loss of balance most commonly occur during movement-related tasks such as walking. Therefore, dynamic rather than static balance assessment is more important for the daily life of patients.^[19] In the study of Chang et al.,^[5] the dynamic balance of the patient was evaluated with the step test and they found that dynamic balance increased significantly in a positive direction. However, in the step test, only the balance of the patients on one leg and the time during stepping were measured. A review of the literature suggests that the step up and over test is the most accurate test to assess dynamic balance in the knee. Besides, it is an objective and commonly used test that assesses stair climbing and descending, common activities used in daily life.^[13] Step up and over is a process that involves carrying the weight with both extremities and then shifting the weight between the extremities. This requires both movement of one limb and interaction between both limbs. The lift up index, the first step of the test, measures the power generated by the legs to lift the body higher. In the present study, in addition to movement time, lift up and impact index were also evaluated under the step up and over test. And in the literature, regardless of the effect of GNRFA, when we evaluate the test specifically; step up and over test parameters were found to be better in healthy individuals than in patients in accordance with the current study.^{[13, 20]}

In the present study, no improvement or worsening was found in postural sway after GNRFA. When the patient group was compared with healthy individuals, the sway rate increased significantly in the patient group in accordance with the literature.^[6] In the present study, not only static balance differences but also changes during dynamic balance were measured under the step up and over test in the patient group and healthy individuals. In the literature, static and dynamic balance comparisons between patient groups and healthy individuals were evaluated only with postural sway and step test. In fact, static balance has been analyzed in most of the studies and there are few studies that analyzed dynamic balance.^{[6, 19, 21, 22]} As far as we know, there is no study comparing all aspects of balance like the tests we used in our study. This is another novel aspect of our study.

There are several limitations in this study. Although the number of patients in the study was higher than the number of patients in similar studies in the literature, a very isolated group was evaluated in the study which make it hard for generalizing in a vast group of patients that constitute knee OA. In order for the results of the study to be more reliable, a larger number of patients and a larger population are needed. Another limitation is the lack of a placebo-control group. This was mainly due to ethical limitations. Therefore, the results in this comparison can only be attributed to the effect of OA on postural balance.

In conclusion, after GNRFA, changes in patients’ postural stability parameters showed that the lift-up index and the movement time improved significantly in both lower extremities 3 weeks and 3 months. These results show that GNRFA treatment contributes positively to postural stability.

Also, GNRFA treatment provided significant improvement in the pain of patients with knee OA in their 3-month follow-up. In addition, significant improvements were seen in the WOMAC OA index, which includes pain, stiffness, and function subheadings, at the 3-month follow-up.