Spasticity is a common motor disorder that occurs after a stroke in approximately 30% of patients. This symptom is a component of the upper motor neuron sign and varies considerably with a velocity-dependent increase in tonic stretch reflexes and exaggerated tendon jerks. It limits the rehabilitation success rate; interferes with posture, sleep, and gait; and can be painful. Various management strategies, including physical therapy, oral medications, chemodenervation, and surgical procedures, have been suggested for spasticity [1]. However, oral medications act on the central nervous system, and chemodenervation has the risk of post-injection weakness includingneutralizing antibodies. Surgical procedures, such as surgical neurolysis, selective neurotomies, and rhizotomies, are usually performed in patients with consistent spasticity after non-surgical procedures.

Radiofrequency (RF) energy, which ablates the targeted nerves or tissues through frictional heating, has been used in various medical and surgical fields. The RF neuroablative technique has been used to reduce the gastrocnemius muscle (GCM) in healthy individuals for cosmetic purposes. The procedure preserves the skin except for a small incision, and ankle movement can be compensated by the soleus muscle [2]. RF energy at different frequencies, known as RF neuromodulation, has been reported to have favorable outcomes in the treatment of pain from an innervated structure [3]. RF, as neuroablation, has been intermittently used to treat intractable spasticity after spinal cord injury or cerebral palsy [4], as well as brain injury [5]. However, no consensus exists regarding neuroablation protocols using RF.

This case study aimed to investigate the effect of neuroablative thermocoagulation using percutaneous RF on spastic foot management in a patient with stroke and determine its safety and persistence.

This study was exempted from review by the Institutional Review Board (IRB) of a tertiary hospital (IRB No. 2020-0581-001). This was a prospective study, and a subject with subacute or chronic stroke was recruited by a clinician during the outpatient rehabilitation period after obtaining approval from the IRB. The inclusion criteria were as follows: 1) hemiplegia caused by unilateral stroke, 2) >3 months post-onset, 3) age >19 years, 4) Modified Ashworth Scale (MAS) score in the lower extremity >grade 1＋, and 5) ability to express subjective gait discomfort. The exclusion criteria were as follows: 1) peripheral neuropathy or neuromuscular disease, 2) history of fracture or operation of the lower extremity, 3) history of chemical denervation within 3 months, 4) secondary bacterial infections of the dermatome, and 5) uncontrolled bleeding.

A 59-year-old man had a history of intracerebral hemorrhage in the left basal ganglia 15 years prior to presentation and hypertension and paroxysmal supraventricular tachycardia managed with oral medications. The patient had no cognitive impairment, and his body mass index (BMI) was 22.1 kg/cm2. His ankle plantar flexor spasticity was assessed as grade 2 using the MAS, and weakness was assessed using manual muscle examination. He had a history of alcohol injection into the lower extremity to reduce spasticity a few years ago and had been taking oral baclofen 10 mg for 5 years. He typically exhibited right hemiplegic gait, circumduction, hip spiking with compensated trunk lateral flexion, and claw toe when asked to walk outdoors with the UD-flex. No symmetric or rhythmic gait pattern was noted because of uncontrolled muscle hypertonia, although he could walk outdoors with a habituated pattern. Functional evaluation revealed gait disturbance with a score of 44 on the Berg Balance Scale (BBS) and 34 m in the 2-minute walk test (considering safety) (Table 1). The circumference of the calf, including the GCM, was estimated to be 337 mm at the infrapatellar 10 cm to confirm the hypotrophy induced by the neuroablative technique. The study was fully explained to the patient, and informed consent was obtained. Antiplatelet agents were skipped for 2 days before the intervention.

Table 1 . Outcomes of the Subject at Baseline and Post-intervention.

Outcome measures		Range	Values
			Baseline	1 Week	4 Weeks	14 Weeks
Passive range of motion (Rt.)	Ankle dorsiflexion with knee extension	Angle	0	10	10	10
	Ankle dorsiflexion with knee flexion	Angle	20	20	20	20
Manual muscle test (Rt.)	Knee extensor		Good	Good	Good	Good
	Ankle dorsiflexor		Fair	Fair	Fair	Fair
Ankle plantar flexor		Trace	Poor -	Poor -	Poor -
Calf circumference (Rt./Lt.)	Infra-patellar 10 cm (including gastrocnemius)	mm	337/355	335/355	320/355	328/355
Modified Ashworth Scale (Rt.)	Gastrocnemius	Grade	2	2	2	1+
2-Minute walk test		m	34	35	33	30
Berg balance scale		Score	44	44	44	44
Physician global assessment		−4∼＋4	.	3	2	0

Another experienced rehabilitation physician performing the neuroablative technique inspected the shape and contour of the medial head of the GCM by resisting the action of inducing maximum contraction of the muscle on the day of the procedure. In addition, the motor nerves innervating the medial GCM were examined using high-resolution ultrasonography (Accuvix V10c system, Samsung Medison Co., Seoul, South Korea) with a linear-array probe (5 to 12 MHz) and surface electromyography/stimulation device (Dantec CLAVIS, REF-9015A0012, Denmark). Mapping the surrounding structures, such as the medial sural cutaneous nerve, popliteal vein, and small saphenous vein of the right leg, was necessary before the intervention. The procedure was performed in the prone position to minimize interference from spasticity (Fig. 1-A). An RF probe connected to a 400-kHz electrical surgical unit (ITC-300D, ITC Co., South Korea) was inserted into the skin below the popliteal crease after administration of local anesthesia using 1% lidocaine (Fig. 2). The motor nerve branches that supplied the medial GCM were continuously ablated using conventional radiofrequency (CRF) (Fig. 1-B). A 120-mm- long, 17-gauge custom-designed bipolar RF-insulated needle probe (Dr. Oppel RF generator by Sometech Medical Co., South Korea and Double Impact by Neomedicals Co., South Korea) with a 7-mm-long exposed active tip facilitated precise ablation of nerve branches with minimal injury to adjacent tissues [2,6]. A temperature of greater than 70°C is produced by alternating current of 100 mA. The subject was administered oral antibiotics and nonsteroidal anti-inflammatory drugs for 5 days. No complications or adverse events were observed.

Figure 1. Intervention approaches. (A) Radiofrequency procedure after local anesthesia in the prone position. (B) The anatomy of the posterior aspect showing the nerve with the popliteal fossa. The motor nerve branch to the medial gastrocnemius was ablated with a radiofrequency probe. The sural cutaneous and communicating nerves run subfascially along the midline of the gastrocnemius bellies. Damage to the sensory nerves must be avoided. N: nerve, P: plantaris, G: gastrocnemius.

Figure 2. (A) Radiofrequency probe (black stick) was connected to a 400-kHz electrical surgical unit. (B) The probe was inserted through a small incision below the popliteal crease to motor nerve branch.

The patient was reported to be satisfied with the hypotonic muscle while walking in the first week of the procedure. He reported an overall lighter sensation in his right leg, and the rigidity caused by the hypertonic muscle was reduced. The physician global assessment, a scoring system from −4 to ＋4 used to measure the overall response to treatment as assessed by the physician, for change as a worsening, no change, or improvement, also showed ＋3 scores. The score represented an improvement in the patient’s life after the intervention [7]. The passive range of motion (PROM) of ankle dorsiflexion with knee extension improved to 10° at that time, and muscle weakness improved. The 2-minute walk test was estimated to be 35 m with a slightly smoother alternative step during the functional assessment. The MAS grade did not change. His claw toe was reduced during the walking test after 4 weeks. The PROM, manual muscle test (MMT), and MAS results were consistent. He reported that his feet felt heavier while walking after 14 weeks; however, his feet were still better than those before the procedure. The MAS score improved to grade 1＋, and the claw toe was reduced. At rest, the inverted ankle improved slightly to a neutral position. The circumference of the calf was 328 mm at the infrapatellar 10 cm, and the PROM, MMT, and BBS were the same. However, the gait speed decreased in the 2-minute walk test (Table 1).

To the best of our knowledge, this is the first report on the effect of an RF neuroablative technique on spastic foot management in a patient with stroke as well as to present its safety and persistence.

The first selective neurotomy for the treatment of spastic deformities of the foot was reported by Stoffel in 1911 and was a surgical procedure involving partial resection of the motor nerve branches innervating the spastic muscles. In the 1990s, the neurosurgical team reintroduced the technique of electrical stimulation. After the extent of nerve sectioning, the sprouting process explains the recovery strength to baseline after 8 to 12 months. Biosignal transmission between the efferent (motor) innervation of extra- and intrafusal muscle fibers and the afferent (sensory) innervation of muscle spindles should be inhibited by selective peripheral neurotomy or neurolysis to decrease spasticity [1]. However, the anti-spastic effects of these procedures vary, and invasive procedures have considerable side effects such as paresthesia and pain.

To date, many studies have investigated the effects of RF on pain, cosmetic management, as well as spasticity [4,5,8]. Two ablation protocols exist: CRF and pulsed radiofrequency (PRF). In CRF application, a 420-kHz high-frequency alternating current of 90∼125 mA generates a temperature of 60°C to 80°C, which results in selective thermocoagulation of nerve fibers. In PRF, the same frequency alternating current at 200 mA produces a temperature below 45°C. Heat is dissipated by thermal conductance into the surrounding area by bursts and quiet phases [9]. However, PRF does not guarantee nonthermal lesions, and appropriate parameters can be determined by considering the relationship between the electric and thermal fields generated by RF [10].

In this study, our patient responded positively to the neuroablative technique of the motor nerve branch without side effects. The application protocol was different from that used in previous studies. In addition, previous studies have focused on deafferentation and targeted the sensory nerve fibers or the muscle itself [2,4,5]. Paresthesia, a common side effect of thermal lesions in CRF, was not believed to have occurred because the electric field was operated differently. The equipment was produced using impedance-dependent thermocouple monitoring [2] and not through a thermosensor on the electrode tip to prevent effective current delivery into the tissues. In addition, physicians must map the precise anatomy and electrical stimulation using US before procedure to avoid damage to the lateral sural nerve and saphenous vein.

The muscle itself with the same CRF parameters, was effective even after 3 months when targeting healthy people according to a previous study; however, a limit to the duration of the effect exists because the state was not tracked afterward [6]. Cryoneurolysis has been reported to be effective among the neuroablation methods even after 1 year in a case report of a patient with SCI. In addition, both CRF and PRF have different mechanisms in previous studies targeting pain; however, nerve regeneration can be explained [8]. Additional RF research on the motor nerves is required.

A previous study revealed that the volume reduction in calves for the same procedure, evaluated using ultrasonography, had an average of 2.5 cm and was consistently maintained after 6 months [2]. This study showed a volume reduction of 1.7 cm at 1 month, which was preserved for three months after the procedure. As his muscle tone was controlled, his ankle plantar flexors and PROM improved at 3 months. Accordingly, he experienced improvement in his symptoms; however, he did not experience much improvement in functioning. The patient with chronic stroke 15 years prior already had hypotrophied muscles, and his gait kinematics and muscle activity responded to dysfunction of the GCM when compared with the pre-procedure pattern. In addition, more elaborate outcomes, such as gait analysis or pedobarography, are needed for comparison before the procedure to identify subtle differences. Increased experience with RF and a better understanding should be investigated to determine optimal parameters.

Surgical treatment is not commonly used for the management of spasticity. However, to date, intraoperative monitoring, technical development, and novel techniques for modulation or ablation of the targeted nerves have been reported [1,6]. This study showed the effect of a neuroablative technique using RF on spasticity in a patient with chronic stroke, as well as its safety and persistence for 3 months. Further studies are required to compare it with conventional treatment during the subacute phase and for long-term follow-up after 3 months to achieve optimal outcomes of percutaneous RF.

No potential conflict of interest relevant to this article was reported.

This study was supported by the Hafis Clinic.