Introduction: Percutaneous radiofrequency (RF) rhizotomy of the cervical medial branches is a palliative treatment that is used to relieve pain arising from cervical facet joints. The currently accepted and validated technique for blocking medial branches and performing RF lesioning involves the precise placement of needles on the waist of the cervical articular pillars under fluoroscopic C-arm guidance, although the use of US offers a potential alternative to fluoroscopic guided needle placement.
The technique of US-guided RF cervical medial branch rhizotomy was well established in our previous study, in which we demonstrated that the hypoechoic band surrounding the lateral curved aspect of the articular pilla was the presumed cervical medial branch and the target point of US-guided RF rhizotomy. However, to date, no studies on US guided cervical medial branch blocks or RF rhizotomy in symptomatic patients have been published. In this study, our objective was to apply US-guided medial branch blocks and rhizotomy techniques in symptomatic patients. To achieve this objective, we performed a two-part study; first, we estimated the accuracy of US for locating the medial branches, and second, US-guided RF rhizotomy was performed on selected symptomatic patients.
Materials and Methods: 1) US visualization of cervical medial branches in normal volunteers
We recruited 20 volunteers from a private, outpatient primary pain center (10 men, 10 women: mean age 47.1 years). We developed a grading system to describe the visibility of medial branches in vivo under US, which we refer to as the subjective US visibility of medial branch (SUVMB) scoring system. SUVMB were scored as follows: 1- poor, 2- good, and 3- excellent. To identify potential technical challenges when utilizing US guidance, SUVMB scores were compared with age, sex, weight, body mass index (BMI), neck circumference, skin thickness, depth of medial branch, and multifidus cross sectional area (Table 1).
Participants were examined in a side-lying position using a 12 to 8 MHz linear probe (Logiq 5; GE Health Care, Milwaukee, WI). The US feature of a presumed medial branch is the curvilinear hypoechoic band coursing from the lateral aspect of an ipsi-segmental articular pillar parallel to the intermuscular plane (figure 1). Determinations of cervical level were confirmed by counting the number of articular pillar waists caudally from C2-3 and/or cephalad from the transverse process of C7 during longitudinal US scans.
2) US guided medial branch blocks and rhizotomy in symptomatic patients
We recruited 20 consecutive patients with chronic non-specific cervical pain of greater than 6 months duration from a private, primary pain center. Prior to US-guided medial branch blocks, individual demographic and cervical physical attributes were reported (Table 1). The presumed symptomatic level was determined by history taking, a physical exam, radiography, CT or MRI, and corresponding zygapophyseal pain maps. In addition, diagnostic block levels were determined usually by using pain distribution patterns: C4, C5, and C6 medial branches were chosen for type A and B pain distributions, and C7 was included for a type C distribution (figure 2).
Zygapophyseal pain was diagnosed based on positive responses to US-guided cervical medial branch blocks using the double comparative block paradigm. A positive response was defined according to the ISIS criteria. Pre-rhizotomy NRS scores were reported. Follow-up data were collected by an independent reviewer during a phone interview with each patient at 1, 3, 6 and 12 months postoperatively. Outcomes were categorized into four functional categories (excellent, good, fair, and poor) based on NRS scores and patient satisfaction ratings .
The US-guided cervical medial branch block and RF rhizotomy techniques
With a patient positioned on his/her side with a pillow under the neck to maintain it in a neutral position, the skin is prepped with betadine and sterile gel. A linear probe of 12 - 8 MHz is then placed inside a sterile plastic cover and applied to the symptomatic side of the neck. After identifying the presumed medial branch, the needle trajectory is planned on axial view. A ¡°free hand technique¡± is used to insert the needle and a 2.5 inch, 25 gauge needle is used to puncture the skin 5-10mm from the lateral end of the probe, and the needle is then advanced transversely at an angle of 0¡Æ to 20¡Æ with respect to the surface of the probe using the ¡°in-plane technique¡± (figure 3a). Prior to infiltrating the local anesthetics, the final needle position is confirmed by fluoroscopy (figure 3b).
Prior to RF lesioning, sensory stimulation is performed to confirm appropriate needle position using an RF generator (Cosman RFG-1B, Burlington, MA, USA). We used a 22 gauge 100 mm RF cannula with a 5 mm exposed active tip (Cosman Curved-Tip RF Cannula, Burlington, MA, USA), and a standard RF coagulation technique. In the present study, the operator held the RF cannula in place during heat lesioning under real time US to ensure constant contact was maintained between the active needle tip and medial branch. Results: (1) US guided visualization of cervical medial branches in normal volunteers
Normal volunteers had the following characteristics (Table 1). The cervical medial branches were clearly visualized by US at C3-7 with an average SUVMB score of 2.65 +/- 0.70 (Figure 4). At C7, the SUVMB score was significantly lower (p¡Â0.05), and the visualization of the medial branches was most challenging in this region. No patient demographic factors or physical characteristics reached a statistically significant correlation with SUVMB scores (p¡Ã0.05).
(2) US guided medial branch blocks and RF rhizotomy in symptomatic patients
Subjects were (8 men, 12 women) of mean age of 51.3 years. Average duration of pain was 21 ¡¾14.13 months with NRS score of 6.6¡¾1.01. No significant differences in demographics/physical characteristics were observed between normal and symptomatic subjects (p¡Ã0.05) except for cross sectional area of the multifidus (p¡Â0.05) (table 1). Seventeen of the 20 patients (85%) showed a positive response to the first US guided block, and became candidates for a 2nd US guided block, in which ten patients (58.8%) showed positive response. Total numbers of selected medial branches were as follows, 37(C4), 37(C5), 37(C6), 10(C7). Ten of the 20 (50%) patients were diagnosed as having cervical zygapophyseal joint pain based on double comparative block paradigm. Three of these 10 patients agreed to undergo US guided RF rhizotomy. Seven patients did not agree to undergo RF rhizotomy; 5 patients were satisfied with pain relief after medial branch blocks, and 2 patients declined RF rhizotomy after recurrence.
Three patients underwent US guided RF rhizotomy (3 females; age:42 ¡¾1.6 years; weight 53.33 ¡¾2.49 kg; NRS 7.33¡¾1.24; pain duration 26 ¡¾14.14 months). In all patients, only medial branches of C5 and C6 positively responded to sensory stimulation and were selected for RF rhizotomy. Average post-rhizotomy NRS scores were 4, 2.6, 2.3, and 3.3 at 1, 3, 6, and 12 months postoperatively, respectively. Excellent results were observed in 2 patients (pre-op NRS: 7.5, 12 months NRS: 2), and good results were achieved in 1 patient (pre-op NRS: 7, 12 months NRS: 4). No complications were observed.
Conclusion: This is the first study to demonstrate that US can accurately locate the C3-7 medial branches in vivo. The study also describes the safe and efficacious use of US guidance for both medial branch blocks and RF rhizotomy in symptomatic patients.
|
