Neuropathies Following an Ultrasound-Guided Axillary Brachial Plexus Block

Introduction

A brachial plexus block (BPB) is used for effective perioperative analgesia in a wide range of procedures involving the upper extremities and can be performed through four approaches, namely interscalene, supraclavicular, infraclavicular, and axillary. Currently, the ultrasound (US)-guided method is the most used when performing BPBs and is superior to methods such as the landmark and nerve stimulation (NS) in terms of an improved success rate, a shorter time for peak effect, and a lower incidence of vascular puncture.

However, several studies have reported that the US-guided method did not contribute to reducing the incidence of postoperative neurological symptoms (PONS).^1–4 Since the injection pressure can be high while administering a local anesthetic, injection pressure monitoring (IPM) may be useful in preventing PONS after a peripheral nerve block.^5,6 Among the four approaches, the axillary block is considered the safest and with fewer adverse events compared to the interscalene (eg, phrenic nerve block, spinal cord or vertebral artery puncture) and supraclavicular (eg, pneumothorax).⁷ However, with regard to PONS, a study has showed its incidence was highest in the axillary block,⁸ while another has reported no difference for each of the four approaches.⁹ It is controversial whether the incidence of PONS after an axillary block was higher than that after non-axillary approaches.

In our daily practice, we had the impression that some patients complained of PONS after axillary block. Therefore, we hypothesized that an axillary block could lead to a higher incidence of neuropathies than non-axillary approaches.

To date, literature comparing the incidence of neuropathies after axillary blocks in UGBPBs with the non-axillary approaches is sparse.^10,11

In this study, we investigated whether the incidence of neuropathies due to nerve injury after an axillary block was higher than that after non-axillary approaches under general anesthesia.

Materials and Methods

Study Design and Data Collection

This single-center retrospective cohort study was conducted at the Juntendo University Shizuoka Hospital in Shizuoka, Japan. The study protocol was approved by the institutional review board of Juntendo University Shizuoka Hospital (ID: S19-0707). Patient consent was not required because this was a retrospective study analyzing data collected in daily clinical practice. All patient data were kept anonymized by the authors of this study. All procedures in this study were in accordance with the Declaration of Helsinki (1964). The patients who were aged >18 years and underwent perioperative UGBPBs between January 2014 (UGBPB was introduced at our institution) and March 2020 (UGBPB has been performed under general anesthesia instead of the awake condition) were included. Patients who were administered nerve blocks through an unknown technique or more than two nerve blocks and patients with trauma and preoperative complaints of neurologic deficits at the surgical site were excluded from this study.

PONS following peripheral nerve blocks has various causes such as the original injury, surgical manipulation, nerve injury caused by block manipulation, neurotoxicity of administered solutions, surgical position, and undetermined etiology.¹² A neuropathy due to a UGBPB in this study was defined as a nerve injury caused by block manipulation with a new onset of paresthesia, hypoesthesia, numbness, or muscle weakness occurring within 48 h postoperatively and lasting for more than 5 days;^8,13,14 the sensory deficit area was consistent with the nerve distribution targeted by a BPB. This period was chosen because previous studies have shown that the time of action after a BPB, using 10 mL of 0.25% levobupivacaine, was of approximately 10 h.¹⁵ As for interscalene blocks, most patients reported symptoms within 48 h postoperatively.^13,14 We excluded cases of neurological symptoms that could be explained by the original injury, surgical manipulation, surgical position, or tourniquet use and those with an undetermined etiology from PONS cases. An original injury was defined as a pre-existing deficit detected before surgery. Surgical manipulation was defined as a deficit that was more likely to be associated with the surgical approach than with the preoperative deficit and UGBPB. Surgical position and tourniquet use were defined as deficits associated with the injured site due to the intraoperative position and tourniquet use, respectively. An undetermined etiology was defined as other neurological symptoms. Cases of PONS were investigated by an anesthesiologist who reviewed the patient’s electronic medical records. PONS factors were classified based on the consensus of the neurologist, attending orthopedic surgeon, and anesthesiologist.

In this study, the overall incidence of PONS and neuropathies for axillary and non-axillary approaches attributable to UGBPBs were investigated as the primary outcomes. Age; body mass index; sex; surgical position; block laterality; surgery location; tourniquet use (time and pressure); local anesthetics (amount and type); etiology; symptoms, and outcomes of patients with PONS were investigated as the secondary outcomes.

UGBPB Procedures

First, nerve pre-scanning was performed using a SonoSite S-Nerve US device (SonoSite, Bothell, WA, USA), with a SonoSite linear probe (HFL50x/15–6 MHz). After disinfection under aseptic conditions, single-dose nerve blocks were administered using a 20 G × 80 mm Tuohy needle (Uniever, Unisys Corp, Tokyo, Japan). All nerve blocks were administered under general anesthesia, which was performed using a parallel approach by an attending anesthesiologist or supervised trainee before the surgery; they were not combined with NS or IPM.

The UGBPB techniques for each approach are as follows:

1. Interscalene: The patient was placed in a semi-lateral position with the blocked side up, and a local anesthetic was administered between the C5 and C6 vertebrae, following US-guided identification of the C6 nerve root.
2. Supraclavicular: The patient was placed in a supine position, and a US probe was placed parallel to the clavicle. The first rib, subclavian artery, and brachial plexus were identified, and a local anesthetic was administered around the entire brachial plexus.
3. Infraclavicular: The patient was placed in the supine position, and the axillary artery was identified using a US probe under the clavicle, just medial to the coracoid process and parallel to the long axis of the trunk. The lateral, medial, and posterior nerve bundles around the axillary artery were identified, and a local anesthetic was administered.
4. Axillary: The blocked side arm was abducted 90° and the elbow flexed 90°. A US probe was applied perpendicular to the long axis of the arm; as the musculocutaneous, radial, ulnar, and median nerves were identified, a local anesthetic was administered around each nerve.

Stastical Analysis

The demographic characteristics, type of BPB applied, and complication details were summarized using descriptive statistics. Categorical variables are described as absolute and relative frequencies. The incidence of PONS and neuropathies were described as means (cases out of 1000) and 95% confidence intervals (CIs). The association of categorical variables with the presence of PONS and neuropathies was analyzed using the Shapiro or Fligner test. Statistical significance was set at P < 0.05. The data were analyzed using the Python software (Python Software Foundation, Delaware, United States of America) and its corresponding packages.

Results

Of the 1022 patients initially included in this study, 12 (1.2%) underwent a combination of two nerve block approaches, and 18 (1.8%), an unknown nerve block approach. Of the 992 patients remaining, 143 (14%) and 849 (86%) underwent axillary and non-axillary approaches, respectively (Figure 1). Table 1 shows the demographic data of these patients. Levobupivacaine was the most common anesthetic used in all procedures. There were 19 cases of PONS (19.2:1000; 95% CI, 18.2–20.1), of which neuropathies caused by UGBPB were reported in four (4.0:1000; 95% CI, 3.8–4.2); the incidence of neuropathies was reported in three cases (21.0:1000; 95% CI, 18.1–23.8) for the axillary and in one case (1.20:1000; 95% CI, 1.10–1.20) for the non-axillary approaches, which included one case for the interscalene approach (Table 2). The incidence of neuropathies after the axillary block was significantly higher than that after non-axillary approaches (P = 0.005) (Figure 2).

Table 1 Demographic Data

Table 2 Incidence of PONS and Neuropathy

Figure 1 Flow diagram of the study.

Figure 2 Comparison of the incidence of postoperative neuropathies between axillary (n=143) and non-axillary approaches (n=849); total (n=992).

Considering the 19 PONS cases, 11 (60%), four (20%), two (10%), one (5%), and one (5%) were attributed to surgical manipulation, neuropathies caused by UGBPBs, original injuries, surgical position, and an undetermined etiology, respectively (Table 3).

Table 3 Report of Cases with Postoperative Neurological Symtoms: Background, Diagnosis, Cause and Outcome

Discussion

In this study of 992 UGBPB cases, there were 19 cases of PONS. Among the four cases of neuropathies caused by UGBPBs, three were with the axillary and one with the non-axillary approaches. The incidence of neuropathies after an axillary block was significantly higher than that after non-axillary approaches.

Of the 19 PONS cases, 11 were classified as secondary to surgical manipulation (cases 2–5, 7–12, and 14), which is generally the most frequent cause of PONS.^16,17 The symptoms of four patients (cases 2–5), who were operated on using an arthroscopic rotator cuff repair, were centered on the ulnar nerve area. The orthopedic surgeon diagnosed these cases to be caused by postoperative difficulty in elevating the affected limb. Though PONS after shoulder surgery is multifactorial, it may occur when the elbow remains extended because of postoperative difficulty in raising the affected limb, resulting in an increased pressure in the elbow canal.¹⁸ Case 7 showed ulnar nerve palsy owing to a fixation pin and fully recovered after it was removed 40 days post-surgery. Case 8 showed numbness in the thumb and index finger due to intraoperative injury to the radial nerve, with full recovery 4 months post-surgery. Case 9 demonstrated ulnar nerve palsy after open reduction with internal fixation of the humerus, with partial recovery 10 days post-surgery. Case 10 showed radial nerve palsy due to an intraoperative radial nerve injury, which was diagnosed as an axonal impairment using electromyography. Case 11 demonstrated paralysis of the ring and little fingers owing to intraoperative ulnar nerve injury; however, some symptoms were relieved 1 month postoperatively. Case 14 showed numbness in the thumb, index, and middle fingers due to intraoperative nerve suture; however, the numbness resolved 10 days post-surgery. Case 1 was attributed to the surgical position, which was the beach chair position, and ulnar nerve palsy was observed. Two other PONS cases (cases 17 and 19) were attributed to the original injury. Case 17 had ossification of the posterior longitudinal ligament before surgery, and case 19 had poorly controlled diabetes mellitus for 35 years, with neurological symptoms observed in the same area before surgery.

It is difficult to compare the incidence of PONS after nerve blocks among different studies as they include other nerve blocks in addition to BPBs. Moreover, the definition of PONS, follow-up periods, and implemented methods differed among institutions. Our study investigated neuropathies after UGBPBs; therefore, it was compared with analogous previous studies.

Pablo et al reported the incidence of PONS after UGBPBs as 2.81:1000 (95% CI, 1.70–4.63). Of this, the incidence of neuropathies due to UGBPBs was 0.0:1000 (95% CI, 0.0–15.4) for axillary, 5.16:1000 (95% CI, 2.21–12.0) for interscalene, 2.79:1000 (95% CI, 1.52–5.13) for supraclavicular, and 0.0:1000 (95% CI, 0.0–13.4) for infraclavicular approaches.⁹

Brian et al reported the incidence of PONS after UGBPBs to be 1.8:1000 (95% CI, 1.1–2.7). Of this, the incidence of neuropathies due to UGBPBs was 23.0:1000 (95% CI, 0.6–125.7), 3.5:1000 (95% CI, 1.4–7.3), and 2.0:1000 (95% CI, 0.4–5.8) for the axillary, interscalene, and supraclavicular blocks, respectively.⁸

A former report⁹ regarding whether UGBPBs were performed under sedation or an awake condition is ambiguous, whereas a latter study,⁸ which presented a higher incidence of PONS in the axillary, performed UGBPBs under sedation, similarly to our study. The ASRA recommends that nerve blocks should be performed under awake conditions to prevent PONS.¹⁹ We hypothesized the following as the possible causes for the higher incidence of neuropathies after axillary blocks compared to that after non-axillary approaches: 1) individual differences in the brachial plexus anatomy; 2) a greater number of needle punctures; and 3) a longer subcutaneous needle length. Of the four BPB approaches, the axillary approach was the most peripheral block. Therefore, the brachial plexus and blood vessels are already branched in the axilla, with significantly varying patterns.^18,20 The branches of the ulnar, radial, median, and musculocutaneous nerves are already present at a level that can be discerned using US. Anatomically, the ulnar, radial, and median nerves are located apart from the musculocutaneous nerve; consequently, more than two punctures are often required to individually block all these nerves using a parallel technique.⁷ In conventional axillary nerve blocks, the arterial penetration method is often combined with NS,²¹ and the block needle is inserted vertically under the axillary artery.²² Therefore, the subcutaneous needle length is shorter than that used in a US-guided block, and one puncture is typically sufficient.

In contrast, in the US-guided axillary block used, the total subcutaneous needle length was visible in all cases, and a parallel technique was used. Accordingly, the subcutaneous needle length was greater than that used in conventional methods, and the number of punctures was likely to be ≥ 2; however, the number of axillary punctures was not recorded. We typically block the musculocutaneous nerve and ulnar, radial, and median nerves separately;⁸ thus, the number of punctures per block was also possibly ≥ 2. Consequently, multiple punctures, long subcutaneous needle lengths, and individual differences in brachial plexus anatomy may cause neuropathies in a US-guided axillary block due to direct nerve injury and unexpected intraneural injections of local anesthetics.

A meta-analysis examining various types of nerve blocks reported that the incidence of neuropathies was not reduced by combining the US-guided and NS methods.² Several randomized controlled trials have compared the success rate of axillary blocks performed with a combination of US and NS guidance against those with NS alone, as well as those with US versus NS guidance; however, these trials have not investigated the incidence of postoperative neuropathies.^23–26 In contrast, a prospective study reported 60 successful axillary blocks with zero incidences of postoperative neuropathies by combining US and NS guidance.²⁷ Additionally, for blocks with difficulty in nerve visualization such as the adductor canal and posterior clavicular blocks, double-monitoring using combined US and NS guidance was effective in preventing neuropathies.^28,29 Since neuropathies are rare, chance variation was possibly detected. Therefore, there may be little differences in the incidence of neuropathies between US with NS and NS guidance alone.³⁰

With a high-pressure injection during the administration of a local anesthetic, IPM may prevent its administration into undesired intraneural spaces, thereby reducing the incidence of neuropathies,³¹ as IPM reportedly prevents nerve injury following an interscalene block.^32,33 Considering the mechanism of neuropathies after US-guided axillary blocks, the use of triple-monitoring as a combination of US, NS guidance, and IPM³¹ may be particularly effective in preventing postoperative axillary neuropathies.

While a uniform block procedure application at a single institution was one of the strengths of our study, there were several limitations. First, this study was designed as a retrospective cohort study, which is known to be associated with potential confounding factors and a lack of randomization. Specifically, the different operator experiences in practicing the blocks, the different local anesthetic concentrations and/ or volume are considered as confounding factors. Second, although the cases of PONS should have been investigated by at least two anesthesiologists to avoid any bias and subjective errors, it have being investigated by only one. Third, the definition and assessment of neuropathies were based on the physician’s diagnosis as determined by the patient’s self-report. Hence, the incidence of neuropathies may have been underestimated, unless it was documented in medical records. Fourth, the etiology of PONS was diagnosed mainly based on clinical symptoms, without using objective measures (electromyography and nerve conduction study). Therefore, the role of the injured nerve may not be accurate. Finally, as UGBPBs were performed under general anesthesia, the incidence of neuropathies may differ from that in studies in which they were performed under awake conditions.

Conclusion

The incidence of neuropathies after US-guided axillary block under general anesthesia was significantly higher than that after non-axillary approaches. Considering the anatomical characteristics of the brachial plexus and subcutaneous needle length in the axillary region, triple-monitoring using US and NS guidance with IPM may be useful for preventing neuropathies following US-guided axillary blocks. The findings of this study may be useful to advance the knowledge of neuropathies secondary to UGBPBs.