Impact of Scalene Muscle Botulinum Toxin Injection With and Without Surgery in Neurogenic Thoracic Outlet Syndrome

Abstract

Objective: 

Scalene blocks are part of both the diagnostic and treatment algorithm for patients presenting with symptoms of neurogenic thoracic outlet syndrome (nTOS). However, there is a paucity of data on the utility of scalene botulinum toxin injection (BTI) before surgical decompression. We sought to determine the impact of BTI with and without surgery at a multidisciplinary referral center.

Design: 

Retrospective cohort study.

Setting: 

Single institution tertiary academic center, 2011 to 2020.

Patients: 

Seventy-seven consecutive patients.

Interventions: 

Scalene muscle BTI for nTOS with or without surgical decompression.

Main Outcome Measures: 

Pain relief and Quick Disability of the Arm, Shoulder and Hand (QDASH) score.

Results: 

Seventy-seven patients, with a mean age of 31.4 years, had BTI for symptoms of nTOS. All patients underwent pretreatment physical therapy through the Edgelow protocol for a mean duration of 3.4 months. However, 72.7% had dynamic vascular compression on duplex ultrasound with provocative maneuvers and 85.7% had a positive physical examination finding. After BTI, 77.9% reported subjective relief, confirmed by an improved QDASH disability score. Thirty-one patients (40.3%) then went on to have further persistent symptoms and proceeded with first rib resection. After BTI + Surgery, 96.8% reported symptomatic relief and had a median QDASH score improvement of 21 (range: 10-40), with all reaching minimal clinically important differences in the QDASH score after combination therapy.

Conclusions: 

In this reported series of chemodenervation in patients with nTOS, BTI is helpful in alleviating symptoms before definitive surgical decompression. BTI followed by first rib resection provides additional symptom improvement over BTI alone.

INTRODUCTION

Neurogenic thoracic outlet syndrome (nTOS) comprises approximately 90% of all TOS cases and is characterized by repetitive motion leading to the compression of the brachial plexus.¹ Patients with nTOS commonly present with vague symptomatology that mimics many other orthopedic issues. Society for Vascular Surgery guidelines were developed to produce consistency in reporting, diagnosis, treatment, and outcomes; however, there are still no objective diagnostic criteria.² Although various diagnostic modalities such as ultrasound, magnetic resonance imaging (MRI), computed tomography (CT), and electroneuromyography of the brachial plexus can help suggest the diagnosis, differentiating between nTOS and other cervicobrachial syndromes and selecting patients who will benefit from surgery remain challenging.

Injection of anesthetics such as lidocaine into muscles adjacent to the brachial plexus, including the anterior or middle scalene or pectoralis minor, has been reported to be useful in selecting patients who are likely to benefit from definitive surgical decompression and can be temporarily therapeutic in patients with nTOS.^3–9 Similarly, botulinum toxin injection (BTI) is thought to cause muscular paralysis from temporary chemodenervation and consequently decrease neural compression, hypothesized to mimic the decompressive effects of first rib resection and scalenectomy. Although previous studies have shown that anesthetics and BTI are safe and well-tolerated, there are mixed outcomes data with the use of longer-acting agents such botulinum toxin.^3,7–11 Some studies demonstrate improvement with the use of botulinum toxin as a diagnostic and therapeutic modality, whereas others show no difference in outcomes.^4,11–13

Reports of surgical outcomes of patients with nTOS preoperatively evaluated and treated with BTI have been limited to small case series. Given the paucity of data on the utility of scalene BTI before surgical decompression, we sought to determine the impact of botulinum toxin scalene chemodenervation with and without surgery at a multidisciplinary referral center.

METHODS

Database

This Health Insurance Portability and Accountability Act compliant study was approved by the local institutional review board. The medical records of patients who underwent BTI for nTOS from 2011 to 2020 were reviewed. The clinical diagnosis of nTOS was made in accordance with SVS guidelines and based on history and physical examination, focused on neurologic symptoms, including focal tenderness at potential sites of compression, radicular pain, paresthesia, and weakness.² Diagnostic testing comprised dynamic testing, cervical spine x-rays, cervical MRI, and nerve conduction study. These patients often would already have had extensive workup by orthopedic surgery, neurology, or sports medicine specialists and were referred to our vascular surgery clinic with nTOS suggested as the most likely etiology. Some of these patients required additional testing for confirmation of diagnosis. Patients with arterial or venous TOS were excluded from this study.

Treatment Algorithm

Initial consultation involved collection of pertinent history, accurate physical examination with provocative maneuvers to elicit symptoms, duplex ultrasound in multiple provocative positions, and other imaging data such as MRI or CT (if already ordered by referring physicians) and a baseline quality of life disability survey (Figure 1). Duplex ultrasound is performed along with arterial plethysmography. If there is obliteration of waveforms with provocative maneuvers, we believed this to be suggestive of a narrowed thoracic outlet. After the initial suggested diagnosis of nTOS in this patient cohort, they are referred for TOS-specific physical therapy (PT), focusing on algorithms described by Edgelow.¹⁴ Our choice of the Edgelow protocol for all patients for their initial PT is based on our previous experience showing improvement in overall success rates documented subjectively and objectively.^15,16 The comparative analysis was performed between the first-time recipients of BTI and those who had relief from BTI and went on to receive surgical decompression. Patients who do not have a response to BTI are not reoffered BTI.

Figure 1.: Highly selective treatment algorithm for patients evaluated for neurogenic thoracic outlet syndrome (nTOS). Of note, all patients are referred to PT and those who are additionally interested in BTI or have a favorable response to PT are referred to sports medicine for chemodenervation.

Patients who do not have a response to BTI and PT are not offered surgery. Figure 1 summarizes our general treatment algorithm of patients with nTOS and is not a schematic of our analysis.

Patients are then asked to return to the clinic after an adequate trial of PT (minimum of 2 months of 3 times per week exercise) and are re-evaluated. Based on the data from initial evaluation/workup, patients are then offered further nonoperative treatment with continued PT or BTI, which we added to our algorithm since 2012. Some patients are at that point offered surgical decompression based on compliance and improvement with PT or symptomatic relief if they can undergo BTI. Patients who did not report any symptomatic improvement with PT or BTI were not offered surgery because we believed those patients would often have other cervicobrachial pathologies (eg cervical disk degenerative disease with radiculopathy, peripheral neuropathy, rotator cuff injury, cervical dystonia, complex regional pain syndrome, etc). Patients not offered surgery or those who preferred nonsurgical management were referred back to sports medicine, PT, or pain medicine specialists. Of note, in subjects having repeat BTI before surgery, only the response to the first injection was considered.

Techniques for BTI and Surgical Decompression

All the BTI are performed by our sports medicine and physical medicine and rehabilitation colleagues. Patients are placed in a supine position with their neck slightly extended. Using typical sterile precautions, a 6- to 15-MHz linear array ultrasound transducer is used to visualize the anterior and middle scalene muscles. A 27G 1.5-inch needle is advanced under direct ultrasound visualization into the anterior scalene muscle and 50 units of botulinum toxin (Xeomin, Merz Pharmaceuticals, LLC. Raleigh, NC) that is previously reconstituted with 0.5 mL of preservative-free normal saline is injected. This is spread throughout the muscle using a single injection site. Post-BTI follow-up is through a virtual or a clinic visit 1 month after the injection.

For those patients who ultimately undergo surgical decompression for nTOS, we perform this under general anesthesia induction without use of regional neuromuscular blocking agents. All decompressions are performed through a standard supraclavicular incision, with mobilization of the fat pad, careful scalenectomy, removal of most of the first rib, and extensive brachial plexus neurolysis.¹⁷ All patients resume PT after surgery. For surgical patients in follow-up, repeat quality of life surveys are given to patients postoperatively (3 months) and yearly thereafter. Postsurgery follow-up is performed through a virtual or clinic visit at least 3 months after surgery.

Variables

The mini-Quick Disability of the Arm, Shoulder, and Hand (QDASH) questionnaire was used to assess the quality of life before and after interventions.^18–21 This survey has been extensively validated in orthopedic and hand surgery literature and has been used in TOS outcomes reports.²² The QDASH survey uses 11 items to objectively measure physical function and symptoms in people with any or multiple musculoskeletal disorders of the upper limb. Scores range from 0 (asymptomatic) to 100 (totally incapacitated).

Those who had any degree of improvement were divided into a dichotomous binary response of reporting “relief” or “no relief.” The degree of relief was graded based on patient's perception of percent improvement in their overall symptoms—significant: >67%, moderate: 34% to 66%, mild: 1% to 33%, or no improvement. Minimal clinically important difference (MCID) change in QDASH was defined as change greater than 10% based on standard error of mean.²³

Competitive athletes were defined by active participation in organized individual or team sports at the varsity high school, collegiate, or professional or semiprofessional levels, as reported by the patients. Positive physical examination finding was defined as positive findings on elevated arm stress, Roos test, and modified upper limb tension test of Elvey.

Statistical Analysis

Descriptive statistics were used, including measures of frequency, median, mean, and SD to compare the baseline characteristics, treatment details, and outcomes. The Shapiro–Wilk test was used to ascertain the normality of the collected data. Univariate differences were assessed using χ² and Fisher exact tests for categorical variables and two-tailed independent sample t test for continuous variables and when appropriate, based on normality distribution, the Mann–Whitney U test was used. The association of treatment outcomes with baseline characteristics was assessed using multilevel regression models. All models were adjusted based on variables that met the inclusion criteria of a P value of < 0.2 on univariate analysis. For all analyses, a P value <0.05 was considered significant. All statistical analyses were performed using Stata version 16, StataCorp-LLC, TX.

RESULTS

Demographics

From 2011 to 2020, 77 patients (47 female, 61%), with a mean age of 31.4 years, had BTI for symptoms of nTOS. Overall, 26% were collegiate athletes, 62.3% had chronic symptoms, and 26% had a history of neck and shoulder orthopedic injuries. Of these, 72.7% had dynamic vascular compression on duplex ultrasound with provocative maneuvers and 85.7% had a positive physical examination finding consistent with nTOS. The median number of BTI was 2 (range 1-4). Of this study cohort of 77 patients who underwent BTI, 31 patients (40.3%) went on additionally to have first rib resection. Based on our algorithm described above and published in the past, those with modest relief, but not full symptom resolution, provided by PT or BTI were subsequently offered surgery, allowing this analysis. There was no significant difference between those who underwent BTI alone (BTI) and those who underwent BTI followed by surgical decompression (BTI + Surgery) regarding sex, age at presentation, race, varsity athlete status, history of neck and shoulder orthopedic injury, history of neck and shoulder surgery, mental health diagnosis, chronicity of symptoms, duration of symptoms, positive nTOS physical examination finding, or dynamic vascular compression on ultrasound (Table 1). However, BTI + Surgery patients were more likely to have a higher median QDASH score at baseline (43 vs 53; P = 0.029) and a smaller anterior scalene area on ultrasound compared with those who underwent BTI alone (1.11 vs 0.94 cm²; P = 0.041).

TABLE 1. - Comparison of Baseline Demographics of Patients Who Underwent Botulinum Toxin Injection With and Without Surgery

Variable	BTI (N = 46)	BTI + Surgery (N = 31)	P
Sex
Female	60.9% (28/46)	61.3% (19/31)	0.97
Male	39.1% (18/46)	38.7% (12/31)
Age at presentation (mean ± SD)	33.2 ± 13	29.9 ± 11.1	0.252
Race
White	78.3% (36/46)	74.2% (23/31)	0.679
Non-White	21.7% (10/46)	25.8% (8/31)
Varsity athletes	26.1% (12/46)	25.8% (8/31)	0.978
History of neck and shoulder orthopedic injury	21.7% (10/46)	32.3% (10/31)	0.302
History of previous neck and shoulder surgery	19.6% (9/46)	25.8% (8/31)	0.517
Mental health diagnosis	30.4% (14/46)	16.1% (5/31)	0.153
Chronicity
Intermittent	69.8% (30/43)	75% (18/24)	0.649
Constant	30.2% (13/43)	25% (6/24)
Symptom duration (years median, IQR)	0.6 (0.02-1.42)	1 (0.05-1.5)	0.436
QDASH score before BTI (median, IQR)	43 (29.5-55)	53 (43-65)	0.029*
Positive physical examination finding	80.4% (37/46)	93.6% (29/31)	0.107
Dynamic vascular compression on ultrasound	78.6% (33/42)	82.1% (23/28)	0.714
Anterior scalene area (cm2)	1.11 (1.02-1.29)	0.94 (0.77-1.2)	0.041*

^*Statistically significant value, P < 0.05.

Treatment

All patients underwent preintervention PT for a mean duration of 3.4 ± 1.4 months. There was no significant difference in the length of PT before BTI, participation rate in the Edgelow protocol, or the amount of botulinum toxin dose administered between BTI and BTI + Surgery groups (Table 2). The median time from BTI to surgery in those who underwent surgical decompression was 164 days (IQR: 114-297).

Table 2. - Comparison of Treatment Details of Patients Who Underwent Botulinum Toxin Injection With and Without Surgery

Variable	BTI (N = 46)	BTI + Surgery (N = 31)	P
Length of physical therapy before BTI (months mean ± SD)	3.4 ± 1.90	3.4 ± 1.98	0.95
Participation in the Edgelow protocol	76.3% (29/38)	89.3% (25/28)	0.177
Botulinum toxin dose (unit mean ± SD)	46.5 ± 7.1	45.5 ± 5.7	0.851
Time since BTI to surgery (median days, IQR)		164 (114-297)

Outcomes

After BTI, 77.9% of 77 patients reported subjective relief, confirmed by an improved QDASH disability score. After BTI + Surgery, all but one patient reported symptomatic relief, with an overall median QDASH score improvement of 21 (range: 10-40), and 96.8% reaching minimal clinically important differences in the QDASH score after combination therapy. Figure 2 graphs individual QDASH disability scores of the patients who underwent surgery from baseline to post-BTI to latest postoperative follow-up.

Figure 2.: Individual improvement in QDASH disability scores among surgical patients (n = 31), with 100 indicating maximal disability. The median QDASH score at baseline was 53, 35 at 1 month follow-up after botulinum toxin injection, and 26 at 3 to 6 months follow-up after surgical decompression.

A higher proportion of patients in the BTI + Surgery group reported relief after the initial BTI than those in the BTI group (90.3% vs 69.6%; P = 0.03). However, there was no significant difference in the MCID QDASH score change after the initial BTI between the 2 groups. Compared with the BTI group, a higher proportion of patients in the BTI + Surgery group ultimately reached overall relief (69.6% vs 96.8%; P = 0.003) and MCID QDASH score improvement (78.3% vs 96.8%; P = 0.032). There was no difference in the overall QDASH score change or distribution of degree of relief between the 2 groups (Table 3).

TABLE 3. - Comparison of Outcomes of Patients Who Underwent Botulinum Toxin Injection With and Without Surgery

Variable	BTI (N = 46)	BTI + Surgery (N = 31)	P
Relief after BTI	69.6% (32/46)	90.3% (28/31)	0.03*
MCID QDASH change after BTI	78.3% (18/23)	80.6% (26/31)	0.6
Overall relief	69.6% (32/46)	96.8% (30/31)	0.003*
Overall MCID QDASH change	78.3% (18/23)	96.8% (30/31)	0.032*
Overall QDASH change (median, IQR)	16 (-6 to 32)	21 (10-40)	0.258
Overall degree of relief
Mild	14.6% (6)	10% (3)	0.803
Moderate	22% (9)	20% (6)
Significant	63.4% (26)	70% (21)

^*Statistically significant value, P < 0.05.

Predictors of Relief After BTI

On univariate analysis, the higher initial QDASH score was predictive of overall relief after BTI (P = 0.026), young age was predictive of significant (>67%) relief after BTI (P = 0.008), and shorter symptom duration was predictive of significant (>67%) relief after BTI (P = 0.027). On multivariate analysis, female sex (odds ratio [OR]: 6.63, P = 0.049) and a positive physical examination finding (OR: 8.97, P = 0.030) were both correlated with greater likelihood of relief after BTI (Table 4).

Table 4. - Adjusted Multivariate Regression Analysis of Baseline Characteristics Predictors of Relief After Botulinum Toxin Injection

Variables	Odds Ratio	95% CI	P
Symptom relief after BTI
Sex (female)	6.6	1.01-43.49	0.049*
Positive physical examination finding	8.97	1.24-64.97	0.030*
MCID change in QDASH
Symptom duration	0.5	0.22-1.13	0.094
Significant degree of relief
Sex (female)	3.22	0.90-11.57	0.073

^*Statistically significant value, P < 0.05.

DISCUSSION

We found in this cohort analysis of patients undergoing BTI for nTOS a significant further symptom improvement with definitive surgery, defined subjectively by symptomatic relief and objectively with QDASH score improvement. All of the patients—with the exception one—who ultimately underwent surgery reported symptomatic relief. These results suggest that BTI is a safe and useful early indicator of patients who might further improve after nTOS diagnosis and definitive surgical decompression. By adding the response to BTI in addition to our previously published PT algorithm, we have noted an even higher likelihood of meaningful clinical symptomatic improvement. Our findings have led to a shift in how we aggressively seek out BTI as both a diagnostic and therapeutic modality in our early workup in patients suspected to have nTOS. The initial treatment for nTOS is conventionally focused on PT to stretch the scalene muscles, improve shoulder girdle mechanics, and change ergonomics. These measures we believe are further augmented by BTI, which has been established as a safe and effective early tool in the management of nTOS. We noted no complications from BTI and a reasonably good rate of temporary symptom relief (80%) in patients being worked up for nTOS. In our and others' experience, this temporarily relief may be dually beneficial in further allowing many patients to more actively participate in a PT program.²⁴

Overall, in this cohort of patients with nTOS, scalene BTI chemodenervation was helpful in temporarily alleviating symptoms of nearly 80% of patients. This relief rate with BTI is marginally higher than those reported in the literature. In a study by Torriani et al,¹⁰ BTI of anterior scalene and pectoris minor produced temporary symptom relief in over two-thirds of patients with nTOS. Rochlin et al²⁵ reported temporary pain relief in 64% of patients after BTI. Jordan et al²⁶ reported significant symptom improvement in 64% of patients with nTOS after BTI. Donahue and Torriani et al reported BTI led to symptom relief in 63% of procedures.^8,27 Our technique of BTI is similar to those of most mentioned studies in that we used ultrasound guidance and low-dose botulinum toxin (50 units). The other described methods may involve electromyographic, fluoroscopic, MRI, or CT guidance and doses as high as 100 units of botulinum toxin.^3,9,26,27 Other nonsurgical techniques to relax the interscalene space are injections of anesthetic agents and steroids.^6,28 However, of these nonsurgical interventions, BTI results in the more sustained symptom reduction, with its effects expected to dissipate after 2 to 3 months, and may offer a therapeutic and diagnostic adjunct to PT.^{5,9,11,26–29} Overall BTI is well-tolerated but the reported adverse events associated with BTI may include pain, bruising, blood clot or bleeding at the injection site, numbness or weakness in arm, skin rash or pruritis, fever or flu-like symptoms, or nausea and vomiting.¹¹ No adverse events were noted in our retrospective chart review, but this study was not designed to assess for such complications.

Owing to a lack of consistent objective diagnostic criteria, nTOS has historically been difficult to diagnose and treat adequately, as other diagnoses should be considered including herniated cervical disks, rotator cuff injuries, peripheral nerve entrapment, and chronic pain syndromes. The challenge with diagnosis and treatment underscores the importance of careful patient selection and use of appropriate adjunctive modalities to optimize the results of surgical outcomes.^22,28,30 Not surprisingly, a higher proportion of patients who eventually underwent surgical decompression had reported relief with the initial BTI than those who had BTI alone and did not undergo surgery. We infer from our findings, although challenging as this was not a randomized trial, that response to BTI may have a predictive value for favorable surgical responses, and therefore, those patients who respond to BTI and have symptom recurrence will likely be reasonable surgical candidates. Characteristics that may be useful predictors of initial relief with BTI include higher initial QDASH, younger age, and shorter symptom duration based on univariate analysis, as well as female sex and positive physical examination finding based on multivariate analysis. Similarly, 2 large retrospective studies reporting 10-year surgical outcomes demonstrated that favorable surgical outcomes in younger patients and those with shorter symptom duration.^4,31 It is not surprising that patients who ended up being selected for surgical decompression (BTI + Surgery) were those with higher QDASH scores and smaller anterior scalene areas (Table 1). A higher QDASH score indicates a higher level of disability and perhaps a better opportunity of symptomatic relief. Moreover, a smaller anterior scalene area may correlate with a more contracted muscular structure, causing worse nTOS compressive symptoms, which would be more likely to respond favorably to chemodenervation. On the other hand, a recent study demonstrated that patients who do not have a response to the local blocks of the anterior scalene muscle tend to have significantly shorter anterior scalene muscle heights (measured as the distance from the superior border of the first rib to the top of C3 vertebrae) than patients who respond, suggesting an anatomic difference in responders versus nonresponders.³ In the same study, it was shown that blocks of the anterior scalene muscle correlated with increasing muscle height. Further studies are necessary to better understand the clinical significance of scalene muscle volume in predicting a favorable response to BTI.

Relief of symptoms after BTI may be a clinical indicator of good surgical candidacy, yet previously reported modest surgical outcomes among some patients with an initially positive response to BTI.^7,26,29 In a recent study, BTI was shown to have a high positive predictive value (99%) and specificity (90%) of response to surgery, suggesting most responders to BTI will also do well with surgery.⁸ However, the negative predictive value was low and only 14% of nonresponders to BTI were nonresponders to surgery.⁸ This indicates that a negative response to BTI alone is not adequate for not offering surgery to patients and that this piece of diagnostic information should be considered along with other prognostication factors such as adherence and success with PT and lifestyle modification. In our study, because only one patient did not respond to surgical decompression, despite initial relief with BTI, the positive predictive value was 96.43%. Negative predictive value cannot be calculated from this study design—statistically limited by lack of any true negatives in the study. Further research will be required to best determine whether subjects with zero response to BTI are truly not surgical candidates. Anecdotally, many of the patients not captured in our experience were denied BTI financially, and we believe studies such as this that document the safety and efficacy of BTI for patients with nTOS will provide impetus for broader utilization in this challenging patient cohort that are often young and having disabling symptoms. Overall, we believe the ability of BTI to further predict which patients are likely to improve with surgical decompression makes it an important tool in distinguishing which patients are best off with surgery. Of interest, BTI often is difficult for insurance coverage due to its “experimental” concept, but we believe this analysis suggests BTI can definitely be used in most nTOS cases.

There are several limitations to this study. First, this is a retrospective study with fewer than 80 patients. Nonetheless, this is one of the few and largest studies to this date to describe, with granularity, the experience with BTI in the management of patients with nTOS. Second, we were not able to collect QDASH scores and follow-up data on some of the patients who did not have surgery and ultimately may have sought treatment elsewhere. Third, all of the patients were seen at a single institution, a tertiary care academic center, where the patients seen in referral may not necessarily be similar to those patients with nTOS seen in general practice. Nonetheless, the care for patients with nTOS is gradually becoming centralized into tertiary centers that could use algorithms similar to this study's to further validate the utility of BTI in improving treatment outcomes. Finally, this study is also limited in that the outcomes are limited to intermediate postoperative period of 6 months. The major barrier to analyzing the data is that a majority of our patients relocated and transferred care as they are student athletes, and those who did not undergo surgery were referred back to sports medicine, PT, or pain medicine specialists for nonsurgical management. It would be of great interest to know if in long terms, these patients required postoperative PT or BTI.

CONCLUSIONS

In this reported series of scalene BTI chemodenervation in patients with nTOS, BTI is helpful in alleviating symptoms before definitive surgical decompression. BTI followed by first rib resection provides significant additional symptom improvement over BTI alone. Despite mixed outcomes of BTI in previous literature, this potentially diagnostic and therapeutic modality may contribute to the traditional PT-based patient selection criteria for surgical management and has become our preferred discriminating factor in patient selection for nTOS surgical decompression. Further research is needed to confirm in which patients initial BTI response predicts future successful surgical outcome.