Arthroscopy-assisted First Rib Resection and Brachial Plexus Neurolysis for Thoracic Outlet Syndrome

The present study included 27 consecutive patients (30 limbs) undergoing arthroscopy-assisted transaxillary �rst rib resection and brachial plexus neurolysis for thoracic outlet syndrome (TOS). To improve visualization, we changed the intraoperative limb position in three stages. We assessed the intraoperative parameters, including the scalene interval width between the anterior and middle interscalene muscles (interscalene base), blood loss, operation time, preoperative and postoperative QuickDASH, patient satisfaction, and complications. The mean intraoperatively measured interscalene base width was 6.2 mm. Appropriate visualization could be obtained at zero-position in the late phase. Intraoperative blood loss and operation time were signi�cantly less in the late phase (p < 0.001). The QuickDASH score was signi�cantly improved (42 before surgery vs. 9 at �nal follow-up, p < 0.001), especially in athletes relative to non-athletes (0.2 vs 14, p < 0.001). The outcome was excellent in 43.3% of cases, good in 43.3%, fair in 13.3%, and poor in none. The present approach achieved complete relief in 43% of cases overall (91% in athletes and 16% in non-athletes). Pneumothorax was present at the early phase in 3.3%. There were no other complications and no recurrences. Arthroscopic surgery is useful for TOS, especially in athletes.


Introduction
Surgical treatment for thoracic outlet syndrome (TOS) associated with a cervical rib was rst described by Coote in 1861 1 .Morley described that brachial pressure neuritis occurred in the absence of a cervical rib, and that resection of a normal rib yielded very satisfactory results 2 .Rogers was rst to describe pain in the upper limb due to lesions of the 'thoracic outlet' 3 .Several surgical treatments for TOS have been reported, including supraclavicular scalenectomy leaving the rst rib intact 4 , supraclavicular rst rib resection with scalenectomy 5 , transaxillary rst rib resection 6 , and endoscopic brachial plexus neurolysis 7 .Transaxillary rst rib resection has become the most common procedure for TOS.However, it is usually di cult to obtain satisfactory visualization under direct vision, and it is sometime di cult to control bleeding.Therefore, this procedure is associated with recurrence and complications to some degree [8][9][10][11][12] .Endoscopically assisted rst rib resection and robotically assisted thoracoscopic rst rib resection have reduced the incidence of these complications [13][14][15][16][17][18][19] .
We hypothesized that an arthroscopically assisted transaxillary approach would be able to improve visualization and allow safe surgery for both vascular and neurogenic TOS, and that the position of the upper limb would be important for transaxillary insertion of the arthroscope.The purpose of the present study was to evaluate the results of arthroscopic transaxillary rst rib resection with scalenectomy and brachial plexus neurolysis for TOS in different limb positions.

Indications and imaging.
We performed surgery only for patients with certain objective ndings, including blood ow disruption (Video 1), low blood ow (Fig. 1a, affected limb; Fig. 1b, contralateral side), and accelerated blood ow (Fig. 1c, affected limb; Fig. 1d, contralateral side) in the subclavian artery demonstrated using Doppler sonography, narrowing of the interscalene base or costoclavicular space demonstrated using Duplex ultrasonography set at an ABER position or CT, or narrowing of the subclavian artery demonstrated by CT angiography (Fig. 2a, preoperative; Fig. 2b, six days after surgery).The mean blood ow was 83.2 cm/s (range, 48-120 cm/s) at rest and 144.1 cm/s (range, 32-301 cm/s) at an ABER position sitting in a chair demonstrated by Doppler sonography (Table 1).The mean interscalene base width was 8.7 mm (range, 5.1-16.3mm) at rest and 8.1 mm (range, 0-14.7 mm) at an ABER position sitting in a chair demonstrated by Duplex ultrasonography (Table 1).The mean costoclavicular space demonstrated by Duplex ultrasonography was 9.6 mm (range, 4.7-13.1 mm) at an ABER position sitting on a chair and the mean costoclavicular space demonstrated by CT was 11.3 mm (range, 6.2-18.1 mm) with the shoulder at full abduction (Table 1).Surgical procedure and intraoperative measurements.
Appropriate visualization could not be obtained simply by inserting the arthroscope at 150 degrees of limb abduction (Fig. 3a) nor at 90 degrees of limb abduction with the arm pulled upwards (Fig. 3b).Therefore, we needed to improve the visualization using some large retractors.However, better visualization was obtained simply by inserting the arthroscope without any retractors at a subordinate pivotal position 20 where the deltoid, the supraspinatus and the infraspinatus were relaxed (Fig. 3c, Video 2, 3).Appropriate visualization was also obtained by applying antifog to the arthroscope and attaching suction to the side of the arthroscope (Fig. 4), thus signi cantly decreasing the amount of intraoperative blood pooling.The mean intraoperative blood loss was 25 mL (range, 3-126 mL): 47 mL (range, 7-126 mL) in the early phase, 33 mL (range 6-101 mL) in the middle phase, and 10 mL (range, 3-20 mL) in the late phase (Table 1).Intraoperative blood loss was signi cantly lower in the late phase (p < 0.001).The mean intraoperative measured width of the interscalene base was 6.15 mm (range, 2-12 mm).The mean operation time was 135 min (range, 60-307 min), being 244 min (range, 166-307 min) in the early phase, 152 min (range, 86-272 min) in the middle phase, and 87 min (range, 60-132 min) in the late phase (Table 1).
Operation time was also signi cantly shorter in the late phase (p < 0.001).
Clinical assessments.
The mean Quick Disability of the Arm, Shoulder and Hand (QuickDASH) score was 42 (range, 11-96) before surgery and 9 (range, 0-48) at nal follow-up (p < 0.001), demonstrating a signi cant improvement (Table 2, 3).The mean QuickDASH score at nal follow-up was 0.2 (range, 0-2) in athletes and 14 (range, 0-48) in non-athletes (Table 3).The QuickDASH score was signi cantly better in athletes (p < 0.001, Table 3).All of the patients were satis ed with their surgical outcomes and were happy with the improvement seen in their limbs.The rating was excellent in 13 patients (43.3%), good in 13 (43.3%),fair in 4 (13.3%), and poor in none (Table 2).Complete relief with the present methods was achieved in 43% of the patients (91% in athletes and 16% in nonathletes).There was one case (3.3%) of slight parietal pleura damage in the early phase with no complaint, and healing was achieved without any additional treatment.There were no other complications and no cases of recurrence.

Discussion
Vascular TOS cases can be diagnosed by Color Doppler and Duplex ultrasonography 21,22 or CT angiography [23][24][25] .We evaluated patients with vascular TOS using similar methods.Blood ow disruption, low blood ow, and accelerated blood ow of the subclavian artery were measured using Doppler sonography in an ABER position.Neurogenic TOS cases can also be diagnosed by Duplex ultrasonography 26,27 .In cadaver studies, the mean interscalene base width and mean costoclavicular space have been reported to be 10.7 mm and 13.5 mm, respectively 28 .The mean costoclavicular space measured by CT was 12.5 mm 29 .Preoperative and intraoperative measures of the interscalene base can predict disorders due to scalene triangular stenosis.In the presence of clinical TOS, the scalene muscles compress the structures of the brachial plexus in the thoracic outlet between the anterior and middle scalene muscles.Therefore, both scalenectomy and rst rib resection provide signi cant functional improvements in patients with TOS.
Arthroscopic surgery requires appropriate visualization, especially when inserting an arthroscope in a place other than a joint.Therefore, we changed the upper arm position in three phases.Better visualization was obtained at the subordinate pivotal position 20 /zero-position 30 .The relationship between the neurovascular bundle and the scalene muscles could be observed clearly using an arthroscope in the zero-position.Furthermore, arthroscopic neurolysis was possible when the brachial plexus and subclavian artery were adherent.Arthroscopically assisted surgery allowed decompression for both vascular and neurogenic TOS.
There are three major procedures for TOS in the absence of a cervical rib: transaxillary rst rib resection 8,31−33 , supraclavicular rst rib resection 2,5,11,34−36 , and supraclavicular release of the anterior and middle scalene muscles leaving the rst rib intact [37][38][39] .Statistically there is no signi cant difference in outcome between the three procedures, fair results being reported in 4-8% of each group 11 .A systematic literature search revealed that both supraclavicular scalenectomy and transaxillary rst rib resection had a high probability of success 12 .In the present study, arthroscopically assisted surgery achieved some degree of improvement in all patients.The mean improvement of QuickDASH was 33, and the complete relief was obtained 43% of the patients.TOS sometimes occurs in throwing athletes 40,41 .Athletes show better improvement than non-athletes after rst rib resection and scalenectomy 42 .Here, complete relief was observed signi cantly more often in athletes than in non-athletes (91% vs 16%).
Transaxillary rst rib resection has a higher incidence of complications than supraclavicular scalenectomy, being 22.5% and 12.6% respectively 12 .Among 594 cases of TOS treated by transaxillary rst rib resection, there were 138 (23%) cases of intraoperative pneumothorax 8 .In the present study, intraoperative pneumothorax occurred in only one case (3.3%) and no other complications or recurrences were observed after arthroscopic surgery.Ohtsuka et al. have reported thoracoscopic rst rib resection 43 .However, as this procedure poses a signi cant potential risk to the neurovascular bundle, modi ed techniques with appropriate instrumentation have been developed 16,18 .Furthermore, endoscopically assisted transaxiallary rst rib resection using a 10-mm endoscope has resulted in a lower incidence of complications 13,14 .In the present study, arthroscopically assisted transaxillary rst rib resection and brachial plexus neurolysis using a 4-mm arthroscope also achieved good results with a lower incidence of complications.

Limitations
The present study had several limitations.First, it was based on a retrospective review with a small number of patients and lacked a control group.Second, most cases of TOS can be cured by conservative therapy.Therefore, there are relatively few cases requiring surgery in our department, and for this reason we accepted TOS patients from other institutions who had not responded to conservative therapy and needed surgery.Because the sample size was limited, a controlled trial would have taken much more time, delaying the publication of the preliminary outcomes.Arthroscopically assisted transaxillary rst rib resection and brachial plexus neurolysis allowed us to obtain satisfactory results and was a safe procedure for TOS.In particular, athletes showed signi cantly better improvement than non-athletes.Third, the diagnosis of TOS is well known to be controversial 44,45 .In the present study we excluded one patient with 'true neurogenic TOS' 22,46,47 associated with a cervical rib.We diagnosed TOS using Doppler sonography adopting an ABER method or CT angiography.

Materials And Methods
Study design.
Twenty-eight consecutive patients (32 limbs) with TOS who underwent surgery in our department between April 2016 and February 2020 were evaluated.We excluded one patient (2 limbs) with 'true neurogenic TOS' on the basis of Gilliatt-Sumner hand 22,46,47 associated with bilateral cervical ribs, which were excised by the supraclavicular approach.Among the remaining 27 patients, 30 limbs were included in the present study.There were 14 males and 13 females, and the mean age at surgery was 28.1 years (range, 15-50 years). Diagnosis.
The diagnosis of TOS is controversial 44,45 , and no speci c set of diagnostic criteria has yet been established 48 .We diagnosed patients as having TOS on the basis of symptomatic presentation, physical examination maneuvers including the Roos test 49 , Wright test 50 and Moley test 2 , and lack of any evidence of a more likely cause.Patients with traumatic TOS were excluded.Color Doppler and Duplex ultrasonography are useful diagnostic modalities in this context 21,22 .The measures assessed included blood ow disruption, low blood ow, and accelerated blood ow in the subclavian artery demonstrated by Doppler sonography, the scalene interval width between the anterior and middle interscalene muscles (interscalene base) 28 , and the costoclavicular space demonstrated by Duplex ultrasonography 26,27 in a resting position with the shoulder in abduction and in the external rotation (ABER) position sitting on a chair (Fig. 5) by a medical technologist (T.K.).Furthermore, enhanced computed tomography (CT) was performed with the shoulder in full abduction to con rm the presence of stenosis of the subclavian artery [23][24][25] and the costoclavicular space 29 .If peripheral neuropathy such as ulnar and median neuropathy could not be ruled out, nerve conduction studies were also performed.

Surgical technique.
The patient was placed in a lateral position with the arm elevated to expose the axilla using a limb positioner (SPIDER2, Smith & Nephew, Memphis, TN) for the upper extremity and operated on under general anesthesia.The upper limb position was set at 150 degrees of abduction in the early phase (n = 3, Fig. 3a), 90 degrees of abduction with the arm pulled upwards according to Roos 33 in the middle phase (n = 15, Fig. 3b), and in the subordinate pivotal position 20 /zero-position 30 in the late phase (n = 12, Fig. 3c).At the beginning, it was di cult to obtain appropriate visualization.Therefore, we changed the position in three stages to improve visualization (Videos 2, 3).
A transverse 4-cm skin incision was made over the third rib between the pectoralis major and the latissimus dorsi muscles at the axillary hairline level (Fig. 3d).Careful dissection was performed to allow con rmation of subclavian artery pulsation with a nger.An arthroscopic incision was made more superior and posterior at the third rib level (Fig. 3d).We used a 4-0-mm 30-degree arthroscope, detached both the anterior and middle inter-scalene muscles from the rst rib, excised the rst rib piecemeal using bone cutting rongeurs, and neurolysis of the brachial plexus was performed with arthroscopic assistance (Fig. 6, Video 4).
Informed consent was obtained from all included patients and/or a parent (age blow 18).
of motion exercise for shoulder abduction up to 90 degrees was allowed immediately after surgery, and unlimited shoulder motion was allowed after four weeks.The patients returned to full activities, such as sports, between two and three months after surgery.

Evaluation of clinical data and de nitions of outcome variables.
A comprehensive review of medical records was conducted by one observer (H.S.).Demographic and surgical data collected included the following: securing visualization in each upper limb position; intraoperative measurement of the interscalene base; intraoperative blood loss; operation time; preoperative and postoperative QuickDASH; patient satisfaction; and complications at a mean of 25.6 months (range, 12-56 months) after surgery.Patient satisfaction was divided into four categories according to Derkash et al. 51 : excellent, complete relief; good, almost complete relief; fair, partial relief; poor, no improvement.We compared the clinical results among these patients according to whether they were athletes or not.

Statistical analysis.
The QuickDASH was compared using Wilcoxon test and Fisher's exact test, while intraoperative blood loss and operation time were compared using Kruskal-Wallis test.Differences at P < 0.05 were regarded as statistically signi cant.All statistical analyses were performed with the EZR software program (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.6.3) 52.

Declarations
Ethical statements.
This study was approved by the ethics committee of Yamagata University (No. 2020-358), and was conducted in accordance with the ethical standards described in the latest revision of the Declaration of Helsinki.
Patient consent for participation and publication.
Informed consent for patient participation and publication was received in the form of an opt-out in-hospital notice.

Figures
Figures

Figure 1 Blood
Figure 1

Table 1
SIW, scalene interval width between anterior and middle interscalene muscles RP, resting position; ABER, shoulder in abduction and external rotation position CCS, costoclavicular space; NA, not available AS, affected side; CS, contralateral side; Rest, resting position