Intermediate and Long-term Outcomes Following Surgical Decompression of Neurogenic Thoracic Outlet Syndrome in an Adolescent Patient Population

Abstract

Background: Although the diagnosis of thoracic outlet syndrome (TOS) is often missed, outcomes from surgical intervention significantly improve patient satisfaction. This article seeks to highlight patient characteristics, intraoperative findings, and both short and long-term outcomes of thoracic outlet decompression in the adolescent population. Methods: A retrospective chart review of patients between the ages of 13 and 21 years with a clinical diagnosis of neurogenic thoracic outlet syndrome (NTOS) who were treated surgically between 2000 and 2015 was performed. Data points including preoperative patient characteristics and intraoperative findings were collected. In addition, patient-reported outcome scores, including Visual Analog Scale (VAS), Single Assessment Numeric Evaluation (SANE), Quick Disabilities of the Arm, Shoulder, and Hand (QuickDASH) questionnaire, Cervical Brachial Symptom Questionnaire (CBSQ), and NTOS index, were obtained for a cohort of patients with follow-up ranging from 2 to 15 years. Results: The study population consisted of 54 patients involving 61 extremities. The most common procedures included neurolysis of the supraclavicular brachial plexus (60, 98.4%), anterior scalenectomy (59, 96.7%), and middle scalenectomy (54, 88.5%). First rib resection (FRR) was performed in 28 patients (45.9%). Long-term outcomes were collected for 24 (44%) of 54 patients with an average follow-up of 69.5 months (range, 24-180 months). The average VAS improved from 7.5 preoperatively to 1.8 postoperatively. The average SANE increased from 28.9 preoperatively to 85.4 postoperatively. The average postoperative scores were 11.4 for the QuickDASH, 27.4 for the CBSQ, and 17.2 for the NTOS index. Subgroup analysis of patients having FRR (28, 45.9%) demonstrated no difference in clinical outcome measures compared with patients who did not have FRR. Conclusion: Surgical treatment of NTOS in adolescent patients has favorable intermediate and long-term outcomes.

Introduction

Thoracic outlet syndrome (TOS) may affect up to 8% of the population, and it refers to a complex constellation of symptoms caused by compression of the neurovascular structures in the thoracic outlet or subcoracoid space. The TOS may be classified as neurogenic, arterial, or venous, with neurogenic thoracic outlet syndrome (NTOS) comprising 95% to 98% of cases.^1-4 Most patients with NTOS are women between the age of 30 and 50 years. ⁵ In NTOS, the compression of the brachial plexus occurs either in the posterior triangle of the neck, formed by the sternocleidomastoid, clavicle, and trapezius, or in the subcoracoid space.^6,7 The clinical presentation can be varied and is frequently misdiagnosed or missed but may include upper extremity pain, paresthesia, and numbness. Physical examination may reveal extremity weakness, a positive Tinel sign at potential compression sites, and reproduction of symptoms with provocative upper extremity maneuvers. The diagnosis of TOS is often one of exclusion and heavily relies on the physical examination of the patient as no criterion standard diagnostic test exists.^8-10 Patients who fail nonoperative treatment including medication and physical therapy are considered for surgery. The relationship between the age of the patient and presentation, diagnostic testing accuracy, and surgical outcomes is also not well studied.

The purpose of this study is to characterize the clinical findings of NTOS, report intraoperative findings, and evaluate both intermediate and long-term outcomes for adolescent patients undergoing surgical intervention. Our hypothesis is that adolescent patients will have favorable outcomes after TOS decompression, including improved pain and functional outcomes, and that there will be more favorable outcomes with first rib resection (FRR).

Methods

Institutional review board approval was obtained from a large tertiary referral hospital. Retrospective chart review was performed on patients surgically treated by a single surgeon (R.D.M.) between 2000 and 2015. Inclusion criteria included age between 13 and 21 years, a clinical diagnosis of NTOS, and failure of nonoperative management (medications, physical therapy, peripheral nerve blocks). Exclusion criteria included arterial or venous TOS as diagnosed with vascular studies, cervical spine injuries or previous cervical spine surgery, and TOS treated previously with surgery. Preoperative patient characteristics and intraoperative findings were collected. The minimum follow-up was set to 24 months.

Initial Evaluation

Each patient was assessed using standard history and physical examination and special tests relevant to NTOS, including the Tinel sign of the neck, Roos test, upper limb tension test, and the Spurling test. Information pertaining to patient-reported pain and paresthesia and positive findings from multiple examination maneuvers were collected. ⁶ Plain radiographs were evaluated for cervical ribs, soft tissue tumors, or other bony abnormalities of the first rib or clavicle. Previous shoulder magnetic resonance imaging (MRI) and electromyography/nerve conduction studies were evaluated. The MRI of the cervical spine was performed to rule out cervical radiculopathy when patients exhibited a positive Spurling maneuver. Patients were treated with surgical decompression after failing extensive nonoperative treatment, including pain modalities and hand therapy. Exercises included scalene stretching and scapular training and included at minimum 1 session with a certified hand therapist knowledgeable in TOS prior to beginning a home program.

Surgical Treatment

Surgical intervention for NTOS focused on releasing any anatomical structure compressing or restricting the plexus through its course in the thoracic outlet. Through the supraclavicular incision between the anterior trapezius and scalene muscles, the phrenic, long thoracic, and spinal accessory nerves were identified and protected. This dissection provided exposure to remove the scalenes and for neurolysis of the brachial plexus and subclavian artery. If the preoperative physical examination suggested plexus compression in the subcoracoid space, an infraclavicular exploration was performed within the deltopectoral interval. This approach allowed for the identification and release of the coracocostal ligament and pectoralis minor tendon. Bony anomalies, such as cervical ribs or elongated transverse processes, potentially causing neurovascular compromise were excised.

Postoperative Follow-up

Gentle overhead range of motion of the shoulder began on postoperative day 0, and patients were evaluated within 2 weeks for wound evaluation. Supervised physical therapy for shoulder girdle range of motion and strengthening was started at 2 weeks and continued for a minimum of 3 months. Physical therapy exercises were continued for a minimum of 6 months. Opiate pain medication was prescribed for 14 days postoperatively.

Patient-Reported Outcome Measures

Postoperative outcomes evaluated include length of postoperative hospital stay, resolution of symptoms, time to return to activities, and patient-reported satisfaction. Each patient was asked to complete the 10-point Visual Analog Scale (VAS) for pain, the Single Assessment Numeric Evaluation (SANE), the Quick Disabilities of the Arm, Shoulder, and Hand (QuickDASH) survey, and the Cervical Brachial Symptom Questionnaire (CBSQ). The VAS assesses pain on a scale of 0 to 10. ¹¹ The SANE is a self-rating of shoulder function ranging from 0 to 100%, with 100% being fully functional. ¹² The QuickDASH is an 11-item survey that measures physical function and symptoms of upper extremity musculoskeletal disorders. Each item is scored 1 to 5, summed, averaged, and converted to a score out of 100, with higher scores indicating greater disability.^2,13 The CBSQ is a 14-item survey scored from 0 to 120, which was developed for evaluation of NTOS and related disorders.^6,14,15 The NTOS index, developed by Caputo et.al, combines the DASH, CBSQ, and VAS scores to help mitigate discrepancies between patient-reported disability and actual functional performance. ⁶ Because previous studies show that QuickDASH is statistically comparable to DASH results, the QuickDASH scores were used for the NTOS index calculation in this article.^2,13

Statistical Analysis

The electronic medical record was filtered using Current Procedural Terminology code 64713 to identify patients who underwent surgical decompression of TOS. Data were collected and stored using REDCap secure database (REDCap, Nashville, Tennessee). Data were deidentified and initially analyzed for frequencies, percentages, and trends using Microsoft Excel (Microsoft, Redmond, Washington). Descriptive data are presented as the mean or frequency and a percentage of total. SPSS version 22.0 (SPSS, Chicago, Illinois) was used for statistical analysis and data modeling. Subgroup analysis of patient-reported outcomes comparing patients with rib resection to those without was performed using analysis of variance (ANOVA). Patient-reported outcomes by mechanism of injury were also compared using ANOVA. Our long-term patient cohort was compared with a cohort of patients from Caputo et al using the Student t test. Values of P < .05 were considered significant. ⁶

Results

The study population consisted of 54 adolescents (61 extremities) with an average age of 17.2 (13-21) years at the time of surgery. The patients were 77.7% women. Right-sided symptoms occurred in 31 patients (57.4%), and left-sided symptoms occurred in 14 (25.9%) patients. Bilateral involvement occurred in 9 patients (16.7%), with 6 patients presenting with bilateral involvement initially, and 3 patients presenting with symptoms in 1 extremity with later development of symptoms in the contralateral extremity. The dominant arm was affected in 32 patients (59.3%). The most common causes were overuse (31 patients, 50.8%) and trauma (13 patients, 21.3%). Overuse commonly involved sports or occupations with repetitive overhead motion. The most common sports reported were swimming, softball, tumbling, and baseball. High-energy trauma (5 patients, 8.2%) included motor vehicle accidents and falls from a height greater than standing. Low-energy trauma (7 patients, 11.5%) included falls from standing or a blow to the chest.

The mean reported age of symptom onset was 15.2 (11-21) years. On average, patients underwent surgical intervention within 1.4 (0-6.8) years of reported symptom onset. Patients saw an average of 2 other health care providers before referral to orthopedics for evaluation. Ten patients (18.5%) had previous surgeries on the affected extremity, including shoulder arthroscopy with labral repair, shoulder capsular fixation, and an ulnar transposition.

Numbness and tingling in the arm (46 patients, 75.4%) and hand (49 patients, 80.3%) were the most common initial symptoms. Arm pain also frequently occurred (39 patients, 63.9%). Shoulder pain and hand pain occurred in only 28 (45.9%) and 20 (32.8%) patients, respectively. Fourteen (22.9%) patients experienced neck pain, in addition to the upper extremity symptoms. Weakness of the arm occurred in 8 patients (13.1%), and weakness of the hand occurred in 6 patients (9.8%). Preoperative physical examination findings included: 44 extremities (72.1%) with positive Roos test, 38 extremities (62.2%) with positive upper limb tension test, and 40 extremities (65.6%) with a positive Tinel sign in the posterior triangle.

The most frequent surgical procedures performed were neurolysis of the brachial plexus (60, 98.4%), anterior scalenectomy (59, 96.7%), middle scalenectomy (54, 88.5%), and excision of the first rib (28, 45.9%). A second incision in the subcoracoid space was used in 28 (45.9%) extremities for exploration of the infraclavicular brachial plexus. The coracocostal ligament was excised in 26 extremities (92.9%), and the pectoralis minor was released in 9 extremities (32.1%) that underwent this approach.

The average operative time was 137.9 ± 39.8 minutes. The most common findings included pleural bands (45, 73.8%), tight scalenes (39, 63.9%), and tethered brachial plexus (29, 47.5%). The average hospital length of stay was 1 day, which is shorter than 4.8 days previously published. ³

Three patients (4.9%) experienced complications, which included pneumothorax in 1 patient (1.6%) and Horner syndrome in 2 patients (3.3%). Recurrent TOS requiring reoperation occurred in 1 extremity (1.6%). In that patient’s initial operation, a rib resection was not performed, as there was a large space between the clavicle and the first rib. On repeat operation, no other areas of compression were identified; therefore, the first rib was removed. There was no recurrence of symptoms following rib resection.

Patients undergoing NTOS decompression showed improvement in symptoms that progressed over a period of 12 months. At 12 months of follow-up, 27 patients (44.3%) had returned to full activities, and 17 patients (27.9%) stated that they had returned to activities with some limitations. Twenty-four patients (44%) had an average long-term follow-up of 69.5 (24-180) months. The average VAS improved 5.7 points, from 7.5 (±3.1, 5-10) preoperatively to 1.8 (±2.1, 0-6) postoperatively. The average SANE score improved 56.5 points, from 28.9 (±22.4, 0-90) to 85.4 (±21, 10-100). Postoperatively, the QuickDASH score was 11.4 (±16.6, 0-54.5), the CBSQ score was 27.4 (±25.8, 0-92), and the NTOS index score was 17.2 (±18.2, 0-63.6). The patient-reported outcomes are found in Table 1.

Table 1.

Patient Outcome Scores.

Outcome Measure	Preoperative	Postoperative
Visual Analog Score	7.5 ± 3.1 (5-10)	1.8 ± 2.1 (0-6)
SANE score	28.9 ± 22.4 (0-90)	85.4 ± 21.0 (10-100)
QuickDASH a		11.4 ± 16.6 (0-54.5)
CBSQ a		27.4 ± 25.8 (0-92)
NTOS index a		17.2 ± 18.2 (0-63.6)

Note. The +/- symbol refers to the standard deviation. The range is being shown in the parenthesis. SANE: Single Assessment Numeric Evaluation; QuickDASH: Quick Disabilities of the Arm, Shoulder, and Hand; CBSQ: Cervical Brachial Symptom Questionnaire; NTOS: neurogenic thoracic outlet syndrome.

^aOnly postoperative measurements obtained.

First rib excision was performed in less than half of the patients (47.62%). Comparing outcomes of patients having a first rib excision with those in whom their first rib was left intact, there was no significant difference in preoperative VAS score (P = .09), postoperative VAS score (P = .17), SANE score (P = .12), postoperative QuickDASH score (P = .24), postoperative CBSQ score (P = .58), and NTOS index (P = .30). These results are shown in Table 2. Table 3 reports comparisons between mechanisms of injury, including overuse, trauma, and idiopathic causes. There was no difference regarding clinical outcome measures in these groups, including VAS, SANE, QuickDASH, CBSQ, and NTOS index.

Table 2.

Outcomes of Rib Resection.

	Rib resection (8 extremities)	No rib resection (16 extremities)	P value
Preoperative VAS	9.0 ± 1.2 (7-10)	6.8 ± 3.5 (0-10)	.09
Postoperative VAS	1.0 ± 1.1 (0-3)	2.3 ± 2.4 (0-6)	.17
SANE Score	95.0 ± 5.3 (90-100)	80.6 ± 24.3 (10-100)	.12
QuickDASH	5.7 ± 7.9 (0-22.7)	14.2 ± 19.1 (0-54.5)	.24
CBSQ	23.1 ± 19.8 (0-49)	29.5 ± 28.7 (0-92)	.58
NTOS index	11.6 ± 11.2 (0-31.1)	20.0 ± 20.6 (0-63.6)	.30

Note. The +/- symbol refers to the standard deviation. The range is being shown in the parenthesis. VAS: Visual Analog Scale; SANE: Single Assessment Numeric Evaluation; QuickDASH: Quick Disabilities of the Arm, Shoulder, and Hand; CBSQ: Cervical Brachial Symptom Questionnaire; NTOS: neurogenic thoracic outlet syndrome.

Table 3.

Meyer Versus Caputo.

	Caputo (6 months)	Meyer (69.5 months) (range, 24-180)	P value
Preoperative VAS	5.8 ± 0.4	7.5 ± 3.1 (5-10)	.02
Postoperative VAS	1.8 ± 0.5	1.8 ± 2.1 (0-6)	1.00
DASH	7.8 ± 3.2	11.4 ± 16.6 (0-54.5)	.46
CBSQ	11.9 ± 4.4	27.4 ± 25.8 (0-92)	.03
NTOS index	10.4 ± 3.1	17.2 ± 18.2 (0-63.6)	.17

Note. The +/- symbol refers to the standard deviation. The range is being shown in the parenthesis. VAS: Visual Analog Scale; SANE: Single Assessment Numeric Evaluation; QuickDASH: Quick Disabilities of the Arm, Shoulder, and Hand; CBSQ: Cervical Brachial Symptom Questionnaire; NTOS: neurogenic thoracic outlet syndrome.

Discussion

There is limited literature on adolescent patients with NTOS. This is the largest study investigating adolescents treated surgically for NTOS. Our study outcomes in this age group, including symptom relief, return to activity within 6 months, and lack of serious complications, suggest long-term benefits of surgical management of NTOS.

Neurogenic thoracic outlet syndrome is difficult to diagnose as there are no criterion standard tests. It is usually a diagnosis of exclusion. It is often unrecognized in patients, especially adolescents, presenting with shoulder, arm, and hand pain.^16,17 This is consistent with our series of patients as they saw an average of 2 other providers before presenting to the operating surgeon, who specializes in TOS. The main complaints are vague paresthesia in the hand and a feeling of heaviness with overhead activities. Many patients anecdotally describe difficulty with fixing their hair. ¹⁸ More than 80% of patients had positive findings on a combined TOS examination, including the Tinel sign of the neck, Roos test, upper limb tension test, and the Spurling test. These history and physical examination characteristics are similar to those seen by Hong et al. ¹⁷ The study population consisted of 42 women and 12 men, with a female to male ratio of greater than 3:1, consistent with the current literature reflecting a female predominance.^4,6,18

In this study, a supraclavicular approach was used for complete visualization of the plexus. The surgical procedures were dictated by areas of compression on examination and intraoperative findings. Using this approach, most of our patients had resection of the anterior and middle scalene, resection of fibrous bands, and neurolysis. Figure 1 shows an example of a fibrous band that is constricting the plexus on its posterior border after removal of the anterior and middle scalene. In a study by Hong et al, ¹⁷ fibrous bands were cited to cause compression in 79% of their pediatric patients treated surgically for TOS. In pediatric patients, it is more common for a bony anomaly to exist in addition to these bands.^17,19 In our series of adolescents, the first rib was left intact in 66.7% of our patients with long-term follow-up with no difference in outcome measures, suggesting that patients who present in adolescence more closely mimic the pathology of adults than younger children.^6,17-19

Figure 1.

A fibrous band constricting the plexus on the posterior border.

Note. Anterior and middle scalene have already been resected at this point. This is a left-sided operation, and the medial aspect of the patient is on the left side of the figure. The arrow is pointing to the fibrous band held within the forceps after it has been pulled away from the brachial plexus.

Despite our hypothesis that patients with FRR would have improved outcomes, there was no significant difference in outcome scores between the groups. However, 1 patient required a reoperation for recurrent symptoms. After the rib excision, his clinical symptoms improved. Revision procedures to remove the rib are more difficult due to postoperative scarring and the change of anatomical landmarks, and there are few studies on complication rates after revision surgery.^18,20,21 In a study by Likes et al, recurrent TOS due to an intact rib or incomplete resection of a rib resulted in improvement in symptoms once the rib was removed. ²² The patients in this series had minimal complications; however, the surgeons used a different approach than the original surgery. ²² There are currently no guidelines on selection of patients, which would be beneficial. The decision to remove the first rib in this patient series was based on the compression observed by the primary surgeon (R.D.M.). Removal of the rib offers the surgeon the opportunity to further decompress any bands around the C8 and T1 nerve roots that are not visible with the rib intact. The morbidity of routine removal of the first rib must be weighed against the risks of a revision procedure in the decision to remove the first rib.

In patients with examination findings consistent with infraclavicular plexus compression, a separate incision was made over the coracoid to address these symptoms in 28 patients. Compression of the infraclavicular plexus by a coracocostal ligament was seen in 26 patients (92.9%). Figure 2 shows an intraoperative photo of the coracocostal ligament as it inserts on the posteromedial aspect of the coracoid. Previous literature does not mention these bands as a source of compression in TOS, yet nearly one-half of our total patient population had this anomaly. Coracocostal ligaments serve no known function in the shoulder, and no negative outcomes have been associated with their removal in our study population.

Figure 2.

The coracocostal ligament is visible on the posterior aspect of the coracoid aiming medially. This is the left side of a patient with medial on the left side of the figure and lateral on the right side of the figure.

Note. The arrow is pointing to the fibrous band held within the forceps after it has been pulled away from the brachial plexus.

Outcomes collected from our adolescent NTOS population were similar to those in the literature. Caputo et al., in comparing a cohort of 35 adolescents with their adult population, found that adolescents had improved outcomes after supraclavicular decompression. ⁶ Table 3 compares our postoperative outcomes with those of Caputo et al., The mean follow-up of our study was 69.5 (24-180) months versus 6 months in the study by Caputo et al. Comparing the patient-reported measures of our study and those of that by Caputo et al., preoperative VAS and CBSQ scores were significantly different. Otherwise, DASH score (converted from QuickDASH), postoperative VAS, and NTOS index were not.

The patients in this study had more pain and disability preoperatively than those in the series by Caputo et al. Despite this, the only postoperative measure that was significantly superior in the cohort of Caputo et al was CBSQ. This difference can possibly be explained by a higher preoperative pain tolerance in our cohort, as evidenced by significantly higher presenting VAS scores. Despite the higher VAS in our series, the adolescents had comparable results long-term similar to the adolescents in the series of Caputo et al. These results suggest that adolescent patients do well after TOS decompression and that they continue to maintain these results over a long-term follow-up period without significant worsening in pain scores and disability.

Several weaknesses exist in our study. This was a single-surgeon series, and intraoperative decisions were based on clinical acumen of the surgeon. This could introduce bias into our comparison of patients who had a rib resection versus those in whom the rib was left intact. There were low numbers within the comparison groups by rib resection and mechanism of injury. The study is underpowered due to TOS being a rare condition, and clinical differences between subgroups may not have been detected. There is a possibility of recall bias in our patient-reported outcomes as patients were asked to quantify their preoperative ability to perform certain tasks after the surgery had been performed. We were not able to have follow-up data on all of our patients.

Conclusion

Surgical decompression is indicated and beneficial in adolescent patients who fail conservative management. The procedure provides symptomatic relief and allows patients to return to desired activities. We present the largest series to date of adolescent patients undergoing thoracic outlet decompression with double the patients seen in the largest study in the current literature. ¹ This study also has the longest follow-up time. We found no difference in outcomes of patients treated with rib resection versus those in whom it was left intact. Overall, patients did well long-term after decompression of the thoracic outlet.