PresentedVascular Surg

DepartmenTexas MedicalInstitute, Hous

CorrespondCardiothoracicSchool and M

Outcomes of Surgical Paraclavicular ThoracicOutlet Decompression

Sapan S. Desai, Mohammad Toliyat, Anahita Dua, Kristofer M. Charlton-Ouw,

Monir Hossain, Anthony L. Estrera, Hazim J. Safi, and Ali Azizzadeh, Houston, Texas

Background: Thoracic outlet syndrome (TOS) is a constellation of signs and symptoms causedby compression of the neurovascular structures in the thoracic outlet. These structures includethe brachial plexus, the subclavian vein, and the subclavian artery, resulting in neurogenic(NTOS), venous (VTOS), and arterial (ATOS) types of TOS, respectively. The purpose of thisstudy was to evaluate the outcomes of paraclavicular surgical decompression for TOS.Methods: A prospective analysis of patients who underwent surgical decompression for TOS ata newly established center was performed. Diagnosis of TOS was based on clinical history,a physical examination, and additional diagnostic studies. The indication for surgery in patientsdiagnosed with NTOS was the presence of persistent symptoms after a trial of physical therapy.Primary outcomes were assessed according to Derkash’s classification as excellent, good, fair,and poor. Secondary outcomes included mortality, complications, and duration of hospital stay.Results: Between August 2004 and June 2011, 40 paraclavicular decompression procedureswere performed on 36 patients (16 men) with TOS. The mean age was 36.5 years (range:15e68). Bilateral decompression was performed on 4 patients. The types were NTOS (n ¼19; 48%), VTOS (n ¼ 16; 40%), and ATOS (n ¼ 5; 12%). In addition to pain, the most commonpresenting symptom was numbness in NTOS, swelling in VTOS, and coolness in ATOS. Ahistory of trauma was present in 22.2%. Two patients suffered from recurrent symptoms afterprevious transaxillary first rib resection for VTOS at another institution. Diagnostic tests per-formed included nerve conduction studies (43%), venogram (40%), and arteriogram (20%).All patients underwent paraclavicular decompression, which included radical anterior and partialmiddle scalenectomy, brachial plexus neurolysis, and first rib resection. The first rib resectionwas partial, through a supraclavicular only approach in NTOS and ATOS patients (60%) orcomplete, through a supra- and infraclavicular approach for VTOS patients (40%). Functionaloutcomes were excellent, good, fair, and poor in 74.4%, 15.4%, 10.3%, and 0% of cases,respectively. One patient was lost to follow-up. Two patients with incomplete relief of symptomsafter paraclavicular decompression for NTOS underwent pectoralis minor decompression. Therewere no deaths. Complications included pleural effusion requiring evacuation (n ¼ 4), neuro-praxia (n ¼ 1), and lymph leak (n ¼ 1) treated with tube thoracostomy. No patients experiencedinjury to the long thoracic or phrenic nerves. The mean duration of hospital stay was 4.4 days.The mean follow-up was 10.3 months.Conclusions: In our experience, surgical paraclavicular decompression can provide safe andeffective relief of NTOS, VTOS, and ATOS symptoms. Functional outcomes were excellent orgood in the majority of patients, with minimal complications.

at the 40th Annual Meeting of the Society for Clinicalery, Las Vegas, NV, March 13e17, 2012.

t of Cardiothoracic and Vascular Surgery, University ofSchool and Memorial Hermann Heart and Vascularton, TX.

ence to: Ali Azizzadeh, MD, FACS, Department ofand Vascular Surgery, University of Texas Medicalemorial Hermann Heart and Vascular Institute, 6400

Fannin, Suite 2850, Houston, TX 77030, USA; E-mail: Ali.Azizzadeh@uth.tmc.edu

Ann Vasc Surg 2014; 28: 457–464http://dx.doi.org/10.1016/j.avsg.2013.02.029� 2014 Elsevier Inc. All rights reserved.

Manuscript received: December 4, 2012; manuscript accepted: February

22, 2013; published online: December 30, 2013.

457

458 Desai et al. Annals of Vascular Surgery

Thoracic outlet syndrome (TOS) is a constellation electively 4e6 weeks after the acute episode. Pa-

of signs and symptoms caused by compression of

the neurovascular structures in the thoracic outlet.

These structures include the brachial plexus, the

subclavian vein, and the subclavian artery. Depend-

ing on the specific structure being affected, TOS is

classified into 3 types: neurogenic (NTOS), venous

(VTOS), and arterial (ATOS). NTOS is related to

compression of the brachial plexus that may lead

to pain, numbness, and weakness in the hand; this

is by far the most common type of TOS.1 VTOS is

caused by chronic compression and scarring of the

subclavian vein leading to arm edema. Acute

thrombosis of the subclavian vein, known as effort

thrombosis or PageteSchroetter syndrome, deve-

lops in many patients with VTOS. ATOS is typically

related to compression of the subclavian artery by

osseous structures, with subsequent dilation, or

aneurysmal degeneration resulting in thrombus

formation and possible distal embolization to the

arm and/or hand.1e6 The purpose of this study

was to evaluate the outcomes of surgical decompres-

sion at a newly established center for TOS.

METHODS

Patient Population

A new program for the treatment of TOS was estab-

lished in the Department of Cardiothoracic and

Vascular Surgery at the University of Texas Medical

School in Houston, TX. A prospective database con-

sisting of patients undergoing surgical decompres-

sion was maintained. We compiled data on clinical

presentation, diagnostic work-up, procedures per-

formed, postoperative care, and outcomes. Based

on these findings, TOS was characterized by type

(i.e., NTOS, VTOS, and ATOS). Diagnosis of NTOS

was based on a thorough history and clinical exam-

ination. We have found the elevated arm stress test

to be a reliable tool in the diagnosis of NTOS. Nerve

conduction tests were ordered. This was primarily

done to rule out other conditions that mimic

NTOS, such as ulnar nerve compression or carpal

tunnel syndrome. All patients with a clinical diag-

nosis of NTOS were offered a trial of physical

therapy. Surgical decompression was offered to can-

didates who had persistent symptoms after physical

therapy. For patients with VTOS, in addition to a

history and physical examination, a venogram was

obtained in supine (arm adduction) and stress (arm

abduction) positions. Patients presenting with acute

PageteSchroetter syndrome underwent thromboly-

sis and were discharged with oral anticoagula-

tion therapy. Open decompression was scheduled

tients with ATOS, in addition to history and physical

examination, underwent a 3-view chest radiograph

(CXR) to rule out a cervical rib or other osseous

anomalies. In addition, an arteriogramwas obtained

to evaluate for a subclavian artery aneurysm as well

as the presence of distal emboli.

Surgical Procedure

All patients underwent surgical decompression

using a paraclavicular approach. All patients with

NTOS and the majority with ATOS had a supracla-

vicular incision only. Occasionally, patients with

ATOS who had a large subclavian artery aneurysm

required an additional infraclavicular incision for

distal arterial control. In patients with VTOS, an

additional infraclavicular incision was made for

resection of the anterior segment of the first rib.

After general anesthesia, the patient was placed

in a supine position with a shoulder roll. The neck

was extended and rotated to the opposite side of

the procedure. A transverse incisionwasmade 1 fin-

gerbreadth above the clavicle, superior and inferior

platysmal flapss were developed, and the lateral

head of the sternocleidomastoid muscle was

divided. The scalene fat pad was then dissected at

the lateral border of the internal jugular vein and

retracted laterally (Fig. 1A). The thoracic duct was

divided in left-sided procedures.

The anterior scalene muscle and scalenus mini-

mus (when present) were circumferentially

dissected and radically resected after protecting the

phrenic nerve and subclavian artery (Fig. 1B). The

anterior scalene muscle was circumferentially

dissected from its insertion on the first rib to its

origin at the C6 transverse process. The subclavian

artery was further mobilized by the division of the

thyrocervical trunk (Fig. 1C). Neurolysis of the

brachial plexus (nerve roots C5eT1) was performed

(Fig. 1D). A partial middle scalenectomy was then

performed with care taken to preserve the

tributaries of the long thoracic nerve. The first rib

was then identified and the posterior portion

(from the transverse process of the cervical spine

to the tubercle) resected. Any cervical ribs present

were also removed.

In patients with VTOS, the anteromedial portion

of the first rib was then accessed via a small infracla-

vicular incision (Fig. 2A) and the rib was resected

(Fig. 2B). The subclavian vein was then mobilized

from the supraclavicular approach and venolysis

was performed, followed by an intraoperative veno-

gram after decompression. In patients for whom

significant stenosis or occlusion of the subclavian

Fig. 1. (A) The supraclavicular incision is seen in this

panel with the scalene fat pad being retracted laterally.

(B) The phrenic nerve is identified and protected; the

anterior scalene muscle is seen in this panel. (C) After

removal of the anterior scalene, the subclavian artery is

mobilized. (D) Neurolysis of the brachial plexus is per-

formed. The middle scalene and the posterior segment

of the first rib are resected. (The subclavian artery here

is encircled in a vessel loop).

Fig. 2. (A) Supraclavicular and infraclavicular incisions

used for venous thoracic outlet syndrome (VTOS)

decompression procedures. (B) The entire first rib is

resected in VTOS procedures. (C) Preoperative venogram

of a patient with VTOS. (D) Postoperative venogram of

a patient with VTOS after decompression and subclavian

vein reconstruction.

Vol. 28, No. 2, February 2014 Thoracic outlet decompression outcomes 459

vein was apparent, venous reconstruction was per-

formed using the great saphenous vein. Satisfactory

flow after subclavian vein repair was documented

by complete venography (Fig. 2D). Patients with

VTOS were restarted on anticoagulation therapy

on postoperative day (POD) 2. Anticoagulation

was continued for 3 months, at which time

a follow-up venography was performed.

460 Desai et al. Annals of Vascular Surgery

In patients with ATOS, the subclavian artery

aneurysm was resected. The artery was recon-

structed using an autogenous conduit whenever

possible. A completion arteriogram was performed,

and any evidence of previous distal embolization

was treated with open thrombectomy through

a brachial artery approach. Patients with ATOS

and distal embolization were restarted on anticoa-

gulation on POD 2. Anticoagulation therapy was

continued for 3 months.

After placement of a drain and closure of all inci-

sions, patients were then extubated and taken to the

postanesthesia care unit for recovery. Most patients

had their drains removed at the time of discharge,

which was typically on POD 4. Patients were seen

in the clinic at 2, 6, and 12 weeks after discharge.

Data Analysis

Patients completed a postoperative survey to assess

their recovery. Derkash’s classification was used to

categorize primary outcomes as excellent, good,

fair, or poor. Secondary outcomes included

mortality, complications, and duration of hospital

stay. The Student’s t-test was used to determine

significance using a P value < 0.05. Descriptive

statistics are presented as either mean ± standard

error of the mean or a range of values when

relevant.

RESULTS

Thirty-six patients underwent 40 successful paracla-

vicular decompressions for TOS between August

2004 and June 2011. The average follow-up was

10.3 months. Based on a combination of history,

physical examination, and diagnostic tests, patients

were classified as having NTOS (n ¼ 19; 48%),

VTOS (n ¼ 16; 40%), or ATOS (n ¼ 5; 12%). The

incidence of each type of TOS varies from the overall

population incidence because of referral bias and

because these were the actual number of patients

who underwent thoracic outlet decompression,

not the number who were diagnosed and treated

nonsurgically. Four patients underwent bilateral

decompression. Table I shows the characterization

of patient symptoms.

The symptoms of pain and numbness were espe-

cially common in patients with NTOS (n ¼ 19) and

ATOS (n ¼ 5; 100% each for pain; 100% and

80%, respectively, for numbness), while the chief

complaint in patients with VTOS was swelling (n ¼15; 94%). Weakness (n ¼ 16; 84%) and fatigue (n

¼ 14; 74%) were also common for NTOS, while

fatigue (n ¼ 4; 80%) and coolness (n ¼ 5; 100%)

were common for ATOS. The only patients who

had ulceration were those with ATOS (n ¼ 2;

40%). All patients reported some difficulty with

motor function. Trauma was relatively uncommon,

with relevant findings in 26% (n ¼ 5) of those with

NTOS, 13% (n¼ 2) of those with VTOS, and 20% (n

¼ 1) of those with ATOS.

Only 2 patients (11%) reported a previous nerve

procedure (ulnar nerve transposition and/or carpal

tunnel release). Ten patients (63%) had previous

thrombolysis, 2 (5%) with previous stent, and 2

(13%) with previous axillary rib resection at an

outside institution. Patients with ATOS were more

likely to have previous thrombolysis (n ¼ 4; 80%),

previous stent (n ¼ 1; 20%), and stent graft place-

ment (n ¼ 1; 20%).

Diagnostic workup included CXR (36/40 encoun-

ters), computed tomography (CT; 9/40), magnetic

resonance imaging (MRI; 17/40), electromyography

(EMG; 17/40), nerve conduction study (17/40),

arteriogram (10/40), and venogram (16/40).

Patients with NTOS had CXR (79%), CT (21%),

MRI (68%), EMG (84%), and/or nerve conduction

study (79%). Two NTOS patients who were sus-

pected of having some component of arterial

involvement underwent an arteriogram (11%); no

patients required venography. All nerve conduction

studies and EMGs were normal for NTOS patients;

none of these patients had a previous scalene block.

An anterior and middle scalenectomy was per-

formed in all 40 procedures. A cervical rib resection

was performed in 7 patients. Cervical ribs were most

commonly found in patients with ATOS (80%)

compared to only 5% for NTOS. Resection of the

first rib was completed in all patients with NTOS,

14 patients (88%) with VTOS, and 2 patients

(40%) with ATOS. Two VTOS patients had previ-

ously undergone a transaxillary first rib resection.

In ATOS patients where the resected cervical rib

was identified as the cause of TOS, the first rib was

spared. Brachial plexus neurolysis was completed

in all patients. Thrombectomy (80%) and arterial

reconstruction (100%) was necessary in ATOS,

while venous reconstruction was necessary in

81% of VTOS patients. No patients with NTOS

required vascular interventions (Table II).

The overall duration of the procedure was 2.9

hours in NTOS, 4.0 hours in VTOS, and 3.6 hours

in ATOS (P > 0.05). The duration of hospital stay

varied between 3.6 days for NTOS, 5.5 days in

VTOS, and 4.2 days in ATOS. Overall, only 2

patients were readmitted, both for shortness of

breath. Two patients with NTOS were readmitted

for shortness of breath (PODs 7 and 14, respec-

tively). Neither patient required any intervention

Table I. Clinical findings by thoracic outlet syndrome subtype out of 36 total patients treated using

a paraclavicular approach over the course of 40 patient encountersa

Symptom, n (%)NTOS (n ¼ 19encounters; 48%)

VTOS (n ¼ 16encounters; 40%)

ATOS (n ¼ 5encounters; 12%)

Pain 19 (100) 9 (56) 5 (100)

Numbness 19 (100) 4 (25) 4 (80)

Weakness 16 (84) 3 (19) 1 (20)

Fatigue 14 (74) 3 (19) 4 (80)

Headache 9 (47) 1 (6) 0

Coolness 8 (42) 0 5 (100)

Ulceration 0 0 2 (40)

Swelling 8 (42) 15 (94) 0

Difficulty with elevation (90�) 19 (100) 12 (75) 4 (80)

Difficulty with reaching overhead (180�) 18 (95) 12 (75) 4 (80)

Difficulty with lifting 17 (89) 9 (56) 4 (80)

Difficulty with typing 10 (53) 6 (38) 3 (60)

Difficulty with driving 13 (68) 8 (50) 3 (60)

Difficulty using the telephone 15 (79) 8 (50) 3 (60)

Difficulty shaving/combing 17 (89) 8 (50) 3 (60)

Relevant trauma 5 (26) 2 (13) 1 (20)

Previous physical therapy 19 (100) 0 0

Previous thrombolysis 0 10 (63) 4 (80)

Previous stenting 0 1 (6) 1 (20)

Previous nerve procedure 2 (11) 0 0

Previous rib removal 0 2 (13) 0

Previous stent graft 0 1 (6) 1 (20)

ATOS, Arterial thoracic outlet syndrome; NTOS, neurogenic thoracic outlet syndrome; VTOS, venous thoracic outlet syndrome.aThe percentages in the first row are of the total number of encounters. All succeeding percentages are of the subset of thoracic outlet

syndrome.

Vol. 28, No. 2, February 2014 Thoracic outlet decompression outcomes 461

after excluding pneumothorax and pulmonary

embolism, and both were discharged a few days

later. Additional complications included pneumo-

thorax (n¼ 3), pleural effusion (n¼ 13), hematoma

(n ¼ 4), and brachial plexus injury (n ¼ 1), a neuro-

praxia secondary to retraction that resolved sponta-

neously (Table III). Most patients experienced

resolution of their chief complaint and were back

to their normal activities within 2e4 weeks after

surgery, with only 1 patient lost to follow-up. Two

patients reported incomplete symptom relief after

paraclavicular decompression and underwent pec-

toralis minor decompression; they reported cessa-

tion of all clinically significant complaints after

their subsequent procedure. TOS and pectoralis

minor syndrome coexist in a small subset of patients

with NTOS, and both present similarly. Our algo-

rithm is to treat for the NTOS first to see if this will

lead to a relief of symptoms. Continuing pain, weak-

ness, paresthesias, and point tenderness over the

pectoralis minor are indications for pectoralis minor

decompression.

All patients had their recovery measured using

Derkash’s classification (Table IV). Fourteen patients

in the NTOS group categorized their recovery as

‘‘excellent,’’ while 2 patients each rated their

recovery as ‘‘good’’ and ‘‘fair.’’ Eleven patients in

the VTOS group rated their recovery as ‘‘excellent,’’

4 as ‘‘good,’’ and 1 ‘‘fair.’’ Four patients in the ATOS

group experienced ‘‘excellent’’ results, and only 1

‘‘fair’’ results. No patient in this study experienced

‘‘poor’’ results or any type of vascular injury.

DISCUSSION

Surgical Approach

Clagett7 described a posterior thoracotomy in 1962

and Roos8 described a transaxillary approach in

1966 for the treatment of TOS. The transaxillary

approach rose to prominence in the 1970se80s

but was later criticized for a higher than usual inci-

dence of nerve injury.9,10 A combined supraclavicu-

lar and transaxillary approach was briefly advocated

in the mid-1980s, but this was replaced by a purely

supraclavicular approach pioneered by Reilly and

Stoney.11,12 Thompson13 and Thompson et al.14

described the paraclavicular approach as an alterna-

tive to the supraclavicular approach, especially in

patients with VTOS.

Table II. Operative intervention by thoracic

outlet syndrome subtype out of 40 total

encounters

Operative intervention,n (%) NTOS VTOS ATOS

Anterior scalenectomy 19 (100) 16 (100) 5 (100)

Middle scalenectomy 19 (100) 16 (100) 5 (100)

Cervical rib resection 1 (5) 2 (13) 4 (80)

First rib resection 19 (100) 14 (88) 2 (40)

Neurolysis of brachial

plexus

19 (100) 15 (94) 5 (100)

Angiolysis 3 (16) 13 (81) 5 (100)

Thrombectomy 0 0 4 (80)

Arterial reconstruction 0 0 5 (100)

Venous reconstruction 0 13 (81) 0

Patch angioplasty 0 8 (50) 0

Arterial or venous

bypass

0 5 (31) 5 (100)

ATOS, Arterial thoracic outlet syndrome; NTOS, neurogenic

thoracic outlet syndrome; VTOS, venous thoracic outlet

syndrome.

Table III. Overall duration of procedure,

duration of hospital stay, and incidence of

complications out of 40 total encounters

NTOS VTOS ATOS

Duration of operation (hrs) 2.9 4.0 3.6

Duration of hospital stay (days) 3.6 5.5 4.2

Readmitted within 30 days,

n (%)

2 (11) 0 0

Pneumothorax, n (%) 1 (5) 1 (6) 1 (20)

Pleural effusion, n (%) 7 (37) 6 (38) 0

Hematoma with evacuation,

n (%)

1 (5) 3 (19) 0

Wound infection, n (%) 0 0 0

Brachial plexus injury, n (%) 0 0 1 (20)

Phrenic nerve injury, n (%) 0 0 0

Long thoracic nerve injury,

n (%)

0 0 0

Lymph leak, n (%) 0 0 0

ATOS, Arterial thoracic outlet syndrome; NTOS, neurogenic

thoracic outlet syndrome; VTOS, venous thoracic outlet

syndrome.

Table IV. Functional outcomes according to type

of thoracic outlet syndrome categorized using

Derkash’s classification

Derkash’s classification

Excellent Good Fair Poor

NTOS, n (%) 14 (78) 2 (11) 2 (11) 0

VTOS, n (%) 11 (69) 4 (25) 1 (6.3) 0

ATOS, n (%) 4 (80) 0 1 (20) 0

ATOS, Arterial thoracic outlet syndrome; NTOS, neurogenic

thoracic outlet syndrome; VTOS, venous thoracic outlet

syndrome.

462 Desai et al. Annals of Vascular Surgery

Compared to the other approaches, we believe

that the paraclavicular approach to TOS offers

clearer views of the brachial plexus and major

vascular structures in an ample surgical field, affords

easy accessibility to cervical and aberrant first ribs,

permits complete resection of the scalene muscles

and associated scar tissue, and allows for a full

brachial plexus neurolysis. In addition, arterial and

venous reconstruction can be performed with ease

under direct visualization. In our series, most

patients with NTOS and ATOS had a supraclavicular

incision only. Patients with VTOS had an additional

infraclavicular incision for resection of the anterior

segment of the first rib.

NTOS

Controversy remains over the diagnosis and

management of NTOS. We consider NTOS a clinical

diagnosis that is primarily based on history and

physical examination. Most of the additional diag-

nostic testing was performed to rule out other condi-

tions that mimic NTOS, such as ulnar nerve

compression or carpal tunnel syndrome. In fact, all

EMG and nerve conduction studies in our NTOS

group were normal. Arteriograms were performed

selectively to rule out an arterial component. We

have found that a good clinical examination is often

sufficient to differentiate NTOS from the other forms

of TOS. All patients diagnosed with NTOS under-

went a trial of physical therapy before being consid-

ered for surgery. Patients who continued to have

clinically significant and lifestyle-limiting symptoms

after physical therapy were offered surgical decom-

pression. Similar selective protocols for surgical

treatment of NTOS have been advocated by

others.4,15,16

Our results mirror what has previously been re-

ported by other centers.3e5 Out of the 19 patients

who underwent decompression for NTOS, all but 2

(89%) reported a resolution in their symptoms

and improvement in muscle strength at follow-up.

One of those patients reported about 80% improve-

ment in symptoms, while the other patient devel-

oped new right upper extremity symptoms that

were attributed to recurrent scarring. In the current

series, the operative time and duration of hospital

stay were shorter for NTOS when compared to

VTOS and ATOS. This is consistent with a more

complex decompression procedure and vascular

reconstruction that is often required for vasculo-

genic TOS (i.e., VTOS and ATOS).

Vol. 28, No. 2, February 2014 Thoracic outlet decompression outcomes 463

The functional outcomes reported by NTOS

patients who underwent paraclavicular decompres-

sion indicate that most patients can expect total or

near-total resolution of their symptoms and imp-

rovement in their range of motion after surgery.

Similar outcomes have been reported by Chandra

et al.4 for NTOS decompression, with 90% of the

patients experiencing resolution of their symptoms

at 1 year of follow-up. A high index of suspicion

for complications, such as pneumothorax, hema-

toma, and pleural effusions, can help reduce the

risk profile of this procedure. Our results, combined

with other recent studies, appear to indicate that

offering carefully selected patients with NTOS

a surgical option can provide long-term and effec-

tive relief.3e5

VTOS

Patients were selected for surgery on the basis

of their history, symptoms, and diagnostic ima-

ging studies. The majority of patients reported arm

swelling (94%), difficulty with arm elevation

(75%), difficulty with reaching overhead (75%),

and other motor defects. All patients with VTOS

had a CXR and venogram. Patients presenting

with effort thrombosis or PageteSchroetter syn-

drome underwent catheter-directed thrombolysis

as first-line therapy.17,18 They were subsequently

discharged on oral anticoagulation followed by elec-

tive formal surgical decompression at 4e6 weeks.

In our study, 10 of 16 patients (63%) had previously

undergone thrombolysis, 1 patient had a previous

stent (6%), and 2 (13%) had previous axillary

rib resection.Wedonot believe that there is a role for

stent placement in patients with VTOS, because the

primary cause is external compression of the vein.

VTOS patients underwent supraclavicular

decompression similar to those with NTOS. In addi-

tion, VTOS patients required an infraclavicular inci-

sion for resection of the anterior segment of the first

rib (paraclavicular approach). Two patients had

previous transaxillary rib resections and did not

require this portion of the surgery. Based on the

intraoperative venogram after surgical decompres-

sion and venolysis, the majority (81%) of the

VTOS group required formal venous reconstruction.

In half the cases (50%), this consisted of a vein patch

angioplasty, while some (31%) required a bypass.

There were no readmissions within 30 days of

surgery. Major complications included 6 patients

(38%) with clinically insignificant pleural effusion

and 3 (19%) with a hematoma that required evacu-

ation. We believe the high reexploration rate was

related to early postoperative anticoagulation.

While early anticoagulation carries a high risk of

bleeding, it is often necessary to maintain patency

in the setting of a fresh venous reconstruction. In

our view, the initiation of the postoperative antico-

agulation has to be tailored to the specific patient

and based on the complexity of the procedure, the

presence and type of venous reconstruction (patch

versus bypass), and the amount of drain output.

Most patients are started on an intravenous heparin

drip on POD 2. Based on the findings of this study,

however, we may delay the initiation of anticoagu-

lation selectively. Functional outcomes based on

Derkash’s classification were similar to the results

seen in the NTOS group. About 94% of the VTOS

patients experienced ‘‘excellent’’ or ‘‘good’’ func-

tional outcomes, indicating that the paraclavicular

decompression appears to be a safe and effective

means for the treatment of this subtype.

ATOS

Only 5 of the 40 (13%) encounters in our study

were for ATOS, and similar to the report by Coote19

in 1861, all of our patients also had either a cervical

or anomalous first rib. While about 13% of our

encounters were for ATOS compared to the 1e6%

reported in the literature, the bias in our study is

caused by referral patterns and the fact that only

patients who underwent surgery were included in

the sample population.6,20e22

All 5 patients with ATOS reported pain and cool-

ness of the affected extremity. Four also reported

numbness, fatigue, and a limited range of motion.

Two patients had extremity ulcers consistent with

distal embolism secondary to a subclavian artery

aneurysm. Four patients had previous thrombolysis

before formal decompression. Our diagnosis was

confirmed by CXR and arteriograms in all 5 patients.

All ATOS patients underwent supraclavicular

decompression with anterior and middle scalenec-

tomy. One patient required an additional infracla-

vicular incision to obtain distal control of a large

subclavian artery aneurysm. Four patients had

resection of a cervical rib and 2 had a partial first

rib resection. We were able to spare the first rib in

selected patients with ATOS in whom resection of

a cervical rib provided adequate decompression.

All 5 patients also had limited neurolysis of the

brachial plexus, angiolysis, and arterial reconstruc-

tion with bypass. We preferentially used autologous

conduits when possible. We have also successfully

used prosthetic conduits, such as Dacron (DuPont,

Wilmington, NC) or polytetrafluoroethylene when

suitable autologous alternatives are unavailable. In

our series, 1 patient experienced a pneumothorax

464 Desai et al. Annals of Vascular Surgery

during surgery because of central venous catheter

placement and required a chest tube to be placed

intraoperatively. This patient also had neuropraxia,

likely related to retraction that resolved spontane-

ously. The other 4 patients had an uneventful

recovery, with full resolution of their symptoms at

follow-up. Four patients reported ‘‘excellent’’

results. Overall, it appears that supraclavicular

decompression appears to be a safe and effective

means of treating ATOS.

CONCLUSION

Paraclavicular decompression is a safe and effective

means of treating all forms of TOS. Functional

outcomes are rated as ‘‘good’’ or ‘‘excellent’’ in the

majority of patients, with few serious complications

after the procedure. Properly selected patients can

be offered paraclavicular decompression as an

option for resolving their symptoms.

REFERENCES

1. Fugate MW, Rotellini-Coltvet L, Freischlag JA. Current

management of thoracic outlet syndrome. Curr Treat

Options Cardiovasc Med 2009;11:176e83.

2. Melby SJ, Vedantham S, Narra VR, et al. Comprehensive

surgical management of the competitive athlete with effort

thrombosis of the subclavian vein (Paget-Schroetter

syndrome). J Vasc Surg 2008;47:809e21.

3. Thompson RW, Driskill MR. Neurovascular problems in the

athlete’s shoulder. Clin Sports Med 2008;27:789e802.

4. Chandra V, Olcott C IV, Lee JT. Early results of a highly

selective algorithm for surgery on patients with neurogenic

thoracic outlet syndrome. J Vasc Surg 2011;54:1698e705.

5. Chang DC, Rotellini-Coltvet LA, Mukherjee D, et al. Surgical

intervention for thoracic outlet syndrome improves patient’s

quality of life. J Vasc Surg 2009;49:630e5.

6. Urschel HC Jr, Kourlis H Jr. Thoracic outlet syndrome: a 50-

year experience at Baylor University Medical Center. Proc

Bayl Univ Med Cent 2007;20:125e35.

7. Clagett OT. Presidential address: research and proresearch. J

Thorac Cardiovasc Surg 1962;44:153e66.8. Roos DB. Transaxillary approach for first rib resection to

relieve thoracic outlet syndrome. Ann Surg 1966;163:354e8.

9. Dale A. Thoracic outlet compression syndrome: critique in

1982. Arch Surg 1982;117:1437e45.

10. Cikrit DF, Haefner R, Nichols WK, et al. Transaxillary or

supraclavicular decompression for the thoracic outlet

syndrome: a comparison of the risks and benefits. Am

Surg 1989;55:347e52.

11. Qvarfordt PG, Ehrenfeld WK, Stoney RJ. Supraclavicular

radical scalenectomy and transaxillary first rib resection for

the thoracic outlet syndrome: a combined approach. Am J

Surg 1984;148:111e6.

12. Reilly LM, Stoney RJ. Supraclavicular approach for thoracic

outlet decompression. J Vasc Surg 1988;8:329e34.13. Thompson R. Venous thoracic outlet syndrome: paraclavic-

ular approach. Opertaive Techniques in General Surgery

2008;10:113e21.

14. Thompson JF, Winterborn RJ, Bays S, et al. Venous thoracic

outlet compression and the Paget-Schroetter syndrome:

a review and recommendations for management. Cardio-

vasc Intervent Radiol 2011;34:903e10.

15. Wilbourn AJ. The thoracic outlet syndrome is overdiag-

nosed. Arch Neurol 1990;47:328e30.

16. Wilbourn AJ. Thoracic outlet syndromes: a plea for conser-

vatism. Neurosurg Clin N Am 1991;2:235e45.

17. Druy EM, Trout HH 3rd, Giordano JM, et al. Lytic therapy in

the treatment of axillary and subclavian vein thrombosis.

J Vasc Surg 1985;2:821e7.

18. Lee JT, Karwowski JK, Harris EJ, et al. Long-term

thrombotic recurrence after nonoperative management

of Paget-Schroetter syndrome. J Vasc Surg 2006;43:

1236e43.

19. Coote H. Exostosis of the transverse process of the seventh

cervical vertebra, surrounded by blood vessels and nerves:

successful removal. Lancet 1861;1:360e1.

20. Makhoul RG, Machleder HI. Developmental anomalies

at the thoracic outlet: an analysis of 200 consecutive cases.

J Vasc Surg 1992;16:534e45.

21. Sanders RJ, Hammond SL, Rao NM. Diagnosis of thoracic

outlet syndrome. J Vasc Surg 2007;46:601e4.22. Criado E, Berguer R, Greenfield L. The spectrum of arterial

compression at the thoracic outlet. J Vasc Surg 2010;52:

406e11.