| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2004;25:173-178
© 2004 Elsevier Science NL
inasi Yavuzer*
Department of Thoracic Surgery, Ankara University Faculty of Medicine, 
bn-i Sina Hospital, 06100, S
hhiye, Ankara, Turkey
Received 23 July 2003; received in revised form 16 November 2003; accepted 19 November 2003.
* Corresponding author. Address: Department of Thoracic Surgery, Ankara University School of Medicine, Koza sokak 114-86, TR-06670 Gaziosmanpasa, Turkey. Tel.: +90-312-310-33-33; fax: +90-312-310-63-71
e-mail: yavuzers{at}hotmail.com
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
Key Words: Thoracic outlet syndrome • Surgery • Surgical outcome
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
Physical treatment may be attempted as the initial treatment in patients with TOS, whereas surgical removal of the first and cervical rib if present has been suggested as the optimal treatment of choice in patients who do not benefit from physiotherapy. However, the optimal surgical approach remains controversial. Several routes such as posterior, transaxillary, supraclavicular, infraclavicular, transthoracic incisions and through the bed of the resected clavicle have been described for the surgical management of TOS. Among these, transaxillary route, which showed comparably favorable results, has become the most popular approach among the thoracic surgeons since its introduction by Roos in 1966 [1,2].
Although numerous studies have appeared in the literature regarding overall surgical results in TOS, we are aware of only a few reports analyzing the impact of particular clinical features with different results on the surgical outcome of patients [3–7]. Thus, the optimal selection criteria and the timing of surgery also remain controversial in patients with TOS. We investigated the clinical features in a series of 127 patients with TOS who underwent operation via transaxillary approach to clarify the clinical predictors on surgical outcome.
|
2. Patients and methods |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
.Y.) in bn-i Sina Hospital of Ankara University School of Medicine between November 1984 and November 2002. All operations were performed via transaxillary route. Posterior thoracoplasty approach was done for recurrent TOS in five cases who were excluded from the study. Nevertheless, the data of surgical outcome was available in only 127 cases. Conclusively, a total of 127 patients with TOS, which consisted of 17 (13%) male and 110 (87%) female patients with a mean age of 32.1±10.0 years (range 14–62 years) were enrolled for this study. The mean duration of symptoms was 38.8±37.1 months (range 1–120 months). Symptoms were arm pain in 49 (32%) patients, paresthesia of the arm in 43 (28%) patients, headache and neck pain in 13 (8%) patients, cyanosis of the hand and the arm in 11 (7%) patients, shoulder pain in 10 (7%) patients, paresis in the arm in 10 (7%) patients, hypoesthesia and sense of coldness in the arm in six (4%) patients, edema of the hand and the arm in six (4%) patients, and atrophy of the hand in five (3%) patients. Four patients with concomitant carpal tunnel syndrome had also previously undergone a median nerve release. Diagnosis of TOS was established on the basis of history, physical examination, electroneuromyographic (ENMG) and color Doppler findings. In addition, angiography was performed in two patients with presumptive diagnosis of vascular TOS. Patients were examined with chest and cervical X-rays to detect bony or other abnormalities. Patients were also examined for a possible diagnosis of cervical disc hernia and carpal tunnel syndrome if symptoms existed. Computed tomographic scan or recently magnetic resonance imaging was carried out in patients with highly suggestive cervical disc hernia. Ulnar nerve conduction velocity (UNCV) was recorded as measured across the thoracic outlet. Operative indications were persistence of symptoms following either physical or medical treatments, or an UNCV value below 60 m/s.
2.2. Surgical procedure
All patients underwent a transaxillary approach to allow thoracic outlet decompression The surgical procedure consisted of an extraperiosteal first rib resection, excision of fibrous bands, and removal of the cervical rib if present. A transverse incision is made below the hairline between the pectoralis major and latissimus dorsi muscles, and deepened to the external intercostalis fascia. Care was taken to avoid a possible injury to the intercostobrachial nerve, and superior thoracic artery and vein. Thereafter, the anterior scalene muscle is divided from its insertion to the first rib. The medial scalene muscle and the intercostal muscle of the second rib were divided with a periosteal elevator. The rib was then divided at its middle portion. The anterior part of the first rib was grasped with an alligator forceps, and the anterior part was divided at the level of costoclavicular ligament. Similarly, the posterior part was grasped and was divided from its articulation with the transverse process of the vertebra. If any bony remnant was left on the vertebra, this was gagged with the use of a pituitary rongeur. If a cervical rib or fibromuscular bands were present, they were removed. The apical pleura was detected for tears, and if present immediately repaired with absorbable sutures. The skin was closed after insertion of a drain, which was removed with an average of 2 or 3 days later.
The patients were encouraged to do stretching exercises in the early post-operative period; however, they were strictly warned not to lift heavy weights for at least 2 months. Patients received physical therapy in the post-operative period.
2.3. Data collection and follow-up
We retrospectively reviewed the data charts and clinical outcomes of the patients. The data of color Doppler examination, Adson test, and history of trauma were available in 96, 119, and 97 patients, respectively. The data of surgical outcome were obtained with the information of referring physicians and telephone calls, or direct clinical examination. Clinical outcome was evaluated at least 1 year after the operation. The median follow-up time was 46 months (range 12–91 months).
2.4. Definitions and statistical analysis
The patients were stratified as neurogenic or vascular TOS according to symptoms, clinical, laboratory and intraoperative findings. TOS type was determined with respect to predominant symptoms and clinical findings in patients showing either neurogenic or vascular symptoms. The surgical outcome was classified into two groups for statistical analysis. Patients were graded as favorable when symptoms improved and the patient returned to work or pre-illness activities. It was graded as poor when symptoms remained unchanged or worsened, or when the patient was unable to work or take part in usual activities as described previously [3,8].
Age, gender, duration of symptoms, history of trauma, Adson test, ENMG findings, the involvement type of the brachial plexus, color Doppler findings, the type of TOS, side of the operation, the presence of fibromuscular bands and cervical rib were included in the assessment of statistical comparisons with the surgical outcomes. The type of TOS was stratified as neurogenic and vascular according to the predominant findings either at operation or laboratory tests. Similarly, the involvement type of the brachial plexus was also dichotomized as upper and lower plexus involvement. Age, duration of symptoms and UNCV were classified as a high or low group relative to the median value.
Categorical variables were analyzed with the 
2 and Fisher's exact tests as appropriate in contingency tables, whereas Student's t-test and Mann–Whitney U-test were performed as appropriate for comparison of continuous variables. The logistic regression analysis along with a stepwise procedure was applied for univariate and multivariate analyses to outline the impact of the factors on surgical outcome. The variables with P-values assigned at 
0.1 were entered into multivariate analysis. Data were expressed as mean±standard deviation (SD). A P-value less than 0.05 was considered statistically significant. All statistical analyses were performed with the Statistical Package for Social Sciences (SPSS, version 11.0, Chicago, IL, USA).
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
The diagnosis was neurogenic TOS in 113 (89%) patients, and vascular TOS in 14 (11%) patients. Upper involvement of the brachial plexus including C5, C6, and C7 nerve roots was found in 21 (19%) patients, whereas lower involvement of the brachial plexus including C8 and T1 roots was found in 92 (81%) patients. The mean pre-operative UNCV was 55.8±8.1 m/s (range 34–75 m/s).
Among the patients with vascular TOS, two (2%) had venous TOS and 12 (9%) had arteriel TOS. Out of patients with arteriel TOS, four had thrombosis in the subclavian artery. Thrombectomy was performed in two patients. Thrombolytic therapy with urokinase in addition to angioplasty and a by-pass surgery between subclavian and axillary artery with a saphenous graft was performed in the remainder.
The rates of favorable and poor surgical outcome were 82.7 and 17.3%, respectively. Morbidity developed in 24 (19%) patients. The most common post-operative complication was an apical pneumothorax in 10 (8%) patients, followed by brachial causalgia in 5 (4%) patients, hemothorax in three (2%) patients, wound infection in two (2%) patients, Horner's syndrome in two (2%) patients, axillary hematoma in one (1%) patient, and chylothorax in one (1%) patient. Mortality was not observed.
3.2. Statistical results
The mean age of patients and the mean duration of symptoms did not show any significant difference between the subgroups of surgical outcome. However, the mean UNCV of patients with favorable surgical outcome was significantly greater that that of patients with poor surgical outcome (P=0.001) (Table 1).
|
24 months) and patients with UNCV values below 55 m/s had a significantly worse surgical outcome. Likewise, anomalous cervical rib showed a significant association with surgical outcome (P=0.003). The rate of favorable surgical outcome (96%) in patients without cervical rib was significantly greater than that of patients with cervical rib (75%). The duration of symptoms (P=0.021), UNCV (P=0.032), and anomalous cervical rib (P=0.010) were the significant predictors in univariate analysis (Table 2). However, multivariate analysis revealed only the duration of symptoms (P=0.017) and the anomalous cervical rib (P=0.026) as independent clinical predictors on surgical outcome (Table 3). Patients with shorter duration of symptoms less than 24 months and those with cervical rib showed a significant worse surgical outcome compared to their counterparts.
|
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
Although it appears that the transaxillary approach is the most convenient route in the surgical management of patients with TOS, the impact of clinical features on surgical outcome remains unclear. Little data exist regarding this issue, and it is very likely that selection criteria are of clinical significance to achieve optimal surgical results. Among the clinical features, psychological factors and educational status have been reported to correlate with surgical outcome [5]. Likewise, Green et al. [7] pointed out the significance of history of trauma and gender on surgical outcome. Our data with high proportion of cases (86.6%) with female gender were comparable with other series; however, we could not reach a statistical difference for surgical outcome with respect to either history of trauma or gender. Upon multivariate analysis, we found that only the duration of symptoms and the presence of cervical rib were the independent predictors of surgical outcome in patients with TOS.
On the other hand, some authors have suggested that pre-operative radiographic and neurophysiologic examinations were not predictive for surgical outcome [6], however, some others have reported that either abnormal ENMG or color Doppler findings had correlated with better surgical outcome [3]. Similarly, the clinical significance of UNCV has been emphasized in TOS. It has been suggested that patients with UNCV values above 60 m/s usually improved from conservative physical therapy and those with UNCV values below this value were very likely to benefit from surgery [10]. Although the results were not compared with respect to surgical outcome, a recent study has shown that pre-operative UNCV values had a significant increase in the post-operative period [11]. Our data showed that the presence of color Doppler findings did not have an effect on surgical outcome, however, the mean of UNCV values showed a significant difference with regard to surgical outcome (P=0.001). Similarly, the groups of UNCV dichotomized relative to the median value, which was 55 m/s, showed a significant difference in univariate analysis (P=0.032). Nevertheless, this significance disappeared in multivariate analysis (P=0.307). Moreover, we also found no significant difference between the groups with respect to surgical outcome (P=0.761) (data not shown) when we analyzed our data dichotomized with regard to the UNCV level of 60 m/s. Thus, our data did not show a significant impact of pre-operative UNCV values on surgical outcome.
We found that patients with shorter duration of symptoms (24 months) had a significantly worse surgical outcome compared with patients with longer duration of symptoms (>24 months). Patients with shorter duration of symptoms had a 5-fold increased risk for a worse surgical outcome compared with patients with longer duration of symptoms. Most of the patients with TOS are admitted to hospitals for surgery when their symptoms are acute or severe [11]. Although we did not stratify the degree of symptoms such as arm pain and paresthesia which were present in 32 and 28% of patients in our series, respectively, we speculate that patients with longer duration of symptoms had mild symptoms which they could tolerate for a longer time. In addition, although it did not reach a statistical significance, our data revealed that the mean UNCV value of patients with symptom duration more than 24 months was 54.2±5.1, which was greater than that (51.3±8.6) of patients with symptom duration less than 24 months.
Among the radiological imaging methods used in the diagnosis of TOS, cervical X-rays and CT are only able to show cervical ribs or costoclavicular anomalies, however, fibromuscular bands, which have been proposed to account for more than 90% of all cases in the etiology, cannot be visualized with these imaging techniques. Thus, diagnosis primarily depends on the laboratory findings rather than the imaging methods in patients without bony anomalies. The fact that the diagnosis of TOS is difficult, the etiology is not adequately recognized and the treatment modalities of TOS may vary from one center to another makes the surgical indication primarily dependent on the surgeon. However, the objective finding of a cervical rib can easily make the diagnosis of TOS if the differential diagnosis has been made. Cervical ribs may be present in about 0.5% of the general population, and 10% of cervical ribs give rise to symptoms. Cervical ribs account for a rate of 7.5–9% in surgical cases [10]. One of the most significant findings of our data is possibly the high rate (63.8%) of patients with cervical rib compared to other series [4,8,11]. This finding may imply less frequent referring of patients without cervical rib for surgery. According to our policy, we consider ENMG and color Doppler imaging neither absolutely accurate diagnostic methods nor absolute criteria for diagnosis and management of TOS. Thus, we sought a more objective finding such as the existence of a cervical rib.
The favorable rate of surgical outcome was 82.7% in our series, and it was comparable with other reports [8,9,11,12], however, we found that the presence of a cervical rib had a significant unfavorable effect on surgical outcome. Similarly, it has been reported that patients with cervical ribs undergoing surgery had a relatively worse surgical outcome, and this finding has been attributed to resecting the first rib, while leaving the cervical ribs in place [8]. The presence of cervical rib may imply the coexistence of fibromuscular bands [1], and hence the necessity of a meticulous operation to avoid any band left behind. Although our data did not reach a significant difference with respect to the correlation of cervical rib and fibromuscular bands (P=0.088) (data not shown), patients with cervical ribs had more fibromuscular bands compared to patients without cervical ribs in our series. Thus, a more meticulous dissection of the brachial plexus and the structures in the thoracic outlet for fibromuscular bands is reasonable in patients presenting with cervical ribs. Although not statistically analyzed and the number of cases was limited by 40, a previous study has reported that no significant difference existed between the surgical outcomes regarding the presence of cervical rib [13]. On the other hand, it has also been suggested that the presence of cervical and anomalous ribs in patients with neurogenic TOS did not improve the success rate from surgery compared with patients having no abnormal ribs [14]. In addition, recurrence is very likely in cases with incomplete removal of either the first or cervical rib as stated by Urschel and Razzuk [10]. Moreover, the presence of a long posterior first rib stump has been shown as the strongest clinical predictor for a poor surgical outcome [4]. Although we excluded recurrent cases and our data did not comprise patients with remnants of the first and cervical ribs, multivariate analysis showed the independent impact of cervical rib on surgical outcome.
In conclusion, among the clinical features those have been proposed to have an influence on surgical outcome in TOS, duration of symptoms and cervical rib appear to be the most potential predictors. Meticulous dissection with excision of fibromuscular bands is very reasonable to achieve favorable surgical outcome in cases with shorter duration of symptoms and cervical ribs. Further studies are necessary to elucidate the above findings in larger series.
|
References |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
