Winged scapula following axillary thoracotomy with long thoracic nerve preservation

Abstract

Winged scapula is a rare condition caused by injuries to the long thoracic nerve (LTN) and accessory nerves. A 69-year-old man underwent surgery for right lung cancer. Video-assisted thoracic surgery was converted to axillary thoracotomy at the fourth intercostal space. The latissimus dorsi was protected, and the serratus anterior was divided on the side anterior to the LTN. Two months after discharge, he presented with difficulty in elevating his right arm and protrusion of the scapula from his back. Active forward flexion of the right shoulder was limited to 110° and abduction to 130°. He was diagnosed with winged scapula. After 6 months of occupational therapy, the symptoms improved. The LTN may have been overstretched or damaged by the electric scalpel. We recommend an increased awareness of the LTN, and to divide the serratus anterior at a site as far as possible from the LTN to avoid postoperative winged scapula.

Background

Winged scapula is a rare condition induced by injuries to the long thoracic nerve (LTN) and accessory nerves. Palsies of the serratus anterior (SA) or the trapezius muscles result in loss of stabilisation of the scapula and reduced strength and range of motion of the shoulder.¹ Damage to the LTN has numerous causes, including non-surgical or postoperative complications. For example, retrospective studies revealed that 27.2-28.1% of patients who underwent surgery for breast cancer showed winged scapula.^{2 3} Winged scapula is likely to develop after thoracic surgeries, like posterolateral thoracotomy,^{4 5} transaxillary thoracotomy⁶ and first rib resection.⁷ However, to the best of our knowledge, postoperative winged scapula after thoracic surgery has not been reported over the past 16 years, except for after posterolateral thoracotomy in paediatric patients.^{1 8} Considering that the size of incisions made during thoracic surgery has decreased over the past 20 years, due to the increased use of video-assisted thoracic surgery (VATS), fewer injuries of the LTN have occurred. We report a case of winged scapula that occurred following axillary thoracotomy with division of the SA at a site far from the LTN.

Case presentation

A 69-year-old man was diagnosed with adenocarcinoma in the right upper lobe by fibre bronchoscopy. His medical history included syphilis infection at age 20. He was scheduled to undergo right upper lobectomy and lymph node dissection. The patient was placed in the lateral decubitus position, the ipsilateral arm was positioned with shoulder in 90° of flexion and the elbow was extended forward. Initially, VATS was undertaken via 15 mm incisions at the middle axillary line at the seventh intercostal space (ICS) as a camera port, with 40 mm incisions at the middle axillary line at the fourth ICS, 20 mm incisions at the anterior axillary line at the fifth ICS and 20 mm incisions at the posterior axillary line at the sixth ICS as working ports, respectively. However, the azygos arch was injured during the mediastinal lymph node dissection. VATS surgery was converted to thoracotomy to stop bleeding. The incision at the fourth ICS was extended to 15 cm. The latissimus dorsi was protected and retracted posteriorly to expose the SA and rib cage. The SA was split along the direction of the fibre and divided along the vertical axis parallel to the LTN, at the anterior side of the leading edge of the latissimus dorsi (figure 1, red dotted line). After exposing the rib cage under the SA, the fourth intercostal muscle was divided to the side of the sympathetic trunk (figure 1, blue dotted line). Following this, a retractor was placed. Ligation and detachment of the azygos arch was performed to stop the bleeding. The operation was completed successfully.

Figure 1.

The splitting or dividing line of the serratus anterior is shown as red dotted lines. The dividing line of the fourth intercostal muscle after exposing the rib cage is shown as blue dotted lines. The diagram is an original creation and was drawn with COPIC MULTILINER 0.1 mm (Too Corporation, Tokyo, Japan) and Adobe Photoshop V.7.0 (Adobe systems Incorporated, San Jose, California, USA) by Tomomi Isono; no material was obtained from others or downloaded from the internet.

Postoperative course was well, and he was discharged from hospital on the 18th day after surgery. He refrained from using his right arm after surgery because he was afraid of a worsened postoperative clinical course resulting from using his right arm. Two months after discharge, he presented with difficulty elevating his right arm and protruding of the shoulder blade from his back (day 0). The medial border and inferior angle of the scapula displayed unusual prominence (figure 2A). Active abduction of the right shoulder was limited to 130° (figure 2B) and active forward flexion to 110°. He was diagnosed with postoperative winged scapula caused by paralysis of the SA. He started occupational therapy on day 0, immediately after diagnosis. After 2 weeks, winged scapula showed little improvement, and his active shoulder range of motion started improving. Therefore, it was predicted that LTN was not completely damaged, because patients with complete damage to the LTN do not experience improvement of winged scapula with occupational therapy. He continued occupational therapy once or twice a week. After 6 months of occupational therapy, mild prominence of the scapula persisted but was improved (figure 2C). The active abduction and active forward flexion improved over time to 180° (figure 2D) and 170°, respectively.

Figure 2.

Physical examinations. (A) The patient’s back on day 0, the day of diagnosis of winged scapula, 2 months after surgery for lung cancer. (B) Active abduction on day 0. (C) The patient’s back after 6 months of occupational therapy. (D) Active abduction after 6 months of occupational therapy.

Differential diagnosis

We established three differential diagnoses. First, winged scapula caused by paralysis of the trapezius due to injury of the accessory nerve was considered. This was excluded because shrugging of the shoulders was normal. Second, palsy of the deltoid muscle was considered. This was excluded because active abduction was 110° on day 0. In addition, the ipsilateral arm was positioned with the shoulder in 90° flexion and the elbow was extended forward in surgery. Injury of the axillary nerve is not so vulnerable to damage in this position. Third, postoperative disuse atrophy was considered. This may have contributed to the limitation of active right shoulder movement for this patient, because he did not use his right arm after surgery for 2 months due to his fear of injury.

Treatment

Occupational therapy was performed once or twice a week for 6 months. The occupational therapist created a special SA strength programme according to his symptoms. First, the patient started bench press exercise. Initially, he used a bar (200 g, 58 cm) and weight was placed to this bar according to improvement of his symptoms. Second, he tried front bridge exercise. He was unable to perform front bridge exercise because he could not maintain the correct position. Third, he tried pushing wall exercise. He was good at this exercise and placed a load strongly on his right side at the final stage of his programme.

Outcome and follow-up

Currently, he has no difficulty performing activities in his daily life.

Discussion

The LTN is formed by anterior branches of the cervical 5,6 and 7 nerve roots and innervates the SA. The main job of the SA is to fix the scapula to the chest wall. A winged scapula is a condition in which the scapula protrudes from the back because of the dysfunctional SA due to damage of the LTN. The first choice for treatment is non-surgical management including physiotherapy and the second choice is surgery such as pectoralis major transfer.^{9 10} Many authors recommend physiotherapy for 1–24 months^{9 10} and after physiotherapy most patients show improvement of winged scapula.³ Speed of regeneration of peripheral nerves is about 1–2 mm/day and the length of the LTN is about 20±2 cm.¹¹ Therefore, it is necessary to undergo physiotherapy for 1–24 months while the nerve regenerates and for a time afterward, to strengthen the affected muscles.

We garnered two clinical lessons from this case.

First, winged scapula can be caused by the fourth intercostal thoracotomy in which the SA was divided carefully to avoid damage to the LTN. In our patient, the LTN must not have been completely damaged because his symptoms improved after 2 weeks of occupational therapy. Two causes for nerve injuries during surgery can be assumed. The first possible cause is overstretching.¹² In our patient, to stop bleeding, we opened the fourth ICS largely and this may have overstretched the LTN. Second, use of an electric scalpel might damage the LTN. As a cause of iatrogenic nerve palsy, use of an electric knife may play a role not only for direct damage, but also for thermal damage by hardening or during coagulation.^{1 13} The thoracotomy procedure was performed urgently to stop bleeding; therefore, the electrical scalpel may have damaged the LTN by strong electrical stimulation or heat. Due to the urgent nature of the thoracotomy, surgeons may not have paid enough attention to the LTN, which can be observable in a thoracotomy. In this patient, overstretching may be a more plausible cause than electrical knife use. Sunderland expanded Seddon’s three classifications of peripheral injuries.¹⁴ According to their classifications, the grade of nerve injury of this patient is axonotmesis, defined as complete interruption of axons with preserved surrounding structures of nerve, and Wallerian degeneration. Axonal regeneration without surgery is possible in this stage. Axonotmesis may be caused, not by an electric knife, but by the tension produced by overstreching damaging the axion and surrounding structures. If an electric knife damages a nerve, this stimulation will interrupt the nerve from the outer layer first. Axon damage occurs after the surrounding structures are damaged. Therefore, nerve injury caused by electrical knife must involve almost the entire layer of the nerve trunk including the axon and surrounding structures including myelin sheaths and other outer structures. This stage of nerve injury is neurotmesis; total disruption of the entire nerve fibre that requires surgical intervention. Our patient’s symptoms improved without surgery, but mild permanent complications remained. Therefore, axonotmesis by overstretching may be the most likely cause of nerve injury.

Usually, we open the fourth ICS, dividing the SA along to the LTN, to preserve the LTN via extended skin incision in conversion from VATS to thoracotomy. This was our first experience of postoperative winged scapula. Even in posterolateral thoracotomy, in which SA and the LTN are absolutely cut, we had no experience of paralysis of the SA. Two retrospective studies found that about 5% of patients showed winged scapula as a complication of posterolateral thoracotomy.^{4 5} The LTN is cut at a relatively peripheral site, the fourth to sixth ICS, which may account for the relatively low prevalence of winged scapula. To the best of our knowledge, there are no case reports of postoperative winged scapula except for paediatric thoracic surgery over the last 16 years. This may be because thoracotomy incisions have become smaller, and fewer LTN injuries have occurred. We recommend an increased awareness of the LTN, and to divide the SA at a site as far as possible from the LTN to avoid postoperative winged scapula.

Second, earlier initiation of treatment may be important to recover functionality and aesthetics. A retrospective study revealed that 96.6% of patients who underwent physiotherapy for postoperative winged scapula recovered, while 66.7% of patients without physiotherapy recovered.³ Moreover, delayed diagnosis may lead to additional complications including traction branchial plexopathy among others.⁹ In our patient, physical therapy was initiated 2 months after surgery. This delay may have contributed to progressive atrophy of the SA and shoulder dysfunction.

In paediatric surgery, much like adult surgery, the same complications may be observed in a similar minimally invasive approach to preserve the LTN that includes a right submammary incision and a right vertical infra-axillary thoracotomy to treat congenital heart disease,¹⁵ or in an axillary skin crease incision to treat oesophageal atresia, patent ductus arteriosus and congenital cystic adenomatoid malformation among others.¹⁶

Winged scapula, especially after thoracic surgery for lung cancer, is so rare that diagnosis can be delayed or underdiagnosed; consequently, more patients may have postoperative winged scapula. Therefore, careful follow-up and physical examination after surgery that includes the ipsilateral back and is necessary to identify postoperative winged scapula.

Patient’s perspective.

‘After surgery, I refrained from using my right arm because I was afraid that use of my right arm could worsen my postoperative clinical course. Certainly, it was hard to lift my right arm immediately after discharge. I thought that this was caused due to the lateral decubitus position assumed for a long-time during surgery, as explained to me before surgery. On the other hand, I felt muscle weakness throughout my body during the hospital stay. In addition, I experienced difficulty in urinating. My main concerns were these two problems, not the shoulder symptoms. Therefore, I did not care about my right shoulder symptoms. Two months after discharge, another person found an unusual protrusion of the scapula from my back, and this was the first time that I wondered if my right arm and shoulder were abnormal. Currently, though permanent damage remains and I cannot elevate my right arm up completely, I have no problem in my daily life. I work part-time as a caddie at a golf link, and I enjoy it.’

Learning points.

• Winged scapula should be considered even after thoracic surgery in which surgeons attempted to preserve the long thoracic nerve.

• Starting occupational therapy as soon as possible is recommended for patients with postoperative winged scapula. Physical function may improve after physiotherapy if the long thoracic nerve is not cut completely. Dysfunction may remain after delayed diagnosis and treatment.

• Careful physical examination as well as careful attention to the patient’s complaint may be necessary to identify postoperative winged scapula.