Thoracoscopic management of posterior mediastinal neurogenic tumours.

Background: This study describes the surgical technique of thoracoscopic resection of posterior mediastinal neurogenic tumours and reporting the surgical outcomes.
Methods: This is a retrospective analysis of 21 patients operated over 7 years in a dedicated thoracic surgery centre. The demographic and post-operative parameters along with complications were recorded and analysed.
Results: Twelve patients had right-sided tumours, while 9 had left-sided lesions, and 9 were on the left side. The most common diagnosis was schwannoma (n = 15, 71.42%), followed by neurofibroma (n = 4, 19.04%). The average surgery duration was 104 min (85-135 min), and the mean blood loss was 120 ml (25 ml-250 ml). The average lesion size was 4.8 cm (2 cm-7 cm). Conversion to open procedure was required in one patient. Two patients (14.2%) developed complications. One patient developed Horner's Syndrome and the other developed post-operative lung atelectasis. The median follow-up was 36 months (6-90 months). No recurrence was observed during the follow-up period. Conclusions: Thoracoscopic approach to posterior mediastinal neurogenic tumours is feasible and allows for low morbidity, short hospital stay and superior cosmesis.

INTRODUCTION

The most common tumours arising in the posterior mediastinum are neurogenic in origin and account for 75% of the tumours in this area.[1] The majority of them are benign, and 50% are asymptomatic, with the diagnosis being incidental on imaging for other reasons.[2] Surgical excision is the treatment of choice. Excision via an open thoracotomy has been the gold standard procedure.[3] The first successful attempt of excision of these tumours by video-assisted thoracoscopic surgery (VATS) was reported by Landreneau et al. in 1992. VATS has gained wide acceptance in the thoracic surgical community owing to lesser surgical trauma, lower blood loss, reduced post-operative morbidity and shorter hospital stay.[4567] We herein describe our surgical technique and medium-term outcome in 21 patients with posterior mediastinal neurogenic tumours resected using a thoracoscopic approach.

METHODS

We did a retrospective analysis of prospectively maintained data of patients who underwent surgery for posterior mediastinal tumour at our tertiary care centre from March 2012 to December 2019. The Institutional ethics committee approved the study. A total of 21 patients were included for analysis.

After a detailed history and clinical examination, all patients underwent a contrast-enhanced computed tomographic (CECT) scan of the chest to assess the location, size and extent of tumour and infiltration, if any, into the surrounding structures [Figure 1]. Patients with neurological symptoms, apical pelvic muscle trainings (PMTs) and those with a suspected intraspinal extension on computed tomography (CT) scan were further evaluated by magnetic resonance imaging (MRI). Pre-operative fitness was evaluated by routine haematological and biochemical tests, spirometry and electrocardiogram. Those with a clinical history of cardiac disease or age >40 underwent a detailed cardiac evaluation. Their demographic data, clinical presentation, pre-operative, intra-operative and post-operative details were recorded. Patients with tumour size <7 cm without any radiological evidence of intraspinal extension or infiltration into adjacent structures were taken up for a thoracoscopic excision.

Figure 1

Contrast-enhanced computed tomographic showing pelvic muscle training on the right side

Chest tube duration, hospital stay, prolonged air leak (>5 days), wound infection and other complications during hospital stay were recorded. All patients were followed up in the outpatient department at weekly intervals for 4 weeks and then every 6th month for the 1st year after surgery. A CECT scan of the chest was done annually to rule out any recurrence for the first 2 years. After that, the patients were followed up annually for 5 years with a chest X-ray only.

Operative details

All the procedures were performed under general anaesthesia with lung isolation achieved by a double-lumen endotracheal tube. A single shot of intravenous antibiotic was administered an hour before induction. The procedure was performed in a lateral decubitus position with 15° of anterior tilt [Figure 2]. An anterior approach with the surgeon standing in the front of the patient was employed in all cases. Three to four ports were placed to achieve a good triangulation of the target lesion. For tumours in the upper third of the thoracic cavity, the camera port was placed in the 5th intercostal space between the mid and anterior axillary line. The thoracic cavity was explored using a 5 mm 30° thoracoscope. The thoracoscopic approach was considered feasible if the tumour could be visualised in its entirety and there was no involvement of the underlying chest wall or adjacent structures. Once the tumour was deemed fit for a thoracoscopic resection, two additional ports were created to achieve a good triangulation to the target lesion, as depicted in Figure 3a. For mid and lower third tumours, the arc of port placement was rotated counter-clockwise, as depicted in Figure 3b. The camera port remains in the 5th intercostal space as described above, but the two working ports were placed differently, as shown in Figure 2.

Figure 2

Position and port placement for thoracoscopic excision of posterior mediastinal tumours in (a) upper third (b) mid and lower third of the thoracic cavity

Figure 3

(a) Thoracoscopic view of an apical pelvic muscle training. (b) Initiation of pleural incision away from the tumour. (c). Tumour Mobilisation being completed. (d). Completely excised tumour ready for retrieval

Carbon dioxide insufflation was started at 6 mm of Hg. The mediastinal pleura were opened 1 cm away from the tumour, as shown in Figure 3b. The tumour was then lifted off its bed by a combination of blunt and sharp dissection. Vessels encountered during dissection were either sealed using an energy device like Harmonic™ Scalpel or Ligasure™ or clipped [Figure 3c]. After complete mobilisation [Figure 3d], the tumour was placed in the specimen bag and retrieved through the camera port, which was enlarged as needed. All the tenets of oncological resection like minimal handling of tumour tissue, adequate margin and avoidance of breach of tumour capsule were strictly adhered to. Bleeding obscuring the vision, non-progression beyond 60 min and difficulty in adherence to oncological principles were considered as indications for conversion.

At the end of the procedure, a single chest tube was placed and connected to an underwater seal. An intercostal nerve block using 0.025% bupivacaine was performed under thoracoscopic vision from the 2nd to 10th intercostal space. The patients were extubated on the table, shifted to the recovery room for overnight monitoring and transferred to the ward the following day.

Post-operative care

All patients were mobilised on the first post-operative day. Adequate pain relief was ensured using IV non-steroidal anti-inflammatory drugs. The chest tubes were removed on postoperative day (POD) 1 after ensuring the absence of an air leak and drainage of <2 ml per kg of body weight over 24 h. All patients were discharged on the day of chest tube removal.

RESULTS

Twenty-one patients with posterior mediastinal lesions underwent thoracoscopic excision of their tumour during the study period and were included in the study. Demographic variables are shown in Table 1. The median age at the time of surgery was 39.89 years (range 10–66 years). The most common presentation was that of an incidental finding on imaging done for other reasons (n = 11, 52.38%) followed by chest pain (n = 6, 28.57%) and unexplained dyspnoea (4, 19.04%). There were 14 PMTs in the upper, 3 in the middle and 4 in the lower thoracic cavity. Twelve patients had right-sided tumours, while 9 had a tumour on the left side. The final histopathological diagnoses of the tumours are listed in Table 2. All patients had a complete resection (R0) with clear margins on histopathological review. The most common diagnosis was schwannoma arising from the nerve sheath (n = 15, 71.42%), followed by neurofibroma (n = 4, 19.04). Lung was densely adherent in three patients who required wedge resection of the adherent lung. Conversion to thoracotomy was required in one case due to bleeding that obscured the surgical field. The procedure's average duration was 104 min (85–135 min), and the mean blood loss was 120 ml (25 ml–250 ml). The tumour's size ranged from 2 cm to 7 cm with a mean of 4.8 cm. The chest tube was removed on POD 1 in all except one patient whose chest tube was removed on POD 3. The average hospital stay was 3.5 days. Two patients (14.2%) developed complications. One patient developed Horner's Syndrome after resection of an apical PMT and the other developed lung atelectasis. The median follow-up was 36 months (6–90 months). No recurrence was observed during the follow-up period.

Table 1

Demography and peri-operative variables

Variable	Frequency
Male	14
Female	07
Mean age	39.89
Symptoms
Incidental	11
Chest pain	6
Dyspnoea	4
Side
Right side	12
Left side	9
Site
Upper one third	14
Middle one third	3
Lower one third	4
Average duration of surgery	104
Average blood loss	120
Mean hospital stay time	3.5
Mean ICD duration	1.14 days
Average tumour size	4.8
Horner’s syndrome	1
Lung atelectasis	1
Conversion	1
Median follow-up	36

ICD: Intercostal Drain

Table 2

Histopathological diagnosis

Diagnosis	Frequency
Schwannoma	15
Neurofibromas	4
Paraganglioma	1
Neurilemoma	1

DISCUSSION

Neurogenic tumours account for 75% of all posterior mediastinal tumours. They originate from different components of the nerves like schwannoma and neurofibroma from Schwann cells in the nerve sheath and ganglioneuromas and paragangliomas from nerve cells.[8] Most of these tumours are benign and 10% can have an intraspinal extension.[35] These tumours usually present as solitary lesions. Nerve sheath tumours are most common in adults, while sympathetic ganglia tumours are more common in children.[9] The median age of patients in our study was under 40 years. Schwannoma was the most common tumour in our series, as is reported in the literature.[8] No sex preponderance has been reported in the literature, though in our series, males were affected twice as often.[9]

The majority of these tumours are asymptomatic and are diagnosed as an incidental finding on imaging done for other reasons. Patients may, however, present with chest pain or respiratory symptoms like dyspnoea or cough due to compression of adjacent lung parenchyma or airways or have neurological symptoms in the form of neuropathic pain or Horner's syndrome.[10] The most typical presentation in our series was an incidental diagnosis, followed by chest pain.

CECT thorax is the most used imaging modality that provides information about the tumour's location, size and relationship with adjacent structures.[11] MRI is required in cases where CT findings suggest intraspinal extension, surrounding infiltration or brachial plexus involvement in apical tumours.[12] Careful radiological evaluation before surgery is mandatory to rule out intraspinal extension (dumb-bell tumours). Careless traction on the lesion with an intraspinal extension may lead to disastrous consequences due to cord traction.

Pre-operative pathological diagnosis is preferred for solid tumours wherever feasible as there are chances of malignancy and pre-operative diagnosis of malignant lesion precludes thoracoscopic resection. The diagnostic yield of CT scan-guided needle biopsy varies from 72% to 100% in various series.[13] We achieved a pre-operative histopathological diagnosis in 17 of the 21 cases (81%). The percutaneous biopsy was inconclusive in the remaining 4 patients, and the final diagnosis was established after surgical excision.

All posterior mediastinal tumours should be excised with clear margins. Even for asymptomatic cases, observation is not recommended because of the risk of malignancy and to avoid problems due to continued growth. The standard posterolateral thoracotomy was, until recently, the gold standard procedure for these lesions.[3] However, video-assisted thoracoscopic approaches have gained wide acceptance due to lower morbidity, reduced hospital stays and better cosmesis.[4567]

Proper selection of cases for thoracoscopic excision is crucial for ensuring good surgical outcomes. Factors such as the size of the tumour, its location and presence or absence of features suggesting malignancy need to be considered. If there is any concern of malignancy on pre-operative evaluation, the thoracoscopic approach should not be undertaken. There is no cut-off limit about the tumour's size that can be taken up for thoracoscopic excision. Tumour sizes of up to 6–8 cm have been successfully operated by VATS.[1415] We believe that long rigid instruments used during thoracoscopic surgery make dissection around the tumour tedious, and an arbitrary limit of 7 cm was used in this study. However, we also believe that the presence of maintained fat planes with the adjacent chest wall, vertebral body and other structures is more important than the tumour's size in selecting cases for the thoracoscopic approach. The average tumour size was 4.8 cm (2 cm to 7 cm). Chen et al.[16] reported an experience of 121 patients of posterior mediastinal tumours, where majority (74 patients) underwent video-assisted thoracic resection (VATS). They concluded that VATS is a safe and effective modality for the resection of most posterior neurogenic tumours. However, in the case of very large tumours, Thoracotomy can be the appropriate surgical approach. A supraclavicular approach can be used for tumours extending in the cervical region, which can be combined with VATS or thoracotomy for larger tumours.

Few of these posterior mediastinal tumours can have intraspinal and neuroforaminal extension, which are known as dumbbell tumours. Chen et al.[17] reported the surgical experience of 20 cases of dumbbell tumours, where they recommended single-stage posterior laminectomy plus VATS/thoracotomy approach for surgical resection.

Another important consideration is the location of the tumour. Tumour at the apex of the chest cavity may be difficult for beginners and should be undertaken only by an experienced thoracoscopic surgeon.[18] Many of such cases can be managed by the VATS approach. Similar observations were reported by Liu et al.,[19] where most (32/56) of thoracic inlet tumours were resected by VATS. However, in cases where mass obscures the 1st and/or 2nd ribs, thoracotomy is preferred. Supraclavicular approach can also be utilised for tumours arising from the brachial plexus. Owing to our experience with robotic-assisted thoracoscopic procedures, we feel that tumours at either extreme may be better approached with the robotic system also where available and feasible. Such similar experience was reported by Shidei et al.,[20] where a non-functional, rib invasive paraganglioma in the left apical chest was successfully removed robotically. The authors also mentioned that the endowrist technology and greater manoeuvrability of the robotic system enabled the tumour resection without organ injury in the thoracic cavity.

We used a semi-lateral decubitus position with 15° of anterior tilt for all patients. It allows for an optimal operative view and naturally allows the lung to fall away from the operative field. The use of CO2 at a pressure of 6–10 mm of Hg is beneficial as it allows the lung to be adequately compressed, pushes the diaphragm downwards and increases the space available, particularly for tumours in the lower third. There has been a report of the use of prone position for posterior mediastinal tumours from Japan.[21] The authors claimed better exposure without the need for retraction due to gravity-assisted anterior displacement of lungs and pericardium. Since retraction is not needed, the surgery could be performed by a solo surgeon without the need for an expert assistant. However, we believe that the prone position has little to offer when compared to the semi-lateral position with anterior tilt. This position allows for adequate exposure to the target area. Moreover, if a conversion is required, it can be better performed in semi-prone or semi-lateral position than the prone position.

Two of our patients developed complications, one case each of Horner's syndrome and lung atelectasis. Intra-operative injury to the stellate ganglion, either due to traction or electrocautery, was the likely cause, and it recovered entirely over the next 6 weeks. It is advisable to be careful with dissection and electrocautery near the ganglion. One patient required conversion due to bleeding, obscuring the vision. It is essential to be more careful with haemostasis during VATS dissection, as any pooling of blood around the tumour makes dissection unsafe.

The average hospital stay in our series was 3.5 days. It is essential to point out that 20 out of the 21 patients were discharged on POD 2. An extra day or two of admission was added because most of the insurance companies in India do not allow the cashless facility for investigations carried out on an outpatient basis, and hence all these patients had to be admitted a day or two before surgery, adding to the hospital stay. Based on our experience, we feel that young patients without comorbidities can be directly admitted on the day of surgery and safely discharged the next day after chest tube removal in uneventful cases. Therefore, there is a potential to reduce the hospital stay further. The study's retrospective nature with small numbers and lack of comparative analysis are the limitations of this study.

CONCLUSIONS

Our study demonstrates the feasibility of the thoracoscopic approach for posterior mediastinal neurogenic tumours.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.