The most common benign tumour of the parotid is pleomorphic adenoma (PA). Some studies have shown that limited surgery to manage these types of tumours is viable option. Extra-capsular dissection (ECD) is one of these minimally invasive techniques. This study aimed to evaluate the morbidity associated with ECD in managing benign parotid tumours and to demonstrate that the technique can be used outside specialist centres. Twenty-six consecutive patients, diagnosed with benign parotid gland tumours were treated by ECD. Their benign status was determined by history, clinical criteria and fine needle aspiration cytology. Inclusion criteria were a discrete mobile lump, more than 3 cm in diameter and clinical/histological evidence of benign biological behaviour. A single surgeon (the author) operated on all cases. Follow up ranged from 2 to 72 months. In 23 patients, the tumour was located entirely within the superficial lobe of the parotid. Postoperative pathology consisted mainly of 17 PA. Fifteen percent had transient damage to the facial nerve. None had recurrence. ECD seems to be an alternative to superficial parotidectomy for most superficial parotid lumps. It appears to be a safe technique in trained hands.
Parotid gland neoplasm comprises 3% of all head and neck tumours . Eighty percent of these are benign, mostly pleomorphic adenoma (PA) . Approximately 90% of parotid gland neoplasms are located within the superficial lobe, lateral to the facial nerve . It is this unique relationship between the parotid gland lumps and facial nerve that has shaped the different surgical approaches.
Before the 1940s, the surgical management of benign parotid lumps was unsatisfactory. Owing to a fear of damaging the facial nerve, the routine operation was enucleation of the tumour contents leaving the capsule in situ . This was associated with a high recurrence rate (35%). B ailey identified and dissected the main trunk of facial nerve through the gland in 1941 . By the 1950s, formal superficial parotidectomy (dissection of the facial nerve) became established as the appropriate treatment for benign parotid lumps. The recurrence rate declined to below 2% .
In the last two decades a general trend adopted by all surgical disciplines has been an attempt to reduce the magnitude of surgery in order to reduce morbidity while keeping outcome results the same. In an attempt to extend this approach to parotid surgery, there is a trend towards limited operations based on extra-capsular dissection (ECD) techniques. This procedure was initially described by G leave and involves careful dissection of the benign parotid tumour in a plane 3–4 mm peripheral to the tumour capsule without identification of the facial nerve. ECD differs from enucleation, because the tumour is removed with an intact capsule as opposed to shelling out the tumour contents and leaving the capsule in situ as is the case with enucleation . ECD has been restricted to three centres (Christie Hospital, Whythenshaw Hospital and Guy’s Hospital, UK). It is unclear whether the technique can be exported safely into a non-specialist environment.
The aim of this study is to evaluate the morbidity associated with ECD in managing benign parotid tumours and to demonstrate that the technique can be exported outside specialist centres.
Material and method
In the period 2003–2008, 26 consecutive patients presented with benign parotid tumours ( Fig. 1 ). All patients ( n = 26) were treated by ECD by a single surgeon (the author). Diagnosis was based primarily on clinical features (discrete lump, mobile in two planes, suggestive clinical benign behaviour) and histological evidence of benign characteristics by fine needle aspiration. The exclusion criteria were small lumps less than 3 cm in diameter, fixed mass, evidence of facial nerve palsy or histological evidence of malignancy. CT scans or ultrasonographs of the parotid region were obtained in all cases.
The study group consisted of 18 males and eight females aged 43–68 years with a mean age of 51.2 years. The following data were collected prospectively: tumour (size, site and relationship to nerve); operative details (operative time, incision and operative complications); postoperative complications (permanent and transient nerve injury, injury to greater auricular nerve, sialocele, haematoma and Frey’s syndrome).
It is necessary when employing ECD to take every precaution to confirm the tumour is benign. Once the preauricular skin is raised to expose the parotid, the lump is palpated again to confirm its mobility. If this is in doubt, conventional superficial parotidectomy is undertaken. If conditions change at any time during the operative procedure, the ECD can be stopped and a traditional superficial parotidectomy performed. In raising the skin flap through a modified Blair incision a careful dissection of the cervical portion of the incision is necessary to avoid injury to the greater auricular nerve. The mass is marked ( Fig. 2 ) and exposed via a cruciate incision made over its surface. It is important to identify the loose areola connective tissue around the tumour lump. Dissection proceeds in this plane in a centripetal fashion ( Fig. 3 ). Any branch of the facial nerve encountered should be retracted gently away from the tumour. When using ECD good homeostasis using bipolar diathermy and adequate illumination are prerequisite to successful dissection of the tumour. Nerve monitoring is an option and loupe magnification (2.5×) can be helpful. After lump removal ( Fig. 4 ) the parotid fascia is approximated using 3 0 vicryl. No drains are used unless there is a large defect in the parotid when a pressure bandage is mandatory (2 days) to avoid haematoma and sialocele. The patients in this series were followed up for 2–72 months (mean 27 months) ( Figs 5 and 6 ).
Twenty-six consecutive parotid tumours were treated by ECD. Preoperatively the patients were assessed clinically and using fine needle aspiration cytology (FNAC) to ensure that only benign tumours were selected for surgery. Histological examination of the specimen postoperatively confirmed no malignant disease was encountered in this series of cases.
All tumours were located in the superficial lobe of the parotid gland except in three patients (12%) in whom the mass extended between branches of the facial nerve into the deep lobe of the parotid. The mean tumour size was 4.4 cm.
The incision used depended on the site of the tumour, but in general the length of the incision was much less than that of the traditional superficial parotidectomy. Superficial tumours immediately in front of the ear could be approached through a small preauricular incision, whereas those more peripherally placed within the parotid gland required a wider exposure than the traditional superficial parotidectomy incision. Lesions in the tail of the parotid could be approached mainly from a cervical incision with limited extension into the preauricular area. The mean operative time for surgery was 56 min (range: 40–90 min). The facial nerve was encountered in 10 patients (39%). In three patients (12%) it had to be physically removed from the capsule of the tumour in order to remove the neoplasm. One intraoperative complication was tumour rupture. This was a controlled rupture without contamination of the wound in general. It occurred late in the operative procedure when dissecting a PA mass that was extending deeply through the retromandibular area to lie over the main trunk of the facial nerve. The macroscopic spillage was controlled by suturing the tear followed by thorough lavage of the area with normal saline. The greater auricular nerve was spared in 20 patients (77%) and if tumours in the tail of the parotid were excluded ( n = 6), then injury to this nerve was avoided in 90% of cases ( n = 18). In the convalescent period, patients were kept in hospital for 1–2 days. The postoperative complications encountered ( Table 1 ) were transient nerve injury in four patients (15%). Nerve paresis recovered within 6–13 months with the help of physiotherapy. Excluding patients who had deep lobe extension ( n = 3), nerve paresis was documented in only one patient (4%). No patients developed permanent nerve injury, sialocele or Frey’s syndrome. Evidence of Frey’s syndrome was sought by questioning the patients about post-prandial sweating. One patient had haematoma that was managed conservatively.