Need for tracheostomy in major head and neck ablative surgery: a paradigm shift during the COVID-19 pandemic

Abstract

Introduction

Management of the airway in the perioperative period for patients requiring major head and neck ablative surgery has commonly included the performance of elective surgical tracheostomy. This has been standard practice in most maxillofacial units across the UK, including ours. However, the COVID-19 pandemic and emerging guidelines on aerosol-generating procedures required us to revisit the need for a perioperative tracheostomy.

Methods

We present our series of 29 consecutive cases, cared for during the first wave of the COVID-19 pandemic, that were managed either using surgical tracheostomy or overnight tracheal intubation.

Results

Out of 29 patients 3 received a surgical tracheostomy. The average duration of tracheostomy use was 8 days. Twenty patients were managed using a period of overnight tracheal intubation. Average duration of tracheal intubation was 1.2 days, with an average intensive care unit stay of 1.7 days. The average duration of hospital stay was 15.8 days for patients managed with overnight tracheal intubation and 30.1 days for patients who received a surgical tracheostomy. The return to theatre rate was 13.8% for reasons including flap failure and neck space infection. There were no airway issues reported in this series of patients.

Conclusions

Our findings suggest that overnight tracheal intubation can be a safe alternative to surgical tracheostomy in the majority of cases.

Introduction

Oral squamous cell carcinoma (SCC) is the 14th most common cancer in the UK. Over 8,300 cases are diagnosed each year, accounting for over 2% of all cancers.^1–4 For human papilloma virus-negative cancers, surgery remains the mainstay of treatment with curative intent.⁵ Tumours graded as stage T2 or worse require significant resective surgery and reconstruction with a free tissue transfer.⁶ The case for a tracheostomy has always been made in relation to the risk of swelling and the potential loss of airway patency in the immediate postoperative period. In addition, it is highly likely that reintubation in these patients will be difficult as a result of altered airway anatomy and the associated swelling.

Methods

This is a retrospective review of 29 consecutive cases of major head and neck resective surgery with flap reconstruction. All cases occurring during the COVID-19 pandemic, from February 2020 to February 2021, were reviewed. A subgroup of 29 patients with free flap tissue transfer were then identified (Table 1). Data were collated for T stage, N stage, site of primary tumour, type of reconstruction, type of neck dissection, patients with tracheostomy, duration of tracheostomy, length of tracheal intubation, duration of intensive care unit (ICU) stay, duration of inpatient stay, day of commencement of oral intake and the need for return to theatre.

Table 1 .

Tumour site and characteristics

Site of primary		T Stage		N Stage		Neck dissection		Reconstruction
Tongue	15	1	2	0	11	Unilateral	21	Radial forearm	13
Mandible	9	2	6	1	3	Bilateral	8	Anterolateral thigh	6
Floor of mouth	3	3	8	2a	3			Fibula	5
Retromolar trigone	3	4a	12	2b	5			Rectus abdominis	3
Buccal mucosa	2	4b	0	2c	0			Pectoralis major	2
Oropharynx	2	Benign	1	3	6
Maxillary antrum	1			Benign	1
Maxilla	1

Results

The highest proportion of SCCs were T4a stage (n=12), followed by T3 (n=8) and T2 (n=6). Two tumours were found to be pathologically T1 stage. One benign tumour (Pindborg tumour) was treated during this period.

The majority of cases were N stage 0 (n=11), followed by N3 (n=6) and N2b (n=5). The remainder of cases were N2a or N1. In total, 72% (n=21) of patients had a unilateral neck dissection performed and 28% (n=8) had a bilateral neck dissection.

The most common primary site of tumour was the tongue (n=15), followed by tumours affecting the mandible (n=9). As expected, the extension of tumours to an adjacent anatomical subsite was noted.

Reconstruction using the radial forearm was the most common (n=13), followed by the anterolateral thigh (n=6) and the fibula for composite tissue (n=5). The remaining cases received alternative free flaps.

Three patients received a surgical tracheostomy. The average duration of tracheostomy use was 8 days. A large proportion of patients were managed using a period of overnight tracheal intubation (n=20). The average duration of tracheal intubation was 1.2 days. The remaining patients (n=6) were successfully extubated in the immediate postoperative period.

Average length of stay in the ICU was 1.7 days, average duration of hospital stay was 15.8 days and average duration of stay for patients who had a surgical tracheostomy was 30.1 days.

Removal of the nasogastric feeding tube was used as an indication of time to recommencing sufficient oral intake, and the average time taken was 9.1 days for 22 patients. Several patients underwent radiologically inserted gastrostomy (RIG) procedures due to unsafe swallow. This was either placed prior to admission (n=3) or postoperatively during inpatient stay (n=4). All patients who had surgical tracheostomy had a RIG placed either before or after surgery. This was due to the size of the resection and not as a result of the tracheostomy.

The return to theatre rate for these patients was 13.8% (n=4). Reasons for return to theatre were: flap failure (n=1), partial flap failure (n=1) and neck space infection (n=2). There were no airway issues reported in this series of patients. Patients who did require a return to theatre were intubated without complication, with the aid of an awake flexible bronchoscopic intubation technique.

Discussion

Free flap tissue transfer is the gold standard for reconstruction of head and neck deformity secondary to cancer ablative resections.⁶ Free tissue transfer delivers a seal between head and neck compartments allowing for early return to functional normality. This surgery is a prolonged procedure and is associated with perioperative complications that may necessitate a return to theatre in the first 24–48h.⁷ In the literature, although free flap failure rates of between 1% and 9% are quoted, return to theatre may be as high as 30%.^8–11 Our own practice reflects a flap failure rate of 7%.

In the current series, three patients had elective perioperative tracheostomy. All three had a total glossectomy for a T4a SCC of tongue. Twenty patients were managed with overnight tracheal intubation in the ICU. Six patients were extubated immediately postoperatively and were observed overnight in the ICU. Patients managed without tracheostomy included those with subsites of tongue, maxilla, mandible and floor of mouth SCC. Reconstruction involved osteocutaneous flaps or soft tissue only flaps. Our series suggests that the nature of the flap used for reconstruction or the subsite of resection is unlikely to have a direct impact on the risk of postoperative airway compromise.

The volume of resection as an indicator may be more useful. We assessed the volume using three-dimensional imaging with either computed tomography or magnetic resonance imaging. If more than half of the organ was to be resected, then a surgical tracheostomy would be indicated. When carrying out soft tissue reconstruction, we aim to overfill the resection defect by 30% to account for tissue atrophy that will occur over the following 12 months.

Arguably, patients with maxillary tumours and lateralised tumours of the bone and buccal mucosa do not require a surgical airway. In addition, patients with T2 and T3 tumours of the tongue can be extubated safely after a soft tissue free flap. There may be some contention regarding the need for a tracheostomy in a medial masticatory compartment resection as this involves both the pharynx and posterolateral tongue. In the four cases in this series, we found that the risk at this subsite was no greater than others.

Preoperative airway assessment was carried out by the anaesthetist on the ‘enhanced recovery after surgery’ clinic two weeks prior to surgery. The airway was assessed immediately postoperatively in theatre by the anaesthetist and surgeon. A steroid protocol is followed in our unit where patients are given a 6.6mg dose of intravenous dexamethasone at anaesthetic induction followed by three equal doses postoperatively.

It has been shown previously that overnight tracheal intubation can lead to shorter mean stay in the ICU, shorter overall hospital stay and reduced time before oral intake compared with tracheostomy.^12,13 The cost-effectiveness of an overnight stay in ICU is superior to a prolonged stay on ward-level care.¹⁴ We have shown that the length of hospital stay can be up to twice as long for patients with surgical tracheostomy.

Surgical tracheostomy carries a risk of complications and morbidity. It has been shown that patients who receive tracheostomy are at higher risk of postoperative airway complications.¹⁵ Castling et al reported that the average hospital stay for patients with tracheostomy-related complications was 25 days compared with 14 days for patients without complications.¹⁶

Early complications include damage to the oesophagus, injury to the recurrent laryngeal nerve, tracheal infections, tracheostomy tube blockages and accidental decannulation.¹⁵ Later complications such as trachea–oesophageal fistula are rare because tracheostomies are removed at an earlier stage in head and neck cancer patients. Patients with tracheostomies will also experience changes to their speech and swallowing. The patients in this series who received a surgical tracheostomy did not return to a safe functioning swallow and this was predicted and managed with placement of a RIG, either pre-placement or postoperatively. As these patients all underwent total glossectomies for T4a tumours, it was expected that their swallow would be severely impaired and that RIG would be required.

There is no national protocol for the management of safe airway in head and neck ablative surgery. Formerly, this was our standard practice for multiple reasons. Patients could receive surgical high-dependency level care postoperatively and therefore avoid the need for an ICU bed. In addition, the practice of elective surgical tracheostomy has cascaded through generations of surgeons and remains the accepted standard, although different units will have variations in practice.

Conclusion

This case series has demonstrated that in selected patients a short period of tracheal intubation postoperatively may negate the requirement for surgical tracheostomy in major head and neck ablative surgery. The nature of the flap used for reconstruction or the subsite of resection does not appear to have a significant impact the level of airway compromise following surgery. Consideration of surgical tracheostomy should be given to larger volume resections and medial masticatory compartment resections. Avoidance of tracheostomy can facilitate earlier return to normal diet and earlier discharge from hospital.