Introduction

Zenker’s diverticulum has an incidence of 2 cases per 100,000 in the UK affecting males 1.5 times more than females, and afflicting persons in the 6th to 8th decade of life(1)(2). It is an acquired condition, thought to be due to increased pressures in the pharynx. Under the stress of the increased pressure the pharyngeal mucosa protrudes through an area of natural weakness, Killian’s dehiscence. This is created by the lack of overlap of the muscles of the inferior constrictor, consisting of the obliquely oriented thyropharyngeus and the transversely oriented crico-pharyngeus (3)(4)(5).

This pathology affects patients’ quality of life, for this reason it is important to have an assessment of the patient’s swallowing preoperatively and post operatively. For this we use the EAT-10 questionnaire. The Eating Assessment Tool-10 (EAT-10) score is a validated measure of swallowing difficulty. The questionnaire consists of 10 questions relating to the quality of their swallowing with a grade of 0 to 4 assigned to each question, with a possible score range of 0-40. Any score greater than 3 is considered abnormal (6)(7).

Figure 1

Patients present with a wide range of symptoms, the most common being dysphagia, regurgitation, halitosis, cough, aspiration and chest infections (1)(8). The gold standard for diagnosis is a barium swallow, which will clearly demonstrate the retention of contrast in a pouch and can also demon- strate cricopharyngeal hypertrophy (9).

Many treatment options exist, these include open excision of the diverticulum with an open cricopharyngeal myotomy and excision of the pouch, transoral stapling and transoral cricopharyngeal myotomy. The literature has shown that minimally invasive techniques have similar success rates in terms of symptom improvement or resolution as open techniques with the added benefit of less associated morbidity and mortality (10).

There are two important aspects to treatment. First, to perform an adequate cricopharyngeal myotomy, which is performed by a CO2 laser in our institution. The second is to minimise the risk of perforation.

Surgical technique

At the Sunderland Royal Hospital from January 2018 to April 2019 we have performed 13 sequential cases. A cricopharyngeal myotomy was performed endoscopically with a CO2 laser and at- tempted closure with titanium clips. All cases were done by one consultant. All cases are per- formed under general anaesthesia using a laser safe endotracheal tube with laser safety precautions in place for all theatre staff. A rigid oesophagoscopy is initially performed to identify and assess the pouch, after which a Weerda bivalve diverticuloscope is placed. The microscope with the laser attachment is then positioned with the criopharyngeal bar visualized. The AcuPulse Duo CO2 laser (Lumenis Ltd, Yokneam, Israel) is then used to perform a cricopharyngeal myotomy. Laser settings are 10 watts, superpulse, on repeat with a linear cutting setting. Ethicon Ligaclips (Ethicon Endo-Surgery, LLC, Guaynabo, Puerto Rico, USA) titanium clips are then used to approximate the mucosal edges, a laryngoforce ii clip forceps (Karl Storz SE & Co. KG, Tuttlingen, Germany) is used to apply the titanium clips. All patients undergo NGT insertion under direct vision intraoperatively. Where satisfactory mucosal closure has been achieved the patient is kept nil by mouth overnight and started on liquid diet on day one graduated to liquidised diet on discharge. If there is incomplete closure patients are fed via the NGT and kept nil by mouth. Per oral is started with clear liquids on day 2 or 3, depending on clinical confidence in closure of the perforation. Patients are then advanced to a liquidised diet and discharge home that day, if no symptoms or signs of perforation. Patients were then followed up at approximately 3 months and asked to complete a post- operative EAT-10 questionnaire.

Technique for application of Titanium clips.

It is difficult to apply the clips when operating down the Weerda diverticuloscope. The access to manipulate the clips into the right place requires some practice. The initial laser incision is longitudinal, and in line with the lumen of the oesophagus. However, the incision is closed in a transverse plane to facilitate widening of the lumen. The application of the first clip is the most important. The distal mucosal edge is gently grasped with a microlaryngosocpy basket forceps or straight graspers and advanced to the opposing edge. While the edges are apposed the clip is applied.

Approximation of the edges can be facilitated by reducing the tension on the oesophagus, by ad- justing the Weerda blades. Sometimes it is easier to hold the forceps in an inverted or upside- down fashion to facilitate an adequate view when applying the clip. Subsequent clips are progressively easier to apply.

Case of previous failed endoscopic stapling, with persistent cricopharyngeal bar.

Figure 2

Image of bar viewed through the Weerda bivalve diverticuloscope. (previous stapling clips are just visible)

Figure 3

Image of cricopharynx following laser cricopharyngeal myotomy and division of the bar

Figure 4

Image of complete closure with Titanium clips

Table 1

A total of 14 patients were consented for the procedure, 5 of whom were male and 9 were female. Most patients were in their 8th decade of life with an average age of 71.6 years, and most had age-related comorbidities, the average ASA score being 2.14. Patients were asked to complete an EAT- 10 questionnaire before and after the procedure, the average decrease in the EAT-10 score was 20.2.

At the time of writing the paper 13 patients had successfully undergone this procedure with Steiner clip closure with no major complications encountered. Minor complications had been encountered in 6 cases. The most common issue was difficult closure with the titanium clips in which case Tisseel (Baxter Healthcare Ltd, Norfolk, UK) had to be used to assist with closure. The most common predisposing factor to incomplete closure was friable mucosa which would not hold the titanium clips. One patient’s procedure was complicated by a small fascial breach, however no evidence of a leak was demonstrated on a contrast swallow that was performed in the immediate postoperative period. This was a first-time procedure for 9 patients, while 3 patients had previous laser crico- pharyngeal myotomies, 1 patient had undergone two prior stapling procedures and 1 patient had a previous unsuccessful balloon dilatation.

Just prior to submitting for publication our fourteenth patient had a major complication, re-presenting on the 4th postoperative day with a retropharyngeal abscess requiring return to theatre, drainage and washout via an external approach. They had a subsequent 4-day admission, kept nil by mouth and fed via NGT for 24 days. In order to provide a clear and honest picture of our experiences this report was re-written to include this data.

This increased the average hospital stay from 2.15 to 2.28 days (range 1-4 days). Despite experiencing this complication the patient has recovered well and has had no further problems, their EAT-10 score had reduced from 24/40 preoperatively to 0 at 1 month post procedure, and they have now been discharged.

No patients have had recurrence of pouches or worsening of symptoms in a 6 month follow up period.

Discussion

In a review by Parker et al (11) a comparison was made of stapling to carbon dioxide laser, the main advantage of CO2 laser is the improved swallow function and reduced repeat procedure rate. This is attributed to the ability of the CO2 laser to do a complete myotomy. The stapling technique invariably leaves a residual cricopharyngeal bar, which can continue to cause symptoms. However, the stapling technique is extremely effective at sealing the oesophageal perforation. As a result, the procedure is far safer with a quicker return to oral feeding and shorter inpatient stay.

Early reports of the use of CO2 laser, describe the incision being left open. Patients were then fed parenterally, and perioperative antibiotics used to reduce the risk of mediastinitis. Inpatient stays tended to be about a week (12)(13).

The issue of whether it is important to close the incision was addressed by Anagiotos(14) . The authors compared 2 groups of patients, those with incisions left open, and those closed by sutures. They concluded there was no significant difference. However, they also kept patients nil by mouth for about one week (range 3-17 days). It seems evident that the longer you keep someone nil by mouth post operatively, the less chance of the complications of perforation (15). However, keeping elderly patients nil by mouth for several days can be quite arduous. The ideal solution would be a laser technique with complete division of the bar with a reliable wound closure or a staple gun that can ensure a complete myotomy. The latter has yet to be invented. Various attempts have been attempted at the former. Others have described Metal U clips (16), Tisseel (17) and sutures (14), but achieving a reliable seal is difficult and there were still reports of the com- plications of perforation.

For full disclosure, following the completion of this series we have since recorded one major com- plication when a patient developed a retropharyngeal abscess after re-presenting on the 4th post- operative day and required a return to theatre for drainage and washout via an external approach. They had a subsequent 4-day admission, IV antibiotic administration, hyoscine topical patch and kept nil by mouth and fed via NGT. NGT feeding continued for 20 days after discharge, at which point a water-soluble contrast swallow demonstrated no leak and normal diet was cautiously reintroduced. Intraoperatively during the cricopharyngeal myotomy there were no issues noted and a good closure had been achieved with clips. There were no factors in the past medical history (asthma, fibromyalgia, depression) or the immediate post-operative period that would have raised alarm for the potential to develop a leak or post-operative collection. The patient made a full recovery, and had a good functional outcome with a reduction in EAT-10 score from 24 preoperatively to 0 at follow-up. Perforation, leak and infection are recognised complications of this procedure (11), and the number of cases performed at this unit having undergone titanium clip closure is still small. It is therefore difficult to demonstrate if this single complication is directly related to the use of clo- sure with titanium clips with any statistical significance. Larger case numbers and ongoing audit of complication rates should add weight to the further analysis of postoperative outcomes.

Conclusion

Our technique is novel and attempts to increase the reliability of closure. It is only a small case series and initially seemed to show promise. The disadvantages are that there is a learning curve to applying titanium clips. The technique can be difficult if there is excessive bleeding or friable mucosa. The advantages are that it appears to be a safe technique with minimal risk. In general patients have shorter hospital stays and faster recovery. A larger study number and further research should eventually demonstrate if this method of closure shows any added benefit of statistical significance.

We wrote the paper with the intention of demonstrating a technique that might shorten recovery times and length of stay, and additionally reduce the risk of retropharyngeal abscess and mediastinitis. However, just prior to publication one patient developed such a complication despite having a straightforward procedure. This is the nature of surgery. Instead of shelving the paper we felt it was important to still describe our technique and to highlight that it is not a guarantee of preventing serious complications.